bunge mente cuerpo

Upload: sebastian-lomeli

Post on 02-Apr-2018

240 views

Category:

Documents


2 download

TRANSCRIPT

  • 7/27/2019 Bunge Mente Cuerpo

    1/9

    Neuroscience, 1977. Vol. 2, pp. Ml-509 Pergamon Press. Printed in Great Brita,in.

    COMMENTARYEMERGENCE AND THE MIND

    M. BUNGEFoundations and Philosophy of Science Unit, McGill University, Montreal, Canada, H3A lW7

    Abstract-This commentary deals with the mind-body problem from the point of view of a generalsystems theory. It starts by elucidating the notions of thing, property, state and process. In particularit shows how the concept of a state space can be used to represent the states and changes of stateof a concrete thing such as the central nervous system. Next the concepts of emergence and of levelare discussed. An emergent property is defined as a property possessed by a system but not by itscomponents. The notion of level and the peculiar relation existing between levels are clarified, onlyto show later on that the mental cannot be regarded as a level on a par with the physical or thesocial. The upshot is a rationalist and naturalist pluralism.

    The second half of the paper expounds and examines the various versions of psychoneural monismand dualism. Dualism is found unclear, at variance with the general framework of science, and untes-table. Eliminative materialism and reductive materialism are rejected for ignoring the peculiar (emer-gent) properties of the central nervous system. A variety of psychoneural monism called emergentistmaterialism is found the most acceptable because of its compatibility with our present knowledgeand because of its heuristic power. However, it is emphasized that emergentist materialism is stilllargely a programmatic hypothesis in search of detailed theories, in particular mathematical ones,of the various emergent functions of the central nervous system and its subsystems.

    THE PROBLEM AND ITS SETTINGTHIS commentary deals with the so-called mind-bodyproblem. This is the set of questions about the natureof the mental and its relations to the bodily. Forexample, are mind and body two separate substances?If so, how are they held together in the livingorganism? These questions are rather difficult toanswer. However, I submit that the difficulty is notwholly intrinsic but has been compounded by hurdlessuch as the following. Firstly, several doctrines con-cerning the mind-body problem have some ideologi-cal bias or other-and ideologies are not particularlyinterested in fostering conceptual clarity and empiri-cal investigation. Secondly, the very formulation ofthe mind-body problem employs certain concepts,such as those of substance, emergent property, stateand event, which are far from clear. (These conceptsoccur in all sciences and are therefore elucidated bynone: they belong in the branch of philosophy knownas ontology or metaphysics.) In fact it is pointlessto engage in an argument about whether or not thereare mental states that are not brain states, or whethermental events have causal efficacy, unless one canmake some sense of the very expressions mental stateand mental event, which in turn contain the philo-sophical concepts of state, event and mind. Let ustherefore start by trying to clarify these and a fewother ontological concepts that occur in the discus-sions on the mind-body problem. (For a detailedAbbreviation: CNS, central nervous system.

    NSC./L-a501

    mathematical treatment of these concepts see BUNGE,1977a.)

    A thing, or concrete object, may be characterizedas whatever can join (or associate with) another thingto form a third thing. On the other hand two conceptscannot always join to form a third concept, e.g. purplenumber is not a concept although purple andnumber are. But things have of course many otherproperties in addition to that of joining to form otherthings. For instance, they can interact and gettogether, forming tightly knit complex things, i.e. sys-tems; they can move about, change in kind, and soforth. We may then assume that every thing, nomatter how simple it may look, has a large numbern of properties. (We are here referring to generalproperties such as that of moving, not to particularproperties such as that of moving from here to therewith such and such an instantaneous velocity relativeto a given frame.)Now, every thing property can be conceptualizedas (or represented by) a function in the mathematicalsense of the term and, in principle, by a real valuedfunction. And the n functions representing the proper-ties of a concrete thing can be collected together intoa single function in conformity with:Def ini t ion 1

    Let each of the n properties of a concrete thingbe represented by a real valued function Fi of time,with 1 I i I n. Then

    (i) IF = ( F1, F1, . . . , F,): T+ IR is called the statefunction of the given thing;

  • 7/27/2019 Bunge Mente Cuerpo

    2/9

    (ii) the value .>= IF(t) of IF at time t is called the,stutcof the given thing at I:

    (iii) the ordered pair (s. 5) of values 01 IF at timest and t respectively is called an rJw/ occurring inthe thing concerned between t and I:

    (iv) the sequence of states joining two states .s ands of a given thing is called the process leading froms to s. or the history of the thing between t and 1.

