Académique Documents
Professionnel Documents
Culture Documents
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=rsl.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
The Royal Society is collaborating with JSTOR to digitize, preserve and extend access to Philosophical
Transactions: Biological Sciences.
http://www.jstor.org
r THE ROYAL
J SOCIETY
John R. Searle
ofPhilosophy,
Department ofCalifornia Berkeley,
University CA 94720, USA
Berkeley,
There is no sharp dividing line between science and philosophy,but philosophical problems tend to have
three special features.First, they tend to concern large frameworksrather than specificquestions within
the framework.Second, they are questions for which there is no generally accepted method of solution.
And third they tend to involve conceptual issues. For these reasons a philosophical problem such as the
nature of life can become a scientificproblem if it is put into a shape where it admits of scientificresolu-
tion. Philosophy in the 20th century was characterized by a concern with logic and language, which is
markedly differentfrom the concerns of earlier centuries of philosophy. However, it shared with the
European philosophical traditionsince the 17thcenturyan excessive concern with issues in the theoryof
knowledge and with scepticism.As the centuryends, we can see that scepticismno longer occupies centre
stage, and this enables us to have a more constructiveapproach to philosophical problems than was
possible for earlier generations. This situation is somewhat analogous to the shift from the sceptical
concerns of Socrates and Plato to the constructivephilosophical enterpriseof Aristotle.With that in
mind, we can discuss the prospects for the followingsix philosophical areas: (i) the traditional mind-
body problem; (ii) the philosophy of mind and cognitive science; (iii) the philosophy of language;
(iv) the philosophyof society; (v) ethics and practical reason; (vi) the philosophyof science.
The general theme of these investigations,I believe, is that the appraisal of the true significanceof
issues in the philosophy of knowledge enables us to have a more constructiveaccount of various other
philosophical problemsthan has typicallybeen possible forthe past threecenturies.
Keywords: philosophy; science; mind-body problem; cognitivescience; epistemology;ethics
Phil.Trans.
R. Soc.Lond.B (1999)
Thefuture J. R. Searle
ofphilosophy 2071
Frege himselfhated ordinarylanguage. He thoughtit was philosophers tend to be very much concerned with
incoherent and self-contradictory, and that we would be science and to see philosophy as aiming for exactly the
much better offwith a logically perfectlanguage of the same sort of objective truththat one gets in the sciences.
sort that he had invented. Russell was no fan of ordinary In my experience, Continental philosophers with some
language, but he thought that the ambiguities and notable exceptions tend to see philosophyas less like the
vagaries of ordinary language could be cleaned up by sciences and more like a branch of literature,or at least
analysing ordinary-language sentences in the predicate closely allied to the studyof literatureand literarytheory.
calculus. The point for the present discussion is this. One furtherfeature of 20th century philosophy needs
Philosophy in the 20th centuryhad three new featuresas to be mentioned. I have said that philosophers in the
a resultof the revolutionbroughtabout by Frege, Russell 20th century showed a special obsession with language,
and Russell's student Wittgenstein,as well as Russell's but the studyof language as a discipline was itselfrevolu-
colleague G. E. Moore. First,Frege'slogic gave us a much tionized by the work of Noam Chomsky and others,
more powerfultool foranalysinglogical relations,and for beginning in the late 1950s. The primary thrust of
the discussion of philosophical problems in general, than Chomsky's work was and still is in the syntax of natural
previous generations of philosophers had. Second, the languages. What exactly are the rules by which humans
philosophical analysis of language itself became a construct sentences in the various natural languages?
central indeed some would say thecentral problem in And what are the rules that are common to all natural
philosophy. What exactly is the relationship between languages, the rules of 'universal grammar'? But philoso-
language and reality? How is it that words can stand for phers were always more interestedin semantics and in
things in the real world? What exactly is the nature of pragmatics, than they were in syntax. Semantics, on a
truth and reference? Third, language was not only a standard definition,deals with the truth conditions of
subject of philosophical investigation,but the analysis of sentences: under what conditions is a sentence true or
language was taken to be an essential tool in investigating false? And pragmatics deals with the use of sentences in
other areas of philosophy. For this reason we need to actual human situations, the use of sentences to give
distinguish between the philosophy of language and orders, make statements,give promises, etc. It seemed to
linguistic philosophy.The philosophy of language deals a number of philosophers of language, myselfincluded,
with certain general featuresof language, such as truth that we should attemptto achieve a unificationof Choms-
and meaning, whereas linguistic philosophy uses the ky's syntax with the results of the researches that were
methods of linguistic analysis to try to solve traditional going on in semantics and pragmatics. I believe that this
problems. So, forexample, a problem I mentionedearlier efforthas proven to be a failure.Though Chomsky did
concerning the nature of causation was treated by 20th- indeed revolutionize the subject of linguistics,it is not at
century philosophers as a matter of analysing the use of all clear, at the end the century,what the solid resultsof
the concept of causation both in the sciences and in this revolutionare. As far as I can tell thereis not a single
ordinary life. What exactly is meant by saying that A rule of syntax that all, or even most, competent linguists
causes B? Can we get an analysis of the causal relation- are prepared to agree is a rule.
