Linguistic Society of America

Knowledge in Context: Naturalized Epistemology and Sociolinguistics by Jane DuranReview by: Scott F. KieslingLanguage, Vol. 71, No. 2 (Jun., 1995), pp. 405-406Published by: Linguistic Society of AmericaStable URL: http://www.jstor.org/stable/416186 .

Accessed: 04/12/2014 13:45

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact support@jstor.org.

.

Linguistic Society of America is collaborating with JSTOR to digitize, preserve and extend access to Language.

http://www.jstor.org

This content downloaded from 128.235.251.161 on Thu, 4 Dec 2014 13:45:33 PMAll use subject to JSTOR Terms and Conditions

BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES BOOK NOTICES

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

likely to be developed for this generation of avi- ation or the next' (97). In the meantime, he of- fers a prototype of a visual interface system, along with an extensive technical description for specialists in computer science. I would like to have seen more attention to the 'immediate fixes' that he mentions briefly, including ways in which controllers, pilots, and crews might be trained to use voice-mediated language less er- roneously. Perhaps others can use C's extensive database as a starting point for developing this type of solution. [KATHRYN RILEY, University of Minnesotca, Duluth.]

Generating referring expressions: Con- structing descriptions in a domain of objects and processes. By ROBERT DALE. Cambridge, MA: MIT Press, 1992. Pp. ix, 277. $39.95.

Research in language generation has tradi- tionally suffered from a lack of theoretical foun- dations, compared to language understanding research. This book, based on Dale's Ph.D. re- search at the University of Edinburgh, consoli- dates and builds upon earlier work, helping to establish a body of standard generation tech- niques. D's contributions include the generation of mass terms and plurals. His work demon- strates how ideas from formal semantics (on mass terms, plurals, event taxonomies, and ad- jective ordering constraints) can be exploited in generation.

Ch. 1, 'Introduction' (1-15), introduces D's overall methodology and his focus on cooking recipes. In Ch. 2, 'The representation of enti- ties' (17-68), D discusses the relevant ontologi- cal distinctions. He argues that whether a quantity of substance is described as a mass, an individual, or a set depends on perspective rather than on intrinsic properties of the sub- stance. Ch. 3, 'The generation of connected dis- course' (69-113), explains the overall model of generation. The generation of a recipe is viewed as a process of generating and executing a plan to request the hearer to carry out various cook- ing tasks. A discourse model is maintained, with discourse structure reflecting plan structure. Requests are stated in a feature-structure repre- sentation at a level called KB structure, which is mapped, successively, to levels of recoverable semantic structure (RS), abstract syntactic structure, and surface string. The motivations for feature choices at each level are not dis-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

cussed explicitly; for instance, RS uses prag- matic, semantic, and syntactic features. D doesn't clarify what's in the world versus what's in the KB structure, or the relation between them.

Ch. 4, 'Generating referring expressions' (115-67), distinguishes various referring expres- sions in terms of their KB-level semantics. Sub- stances described as individuals (or mass) map to singular (or mass) NPs. When accompanied by quantity specifications, individual and mass descriptions map to pseudopartitives and mea- sure phrases, respectively. Sets map to plural NPs. A pound of carrots is treated as describing a set (65); unfortunately, in D's system, this will be realized as a plural NP (165).

In Ch. 5, 'Generating anaphoric references (169-232), D considers when it is appropriate to generate various forms of subsequent reference. This depends in large part on the entity's cogni- tive status. New entities become indefinites; centered or recently mentioned entities are either elided, pronominalized, or generated as one-anaphors; otherwise, definite NPs are gen- erated. D introduces an algorithm to determine what properties must be included for a descrip- tion to adequately distinguish an entity in a given context.

In Ch. 6, 'A worked example' (233-58), D compares a generated example with an authen- tic recipe. The restriction to single-clause con- structions makes for slight awkwardness in the output, but otherwise the results are impressive. It would have been useful to compare different recipes generated from a given input.

The book provides an excellent window into representational techniques used in engineering a highly sophisticated language generation sys- tem. [INDERJEET MANI, Georgetown University and The MITRE Corporation.]

Knowledge in context: Naturalized epistemology and sociolinguistics. By JANE DURAN. Lanham, MD: Rowman & Littlefield, 1994. Pp. xiii, 238. Cloth $55.00, paper $22.95.