    Note that we have not been talking about proper-ties, states or changes in themselves: every propertyis a property of (possessed by) some thing or other:likewise every state is a state of some thing. and everychange of state is a change of or in some thing orother. Thus physical states are states of physicalthings; chemical states, states of chemical systems:biological states, states of organisms; social states.states of social systems; and so on. This manner ofspeaking, which is entrenched in modern science andwhich ignores the Platonic forms hovering abovethings, will prove of decisive importance rn our dis-cussion of the mind-body problem.

    Now, the state function IF describing the states andchanges of state of a thing is not o priori: it is deter-mined by the laws possessed by the thing. In otherwords, there are laws that restrict the possible formsof IF. These laws may take the form of mere restric-tions on the range of IF, or of algebraic relationsamong the components of IF, or of differential equa-tions satisfied by them, or what have you. By virtueof such restrictions, the tip of IF spans not the totalityof its codomain lR but only a subset of it. This subsetof the set of all logically possible states of the thingwill be called the .staw spuceof the latter or. moreprecisely, its law@! srute space. We designate it S,t,(.u),where IL is the set of laws possessed (obeyed) bything x. (See Fig. I.)

    So much for the concepts of thing. property, stateand process. Let us now take a closer look at proper-ties of a particular kind. namely emergent properties.They are of special interest to the neuroscientist who.while acknowledging that feeling, recalling, imaginingand reasoning are emergent properties of the brain.would like to explain them in terms of events occur-ring in certain subsystems of it.

    RESULTANTS. EMERGENTS ANDLEVELS

    Temperature and entropy are properties of anatomic aggregate, not possessed by any of its atomiccomponents. Likewise the capacity to self-duplicateis a property of deoxyribonucleic acid molecules thatnone of their components (i.e. the nucleotides) pas-sesses. These are examples of emergent properties, orproperties characterizing a system as a whole andwhich the system components do not have. Emer-gence is conspicuous at all levels and u fortioribetween levels. This much seems clear.

    What are not clear at all are the various notionsof emergence and level. There are several reasons for

    this. One IS that nlost ratronultst philosoptrc~ ., ,,r_radical reductionists and so IKIVL claimed th,li nr2:gencc IX it myth. ~\nother I> that JIKN erncl-g:JitI-.iphilosopherc are irrationalist~ and so have hclc: th,r:thcrz is nothing to bc explained: that cmergi7io~ i-1as mysterious 2s it 15 real II third rc;tson 1~ lb.obscurity of the notion ts that ccienttsts at-c I

  • 7/27/2019 Bunge Mente Cuerpo

    3/9

    Emergence and the mind

    STATE SPACE OF THING

    503

    FIG. 1. The states and changes of state of a thing (e.g. neuron, neuronal circuit, subsystem of theCNS or entire CNS are representable in the state space of the latter, which is the space spannedby its state function-the ordered n-tuple of functions representing its various properties. (On theother hand according to eliminative and reductive materialism of the radical sort, i.e. mechanism,the states of any thing should be representable as points in spacetime.) In the diagram only twoproperties, represented by functions F, and F, (or rather their ranges), are shown. Actually any realisticmodel of a complex thing will involve a state function with many more components. So, try to imagine

    the state of a thing as the tip of a vector in an n-dimensional state space.