ship in termsof more fundamentalfeatures?This was felt In the middle years of the century,in the decades after
by many philosophers of the 20th century to be not so the Second World War, optimism about using logic and
much a revolutionarychange in philosophy,but rather a language as the primary tools of philosophy ran at its
matter of making patterns of analysis that had already highest. And indeed it seems to me that a great deal of
been presentin philosophyclearer and more precise.Thus progress was in fact achieved in those decades. Much of
Hume tried to analyse the notion of causation by exam- the optimism and self-confidenceof the period derived
ining the ideas about causation that he had in his mind. from the belief in two linguistic distinctions.These are
The 20th-centuryphilosopher also proceeds by analysis, (i) the distinction between analytical and synthetic
but instead of analysing the ideas of causation in his propositions,and (ii) the distinctionbetween descriptive
mind, he analyses the language we use in stating causal and evaluative utterances.If you accept these two distinc-
factsabout the world. tions in their pure forms and many philosophers did-
I would not wish to give the impressionthat philosophy they seem to definethe nature of philosophyand to deter-
has been or is now a unified subject. There are many mine its specificresearchprogramme.The firstdistinction
differentschools, methods and approaches in philosophy, between analytic and syntheticstatementsis between, on
and the one that I have described is usually called the one hand, those statementsthat are true or false by
'analytic philosophy'. It is not the only way of doing definition,such as statementsin logic and mathematics
philosophy,but in Britain, the USA and other English- and such commonsense tautologies as All bachelors are
speaking countries, there is no question that it has unmarried'; and, on the other hand, statementsthat are
become the dominant approach to philosophy,and it is true or false as a matterof fact in the world, such as the
the prevailing approach in literally all of our major statements of the natural sciences and such statements
universities.There are other approaches, such as existenti- about contingentfactsin the world as, forexample,'Most
alism and phenomenology. Indeed, phenomenology and bachelors drink beer'. The second distinction, between
its successorscan properlybe described as the more influ- descriptive and evaluative statements,is between those
ential method of doing philosophy in certain European statementsthat describe states of affairsin the world and
countries,especially France.This is not the place to tryto thus can be literallytrue or false, and those that serve to
explain the diffierences between so called 'Continental' express our feelings,attitudesand evaluations, and, thus,
philosophy and analytic philosophy, but one crucial according to the theory,cannot be literallytrue or false.
diffierencefor the purpose of this article is that analytic An example of a descriptive statement would be 'The
R. Soc.Lond.B (1999)
Phil.Trans.
2072 J.R. Searle Thefuture
ofphilosophy
incidentsof crimes of violence have decreased in the last extent of human knowledge?' Hume ended up with a
decade'; an example of an evaluative statementwould be much more radical scepticism than Descartes ever envi-
'It is wrong to commit a crime'. The descriptive class saged, but he feltthat we could live with this scepticism
includes both analytic and synthetic statements. by adopting a completely naturalistic attitude toward
According to those who accepted this theory,statements ourselvesand the world.We just have to accept the limita-
in the sciences and mathematics are descriptivebecause tions of our knowledge, recognize that we do not really
theydescribe mattersof objectivefact; whereas statements know very much and go on as if we did know a great
in ethics and aesthetics are evaluative because they are deal, even though we can offerno justification for the
used to express feelingsand attitudes,and to guide beha- assumptionswe make about the world. Kant read Hume
viour rather than to state facts. For those who accepted and felt that it awakened him from his 'dogmatic
these distinctions,and they were the mainstreamviews in slumber'. He made a heroic effortto overcome Hume's
the middle decades of the century, the distinctions scepticism,but it too, I think,was a failure. In the 20th
defined the nature of philosophy. Philosophers aimed at century,as I have said, the primary interestof philoso-
the truth and so were not in the business of making phers was in language and meaning, and not with knowl-
evaluations or value judgements of any kind. Telling edge and its justification. In short, Descartes' question
people how to live is not thejob of the professionalphilo- was 'How do you know?' and later Russell and Moore
sopher. But the truths of philosophy are not contingent turned that into the question'What do you mean?' None-
synthetic truths of the sort one finds in the natural theless,philosopherssuch as Russell, Moore, Wittgenstein
sciences either.They are necessary analytic truthsabout and Austin devoted a great deal of intellectual effortto
concepts. The philosopher's task, like that of the logician trying to overcome scepticism using linguistic methods.
