You are on page 1of 86
BOSTON STUDIES IN THE PHILOSOPHY OF SCIENCE. HUMBERTO R, MATURANA and FRANCISCO J. VARELA [EDITED BY ROBERT 8, COHEN AND MARX W. WARTOFSKY AUTOPOIESIS AND COGNITION The Realization of the Living With prafice to “Autopoeis by Si Stafford eer = D. REIDEL PUBLISHING COMPANY DORDRECHT : HOLLAND / BOSTON: US.A. VOLUME 42 LONDON: ENGLAND O47 dye ay of Congress Ctgingin Plena Dat (@oson atin the pony of ee 42 2, Copan. 3. ie nla) So ison) 19-24iae ua? (88) ‘pi by Rs ating Coty, 601 Bey Stet, Higa A 02083, 0S. Ina ihr count, ed eat 1.0 Box 322,200 All Dordrecht, Holand Rese Pbting Company ia mens fh Klar Grup Apotess The Orton of the Ling was cgay bie in Cb dest i Brag Ss Ys © 172% Hl a pts Reserved ‘coneeh 190 by. Reel Puig Company Done, Holand ‘No pao the mates tet by hs opt ote mayb poet or Uhl nay foo aby ty nen, ecco or mecha, Ince potorsping, enti ty any nortan nd ‘Str! yen, hou writen peemaon fom th coy met EDITORIAL PREFACE ‘This sa bol, ballant, provocative and puzaing work. It demands a raat shit in standpoint, an almost parsdoicl posture in which ving systems are ‘ested in terms of what Lies out the domain of dsciptions. Profesor Humberto Maturans, wth his colleague Franlco Varels, have undectaken the consrvtion ofa syremati theoreti biology whch attemptsto define ving fystems not a they ae objects of observation and description, no even a i {enacting stems, buts selfcoataned unites whom only reference sto then selves Ths, the standpoint of description of such unites from tho ‘outside’, Uy by an observe, sceady stems to lla th fundamental requirement ‘which Matrana and Varela post forth characterization of such systems pramely, that they ae autonomous, eieferrng and slf-constracting closed ‘stent In short eutopoeic ystems in her erm, Yet, on the bas of such ‘conceptual method, and sucha theory of living yes, Maturans goes onto ‘efine cognition a 2 biologi phenomenon; as, i eect, the very nature of al lng systems. And on this bas, o gnerate te very domains of interac tion mang sch ysems which constitute languags, description and thinking. “The rade! shift in sandpolnt hore requles an imaginative leap andthe sbandonment a the outset ofthe standard chaacterzations of ving systems in tenn of funeton or parpot, or of expaismenvironment rations, oF of ‘cual interactions with an extemal weld, or een in terms of information, ‘coding and transi, I effect, Maturasa and Varela propose a theoretical Diology which I topological, and a topology in which elements and ther relitlons constitute a claed system, or more radial ail, one which fom the pont of view’ ofthe system ite, entirely selfsefrenti and has no ‘outside, Lebnizian for our day. “The work demands and deserves careful reading. It is technica, forma, tial, pilsophial and boldly imaginal, Its gorously constructed, and insofar as ts theoretical biology, It remains uncompromising) abstract tnd formal. Yet it smalls of the medical laboratory and of the working do- ima ofthe neurophysologis. Where the interpretation ofthe formal theory ‘maps It into the domain ofthe nervous system, the insights and suggestions for ferther interpretation ate exciting indeed. And we expect nating ls, here and to come. ’ LETTS " [EDITORIAL PREFACE Maturan is pethaps best known to the pliloopher andthe sientifc onspeialst as 4 coauthor of the dase 1989 paper What the Frog's Bye Tals the Frogs Bain’ (with Lettin, McCulloch snd Pits). Slace thea, he ‘us worked on the anatomy and neurophysiology of vison, especially on olor ria. He has aso been teaching medical students, The problems and Pzles which emergol in his earch and teaching led Maturana to develop distinctively atemative theoretic! framework in order to answer te ques- tions, What isa Living Syste?” and “What ls Copation?” Tae consequence Df his invertigntions, and of his constuction of living systems selfing, ‘esefecrng autonomous unites, s tht he dicovered thatthe wo guestions have e common answer. He writes, "Living systems ae cognitive systems and living as a process a process of cognition.” ‘We ae very pleased to introduce this major theoretical workin the Boston Studien the Phlsophy of Science. Te integration of biological theory, formal corstuction, epistemology (end, further, Maturan's suggestions of the nature of interacting systems a8 Kind of biological sociology, and his sketch of the ethical inplieatlons of such a construction) ~ all mark these ‘vo studies as among the most engnal attempts ata systemuti biology in Adacades, and 5» profoundly phlosophical wosk Center for Pitosophy and tory of Science Boston University July, 1978 GENERAL TABLE OF CONTENTS [EDITORIAL FREFACE GENERAL TABLE OF CONTENTS INTRODUCTION (by Professor Maturans) BIOLOGY OF COGNITION Dedieation Table of Content 1. Introduction U, The Problem IIL Cogatve Function in General A. The Obeerer BL The Living System €. Evolution D._ The Cognitive Process 1V. Cognitive Function in Partial ‘Nene Cells Achitectre Function Representation Description ‘Thinking Natural Language Memory and Leaning ‘The Obserer \. Probl in the Nurophysalogy of Cognition VI. Conclusions ‘VIL Post Seriptam vi GENERAL TABLE OF CONTENTS AUTOPOIESIS: THE ORGANIZATION OF THE LIVING tac (by Sir Stafford Beet) Introtuetion 1. On Machines, Living and Otherwise 1 Machines 2, Living Machines 1, Dispnsbiity of Teleonomy 1. Prposelesess 2, Indiviusity Il, Embodiments of Autopoiesis 1. Descriptive and Causal Notions 2, Mocs Embodiments 3. Origin 1¥._ Dy of Antoine Subordination tothe Condition of Ualy Plasicity of Ontogeny Reproution, a Complication of the Ulty voltion, Historie Network Second and Third Onder Autopoetic Systems . Preece of Autopoless 1. Biological Implications 2, Epistemoloiealiplcations 5, Cognitive implications Appendix: The Nerous Systm Giowary pinLioGRarHY INDEX OF NAMES mi FOREWORD Everything that Lsty in the introduction is my exclusive responsibly Franco Varela hasbeen ganerovs enough to grant me thi eadom in what concerns the esay that ws wrote together. His views about it he expreses fully and independently in his book Prindples of Biological Autonomy, publihed by Euovie-North Holand, New York, 1979, ‘Also, [wish to acknowledge the subte det which the contents ofthis book owe to the any dlumiastng conversations about all the topics here contained that Ihave had over more han fourteen years with my dear end Profesor Felix Schwartzmana, ofthe University of Chile Repntton fh car stop neve ‘Al aows tht do at ene he Yous ofthe erent ity nate production ‘Maton The wary Sadd sey ining te bowdary ine ane eas, {eet ihe names inte coe etn, Th eer farm close wth pet Co odacten ad sonsubon Ina