BioSystems 60 (2001) 1–4

Editorial

The physics and evolution of symbols and codes:reflections on the work of Howard Pattee

Keywords: Physics and evolution; Symbols and codes; Howard Pattee

www.elsevier.com/locate/biosystems

The work of Howard Pattee has been the chiefintellectual influence in my career so far, but itwas by no means a sudden realization. None ofhis influential ideas is received with an immediate‘eureka’, rather, his positions result from a thor-ough observation of the essential characteristics ofbiological systems. Unlike ‘headline’ positionswhich send ripples through science by defendingextremist views, Pattee seems to have always posi-tioned himself in the middle of controversies suchas dynamics versus symbols (or computation),selection versus self-organization, and the like.But his consensual approach is by no meanspursued for the sake of pleasing all audiences or alack of vision. Rather, it stems from a long,thorough, and attentive observation that complexevolving systems cannot be fully described by anyone single, neat, extreme theory or another, butrequire instead several complementary models —a characteristic that indeed grants them the statusof complex.

While such a complementary approach doesnot cause ‘revolutions’ or immediate attention, itdoes grow on you. The more evidence one gath-ers, the more single models of complex systemsfail, and one is catapulted back to the kinds of

questions asked by Pattee. This mirrors my owndevelopment as one of his students. Initially inter-ested in artificial intelligence and cognitive sci-ence, I never expected to study under his directionbut simply to take his graduate courses. However,I found myself asking him to sit in on these samecourses, semester after semester, as his ideas be-came increasingly more relevant and coherent inmy mind. During this process I realized that thequestions that bothered me about artificial intelli-gence, could not be treated solely from a cogni-tivist point of view, but demanded an inclusion ofevolutionary ideas (both selectionist and develop-mental). Simultaneously, I became aware of thediverse community of scientists and philosophersinspired by Pattee’s thought. It was with greatpleasure that I saw many of these responding tothe call for this special issue. I hope their contri-butions demonstrate the merits of Howard’s com-plementarity approach to complex systems, and inparticular his insistence on symbolic and materialmodes of description. Most of all, I hope thiscompilation leads others to miss a profound eu-reka, but somehow be puzzled enough to embarkon a slower but deeper examination of the com-plex, complementary nature of living systems.

0303-2647/01/$ - see front matter © 2001 Elsevier Science Ireland Ltd. All rights reserved.

PII: S0303-2647(01)00103-4

Editorial2

After conversations with Michael Conrad, theidea for this special issue was to mark the occa-sion of the retirement of Professor Howard Pat-tee. The goal was to do justice to theinterdisciplinary nature of Pattee’s work, whilemaintaining a directed focus to the area that heconsiders his most relevant contribution to scienceand philosophy of science: The Physics and E�olu-tion of Symbols and Codes. This theme encom-passes theoretical and modeling aspects ofbiological and evolutionary systems. Indeed, atthe core of biological and evolutionary systemslies the notion of codes used to construct organ-isms from inherited information. The study andmodeling of the origin, physical constraints, evo-lutionary potential, and taxonomy of naturalsymbols and codes is extremely relevant to thedisciplines of biology (particularly for the originof life problem), complex systems research, artifi-cial life, cognitive science, and artificial intelli-gence. Authors were invited to relate their workto Howard Pattee’s own description of the prob-lem area:

The two great scientific disciplines of physicsand evolution theory have traditionally beentaught as disjoint subjects. Yet some billions ofyears ago, certain collections of physicalmolecules reached a level of complexity thatbegan open-ended evolution by heritable (sym-bolic memory-based) variation and naturalselection.

Von Neumann was the first to propose ex-plicitly why this ‘threshold’ of complexity re-quires description-based reproduction (takenfor granted by biologists), but his argument wasfocused on the logical, not the physical require-ments. He did not discuss the organizationalrequirements that would allow normal physicalmolecules to function as descriptions, nor washe clear about his logical distinction between‘active’ physical dynamics and ‘quiescent’ sym-bolic descriptions. He did not mention theorigin problem except to say it was ‘a miracle ofthe first magnitude’.