and mathematician,is to state necessary analytic truths. Though the primary thrust of their analyses was on
His truths are conceptual, explicating puzzling philo- language and meaning, much of the point of the analysis
sophical concepts such as causation, knowledge,justice or of language and meaning was to explain and justify the
truthitself. notions of truth,evidence and knowledge. I believe, and
Philosophy so construed is defined as conceptual I sincerelyhope, that this whole epoch has finallycome to
analysis, and much of the optimismof the middle decades an end. Of course in philosophy nothing is ever finished
of the century derived from the conviction that philo- once and for all, but my interpretationof the present
sophy now had a well-definedresearch project and well- intellectual scene, and my hope for the next century,is
definedmethods forachieving results. that we may simply relinquish our obsession with scepti-
Confidence in these distinctions is now seriously cism and get on with the more constructiveaspects of
weakened. Language does not seem so neat or simple that philosophy.
we can divide utterances into these simple categories of The obsession with epistemology, and its endemic
analytic and synthetic,descriptiveand evaluative. In part obsession with overcoming scepticism, led to a second
because of a loss of confidencein the adequacy of these feature of philosophy in the three centuries after
distinctions,the general optimismthat we mightsolve all, Descartes. For many philosophers real progress required
or even most, philosophical problems using the methods logical reduction.To understanda phenomenonwe had to
of conceptual analysis has now abated. The upshot is that reduce it to simpler phenomena in the sense that we had
philosophy is less self-confidentthan it was in the 1950s to show how statementsabout the puzzling phenomenon
and 1960s, but at the same time, it is much more inter- could be logically derived from statementsabout episte-
esting. All sorts of questions that were not regarded as mically simpler and more primitive phenomena. Thus
really possible philosophical questions in the heyday of many empirically minded philosophers thought that the
language analysis have now become possible, and I will only way to understand human mental states was to
say somethingmore about these shortly.But at the same reduce them to behaviour (behaviourism). Analogously
time there is less confidence about the possibility of many philosophers thought that in order to understand
getting definitive solutions to traditional philosophical empirical reality we had to reduce it to sensory experi-
problemsusing the methods of linguisticanalysis. ences (phenomenalism). A natural consequence of the
There is another important development of 20th- obsession with epistemologywas to see the solution to the
century philosophy that I am less confidentabout, but sceptical problem in reductionism.So there were, in my
which in the end may be its most important result. For view, twin errors that pervaded philosophy and which I
the three centuries afterDescartes, fromthe middle 17th hope we have now overcome.These are scepticismand an
to the late 20th century,the single greatestpreoccupation inappropriateextensionof reductionism.
of philosopherswas with the problems of knowledge and I cannot overestimatethe extentto which the epistemic
scepticism. Descartes made epistemology the theory of bias has infectedthe practice of philosophyfornearly 400
knowledge central to philosophy. For Descartes the years. Even in subjects that would appear to have only a
primary question was what sort of solid foundational fairlyremote connection with epistemology,the epistemic
groundingcan we give to our claims to knowledge,in the question became central to the entire subject. This was
sciences, in common sense, in religion, in mathematics, nowhere more obvious than in the case of ethics and poli-
etc. Subsequent great philosophers, such as Locke, tical philosophy.You might think that the question 'How
Berkeley, Hume, Leibniz, Spinoza and Kant, felt that do we know?' would not figurelarge in these disciplines,
Descartes' attempt to answer scepticism was inadequate, but in fact the central question of ethics in this epoch has
but Descartes' problem remained uppermost in their been 'How can we have objective knowledge in ethics?
philosophical work. Locke, for example, took the main How can we get the kind of epistemic certainty in our
question of philosophy to be 'What is the nature and ethical judgements that we strive for in our scientific
Phil.Trans.