recone network raed a4 cnce'e ttt te pees elon of the companents isthe ter, nse ‘elton dhe coment clags of he twa a ig ttm Ito. The whole epet Sond natok pada b he aw ts ‘he deiced cose bacnay of te network inde ie coy atc Spots ofthe stun tron the pol pace INTRODUCTION In hs intoducton 1 with 19 make sme comment in station to the two ‘aye that make thi book, but in onder todo so T mut, at lest apart wie bout how thny came ito bing. wistoRY Since my cildhood 1 have been intrested in animals an plants, and I fe ‘wenty asked myself what made them ving. Tus, in 1948, in my fst year tsa media student, [wrote «poem whose fist sana was “qué ese muerte para el que la mir? Qube la met pra el quelasent? Pesadezigots,incompronsble, dolor que el egolmo tae, par be; Sect, puzy nada, para te. Sin embargo el no siente ‘ue au orgllo se rela, que sa mente no soporta que tsa muerte nada quede, (que tras a muerte et la moecte Lotro, en supa, ens enc, ensu majested inconsiente sent, ‘ada sone, nada sabe, porque la ert esa muerte Y tas a moet est la vide ‘ue sin la muerte so es muerte What x death for the beholder? / What b death for the dying? ( A weight beyond knowledge or understanding, / A pain forth slfasating 30, for the one; / For the othe, silence, peace, ad nathngness. Yet the one fel, bis pie in anger J And in is mind he dos not seept | That beyond death nothing should srs, | And that beyend death / There shoud be only death 1 The other, inhi enc, | In ht unknowing majesty fea, He fel oth. ing, he knows noting] Because death death / And life without des s ‘only emptiness. jf Pn merRopucrion “The poem snot avery good one, yt it contains the implit question: ‘What kindof ystems ae ving eyteme that they may die and how come that they oie in 1954 1 wont to study biology in Baglaad and the US.A., and when | returned to Che im 1960, afer SX years of study and research abroad, 1 tegun to work tthe Medial School of the Univers of Chile in Satigo as 2 research autos in the Department of Biology. Thve I was inveived in two kind of acti; I collaborated in teaching a cousin general biology forthe medial students, and Idi serch inthe Hels of neurophysiology land nevroanatomy. In my teaching my responsibilty was to convey tothe Inet stunts some understanding ofthe orpnization of ving systems 5 futonomous ene, as well at some understanding of thelr posse origin fon earth, In my revere T wanted to apply to the stody of form and color ‘ison in birds the same approach that J. Y. Letvn and [had used in the "Hady of form vison nthe Fog To discovered through my teaching tht the conta question that the students woul alwys ask wat: ‘What & proper oling systems that ha its fovign when they ofginated, and has remained invariant sac then in the Succession of ther generation” At the sume time Troon realized ia my research that my cetal purpos inthe stay of color vison eould not be the study ofa mapping ofa colorfol world onthe nevous system, but ater that Ht had to be the understanding of the parpation of the setina (or ‘etvous sytem inthe goneration ofthe color spac of the observer. “Ara ewlt of these different acts Ientored a situation in which my academe was vided, and I rented myself in each of the answers (0 to queitions tht seemed to lead in oppose directions, namely: "What i ‘he organization ofthe living” and "What takes pce in the phenomenon of pereptin? Tat me speak about how I acd thm ‘ist Question: What isthe ompanization ofthe ving? ‘When this queton was fst asked by the students, although twas the same quit tat had ben uckiag in my mind for many yeas, cold not answer (Thad prepied myself for this moment, but when jt came and I ted to answer itn a manner stsactory for the students I eaized that Thad to think everything anew, I could speak about form and function and astonish ‘my sunents and myself atthe humony of nature, exating the ines ofthe Svtonment and the Anse ofthe indivi, ooo chim tat the question ‘wav a wry difficult one and that it could not yet be anevered dv to oor metropuction su Insufficient knowledge. We had to accep that we could recognize ving ‘ystems when we encountered them, Dat that we could not yo say whet they tree, I could enumerate features of ving systems such at reproduction, heredity, growth, italy, and soon; but how lng a ist was necessary, sen would thelist be completed? Ia order to now when the is was com ‘lied Ud to know what ving system was, which was, infact, the question {hat wanted to answer in the Rint place by producing such 3 lis. T could {zea shout adaptation and elution, about development and differentiation, fd show how al these phenomena were ed together by the phenomenon of fara seleton; but the question: "What was the invarant feature of ling ‘Sistem around which natural selection operated”, remained unanswered [Every approach that I could attempt and that Td attempt left me a the stating point, "Yet Tobvously ad some ikling of what waste correct answere, because 1 rejeeted the unsatisfactory ones. Afar several yeats of these vaio at tempts I realized thatthe difficulty was both epistemological and linguist, and tht both my wif and my old professr, J. 2, Young, were righ: one ea ‘oni say with 3 given language what the language permits, Thad to stp look ing a lng syten x open systems defined ina envizonment, and I nesded 4 anguage that would permit me to deseribe an autonoreus sytem in 9 ‘manner that rained autonomy’ a feature of the system or entity specified by the desrption. In other words, ny attompt to characterize living stems ‘with notions of purpse of fancton was doomed to fal berase these notions {he intinsclly referential and cannot be operaonally used to characterize ‘ny stem azansutonomous entity. Therefore, notions of purpose, pal se ‘or function, a oe ejected, but inal Tdi nt know how Aecordinaly, Tei in my lectures several approaches in oréer to find 3 way of speaking ‘shout living systems ina manner that would paip thelr autonomy asa phe ‘nomenon of thelr operon ee unitary systems. Thus, eventually, 1 made & ‘tinction between what Iclled sexeferred and allowetered systems, & ‘tintin that separated systems that could only be characterized with reference to themoelves, suc a ving systems, fom systems that could only ‘be characterized with ference co a context. di this in onder to emphasize that whatever took plas in ving rystems as living systems, took place as ecesualy and costiutvely detemined in relation to themselves because thelr being defied at units though selfseference wae thir msnaer of futonomy; and that whetver took place in other systems took place at fonstutvely detnlned in relation tothe context