    We recognize the fact of emergence but assume thatevery emergent can be accounted for in terms of asystems components and the couplings among them.For example, refraction is not a bulk property oftransparent bodies: it is au emergent property rela-tive to the atomic (or molecular) components ofsuch bodies, for none of those components possessesthe property of refrangibility. Yet this emergent pro-perty of the whole is explained by electrodynamicsin terms of the electrical properties of atoms (or mol-ecules) and light. However, this explanation is notreductive in a simple sense, as it does not consistin attributing refrangibility to individual atoms: it isreductive in consisting in the deduction of the formulafor refractive power from premises concerning the in-teraction between electromagnetic waves and atomiclattices.What holds for physical systems holds a jbr t ior ifor chemical, biochemical, biological and social sys-tems. For example, enzymatic catalysis is an emergentproperty of biochemical systems, sexuality an emer-gent property of some biosystems, and social cohesionan emergent property of sociosystems. However, theseare not unintelligible properties: they can be and are

    being explained. (That no scientific explanation islikely to be definitive is beside the point.)

    The foregoing assumptions can be compressed intotwo postulates, one ontological or concerning reality,the other epistemological or concerning our knowl-edge of reality. Here is the emergence postu l at e:

    Postulate 1. Some of the properties of every systemare emergent.

    And here is the rat i onal i ty postulat e:Postul at e 2. Every emergent property of a system

    can be explained in terms of properties of itscomponents and of the couplings amongst these.These two postulates constitute the kernel of whatmay be called rat io nal emergent i sm, a doctrine differ-ing from both the irrationalist emergentism of theholists and the rationalist flattening (or leveling) bythe mechanists, energetists and idealists.The last ontological concept we must handle beforeturning to the mind-body problem is that of level,particularly in view of the popular assumption thatthe mental constitutes a higher level than the biologi-cal one. First the intuitive idea.Most biologists seem to agree that things, and inparticular things of concern to biology, are found not

  • 7/27/2019 Bunge Mente Cuerpo

    4/9

  • 7/27/2019 Bunge Mente Cuerpo

    5/9

    Emergence and the mind 505us examine them briefly, starting with the main var-ieties of psychoneural dualism.

    We need not consider the independence thesis Dl,as both introspection and neuroscience tell us thatthe bodily and the mental-whatever the latter maybe-are interdependent. As for the parallelism orsynchronization thesis D2, upheld by the Gestaltschool, it begs the question instead of answering it,for what we want to know is precisely the mechanismresponsible for the parallel sequences of mental andphysiological states. To say that mental events haveneural correlates is fine but not very informative un-less one states what a mental event is and the natureof its correlation with its neural correlate. For thesereasons D2 is vague to the point of being confirmableby all possible empirical data. Hence D2 is not ascientific hypothesis.On the dualist side we are then left with eitherof the thesis acknowledging one substances actingupon the other. However, in this case too only thephysical is supposed to be knowable, whereas themental is left in the dark or, at best, in the care ofphilosophy or even theology. We do indeed under-stand what it is for a given neuron, or neuron assem-bly, to be in such and such a state: a state of a thingis always an ordered n-tuple of the n properties wecare to assign to it. (Cf. The first section in this com-mentary.) And we understand what is a neural eventor process, namely a change in the state of a neuralunit neuron or neuron assembly). Consequently weknow what it is for one neural unit (neuron or neuronassembly) to act upon another: A acts on B if thestates of B when it is connected with A are not thesame as those of fl when it is not so connected. Inshort we have some idea of neural functions (states,events, processes). Recall Fig. 1.

    But these ideas-common to all sciences-are nottransferable to the mental substance. If they are,nobody has shown how. In particular, attention,memory and ideation have not been shown to beproperties or changes of properties of a mental sub-stance (mind, soul or spirit). In sum, the conceptsof mental state, event and process do not fit within

    the general framework of contemporary science unlessthey are construed in neural terms, i.e. as, respectively,a state of the brain or an event or process in a brain.This is one of the reasons for dualisms inability togo beyond the stage of verbal and metaphorical for-mulations. This is why there is not a single dualisticmodel-in particular a mathematical model-inphysiological psychology.

    In short, interactionism is just as imprecise as par-allelism-which is to be expected of a popular, i.e.nonscientific, view. (Recall that ordinary knowledgeis largely popular superstition.) And, not being a pre-cise hypothesis, it can hardly be put to empirical tests.Moreover even if parallelism and interactionism wereto be formulated in a precise manner, it might notbe possible to decide between them on the strengthof empirical data. Indeed, it would seem that everypsychological experience and every psychophysiologi-cal experiment could be interpreted (or misinter-preted) either in parallelist or in interactionist terms,since neural events are simultaneous with their mentalcorrelates.