R. Soc.Lond.B (1999)
Thefuture J. R. Searle
ofphilosophy 2073
judgements?' It did not even seem possible to our phenomena. The way to deal with scepticismis not to try
philosophical parents and grandparentsthat there could to refuteit on its own terms,but to overcome it in such a
be a more fundamentalquestion in ethics than this one. way that we can go on to deal with the problems at hand.
Indeed, for those who accepted the distinctionbetween As I said earlier, I am not certain that this is where we
descriptive and evaluative, the result of philosophical are, but it certainlyis where I am in my own intellectual
analysis of ethical discourse was sceptical. According to development. On my interpretationof the contemporary
this view it is impossible to have objective knowledge in philosophical scene, scepticism has finallyceased to be a
ethics because ethical statementscannot be objectively primaryconcern of philosophers,and reductionismhas in
true or false. A similar epistemic bias affectedpolitical general failed.The situationwe are in is somewhat analo-
philosophy. The question again was 'How can we be gous to the situationof the Greeks at the time of the tran-
certain, how can we have epistemic objectivityabout our sition from Socrates and Plato to Aristotle. Socrates and
political judgements and our claims of political obliga- Plato took scepticism very seriously and struggled with
tion?' Just as ethics was afflictedwith a form of scepti- piecemeal issues. Aristotle did not regard the sceptical
cism, so political philosophy fell into the doldrums paradoxes as a serious threat to his overall enterpriseof
because of the same sort of scepticism. Political philo- attempting to do systematic, constructive, theoretical
sophy was revolutionizedand revitalized by the publica- philosophy.I think we now have the tools to move into a
tion ofJohnRawls's A theogy ofjustice(1972), about which I 21st-centuryversion of an Aristotelian phase. Wittgen-
will say more later. stein,one of the most importantphilosophersof the 20th
Nowhere was the epistemic bias more blatant than in centurythoughtthat general theories in philosophy were
the philosophyof language. Frege did not have primarily impossible. Paradoxically, by helping to clear the field of
epistemicworries about meaning, but his followersin the sceptical worriesWittgensteindid as much as anybody to
20th centuryturned questions about meanings into ques- make general philosophical theoriespossible.
tions about knowledge of meanings.This was, in my view,
a disastrous error,but it is an errorthat continues to this
3. SIX PROBLEM AREAS
day. There is an entire movement in the philosophy of
language that thinksthe central question is: What sort of Because of the nature of the subject, I do not believe it
evidence does a hearer have when he attributesmeaning is possible to project a futurecourse of philosophy with
to a speaker of a language? What sort of evidence do I anything like the confidence that one can project the
have that when you utter the word 'rabbit' you mean futurecourse of the sciences though, of course, that is
what I mean by 'rabbit',for example? And the answer to not at all an easy thing to do in itself.What I will, there-
this question is, again in my view mistakenly,taken to be fore,do here is take about a half a dozen areas of philoso-
not merelyan epistemicpoint about how we decide ques- phical investigationwhich are very much alive at the
tions of meaning, but the key to understandingthe very present moment, and discuss their present status and
nature of meaning. Meaning is analysed completely into futureprospects. In some cases I feel confidentenough to
the sorts of evidence that hearers can have about what make some guesses about what I think will happen, in
the speaker means. Many influentialphilosophers have others I can only make critical remarks and expressions
thought that the epistemic question already gave us an of hope forfutureresearchin the coming decades.
answer to the ontological question, that the facts about
meaning were entirely constituted by the evidence we (a) The traditional mind-body problem
could have about meaning. I believe this view is as I begin with the traditional mind-body problem,
mistaken in the philosophy of language as it is in the because I believe it is the contemporary philosophical
sciences and philosophygenerally.It is as if knowledge in problem most amenable to scientific solution: What
physicswere supposed to be knowledge entirelyof experi- exactly are the relations between consciousness and the
ments and meter readings, since we use experimentsand brain? It seems to me the neurosciences have now
meter readings to test our knowledge of the physical progressed to the point that we can address this as a
universe. Analogously, it is equally a mistake to suppose straight neurolobiological problem, and indeed several
that factsabout meaning are factsabout circumstancesin neurobiologists are doing precisely that. In its simplest
which people utter expressions, since we use circum- form, the question is how exactly do neurobiological
stances in which they utter expressions as evidence to processes in the brain cause conscious states and
make judgements about what they mean. I believe this processes, and how exactly are those conscious states and
epistemicbias is nothingless than the philosophical error processes realized in the brain?