with rexpect t which they were defined as unl. This way of speaking wasnt ful tifctory xv IrropuctION ‘but led me to sel that what was indced needed was the characterization of 1 king of system which, if allowed to operate, would operate ina manner Inditngutable from the operon of living sytem, and that one shuld do so ting only neighborhood relation razed trough the propertie of the components ofthe system. Iwas with such am that Fepoke for he fet time in 1969 of tiving systems a ystems defined a unite through the baie tirularty of ther production of ther components Second question: What takes place inthe phenomenon of perception? ‘When Jery Y.Letvin and I wrote our several atleson fro vison (Leti, J. Yo, H.R, Maturana, W. &. MeCalloch and W. H, Pits 1959 Marana, H.R J. Letvin,W. 8. MeCulloch and W. H. Pits 1960], we with ‘the impli assumption that we were handling « clearly defined copie situation: thece was an objective (asote reality, extemal fo the animal, and Independent of i (aot determined by 1), which it could percsve (coiz), and the animal could use the infomation obtained ini perception to com ‘ue 4 behavior adequate to the perceived station. This asumption of outs appeared clearly in our language. We deserbed the various kinds of retina ‘angio cell as feature datectors, and we spoke about the detection of prey tnd enemy. We knew that was not the whole europhysoiogal ony, 2 ‘vat apprent partcly in the dcusson ofthe article called “Anatomy snd Physiolony of Vision inthe Frog (Rana pipiens). But even there the epistemology that guided our thinking and wating we that ofan objective reality independent ofthe observer. Thus, when Samy Fre and T bapa {o work with pigeons in 1961, fast studying form won, we approgched ‘that study with the same fundamental view. No problem aroee then ane without any difeulty we ould characterize many cases of retinal gnglion calls. Yet, when Gabrisla Uribe joined us and we infact began to sty ool ison in 1964, it soon became apparent to us that that approach lads to deep trouble. Newroplysiologieally we did not see anything fundamentally diffrent fom wat other scholars had aleady seen. We found the case ‘ypes of ganglion ells with separate, concentric or ovelapping opponent special preferences. Bul we ali found: (#) that although the geometry ‘ofthe receptive Hels of the ganglion cells with opponent spectral prefer: ness hal nothing todo with the gomety ofthe val object, the geome: try ofthe vial bject had todo withthe vsponse of those cls; end (S) that we could not account for the manifold ehvomatic experienc of the observer by mapping the visible colorful world upon the activity of the nervous sem, because the nervous system Seed ture oometi lations inrropuction ” to specify color distinctions. ferent logy were necessary. ‘There are many vial configurations, with uniform and vasasted spectral compositions, in simple ané complex geometrical forms, that ge rise to {ndbtingushable color experience. How thould one, then, look forthe Invaianoes inthe sty of the nervous system, If any In elation tothe ereption of color! Afer we reaied thatthe mapping of he extemal wold ‘vasa inadequate approach we found that the very formulation of the ques- ton gave us the cue, What if stead of attempting to correlate the activity ‘nthe retina with the pysical stl exteeal tothe organism, wedi other ‘wise, andre to corelatethe activity inthe retin withthe color experince ofthe subject? Such 4 sep entailed wo iffiulies, On the one hand it required the efinion of 4 eference which would parmit the chsracterizaton of the stinty of the ctinl ganglion calla independetly of the stimulus 8 such; fon the other hand it required us to close the nervous system and tet the report ofthe color experience as if t represented the state of the nerous system a 2 whole. In other words, the new approsch required to tet Seriously th activity ofthe aervous system at dotemined by the nervous system Itself, and not by the extemal word; ths the extorsl world would ‘only have atigering role in the eles of the internally determined activity fof the nonaus system, We did this rigorously, and showed that sich an approach di indeed penmit us to generate cho whole colorspace of the observe, That was a Yery fundamental result that we published in very unknow ste [Maturan, HR. . Unb, and S, Frenk, 1968}. But what was sill more fundamental was the discovery that one had to lose off the nervous sytem to acooent forts operation, and that perception should not be viewed asa grasping of an extemal reality, but rather a the ‘specication of one, because no dtintion was pose between perception {nd hallucination ithe operation of the nervous system a a closed network, ‘Although we artved at thi conlision through the study of color vin, ‘there ae any exer expedmental studies (ch a those of Stone on the rotation of the eye ofthe salamander inthe easy "ories) that could also Ihave led to an understanding of the nervous system asa closed network of Interacting neurons. Whether they aid or not, Ido not know; but If thay Aid i seems thatthe implications were not pursued to thelr Ulinate conse- uence Whatever the case, for me this nding had grat sgniienoe and plunged sme into the study of cognition at a lptinate biological problem. Two proach and a diferent elton: Inmedinte consequences arose fom this: the Fist one was that a my neo pysologial studies Thad to take seriously the indstinguishabiy inthe foptration of the nervous system betwen pereption and hallucination he second one was that [needed anew language to ak about the phenomena of perception and cognition. The Mist cossquence required that the questo: "How docs the ofgnism obtain infomation about its envionment” be fanged to: "How does it happen that the organism has the structure tht Demis it to operate adequtelyin the medium in which it exis” A seman {le question had tobe changed into a sruetual question. The second ques ‘lon requited the actual attempt to desea the phenomena that take place In the organi during the oreurrene ofthe phenomens of peroption and ‘cognition ina langage that retained them a phenomena proper to closed servos system. A Congress in Anthropology Ely in May of 1968 the Unversity of Chile entered sat of revolaton ‘The students took over the University in an attempt to reformulate the pilosopy dat had ingpied i organization, [joined them. all standard Zcademie active stopped and studeats snd some members ofthe faculty teed to sy somthing new. It wat not ety. Language wat a tap, but the whole exptince as a wonderful school in which one cou drcoer how ‘mute, dat and bid one was, twas eay tobe caught