Haskins Laboratories Status Report on Speech Research1992, SR-109/110, 221-224
Starting on the Right Foot*A review of Marilyn Jager Adams' Beginning to Read:
Thinking and Learning about Print**
Marilyn Jager Adams has performed a valuableservice to all who wish to improve how reading istaught. Her book presents a comprehensive andscientifically responsible treatment of problems ofimmense social importance-problems that partlybecause of their very complexity are too oftentreated cavalierly. This book is required readingfor professionals engaged in research on designand assessment of programs of reading instructionand research on diagnosis and treatment ofreading disability. It is also a valuable resourcefor a wider readership in psychology, cognitivescience and education. Indeed, anyone who needsa clear-headed synthesis of relevant researchfindings bearing on the problems of learning andteaching to read can profit greatly from this book.With unusual thoroughness, Adams has reviewedthe mass of research literature that bears on thedebate between advocates and adversaries of thecode emphasis in reading instruction. The tone isalways constructive. She avoids the rancor that sooften accompanies discussion of these issues.Though even-handed in her treatment, Adamsdoes not wrap herself in the cloak of the eclectic;after sifting the evidence, she draws strongconclusions and states them boldly.
This book originated with a mandate from theUnited States Congress for a new appraisal of theplace of phonics in teaching children to read.Inundated with complaints about the performanceof the schools in imparting literacy, and confusedby the welter of conflicting voices from theexperts, Congress enacted legislation that ledultimately to the U.S. Department of Education'scommission of this report. Responsibility forproducing the report was placed in the hands ofthe Center for the Study of Reading, University ofIllinois at Urbana-Champaign. Adams, a cognitive
and developmental psychologist at the Center'sbranch at Bolt, Beranek and Newman inCambridge, Massachusetts, was chosen for thetask.
Given Adams' extensive background ininvestigation of basic reading processes, she was alogical choice and the choice proves to have beenan excellent one. Charged with the responsibilityfor presenting a thoroughgoing clarification of theissues that divide the two sides in what JeanneChall has called "the great debate," Adams wasgiven a free hand to shape the report. A panelconsisting of well-known reading experts fromaround the nation was assembled to offer adviceand criticism of interim drafts, but the book waswritten by Adams, not the committee. And to hergreat credit, the book is highly readable. It hasnone of the dryness one often finds in a technicalreport. The book displays a graceful and informalwriting style and betokens an uncommon abilityto use the language well.
As Adams points out, this book has apredecessor: the task of reviewing the relevantresearch literature was undertaken in the 1960sby Jeanne Chall whose report was publishednearly 25 years ago (Chall, 1967). Appropriately,Adams often refers to the earlier work. It, too, wasa praiseworthy review, but time does not standstill. The unprecedented technological explosion inthe work place presents ever greater demands onreading skills. Moreover, the crisis in the schoolshas intensified, consensus on a remedy for theunacceptably high rate of illiteracy in our societyseems as elusive as ever.
In the meantime, research activity hasmushroomed both in quantity and in variety. Animportant new development since ChaIrs bookappeared is the rediscovery of reading as a central
problem for investigation by mainline psychology.No less significant, reading and orthography havebecome major concerns within the fast-growingfields of applied linguistics and the psychology oflanguage. One consequence of the remarkablesurge in research on reading is obvious: Anyonewho would undertake to review the literaturemust be prepared to digest and critically evaluatean enormous range of material. Accordingly,heavy demands are placed on a reviewer'sknowledge and critical judgment. On the whole,Adams proves more than equal to the task.
The report has five parts. Part I deals with thenature of writing systems, the origin of thealphabet and the place of word recognition inreading. Part 2 presents the rationale forapproaches to instruction that emphasize phonics,and it reviews research that attempts to comparethe efficacy of this approach with otherapproaches. Part 3 presents conceptions of readingfrom the standpoint of laboratory analysis of whatskilled readers do. It presents a model of thereading process that encompasses each of thecomponents of reading skill and their integrationin the act of reading. Part 4 articulates the goalsof instruction in reading from the standpoint ofthe analysis of the skills of the mature readerpresented in Part 3. Part 5 discusses research onthe processes involved in learning to read. Part 6summarizes the conclusions reached from thereview of the research literature and discusses theimplications for teaching and leaming to read.