    We are led to the conclusion that the two mainvariants of psychoneural dualism, namely parallelismand interactionism, though conceptually different, areequally fuzzy and are empirically equivalent in so faras they accord (much too easily!) with the same em-pirical data. For these reasons dualism is not scientti-tally viable. It is barren double talk and, as Spinozacharacterized it, a disguise for our ignorance. We arethen left with psychoneural monism as the only scien-tifically and philosophically viable alternative.But, as shown in Table 1, psychoneural monismis a whole class of doctrines. Let us start with Mlor subjectivism. We can write it off without furtherado because it is incompatible with physics, chemis-try, molecular biology and social science, all of whichare busy hypothesizing and manipulating unobserv-ables such as atoms, ecosystems and societies. More-over, all of these disciplines are supposed to abideby the scientific approach, which includes objectivity.As for neutral monism, it has yet to be formulatedclearly and in agreement with the natural sciences.

    TABLE 1. TEN VIEWS ON THE MIND-BODY PROBLEMPsychoneural monism

    Ml Everything is I(, phenomenalism,idealism).M2 cpand II, are so manyaspects or manifestationsof a single entity(neutral monism).M3 Nothing is 1(1 eliminativematerialism).M4 Ih~ mind is corporeal(reductive materialism).

    M 5 $ is a collection ofemergent functions of cp(emergentist materialism).

    Psychoneural dualismDl p and $ are independent.D2 cp ( , i.e. cpand $are parallel (psycho-

    physical parallelism).D3 cp-+ $, i.e. cpaffects(or even secretes) $(epiphenomenalism).

    D4 $ + cp, .e. 1(1 ffects(e.g. controls) cp(mentalism).D5 cp -t $, i.e. cpand $interact (interactionism).cp, body (or the organi c); $, mi nd (or the mental).

  • 7/27/2019 Bunge Mente Cuerpo

    6/9

    Even the least obscure and mystical of its versions.namely Ostwalds energetism, is vague. (~orc~~cverit rests on the mistaken reification of energy. whichis actually a property of physical objects not a thing.)We may therefore dismiss Ml and M2, and examinematerialism.

    WC distinguish three kinds of materialism, to wit.eliminative, reductive and emergent& (Recall TableI .) Eliminative materialism holds that there is no suchthing as the mental: that everything is material. Thereare two different versions of this thesis: the ancientthesis that all subjective phenomena are composedof particles. and the modern thesis that there are onlyneural facts (states. events. processes). Neither of thesetheses is capable of distinguishing between appear-ance and reality. i.c. between facts as perceived bya sentient being (i.e. phenomena) and facts as theyare independently of the organism. Nor does elimina-tive materialism distinguish between Nom supirr7.sand its nearest cousin the amazing chimpanzee. sosimilar at the cellular level and yet so different atthe organismic level. In short, eliminative materialismcan be eliminated.The thesis of reductive materialism may be formu-lated thus: Every mental state (or event or process)is a state (or event or process) of the central nervoussystem. Therefore the mental is no diRerent from thephysical. While I have no quarrel with the premise,I submit that the conclusion is a uou s+&ur. Butbefore arguing for emergence let us examine this argu-ment of the reductive materialists. It is an argumentfrom analogy with other macrofacts rather than anindependent examination of the models and the em-pirical evidence in physiological psychology. Let usdwell on it for a while.