in ones own ego, Ut 5 one suceeedod in attaining atleast som dere of freedom fom it, one ‘began to listen and one's language began to change; and then, but only then, ‘nw things ould be ai, Thi sted for several mont, "In Septembu ofthat year, I aaypted an invitation to go to the University ‘of Maas at Utbana to the Bolopial Computer Laborstory of Profesor Heinz von Foerster. Furthermore, Profesor von Foaster ited me t0 purtipate ina symposium spoasoied by the Wenner—Gren Foundation for ‘Antnopologial Research tbs held dung March 1969 in Cheng, with the purpess of considening the anthropology of cogatio, The inition was {o spesk on “The state of the at of the neurophysiology of cognition’. I scexpted this nvatlon and deeded not to speak about neuron iets, ‘ere impulses o synapses, but ater I decided to consider wAst should ake place inthe eranism during cognition by considering cogation a3 boli Dienomenon. In doing this I found that my two apparently contradictory teams actives were not contradletory, and that they were, infact d+ essed to the same phenomenon: cognition andthe operation ofthe ling system — its nowous system ince when present — were the same thin From this understanding the ey ‘Biology of Cognition’ arose a an expat on of my presentation Ia that symposium. The Word Autopoeal “The second esny Included in this book was writen in 1972, aan expansion ofthe scton on “Living Sytem in the “Biology of Cognition. The writing (ofthis essay was i at tapered by a conversation that Francia Varsa and had in which he slé: “IF indood the cirelarorpnizton is sufi to charactorce living systems a8 unite, then one should be able to put itn ‘more formal tems 1 speed, but si that a formaliction culd only come after acomplate lngubtie description, and we immediatly begun fo work on the complete desrption, Yet we were unhsppy wit the expression ‘relat ‘organization’, and we want a word that would by il convey the cent feature of the organization ofthe ving, which ls autonomy. I wa in thee clteumstanes that one diy, while talking mith «fend (ost Buln) aboot tn esay of his on Don Quitote de Ia Mancha, in which he analyzed Don uixote’saiemma of whether to follow the path of arms (ri, action) or the path of eters (pote, eration, production), and his eventual choice of the path of praxis defering any aitempt at pote, understood for the frst tine the power of the word ‘pies’ and Invented the word that we noaded: auopotexs. This was a word without «history, « word that cou recy mean what takes place inthe dynamic of the autonomy proper 12 living systems. Curiously, but not surprisingly, the invention of tht ord proved of great value It sinplifedenormousy the ask of taking about the ‘organization ofthe lving without falling into the always gaping trap of not Saying anything new because the Inguage dose not permit i. We could not ‘ape being immersdin tradition, but with an adegute language we could vient ourselves difereatly and, petaps, rom the new perspective generate ew tradition, [Let me now say something about the esas themes, srotogy oF cooxrrion ‘When I wrote this esay 1 id not yet have the word ‘eutoplest, aor had I come to the more formal expresion of the living organization gven in the next esay. Yet, these shortcomings do not detract fom what i sid because the basic relations embodied inthe notion of autopoiesis ar fully implied, ssi inrRopuction skthovghcmbesomety si hy expressions och er xiao! and ‘felon! geen Fitemore, when {wot he ony Tec not to mike any conan fo enting nto tht sonidred ints ot iding een ith ere to nae th fet pastel bso. Ho tre, Imad» cocsion ni he ava opted {ved the rere of my nnd ale shout nal stiongwhen pean bot ths cealrernzaton of lvg sytem Todo this wa oth ngs Sn mieatng Ie was naegute bch notion of xual i tion {tt pertains to the domain of deserpton, sn sch tila el in {he stadia in which the Seer mesh ones nd ena te deumed to beer inthe phenomenal dome, eee of Gestion. Its ding bene cbtcued the acl opecon of the stcnncy ofthe ation of propery afd by the dete ope ton performed bythe over when pcping ny forthe dso ot the phenomenal rin genet bythe meted units mending tector obscured the undestnting ofte dependency of ny of the ny on the tincve cpraton tat pei I ws ilenig tecase It shscured both he endemtaning ofthe enrol Gomi Gerrmined by th prope of the unis that ent they a tonitonectin of the peroneal domaine sneated bythe option of {compote anit 4 snl uty ins nem and by the epeton ot ‘ompmentsorconporents Ther i nothng de hat I wiht adda a commentary to hi. $14 comolopy anda uch te colt, Ply Tw tony tht it enaing my views and unestanding of eythane thas ben hy way to tensendetlexperne:t he disoey tht mati, mot hove Seating, the cretion fhe it (he mae of exten oth biome fa dnmin of Gacune) and tte ats the eto of th tater etn This 0 pc, bt tthe expreion of ere Cove domain of option i which th content of eaten eaten Sir teyod tht othing canbe a Avrorousis ‘This article was written as an expansion of two pages of the ‘ology of Cogition’, in an attempt to show that autopoeis necessary and sufficient to characterize the oxganization of ving systems, and that gen the proper Ihstorcal contingencies one can derive all the bolle phenomenology from the characterization of living systems as avtopolte eyes in the physi space. Notions of purpose, function or goal ae unnecessary and misleading ‘This the es does, and in this respect no consmentary is needed; the essay stands by itll, Yet, when we wrote it we were just beplanig o realize the fundamental dsinction between orzantzation and structure, and we do not separate the tems with complete rigor. Also, because it was not written "under the suppodton tat the reader bad ead the ‘ology of Cognition’, the say Isnt fully clear concerning the valty ofthe statement “Everything Said ssid by an observer” in relation fo the dtinction between charter: 2ation snd deserption of system, I shall make some comments sou these Unity, Organization and Structure Unity. The basic copitive operation that we pefonm as observers i the ‘operation of ditinction, By aneans of thi operation we spocfy aunty ae an entity distinct fiom a background, characterize bth unly and background ‘with the propertie with which tis operation endows ter, and speiy thelr feparabity. A unity thus spelid fea spe unity that defines through is ‘properties the space in which texsts and the phenomenal domain which ‘may geezato nits interctons with ober unites Tr we recursively apply the operation of diineton to aunty s0 tht we even in maa, [Cs tue that unde the laws of tural selection the indidals mot apt inthe features which are favorably