Even if we still do not have a clear picture ofthe origin of life, the significance of this funda-

mental distinction between descriptions andconstructions, that is, between semiotic pro-cesses (rules, codes, languages, information,control) and physical systems (laws, dynamics,energy, forces, matter) reaches to all levels ofevolution. This is an essential distinction fromthe earliest genetic control of the synthesis ofproteins, to the codes and languages of thebrain, to the distinction between the mind andthe brain (the knower and the known, the epis-temic cut), and even to physical theory itselfthat requires a clear distinction between univer-sal physical laws and the local semiotic processof measurement — an area in which there isstill no consensus. This distinction between lawsand semiosis, as well as how they are related,needs to be made more clearly at all levels if weare to fully understand evolution, physical laws,and the languages of the brain.

In biology, the basic physics and chemistry ofelementary life processes as they exist on earthis well-developed. However, our knowledge ofthe semiotic controls and interactions withinand between organisms and in some cases evenin single cells is far from complete. In evolutiontheory it is still not clear that blind variation ina virtually infinite semiotic search space is ade-quate to explain so many successful species.(Howard H. Pattee)

Eleven contributions were accepted for inclu-sion in this special issue, ranging from physics tophilosophy of science. Howard Pattee’s paper Thephysics of symbols: bridging the epistemic cutopens and grounds all subsequent contributions.It discusses his own motivations throughout hiscareer. Pattee notes that the questions physicistsand biologists ask about life are quite distinct,and have lead many to think of living systems asparadoxical. Pattee chose to work on theboundary between these two branches of science,and pursue a theoretical biophysics. From thisinterface position, it became clear to him that tostudy life and its origin one needs to study theorigin of the genotype–phenotype distinction,which he observes, is an instance of the origin ofsymbol systems from material components. This

Editorial 3

symbol–matter or subject–object distinction oc-curs at all higher levels where symbols are relatedto a referent by an arbitrary code, and instantiatesan epistemic cut, a concept he thoroughly devel-ops here.

Michael Conrad’s paper Unit of measurementand motion discusses the implications of his fluc-tuon model for the origin of cellular life and thedevelopment of symbolic systems, as it applies tothe quantum measurement problem, so importantto Pattee’s theory. In particular, Conrad dealswith the concept of complementarity and choice,and what this means for physics theories of lifeand Pattee’s motivations in his origin of life lab inthe 1960s.

Vahe Bedian’s paper Self-description and theorigin of the genetic code describes the model ofthe origin of the genetic code he developed as astudent of Howard Pattee in the 1970s. Thismodel is based on simulation and analysis ofcompetitive code assignments with arbitrary de-scriptor–catalyst relationships. Bedian proposesthat the efficiency of utilization of raw materialsfor the production of a coding family of catalystsis a selection criterion that drives such systemstowards a coded state, prior to the establishmentof Natural Selection.

Peter Wills’ paper Autocatalysis, informationand coding develops Bedian’s origin of life modelfurther. He notes that some genetic sequencespossess the semiotic property of reflexivity be-tween structural components and the functionaloperations they perform to synthesize themselves.His model investigates the embedding of catalyticfunctions in the space of polymeric structuresleading to such reflexivity, which he considersessential for autocatalytic self-construction inmacromolecular systems. He proposes that suchreflexive sequences may serve as the basis for theevolution of coding as a result of autocatalyticself-organization in a population of assignmentcatalysts.

Peter Cariani’s paper Symbols and regenerati�eprocess in organism and brain discusses how Pat-tee’s work can be applied to a high-level under-standing of the brain. Cariani presents anoverview and history of scientific models dealing

with dynamics and symbols in different domainareas, but focusing specifically in neuroscience. Healso discusses Pattee’s notion of epistemic cut inthis context. Finally, Cariani maps the commonroles that symbols might play in the self-produc-tion of organisms and in the regenerative self-pro-duction of neural signaling patterns.