    Reductive materialists claim that the body-mindrelation is just a particular case of the macroscopic-atomic relation, and that in both cases it is one ofepistemological urrd ontological reduction. While Iagree with the first contention. 1 disagree with halfof the second, namely concerning the ontologicalreduction of the mental to the neural, The first thesisseems rather plausible: however localized some oreven all mental facts may be. they always involve alarge number of neurons. not to speak of blood cellsand other non-neuronal components of the nervoussystem. As for the reducibility thesis. let us discussit in the Iight of the stock-in-trade example of thereductionists, namely the alleged reduction of waterto water molecules.The microreductionists claim that water is justH,O, which in turn they claim to be nothing butan aggregate of two hydrogen atoms and one oxygenatom. This they take to be a paradigm of microreduc-tion. Epistemologically maybe so. (And I hope so,although as a matter of fact there exists no adequatetheory of liquids, and (I ,fbrtiori none has beendeduced from quantum mechanics.) But the thesis ofontological reduction is obviously false. Indeed, tostate that the ccrrllposifion of ;I body of water is a

    set of H,O molecules IS not to state that [hi I( rmi~is nothing but the I;~ttcr. an! mol-c tit:~n I

  • 7/27/2019 Bunge Mente Cuerpo

    7/9

    Emergence and the mind 507

    FIG. 3. Mental states form a subset of the collection of all btain states, which in turn are includedin the set of possible states of the whole organism. Awareness (or the self) is conjectured to be adistinguished subset of the mental states and therefore a subset of the totality of organic states. Thearc of curve represents a mental process, such as recollecting an experience, which is partly conscious.The diagram is programmatic: we still have to identify the properties represented by the blanks F,and F, in the state function of the CNS.

    It also makes good scientific sense to speak of psycho-somatic nteractions, because these are now construedas reciprocal actions ~ etween different subsystems ofone and the same organism, such as the neocortexand the sympathetic nervous system. For example,rather than say that love can color our reasonings,we may say that the right brain hemisphere affectsthe left one, and that sex hormones can act uponthe cell assemblies that do the thinking. In short,ironic as it may sound, the dualistic modes of speech,which encapsulate our undigested introspective ex-perience and which are but metaphorical and vaguein the context of psychoneural dualism, become literaland precise in the context of emergentist materialism.The latter salvages whatever can be salvaged fromthe dualist myth.

    Emergentist monism has many attractive features,the most important of which are that (i) it squareswith the natural sciences by postulating that mentalfacts, far from being affections of an immaterial sub-stance, are states of, or events and processes in, con-crete organisms, whence (ii) they can be investigatedthrough the normal procedures of science-a featurewhich turns psychology into a natural science insteadof a supernatural one.

    Emergentist materialism holds then splendid pro-mise and moreover has already rendered dis-tinguished service by being the driving force behindphysiological psychology. However, it has one impor-tant shortcoming, namely that it is still immature. Infact emergentist materialism is not a theory proper,i.e. a hypothetical-deductive system containing pre-cisely formulated and detailed hypotheses accountingfor a wide range of psychoneural facts. It is insteada programmati c hy pothesis--one both scientific andphilosophical-in search of scientific theories embo-dying it. So much so that emergentist materialism canbe summed up in a single sentence, to wit: M entalstat es orm a subset (al bei t a v ery di sti nguished one)of brain stat es (which in turn are a subset of the statespace of the whole organism). This, however sugges-tive, is so little as to be representable in simple dia-gram: see Fig. 3. (Dualism on the other hand cannotbe diagrammed at all, except metaphorically, so itis even poorer.)What is needed for implementing the program ofemergentist materialism, i.e. for developing it into amature scientific enterprise? Obviously, not more tm-digested data. What we do need are two differentthough complementary batches of theories: (i) ex-

  • 7/27/2019 Bunge Mente Cuerpo

    8/9

    rrr~nrl~ yrr~tri theories (not just stray hypotheses OIprograms) of the mental conceived of as ;I collectionof functions of the CNS: and (ii) spc~ific~ theoriesaccounting for the functioning of the various subsys-tems of the CNS.

    The general theories of psychoneural activity wouldbelong to the intersection of ontology and psy-chology. while the specific theories of the psy-choneural would be the exclusive property of physio-logical psychology. And all of them should be statedin precise terms. i.e. should be mathematical in form.

    It may be argued that the preceding plea for inten-sifying theoretical work in the fields of psychophiloso-phy and physiopsychology are impertinent becausethere is no dearth of theories in both fields. Let ussee about that.