fslected survive, or ave reproductive advantages over the others, and that, those which do not sursve or ae Test successful inthe reproductive sense fo not contbate or contribute less the historical destin ofthe speci, ‘has, from the Drwinanperspoctive it semed thatthe roe ofthe individual we fo contribute (othe perpetuation of the species, and that all hat one had to do forthe wellbeing of mankind wat to let the natural phenome follow their cour. Sconce, biology, appeared to jusly the notion ‘anything for the benefit of mankind’, whatever the inention or purpose of whower 5 HUMBERTO R. MATURANA AND FRANCISCO J. VARELA, tered it fist, We have shown, however, that these arguments ae not val to jurify the subordination of the Individual tothe specie, beaute the Dologel phenomenology is determined by the phemomenslogy of the indiidots, and withowt individuals there is no bilopel phenomenclory whatsoever. The organization of the individuals autopoietie and upon thi Tact ress al slime: It becomes defined throvgh its exiting, end it exiting autopoiatic, Ths, biology cannot be aed anymore to joy the Alepenabilty of the ladiiduls for the bene of the specie, socety or Imankind under the pretense that ite oe so perpetuate ther, Biology the individuals ae not dispensable, (Gv) Bologlat phenomena depend upon the autopotss of the adidas involved; thus, there ae biological systems that ase from the coupling of autopreticunties, some of which may even constitute autopeetic systems fof higher erder, What sbout human societies, are hey a8 ystems of coupled human beings, aio bilopcl system? Or, ia other word, to what extent do the relation hich characterize human society ae system constittinely dapend on the autopoiede ofthe individual which integrate It IF human focletis are blloglal systems the dyramie of a human society would be Aetermineé through the autopeies ofits components. If human societies tre not bilogial systems, the socal dynamics would depend on lw and relations which are Independent of te autopoies ofthe individuals which Integrate them. The answer to this question is ao trv and reqies con siderations which ination to thee biological ignficance have ethical and polieal implletions, Tis Is obviously the cate, bechuse rich an answer Fequles the charactenzation of the elation whieh define a socety 89 unity (@ system), and whatever we may sy bologialy wil ppl inthe domain of human interactions dieey, either by use or abuse, as we ssw It happen ith evolutionary notions. I fact no postion or view that has any relevance in the domain of human relations cin be deemed free from edial and polieal implications nor can a scientist consider himself alien to these Aplications. This esponsibility we are eady to take, yet since we Maturans and Varela ~ do at fully agree on an answer tothe question posed By the Dlolopel character of human societies fom the vatage pon ofthis hare. terization of the bologial oxgulzation, we hate deided to postpone this scason, 3, COGNITIVE IMPLICATIONS ‘The domain of interactions of an autopoietic unity isthe domain ofall she PRESENCE OF AUTOPOIESIS: us eformations that it may undergo without ls of autopolss. Such a domain is determined foreach ity by the parlcular mode through which its auto- poles is elized inthe space of Is components, dat by its structure. 1 fellows thatthe domain of inertctone ofan astopoetc unity snes bounded, and that autopoiete unites with different structures have diferent, domains of intectons. Furthenoore, an observer can consider the way in which an autopoeic stem compensates its deformations aa description of the deforming agent that he ses ating upon it andthe deformation suffered ty the system a representation ofthe deforming agent, However since the domain of interitons ofan autopoietcrystem bounded, an abserver of an fulopoaticeytem can dvctibo enies extemal to It (by laercting with Them) which it cannot scribe because i cannot interact with them or it cannot compensate the deformations which thes case init. The domain of all the interactions in which an avtopoitic ayer can ener without los of ‘dentity i comntive domain; or, in ater words, the cognitive domain of fm autopoetie system Je the domain of all the descriptions which ft can Posibly make, Accordingly, for any zutopoletis rst ts particular mode Of autopoiss determines is copatve domain and hence its behavioral ‘ively, and 1 fllows that the cogitive domaln of an autopoletc sytem hangs along its ontogeny only tothe extent that its mode of autopoies changes ‘We shall not explore inthis book all he implications thatthe proper characterization ofthe biological phenomenology ha within the domain of opriton, but we sal make our remarks inorder show the dependence ofthis dorsain upon the autopoetic organization ofthe individual For any autooieti system its cognitive domain i necastrly relative to the parcaar way in which i autopoesis is ealzed Alo if knowedge 5s descriptive conduct, itis relative to the cogitive domain ofthe knowes. ‘Therefore, if the way in which the autopoesis of an organism i realized changes during itt ontogeny, the actual knowledge of the organi (ite conduct rpertoie) aio changes; kaowledge, then, is necesarily always & ‘election of ontogeny of tho Knower because ontogeny at a process of continous structural change without lst of autopotet is 4 process of entinoous rpeciication of the behavioral capacity of the oranism, an, hence, of itt actual domain of interactions. Inrnsely, then, no absolute nowise is posible, andthe validation ofall posible tative knowledge i attained trough succesful autopoiess. (@ Avtopoietic systems may interact with each other under conditions 120 NUMBERTO R. MATURANA AND FRANCISCO J. VARELA ‘hat raul in behavioral opupling In thi coupling, the sutpoitic conduct of tn orgasm A becomes a souce of deformation for an organism B, andthe ‘compensatory behavior of organism B acts, tum, sa source of deformation fof orgasm A, whose compensatory Sehaior acs again asa sour of de- formation of Bands on ecuravely uni the coupling i interrupted. In hs smaaner, + chain of interlocked fatrations develops such that, although in ch interttion the conduct ofeach organism i constitutively independent in its generation of the conduct of the other, because its internallydeter- mine by the strstore of the behaving ergs only, ts for the other ‘organism, while the chun ast, a sours of compensable deformations which ‘a be decribed at mearingfl in the context ofthe coupled behavior. These St communicative interactions. In other words f sho intrctng organism: ft dynamic aytems have continuously chagglag structures, and if