Eileen Way’s paper The role of computation inmodeling e�olution deals with the role of computa-tion in artificial life models of evolution. Shediscusses the notion of scientific models and howthey are used to explain and predict in philosophyof science. In particular, she discusses how com-putational models in artificial life can producescientific discovery. Finally, she criticizes Pattee’snotion of epistemic cut, particularly its quantumtheory interpretations, while exploring how onecould implement a cut between description andconstruction in artificial life models, as observedin biological systems.

Luis Rocha’s paper E�olution with materialsymbol systems presents a theoretical and compu-tational study of the inter-dependencies of symboland matter. He first uses Pattee’s semantic closureprinciple to identify the requirements for open-ended evolution, framing this principle in biosemi-otic terms and offering a definition of the conceptof representation for both biological and cognitivesystems. The second part of the paper describesartificial life simulations that contrast two types ofevolution observed by different populations ofagents: those that reproduce via genetic variationand those that reproduce via self-inspection ofself-organized components. From this study,Rocha concludes that symbols are necessary toattain open-ended evolution, but only if the phe-notypes of agents are the result of a material,self-organization process.

Jesper Hoffmeyer’s paper Life and referenceproposes a broadening of Howard Pattee’s dis-tinction between dynamic and symbolic modes ofdescription of living systems. He suggests a moregeneral biosemiotic explanatory framework whereeven the dynamic aspects of living organisms pos-sess semiotic characteristics, although indexicaland analogically coded rather than symbolic anddigitally coded. He further discusses and specifies

Editorial4

the function of analog and digital codes in evolu-tionary systems.

Cliff Joslyn’s paper The semiotics of control andmodeling relations in complex systems positionsPattee’s semantic closure principle in the discourseof systems theory and cybernetics. He maps dif-ferent kinds of control systems, particularly thoseendowed with semiotic relations, as well as theconstraint relations they entail. Casting Pattee’ssemantic closure principle in this setting, allowsJoslyn to define selection as a meta-level con-straint necessary for obtaining semantic relationsin control systems.

Arantza Etxeberria and Alvaro Moreno’s paperFrom complexity to simplicity: nature and symbolsreviews Pattee’s ideas from a philosophy of biol-ogy perspective. They elaborate on the conceptsof constraint, record, and symbol used by Patteein his semantic closure principle. In particular,they relate Pattee’s observation of symbolic as-pects of matter to the self-simplifying processesadhered to by certain hierarchical systems, such asliving systems. They also discuss the notion ofcomplementarity both as an epistemological andas an ontological principle. Finally, they discuss ifbiological symbols can be regarded as properdescriptions.

Finally, Jon Umerez’ paper Howard Pattee’stheoretical biology. — A radical epistemologicalstance to approach life, e�olution, and complexityprovides a historical overview of Pattee’s contri-butions to and influence in science and philoso-phy. Umerez emphasizes in particular what he

considers to be Pattee’s chief contribution: theelaboration of an internal epistemic stance tobetter understand life, evolution and complexity.This paper offers a perfect closure to a specialissue devoted to a reflection on the work ofHoward Pattee.

As an appendix, this special issue includes acomplete bibliography of Pattee’s writings com-piled by Jon Umerez. I wish to thank Jon formaking this bibliography available here. Grati-tude is also due to all the reviewers who commit-ted their time to this effort. Naturally, deepestgratitude and respect is due to Howard Pattee forproviding us with so many stimulating and pro-found ideas.

Finally, I want to especially thank MichaelConrad posthumously for his involvement andsupport for this issue, as well as his editing guid-ance. He is a tremendous influence for most of thecontributors in this issue, and will be deeplymissed. Many thanks also to Deborah Conrad forher help and support in the last editing stages ofthis issue.

December, 2000 Luis Mateus RochaModeling,

Algorithms, and Informatics Group (CCS-3),Los Alamos National Laboratory,

MS B256 Los Alamos,NM 87545,

USAwww: http://www.c3.lanl.gov/�rocha