    Certainly. much has been written about the so-called idrr~tit~, theor!. over the past two and a halfmillenia. But none of the theories of the psy-choneural that agree with the materialist hypothesisare theories proper, i.e. hypothetical -deductive sys-tems, let alone mathematical ones. They are insteadsingle and stray hypotheses. And they are verbal andoften verbose. (This may bc one of the reasons thatmost mathematical psychologists have not beenattracted to materialism.) In other words we still donot have a general materialist theory of the mind.

    As for specific theories in physiological psychology,there is no doubt that many have been proposed, par-ticularly over the past quarter of a century and largelythanks to Hebbs influence (HEBB 1949; MILNER,1970; BINDRA, 1976). However, (a) there are too fewof them, (b) those which are close to experiment arefor the most part verbal, and (c) those which aremathematical are usually far removed from experi-ment. (Moreover most theories in mathematical psy-chology are either (a) neobehavioristic learningtheories disregarding the CNS or (b) information-theoretic theories regarding the CNS as a computerrather than a biosystem. Both skip chemistry and bio-logy.)So much for the shortcomings of emergentistmaterialism in its infancy. However many and gravethese may be, the emergentist materialist philosophyof mind seems to be the best we have. and this forthe following reasons:

    I. Because it eschews the mysterious mental sub-stance without thereby denying the mental, emergen-tist materialism is cornputihle \~itll the scirntjficapproach far more than either dualism or eliminativeand reductive materialism.

    2. Emergentist materialism isfieefiom the jii~~inessthat characterizes dualism with its talk of correla-tions between the mental and the physical relationsthat dualists do not care to clarify. perhaps becausethey cannot.

    3. Unlike dualism, emergentist materialism is cotl-sistent with the general concepts ofstate und event thatcan be gleaned from all the sciences. (On the otherhand according to dualism mental states would be

    the o~~IJ~ Lates that liui to lx \ txtcs OI sonic ::$;I+!and mental cvcnts would be the r,~r/l, c\cntz h! l;iiito be changes of state of some thing this being uh~dualism agrees more closeI! \\ith theology th;ln 8:. 1lrscience. i

    4. Unhkc duahsm. cmergentist materialism fo5rc.r\illtCI.u(.tioll /wtwc~rl p.sJ~chol~~~/yrrd r/w crth XlCiill.in particular neuroscience, arid this precisely hecauscit regards mental events as special biological cvcnts.

    5. Unlike dualism. which digs an unbridgeablechasm between man and beast. emergentist matel~ia-lism jibes nith r~~/utior~~~ hioloy~~. which - 1~) c\h-biting the gradual development of the mental Rlcultic:,along certain lineages refutes the superstition thatonly Man has been endowed with a mind.

    6. Unlike dualism, Ljhich postulates an unchal~gingmind, cmergentist materialism LIWOR!.V irh tlcr:&~~~-r~r~t(l/ ps~vholog~ crrltl ~~~~u~~oph~~sio/o~~~.hich cshi bitthe gradual maturation of the brain.

    None of the rivals of cmcrgentist materialism canboast of so many and important supports. direct andindirect. scientific and philosophicaL Therefore it isworth while to try and implement the program ofemcrgentist materialism, i.c. to attempt to build theor-eies of various degrees of generality. mathematical inform and agreeing with the known facts. that construethe mind as a distinguished subset of the set of neur:tlstates and events.

    (ON

  • 7/27/2019 Bunge Mente Cuerpo

    9/9

    Emergence and the mind 509Our espousing emergentist materialism does not

    require affirming that it has in fact solved the mind-body problem. It has not. But it is working on it:witness the progress of physiological psychology.Moreover we submit that emergentist materialism isthe only philosophy of mind that enables a break-through in the scientific investigation of the mind-body (or rather brain-rest-of-me-~dy) problem. Infact it is tbe only one that enjoys the support of allthe life sciences, that does not promote a quixoticreductionism, and that defends neuroscience againstobstruction by obsolete philosophies and ideologies.

    Finally, three caveats are in order. The first is thatto espouse emergentist materialism is not to denysubjective experience or even to disallow employingintrospection as a tool in the scientific inv~tigationof the mental. To espouse emergentist materialism isto favor the understanding of subjectivity in neuralterms, and to encourage the control of subjectivityinstead of allowing the latter to control the courseof research. (In particular, hunches got by introspec-tion must be regarded not as self-evident but as hy-potheses to be subjected to objective tests.)