of our epoch, and I will have more to say about it in the So stated,thislooks like an empirical scientificproblem.
next section. It looks similar to such problems as 'How exactly do
I have a specific intellectual objective in making the biochemical processes at the level of cells cause cancer?'
proposal that we should abandon scepticism and reduc- and 'How exactly does the genetic structureof a zygote
tionism. I believe we cannot get a satisfactoryconstruc- produce the phenotypicaltraitsof a mature organism?'
tive analysis of language, mind, society, rationality, However, there are a number of purely philosophical
political justice, etc., until we abandon our obsession with obstacles to gettinga satisfactoryneurobiological solution
the idea that the presuppositionof all investigationis first to the problem of consciousness, and I have to devote
to provide a justificationforthe very possibilityof knowl- some space at least to tryingto remove some of the worst
edge, and that real advances in philosophical knowledge of these obstacles.
in general require the reduction of higher-level The single most important obstacle to getting a solu-
phenomena to more epistemically fundamental tion to the traditional mind-brain problem is the
R. Soc.Lond.B (1999)
Phil.Trans.
2074 J. R. Searle Thefuture
ofphilosophy
persistenceof a set of traditionalbut obsolete categories of in modes of existence, the ontological sense of the objec-
mind and body, matter and spirit, mental and physical. tive-subjectivedistinction.
As long as we continue to talk and think as if the mental Science is indeed epistemically objective in the sense
and the physical were separate metaphysical realms, the that scientistsattemptto establishtruthsthat can be veri-
relation of the brain to consciousness will foreverseem fied independentlyof the attitudes and prejudices of the
mysterious,and we will not have a satisfactoryexplana- scientists.But epistemic objectivityof method does not
tion of the relation of neuron firingsto consciousness.The preclude ontological subjectivityof subject matter.Thus
first step on the road to philosophical and scientific there is no objection in principle to having an epistemi-
progress in these areas is to forgetabout the tradition of cally objective science of an ontologically subjective
Cartesian dualism and just remind ourselves that mental domain, such as human consciousness.
phenomena are ordinary biological phenomena in the Another difficultyencountered by a science of subjec-
same sense as photosynthesisor digestion.We must stop tivity is the difficultyin verifyingclaims about human
worryingabout how the brain couldcause consciousness and animal consciousness. In the case of humans, unless
and begin with the plain fact that it does.The notions of we perform experiments on ourselves individually, our
both mental and physical as theyare traditionallydefined only conclusive evidence for the presence and nature of
need to be abandoned, as we reconcile ourselves to the consciousness is what the subject says and does, and
fact that we live in one world, and all the featuresof the subjectsare notoriouslyunreliable. In the case of animals,
world, from quarks and electrons to nation-states and we are in an even worse situation,because we have to rely
balance of paymentsproblemsare, in theirdifferentways, on just the animal's behaviour in response to stimuli.We
part of that one world. I find it truly amazing that the cannot get any statements from the animal about its
obsolete categories of mind and matter continue to conscious states.
impede progress. Many scientistsfeel that they can only but I would point out
I think this is a real difficulty,
investigatethe 'physical' realm and are reluctant to face that it is no more an obstacle in principle than the diffi-
consciousness on its own termsbecause it seems not to be culties encountered in other formsof scientificinvestiga-
physical but to be 'mental', and several prominentphilo- tion where we have to rely on indirectmeans of verifying
sophers think it is impossible for us to understand the our claims. We have no way of observingblack holes, and
relations of mind to brain. Just as Einstein made a indeed, strictlyspeaking, we have no way of directly
conceptual change to break the distinctionbetween space observing atomic and subatomic particles. Nonetheless,
and time, so we need a similar conceptual change to we have quite well-establishedscientificaccounts of these
break the bifurcationof mental and physical. domains, and the difficultiesin verifyinghypotheses in
broughtabout by accepting the
Related to the difficulty these areas should give us a model for verifyinghypoth-
traditional categories is a straightlogical fallacy that I eses in the area of the studyof human and animal subjec-
need to expose. Consciousness is, by definition,subjective, tivity.The 'privacy' of human and animal consciousness
in the sense that fora conscious state to exist it has to be does not make a science of consciousness impossible. As
experienced by some conscious subject. Consciousness in far as 'methodology' is concerned, in real sciences meth-
this sense has a first-personontologyin that it only exists odological questions always have the same answer.To find
fromthe point of view of a human or animal subject, an out how the world works, you have to use any weapon
'I', who has the conscious experience. Science is not used you can lay your hands on, and stick with any weapon
to dealing with phenomena that have a first-person that seems to work.