Adams begins with a discussion of the nature ofwriting. It is noted that true orthographies, unlikepicture writing, represent words, and not mean-ings directly. This is an appropriate starting pointbecause it underscores the key 8
nnificance of theword in reading. The importanc'. lpprehendingeach and every word in the text ..mot be takenfor granted, because it is unfortunately true thatsome popular programs of beginning reading in-struction encourage the novice to skip words or toguess in the search for meaning. Adams leaves usin no doubt where she stands: This is bad advicefor a beginning reader or anyone else. "Unless theprocesses involved in individual word recognitionoperate properly, nothing else in the system caneither (p. 3)." The ability to identify printed wordsis necessary but not sufficient for reading; it mustbe backed up by well-oiled mechanisms of lan-guage comprehension. Reading depends on a sys-tem of skills whose components must meshproperly.
Alphabetic forms of writing are codes on thephonological structure of the language, or more
properly, the morphophonological structure. Byusing letters to represent the several dozenconsonant and vowel sounds of the language,alphabets achieve their great advantages overother forms of writing: First, economy-a smallset of symbols is sufficient to represent any and allwords in the language; second, transparency-auser who knows how the system works canusually recognize words in print that werepreviously known only through spoken language.Adams' account notes that these advantages comeat a cost that must be borne by the beginner.Every alphabetic system presents its users with aproblem of cognitive penetrability. Because vowelsand consonants are co-produced and overlapped intime, these abstract phonemic units are notrealized in speech as physically separable chunksof sound. That is probably one reason why theyare often difficult to apprehend consciously(Liberman, Shankweiler, Fischer, & Carter, 1974).For the purposes of speaking and listening,language users need not attain awareness ofphonemes. But to grasp the principle (by whichalphabetic writing represents the phonemes andmorphophonemes of the language), a would-bereader must first identify the speech units thatthe letters represent. Consequently, the grasp ofthe alphabetic principle is a rather sophisticatedintellectual achievement.
Because the orthography of English is complexand often irregular, some commentators haveoverlooked that it is, nonetheless, essentiallyalphabetic. Adams does not make that mistake.Yet to dwell on the irregularities, as she does atthe end of Chapter 2, is to invite a reader who isless than astute to draw the wrong conclusion andto miss the larger point: that there is a system tobe learned and that, even in English, knowledge ofthe orthography is productive.
The chapters that follow present a much neededand thoughtful analysis of the pertinentinformation on phonics and reading. As forphonics, the term itself has long been a source ofconfusion. For the most part, Adams uses the termsimply to denote instruction aimed at instillingthe alphabetic principle. Well and good. Butunfortunately the term has other connotationsthat are hard to shake off: In the minds of somepeople, phonics denotes an old-fashioned anddiscredited method of teaching reading by havingchildren attempt to recognize a word by speakingthe "sound" of each letter. The method impliesthat what a reader does is to approach wordspiecemeal by translating the letters that make upa word into their phonetic equivalents, letter by
Starting on the Right Foot 223
letter, as though reading were simply spellingaloud. Thus the term phonics has come torepresent an inapt caricature of the readingprocess. Accordingly, Liberman and Liberman(1990) recommend substituting for phonics Chall'sterm, code-based approach.
As Isabelle Liberman (who is cited by Adams onthis point) often explained, letter-by-Ietterencoding is assuredly not what a successful readerdoes. The word bat contains one syllable, notthree; the word is not buh-a-tuh but bat. Yet somebeginning readers will say something like "buh-a-tuh when asked to read the word and will nevermanage to discover that the word is bat(Liberman, 1973). In Adams words, "It is asthough these children can find no connectionbetween the sequence of sounds they haveproduced and the highly familiar word which theyhave 'read.' It is not enough to have memorizedthe sounds that go with each letter. To make useof those sounds, the child must realize that theyare the subsounds of language" (p. 208).Beginners who are stuck in this way can be helpedto develop phonological awareness, that is, tobecome aware of the phonological structure ofwords, by identifying their phoneme and syllableconstituents. Then they are prepared to grasp thealphabetic principle and can begin to build wordrecognition skills on a solid foundation. As Adamsnotes, experienced readers parse the letterstrings, ordinarily apprehending sequences ofletters that correspond to a demi-syllable atminimum. According to laboratory researchdiscussed in Part 3, such sequences constitute themajor spelling patterns that experienced readersimplicitly recognize as wholes.