they ‘eciprouly elect in each othe ther respective paths of entogene srvctral ‘Changes through ther interactions without loss of antopoiss, then they feerat, a a ocurive or expanding domain of communicative ntrctions, Intelocked ontgenise that together constitute =. dona of mutually ‘wiggeing conreinl conducts that becomes specified during its generation. Sush a sonsennal domain of communicative interactions in which the behaviorally coupled organisms orient each other with modes of behavior ‘whose lniemal determination hat become spciied during ther coupled ‘ntogeies, ra ngstic domain, In such a consensual domain of interactions the conduct ofeach oraniam may be treated by an obser as constituting = conotative description of the conduct af the other, of, im his domain of esciption eran observer a8 a consensual denottion of it. Communicative tnd Hague interactions ar tinal not informative; organism A does fot and cennot determine the conduct of organism B because due to the ature ofthe autopoitic organization isl” every chang that an organist. tindergos is necesriy and unavoldsly determined by its own orpanization. ‘Alingatic dora, then, a consensual domain that aris from the copling of the ontoenies of otherwise Independent zutopieic systems, i inten Sally nonsnformatve, eventhough an observer, by necting te internal ‘eteminaton of the sutopoltc systems which generate i, may describe it ‘sit it were #0. Phenomenology the linguistic domain and the domi of Sopoels are dfeent domains, nd although one generates the elemonts of the other, they donot intersect. (i) An auopoiee system capable of interctng with is own sats (as an organism with a nervous system can do), and capable of developing with PRESENCE OF AUTOPOLESIS m others linguistic consensual domain, can treat ts own linge sates 382 ours of deformations and dhs interact Lnguistaly ina closed linguist domain Such a sytem has two remarkable prope: 1. Through recursive Interactions with ts ingustclly generated states 4 can teat some ofthese stats as objects of further interactions, ging tse to e meadomain of eonsensaldstintons that appears to an obser asa fomain of interactions with representation of iteacions, Whea this hap- pew the system operate a an observer, The domain of such recuse intr Tetion i in pra, infinite becuse once the system has etiained the mechani for doing so there imo moment Sa which it wil ot be in the postion of recursively interacting withits own slates, unless autopoesi is owt, Whether an sutopoetic system with this capactty does in fact generate an endless series of efferent sates during is ontogeny depends, obviously, ‘on whether ite history of linguist ntractons inthe metaiomain of descrip tons has significance for the ezcunstantial realization ofthe atopoiss of {he interacting organo 2. Atvng system capable of began observer can inter with those ‘ofits own descriptive states which ave nguitledesrptons of itll. By Aoing soit generates the domain of sllinguistedesciptons within which it isan obsrver of tislf ax an observer, a proces which can be necessarily repeated in an endles manner, We eal this domain the domain of sf ‘bveration and we consider that selfconsous behavior i selrobseng ‘chao, that, behavior within the domain of elfoburaton, Thecbseret san obtererneceiarlly lays mans in 2 descriptive domain, that in 4 lative cognitive domain, No desrpion of an absolute reality is possible. ‘Such a description would require an interaction with the absolute to be ested, but the representation which would as fom such am interaction would necrsnly be determined by the atopoitic organization of the Dbeerer, not by the deforming sgeat; hence, the cognitive reality that it ‘woold generate would unavoidably be lative to th knower. In every explanation, be this an actual concrete reproduction, «formal teptesntation oc = purely rational description, the reformulation of the ‘enomenon to be explained resorts to the same notions (identity, exclusion, fvcenion, ete). Thete i, then, a univer lope, valid for all phenomeno logical domain, that eee tothe relations possible between the wis that eerate these domains, and not tothe particular properties ofthe generating Uni, We have applied the loge (it coulé aot have been otherwise) in this 122 HUMBERTO R, MATURANA AND FRANCISCO 1. VARELA, book, and the validity of our arguments, asthe validity of any rational anpument or concrete phenomenological ralaton, rests on its validity. Furthermore, we have in principle shown through is application that the rihenomenclgy of sutopetc systems generates observes, and through them the phenomenology of description within which thi logic sao wall For epistemological reton, in orde o say all that we hve sald about ving systems, we had to siime a space (the physical space) within which the Bhenomencogy of autopoiess of ving systems takes place. To the extent fiat we have been suceetol (fee from logical and expecential contrat tions), we can conclude that sich a spect fs ontological a space within ‘which the ole that we have applied in our deserption is intinsaly valid If this were not the cae we could nt have done what we have done in terms of characereng living stems, o of showing how these may nem systems apable ofthe own description We cannot characterize this sac in absolute ferns, In ngs interactions all that we ean dois to deseribe dough Tingle babavor and construct further desrpions based on these desrp- tions which alvaye remain in the sane domain of operations defined in ‘lation tothe operating eye. 'A rediton i a estement of cso within relational mais i Isa cognitive saetnent, and a such stakes place within a deseitive domain, ‘Ths, unless mistakes ae made, i all the eitios that define the particular ‘nati within which the prediction is made are propedy taken, the prediction [S wid. Eno of interpretations may arse only by msapplcatin, that i by preending thatthe observer makes prediction in one matrix when he is emaking in another, In particular, predictions in the physical space are posible, because a description, a8 an actual behavior, exist in mati of Interactions which (by constitson) bas algal matix necessarily isomor ple withthe robstatum matrix within which i takes place, not eens we fave an absolute knowledge of the universe. These cogaitie relations are ‘lid for the posible cognitive phenomenology generated by any closed ‘stem. Living systems are an existential proof Chey exist only to the extent that they can ext The fentty of our imagination cannot dey thi, Living ‘ystems are concatentions of procesies in mechanistic domain fantasies are Cneatnatons of descriptions in a ingistc