ontology.By tradition,science deals with phenomena that Assuming, then, that we are not worried about the
are 'objective', and avoids anything that is 'subjective'. problem of objectivityand subjectivity,and that we are
Indeed, many philosophers and scientists feel that prepared to seek indirect methods of verification of
because science is, by definition,objective, there can be hypotheses concerning consciousness, how should we
no such thing as a science of consciousness, because proceed? Most scientificresearch today into the problem
consciousness is subjective.This whole argument rests on of consciousness seems to me to be based on a mistake.
a massive confusion,which is one of the most persistent The scientistsin question characteristicallyadopt what I
confusionsin our intellectual civilization. There are two will call the building-block theory of consciousness, and
quite distinct senses of the distinctionbetween objective they conduct their investigation accordingly. On the
and subjective. In one sense, which I will call the episte- building-block theory,we should think of our conscious
mological sense, there is a distinctionbetween objective fieldas made up of various building blocks, such as visual
knowledge and subjectivemattersof opinion. If I say, for experience, auditory experience, tactile experience, the
example, 'Rembrandt was born in 1606', that statementis stream of thought,etc. The task of a scientifictheory of
epistemicallyobjective in the sense that it can be estab- consciousness would be to findthe neurobiological corre-
lished as true or false independentlyof the attitudes,feel- late of consciousness (nowadays called the NCC) and, on
ings, opinions or prejudices of the agents investigatingthe the building-block theory,if we could find the NCC for
question. If I say 'Rembrandt was a better painter than even one building block, such as the NCC for colour
Rubens', that claim is not a matter of objective knowl- vision, that would in all likelihood give us a clue to the
edge, but is a matterof subjectiveopinion. But in addition building blocks for the other sensory modalities and for
to the distinction between epistemically objective and the stream of thought. This research programme may
subjectiveclaims, there is a distinctionbetween entitiesin turn out to be rightin the end. Nonetheless,it seems to
the world that have an objective existence,such as moun- me doubtfulas a way to proceed in the presentsituation,
tains and molecules, and entities that have a subjective for the following reason. I said above that the essence
existence,such as pains and tickles. I call this distinction of consciousness was subjectivity.There is a certain
R. Soc.Lond.B (1999)
Phil.Trans.
Thefuture J.R. Searle
ofphilosophy 2075
Phil.Trans.
R. Soc.Lond.B (1999)
2076 J.R. Searle Thefuture
ofphilosophy
rather then in series. But if we claim that the remarked, during the first three-quarters of the 20th
connectionist architecture rather than connectionist century,the philosophyof language was taken to be 'first
computationsare responsibleformental processes, we are philosophy'. But as the century comes to an end that is
no longer advancing the computational theory of the changing. Less is happening in the philosophy of
mind, but are engaging in neurobiological speculation. language now than in the philosophy of mind, for
With this hypothesis we have abandoned the computa- example, and I believe that some of the currentlymost
tional theoryof the mind in favourof speculative neuro- influentialresearch programmes have reached a kind of
biology. dead end. Why? There are many reasons of which I will
What is actually happening in cognitive science is a mentiononly three.
paradigm shift away from the computational model of First,the more successfulbranches of the philosophyof
the mind and towards a much more neurobiologically language are now passing into the science of linguistics.
based conception of the mind. For reasons that should be The sortof researchthat I and othersdid 30 years ago on
clear by now, I welcome this development.As we come to the theory of speech acts and on the use of language is
understand more about the operations of the brain it now becoming a part of linguistics called 'pragmatics',
seems to me that we will succeed in gradually replacing which has its own corner in linguistics,with its own jour-
computational cognitive science with cognitive neuro- nals, annual meetings,etc. In short,this part of the philo-
science. Indeed I believe this transformationis already sophy of language is gradually being kicked out of
taking place. philosophy,upstairs into the social sciences. I welcome
Advances in cognitive neuroscience are likely to create this development, and I believe that it is an example of
more philosophical problems than they solve. For the sort of phenomenon that I described in the early part
example, to what extentwill an increased understanding of this article, where I explained that as areas of investi-
of brain operations force us to make conceptual revisions gation arrive at established methodologies for their
in our commonsense vocabulary for describing mental research,they tend to be thoughtof as more scientificand
processes as they occur in thought and action? In the less philosophical.