Spelling patterns must be not only apprehendedbut also overlearned to the point that wordrecognition can become unhesitating andautomatic. Speed, as well as accuracy, isimportant because the fast-fading short-termmemory forms the stage for the integration ofwords into syntactic units. If word decodingroutines work poorly, all other aspects of readingwill be hampered and comprehension will becorrespondingly poor, a point often stressed byPerfetti and his associates (Perfetti, 1985). Thus,although word recognition per se is not the goal ofreading, getting the meaning of the text dependson it. And word recognition, in tum, depends onaccurate identification of the lower-level buildingblocks: the letters and the spelling patternsformed by letter combinations.
In Part 3, Adams sketches a model of readingthat derives largely from the work of Seidenberg
and McClelland. The chief characteristic of thismodel is that information the reader derives fromprint interacts freely and at every level withstored knowledge. Thus the model contrasts witha hierarchical model in which information flow islargely unidirectional and bottom-up. Otherresearchers have maintained that an interactivemodel does not readily account for the importantdifferences between reading and speechperception. Above all, it offers no explanation ofthe fundamental fact that speech is acquired byevery neurologically normal child whereas readingskill is far from universally acquired. For someresearchers, a unidirectional model seems dictatedby the modular nature of the language apparatus(see Crain, 1989; Fodor, 1983; Shankweiler &Crain, 1986). Of course the question is notwhether linguistic input (whether speech or print)must make contact with stored knowledge, buthow and when. The modular view supposes thatprocessing within the language module isaccomplished before the linguistic input isintegrated with other aspects of cognition. On thisaccount, it is emphasized that word recognition byear is privileged in the sense that it is served bymechanisms that evolved in our species and thatform part of a coherent biological specialization forlanguage. In contrast to speech, the alphabet is anartifact. Learning to use it is a cognitive task in away that primary language acquisition is not. Ithas been argued that an adequate theory ofreading would have to explain the difficulty ofreading and the comparative ease of acquiring aspoken language (Liberman, 1989).
After examining the myriad studies comparingprograms for the teaching of beginning reading,Adams concludes that the great majority ofprogram comparison studies indicate thatapproaches that incorporate code-basedinstruction"...result in comprehension skills thatare at least comparable to, and word recognitionand spelling skills that are significantly betterthan, those that do not" (p. 49). This, she notes, isexactly the same conclusion that Jeanne Chandrew 25 years earlier. Code-based approaches thathelp the beginner to appreciate that words havean internal phonological structure and torecognize that word spellings represent thatstructure have the edge over programs that passover these aspects.
While stressing that these program comparisonsare essential, and have been highly informative,Adams is sensitive to the limitations of theseresearch studies and in Chapter 3 sheknowledgeably discusses the reasons why they so
often yield noisy data. The classroom teacher, whois charged with implementing the program, isoften the weak link. Adams' conviction thatsuccessful readers must grasp the alphabeticprinciple and that code-based teaching is the bestway to help beginners to grasp it stems only inpart from such program comparisons. At least asimportant are other research findings which arediscussed in detail in this book. The pertinentevidence comes from a variety of sources: Itincludes the findings of research on prereaders,prediction studies seeking to identify thosepreschoolers who are at risk for reading failure,follow-up studies on the long-term educationalconsequences of failing to crack the code in theearly primary grades, studies identifying theshared characteristics of unsuccessful readers,and finally, the picture of reading derived fromresearch on the skilled reader. Adams concludesthat all these lines of evidence converge inunderscoring the vital importance of helpingchildren grasp the alphabetic principle from thebeginning. This entails giving prereadersadequate preparation for learning to read byinstilling phonological awareness (introducing,through well-chosen word games, the fact thatwor...