domain, In the ist caso, the Coneatnated unites ate preter; in the second ee, thay ae modes of linguist behavior. “Autopoesis soles the problem ofthe biological phenomenology in peer by defining it, New problems ae, and old ones ppea ins different perspec tive; in patcular, thse which refer to the org of Ivig systems on earth, PRESENCE OF AUTOPOLESIS 125 (cobiogness and neabiogenes), and those which refer to the particular feznizaton through which recursive deseripie Interactions take place in ‘snimals (te nero sstem), Autopletic ystems define the wor in which they cn exe in elation to thelr atopoiess, ani some interact recursively ‘with this world through ther descriptions, being imposible fr them to step ‘ut of thi relative descriptive donala through descriptions. This demands an ‘entirely new eopnitve outlook: here is a space in which diferent phenome nologies can tke pce; one of theses autopoless; atopoiesis generates & ‘Phenomenologcal domain, this is cognition. APPENDIX ‘The phenomenology of an onganism as aunty is the phenomenology of is sutopoies. The changes thst an organism undergoes while maintaining it tutopoies constitute is conduct. The conduct of an organism is revealed to an observer by the changss that It eauses inthe ambience (ieidng the sere) in which It exes, Accordingly, the condct which an observer bolas in aay organi, however complex it may sem, i always an ex ‘ession of the autopoei of th observed organism, ands such, 1 alays fries through « phenomenology that takes places inthe present because fistory isnot a cused component inthe mechanism of autopoia (68 Chap- ler IV), Yet lt spears to wr a subjects of el observaton and a observes of the conduct of other organisms tht past experiences determine our and thet onduct inthe present wif, embodied in nodications ofthe nous system, they were casa components inthe mechanism which generates behavior. It appears, therfore, a ifthe operation of the omanism asa state-determined system in which time Is not a component were determined by temporal Phenomena, and we speak of letmnig, memory and real asembodinents of the pat, We consider that thir costradition argos from not distinguishing what pertains tothe phenomenalgy ofthe autopoess rom what pertains {o the domain of interactions of the organo a8 unity, and, dus, from an Inadequate eration of the couplag ofthe sruture of the merous system to the ontogeny ofthe organism. Accordingly, our purpose inthis Append ‘tout the neous sytem et considers organization ssa neuronal network feta evaote this coupling in which past and present ase asnew dimensions from the recurve interactions ofthe organi with sown state. [A. THE NERVOUS SYSTEM AS A SYSTEM ‘The nervous system ‘0 the ogni of wi network of interacting neurons coupled in ee ways tis x component () The oxpanion, inching the nervous syste, provides the py! and ne APPENDIX 1s bochemial environment forthe autopoiess of the neurons aswell a for all other els, and, hence, is posible soueeof physical and bochomial perturbations which may alter the properties of the neurons and thus lead to or Gi, (i) There are sites of the organism (physi and biochemical) which change the sat of activity ofthe ners system a whole by acting upon the reosptor surfaces of vom ofits component neurons and thus ead to (i). (Gi) There are states ofthe nervous systom which change the state ofthe ‘orpnism (pityscal or biochemical) and lead wecursivly to) and (3), ‘Though this coupling the nervous system partite inthe generation of the avtopoiteclations which define the organism which it interes, en, sccoringly, ts srvture is subordinated to ds ptelpation, 1, Thesouon Newrons determine their om Boundaries though tel sutopoise therfore they ae the anatomical units of the nervous system, Thore ae many’ clases of neurons tat can be dtingished by thei hapes, But all of them, egard- les ofthe morphological clas to which they belong, have branches which put ‘hem in diet or indirect operational elation with other otherwise separated neurons. Functionally, that, vewed stan allooietis component of the ‘nervous system, 2 neuron has © collector surface, a conducting element, and an effector surface, whose relative postions, shapes and extensions are ferent indifferent elses of neurons The collector surface i that part of te surfee ofa neuron where it esives afferent inluenoes (ynaptic oF not) Som the effector surfces of other neurons or Is own, The effetor surface of a neuron is that part ofits surface which eltiardielly by meane of synaptic contact) or indirect (rough Is synaple oF nonsynapti ation on other kinds of el) afects the ealctr surface of other neurons or itt ‘own. Depending on its Kind, a neuron may have its eollactor and effector surfaces complotly of partly seperated by a conducting element (absence fr prseace of presynapi inhibition), ori may have both cllstor and effector suficet completely intespaced, with no conducting element be- ‘ween them (amacrine cls) Th interactions between cllctr ani effector surfaces may be excitatory of inhibitory according to the kinds of neurons involved. Excltatory afore ifluness cae «change inthe state of activity ‘ofthe ellector surface ofthe ceeiving neuro which may lead oa hangs in 126 HUMBERTO R. MATURANA AND FRANCISCO J. VARELA the mate of actly of ts effector surface, while he inhibitory influences Iipingng it shunt off the effect ofthe afferent influences os ooeptor sorfae ao that this effect does not a al rach ts effetor surface, or exces thisurace with reduced effectiveness. (Operationally the ste of activity of«nowton, characterized by the state ‘of activity of is effector surface ls determined by oth it nterel structure (rambrane properties, relative thickness of branches, and in genral all Structural lations which determine its possible stats) and the alert lnflences impinging on its rexptor surface. Conversely, the effectiveness of ® nevron in changing the state of activity of other neurons depends both on the intemal strueture ofthese, and on the relative effetvenes of tis actin fn thle receptor surftes with respect tothe other afferent lalunces that these neurone ectv, Thi 690 beens excistory and inhibitory influences do nota linerly i the dateminston of the stte of actviy of a neuron, but rather have effects whlch depend onthe relative poston of their points tt action with respect to each other and with respect to the efector surface ofthe rceiting call, Furthermore, the ineralstuctue ofa neuron changes tlong its ie history, bot ae are ofits autonomous genetic determinations And ssa ves of the ccumtanoes of ts operations during the enlozeny of the ovgnis, Thus, neurons