    Second caveat: Emergentist mate~ali~ does notrequire one to investigate the mental in exactly thesame way as one would investigate earthquakes orinfections. Indeed the psychologist is the luckiest ofscientists in that he can tap a number of sources:he can learn from neurophysiology as it deals withthe levels of the neuronal circuit, the brain subsystem(e.g. the brain stem), and the entire CNS; he can learnfrom introspection and the study of behavior, fromneurosurgery and psychiatry, from com~rativezoology and the study of cultures. He can command,then, many sources of hypotheses and data and justas many ways of checking his hypotheses. In thisregard, then, the study of the psyche is unique. (It

    is not unique in its being subject to the standardcanons of scientific research.)

    Third and last caveat: To explain the mental interms of the neural is not to rule out that the mentalis a set of emergent functions of the brain, any morethan explaining the formation of a liquid vortex rulesout that it possesses properties beyond the propertiesof the indi~dual atoms that take part in it. In otherwords, the ideal of rationality is consistent with plura-lism: to explain is not necessarily to explain away.Besides, the explanation of emergence is anything butstraightforward: it is a matter not of deducing conse-quences from a theory concerning some lower level,but of suitably enriching the latter with new assumptions and data. Thus the theory of neurons does notentail the theory of neuronal circuits, nor does thelatter entail the theory of the reticular formation, andso on. (For the logic of reduction see BUNGE, 197%)A psychophysiological theory, though concerned withsome of the physical or chemical processes in theCNS, deals not just with them but also with a dis-tinguished subset of biological processes going on inneural assemblies, namely those processes that arecommoniy called mental. To explain the mind indepth is to know it, not to ignore it. And to knowsomething is to have adequate theories about it.Acknowledgements-1 gratefully acknowledge stimulatingexchanges with Dr BERNARWDUBROVSK~Allan Memor-ial Institute, McGill University), Professor ROWLFOLLINAS Department of Physiology and Biophysics, NewYork University Medical Center), and Dr RAFAELP&EZPA.SCUAL (Institute de Fisica, Universidad NationalAut~noma de Mexico). And I am indebted to the CanadaCouncil for its continued support of my research as wellas to the Humanities Research Grants Subcommittee ofthe Faculty of Graduate Studies and Research of McGillUniversity.

    REFERENCESAR~~ONG D. M. (1968) A ~~rer~~~~s~Theory of the Mind. Routledge & Kegan Paul, London.BINDR~ D. (1976) A Theory of Intelligent Behauiot. Wiley, New York.BURSTC. V., ED. (1970) The Mind-Body Identity Theory. Macmillan, London.BUNGEM. (1977a) The Furniture of the Wor ld. Reidel, Boston.BUNGEM. (1977b) Levels and reduction. Am. J. Phy siol .: Regul ., Integr. & Compar. Physiol. Z(2).FEXGLH. (1967) The M ental and the Physical. University of Minnesota Press, Minneapolis.FEYERABEND. K. & MAXWELLG., Eds (1966) Mind, Mutt er, and M ethod. University of Minnesota Press, MinneapolisHEBBD. 0. (1949) The Or ganizati on of Behaviour. Wiley, New York.HOOK S., ED. (1960) ~i~e~s~o~s o~~~~d. New York University Press, New York.KOES~LERA. & SMITHIES . R., Eds (1969) Beyond Red~ctio~~s~. Beacon Press, Boston.MILNERP. (1970) PhysioZogica~ Psychology. Holt, Ruehart & Winston, New York.@cONNOR J., Ed. (1969) Modern M ateriali sm: Readings on M ind-Body identit y. Harcourt, Brace & World, New York.PLACEU. T. (1956) Is consciousness a brain process? & it . J. Psychol . 67, 4450.ROSENBLUETH. (1970) M ind and Body. M.I.T. Press, Cambridge, Mass.SMART . J. C. (1959) Sensations and brain processes. Phil. Rev. 68, 141-156.

    (Accept ed 14 Apr il 1977)