The stated goal of this book is 'to heighten awareness of the social sciences ... among those interested in contemporary theory of knowl- edge' (xiii). D also wishes to explain to social scientists 'what philosophers mean by matters epistemologicai' (xiii). She traces the develop-

405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405 405

This content downloaded from 128.235.251.161 on Thu, 4 Dec 2014 13:45:33 PMAll use subject to JSTOR Terms and Conditions

LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995) LANGUAGE, VOLUME 71, NUMBER 2 (1995)

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

ment of epistemology from a formal discipline concerned with constructed examples to one concerned with the everyday processes that allow people to 'know'. This development par- allels the similar evolution of sociolinguistics. D shows how both disciplines have moved from the ideal to the real, largely by taking the social context into account. While she does reveal the parallels, D does not fully succeed in making her points accessible to nonepistemologists.

Part 1, 'The development of sociolinguistics' (3-81), includes an overview of the field, a his- torical account of its evolution from classical linguistics, and one chapter each on influences of the work of William Labov and Erving Goff- man. The chapter on Labov focuses mainly on methodology; although D gives somewhat cur- sory treatment to Labov's concern with explain- ing language variation and change, her concern is to show how the assumptions of sociolinguis- tics differ from those of Chomskyan linguistics. Her review of Goffman's interaction theory highlights the importance of social processes, both for how we use language and how we un- derstand our world. Notably, however, she does not refer to Goffman's notions of alignment and frame, which have been widely adopted by interactional sociolinguists.

In Part 2, 'Sociolinguistically naturalized epistemology' (83-145), D provides a twentieth- century history of the field of epistemology and an overview of the field, and presents her model combining sociolinguistics and epistemology. Because this part is the most epistemologically oriented and contains unexplained jargon, it is the most difficult. But it is also the heart of the book, in which she outlines her synthesis of so- ciolinguistics and naturalized epistemology. The model uses sociolinguistic constraints to explain how humans justify their knowledge. While flattering to sociolinguistics (because the field is seen as central to human cognition), D's sociolinguistic model is somewhat disappoint- ing, mainly because she proposes few con- straints. Moreover, these constraints do not seem to have been chosen systematically. She acknowledges, however, that the model is 'al- most purely exemplary' (120). More unfortunate is her use of constructed data to explain her model; given her argument for 'natural' data, the model would have been more convincing if it had been motivated by 'natural' interactions.

D discusses sociolinguistic research that adds to her model in Part 3, 'Extensions of the view' (147-85). Her choices of illustrative research

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

are excellent; she brings together some of the best work from different corners of the field. But exactly how the research supports her model is not clear. A more explicit connection would have been useful, for epistemologists as well as for sociolinguists. The final section, which presents a summary of the book, clarifies the main themes and points of the earlier chapters.

These points were sometimes lost in D's style; she often gets ahead of herself, failing to make explicit connections between smaller and larger points. Although this style makes for diffi- cult reading, the book is a valuable, truly inter- disciplinary volume; few scholars attempt to understand more than one field, let alone try to find a synthesis. [SCOTT F. KIESLING, George- town University.]

Talking data: Transcription and coding in discourse research. Ed. by JANE A. EDWARDS and MARTIN D. LAMP- ERT. Hillsdale, NJ: Lawrence Ern- baum, 1993. Pp. vi, 325. Cloth $59.95, paper $27.50. In the lead essay in this collection, JANE ED-

WARDS ('Principles and contrasting systems of discourse transcription', 3-31) defines the focus of the collection as the interaction among theo- retical frameworks, research methods, and tran- scription/coding systems, claiming that 'to the degree that transcription conventions influence perceptions of the data, it seems important to consider them explicitly, and to systematically enumerate their underlying assumptions and im- plications for research' (4). Following Johans- son (Some thoughts on the encoding of spoken texts in machine-readable form, University of Oslo, 1991), Edwards reiterates two general principles of transcription design: authentic- ity-preserving information needed in research in a manner which is true to the nature of the interaction, and practicality-developing con- ventions of readability, applicability to new data, and expandability for other purposes such as electronic preservation (4). She then dis- cusses some specific principles of transcription design, such as the importance of systematically discriminable, exhaustive, and contrastive cate- gories.

The remaining papers in this first section de- voted to transcription systems arise from differ-

406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406 406

This content downloaded from 128.235.251.161 on Thu, 4 Dec 2014 13:45:33 PMAll use subject to JSTOR Terms and Conditions