are not static enties whose propertis remain levtlant, On the contrary they change, Tha thre general eosequence: (© There are many configurations of afferent (put) influences on the receptor surface of newon which produce the se configuration of efferent (utp aetvty at selector surface, (i) Changes in the intemal structure of «neuro (egmdes of whether they abe saactod by the autonomous transformation of the ell or by it history of lateractions inthe neuronal network) by changing the domain of states of actlty thatthe neuron ean adopt change ts domain of inpu-outpat ‘elton, that i changeit transfer function, (Gi) No single cell orcas of ells can alone determine the properties of the neal network which iterates, GGeneraly then, the structure of x neuron and its role ia the neuronal rcwork which It integrates doesnot stay invariant, but changes along. its fontogey In a manner subordinated tothe ontogeny of the oranim which Is botha result and source ofthe changes thatthe neuronal network andthe anism oder, APPENDIX m7 2, Organization: The Nerous System Asa Closed System From the descriptive point of view itis possible to say thatthe properties of ‘the neurons, thei Internal suet shape and relative position, determine the connetvty ofthe nervous sytem and consiute ia a dyaamienetwork ‘of neuronal interactions. This connectivity, that i, the anatomical and ‘operational clons which hold between the neront which consitate the rarvous syster as network of lateral, pre, sequential and recursive Inhibitory and excitatory Interactions, determines ite domain of posite dynamic tates Since the propertisof the neurons change along he ontogeny ofthe organism, both due to their internal determination and os a result of their interactions as components of the nervous system, the connectivity of ‘he nervous system changes slong the ontogeny of the organism ina manner ‘eursively selected during this ontogeny. Furthermore, since the ontogeny of the organism Is the history of ls autopole, the connectivity of the nervous system, through the neurons which constitute It, s dynamically ‘ubordinstod to th autopois of the onganiim which it integrates (Opeatonlly, the aorous system ie « closed network of interacting neurone rich thats change of activity ina newon always lads to s change of sett nother nerons, either dretly through synaptic ction, orndietly ‘through the participation of some physcl or emia intervening eltent, ‘Therefore, the ongaization ofthe nervous sytem ae init nevrona stork 4s defined by elations of closeness inthe neuronal interactions generated in the network, Sensory and effector neurons, at they would be dteribe by an ‘observer ho behold an organism in an envionment ae not an exception to this because all sensory etvty in an organism leads fo att init efector surfaces, and all effector activity init esto changes in is sensory surfaces, ‘That at this point an observer should se environmental elements intervening ‘between the efector and te sensory sass ofthe ongnism,f ielrant ‘because the nervous systems defined asa network of euronalinteractionsby the interactions ofits component neurons realest of intervening elements. ‘Therefore, song a the neural network closes on fits phenomenology is the phencmanology af «closed system in which nearonal activity always leads 10 nouroal aetvty. This es eventhough the ambience can perturb the nervous systom and change its satus by coupling to it ean independent ‘goat at any neuronal ecepor surface, The changes thatthe neroue system an undergo without disintegration (loss of defining tlstions sta closed neuronal network) ata rent of these or any othe perturbation are fly specifi’ by Hts connectivity, and the pertwbing agent only constitutes & 128 HUMBERTO R. MATURANA AND FRANCISCO J. VARELA histo determinant for the concurrence of thse changes. As a dosed ‘euronal network the nervous system has ao input or output and thee ino Intaai feture in ite organization which would allow it to diciminate {trough the dynamics of Hs changes of stato between posible internal or extemal cases for these changes of stats. This has two fundamental con- Sequences () The phenomenology of the changes of sate ofthe nervous stom ie cxclusvaly the phonomenology ofthe changes of sate ofa closed neuronal intworks thst for the merous aystem ats neuronal notwork theres no inside or outside (The distinction between intecral and extemal causes inthe exgin of the changes of sate of the nervous sytem can only be made by an observer that Bebe th organism (the neryous sytem) a€ 3 unity, and defines its Insie and ouside by speifying Its boundaries, 1 flows that is only with respect othe domain of neractons of the ‘organism ara nity thatthe changes of tae of tho nervous system may have fan internal or an external igi, and, bance, thatthe history ofthe causes of the changes of sae ofthe nervous systole ina diferent phenomencogisl domain thn the changes of sate thamslvs, 3. Ounge “Any chang in he structue of the nervous system arises from 2 change inthe ‘properties of component neurons What changin fat thes pace, whether ‘orphologcl or biochemical or both, relevant forthe present aicuson. ‘The significant point is tha hase changes asin the coupling ofthe nervous system andthe organism through ther homeostatic operation subordinated to the aulopoie of the latter, Some of the changes det affect the ‘operation of the nervous system because they take place though is working tra clsed networkother affect indie bec they tke pace through ‘the biochemical or genetic coupling of the neurons to the ornism and ‘change the properties of the neurons in a manner unrelated to the actual working ofthe network. The results are twofold on the one han, ll changes lead to the same thing, that i, changes inthe domain of posible states of the nervous syst; on the ater hand the nervou system coupled 10 the congnism both in its domain of interactions and in its domain of internal transformations APPENDIX 9 4, Arcitecrure ‘The connectivity of the nervous system is determined by the shapes of 4s component neurons. Accordingly, every nervous system has definite sschteture determined by the kinds and the mumbers ofthe neroos which compose it; therefore, members of the same species have nervous esteme ‘vith sma architectures to the extent that they hive smiley kinds and ‘numbers of neurone, Conversely, members of eifrent species have nervous systems with diferent architectures according to thelr spell differences in ‘neuronal composition, Therefore, the closed organization of the nervous system is vealed in diferent spels in diffrent ranners that have been

You might also like