simplest and easiest cases we can simply assimilate the Second, one of the main research programmes in the
cognitiveneuroscience discoveriesto our existingconcep- philosophy of language suffersfromthe epistemic obses-
tual apparatus. Thus, we do not make a major shiftin our sion that I have been castigating. A commitment to a
concept of memory when we introduce the sorts of certain formof empiricism,and in some cases even beha-
distinctionsthat neurobiological investigationhas made viourism, led some prominentphilosophersto try to give
apparent to us. We now in popular speech distinguish an analysis of meaning according to which the hearer is
between short-term and long-term memory, and no engaged in an epistemic task of tryingto figureout what
doubt as our investigationproceeds, we will have further the speaker means either by looking at his behaviour in
distinctions. Perhaps the concept of iconic memory is response to stimulusor by looking at the conditionsunder
already passing into the general speech of educated which he would hold a sentence to be true. The idea is
people. But in some cases it seems we are forcedto make that if we could describe how the hearer solves the epis-
conceptual revisions. I have thoughtfor a long time that temic problemwe will therebyanalyse meaning.
the commonsense conception of memory as a storehouse This work, I believe, is going nowhere, because its
of previous experience and knowledge is both psychologi- obsession with how we know what a speaker means
cally and biologically inadequate. My impression is that obscures the distinctionbetween how the hearer knows
contemporaryresearch bears me out on this.We have to what the speaker means and what it is that the hearer
have a conception of memoryas a creative process rather knows. I think that epistemology plays the same role in
than simply a retrievalprocess. Some philosophers think the philosophy of language as it does, for example, in
even more radical revisionsthan this will be forcedupon geology. The geologist is interested in such things as
us by the neurobiological discoveriesof the future. tectonic plates, sedimentation and rock layers, and will
Another set of philosophical problems arises when we use any method that comes to hand to tryto findout how
begin to examine the relationshipsbetween the develop- these phenomena work. The philosopher of language is
mental evidence regarding mental phenomena and the interestedin meaning, truth,referenceand necessity,and
mental phenomena as they occur in mature adults. Very analogously should use any epistemic method that comes
young children apparently have a differentconception of to hand to tryto figureout how these phenomena work in
the relation of belief to truthfromthat which adults have. the minds of actual speakers and hearers. What we are
How seriously should we take these differences?Do we interestedin is what are the factswhich are known; and
need to enrich our theory of intentionalityby incorpor- to a much lesser extent are we interestedin the question
ating the developmental data? We do not yet know the of how we come to know these facts.
answer to any of these questions, and my point in raising Finally, I think the greatest source of weakness in the
them here is to call attentionto the factthat once we have philosophy of language is that its currentlymost influen-
removed the philosophical error of supposing that the tial research project is based on a mistake. I said earlier
brain is a digital computer, and once we have a more that Frege was the founderof the philosophyof language,
mature and sophisticatedcognitive neuroscience, we still but Frege had a conception of meaning that placed the
have to deal with a numberof philosophical questions. meanings of words inside the heads of the speakers of a
language. Frege was anxious to insistthat these meanings
(c) The philosophy of language were not psychologicalentities,but he did thinkthat they
I said that the philosophyof language was the centre of could be grasped by speakers and hearers of a language.
philosophy for most of the 20th century. Indeed, as I Frege thought that communication in a public language
Phil.Trans.
R. Soc.Lond.B (1999)
Thefuture J. R. Searle
ofphilosophy 2077
was possible only because there is an ontologically tive agreement or a belief that they exist? How is it
objective realm of meanings, and that the same meaning possible that there can be an objective social reality that
can be grasped equally by both speaker and hearer. A existsonly because we thinkit exists?
number of authors have attacked this conception. They When questions of social ontology have been properly
believe that meaning is a matter of causal relations sorted out it seems to me that the questions of social
between the utterancesof words and objects in the world. philosophy,namely the nature of explanation in the social
So the word 'water', for example, means what it does to sciences and the relation of social philosophy to political
me not because I have some mental content associated philosophy,will naturally fall into place. I attempted to
with that word, but ratherbecause there is a causal chain begin this research project in The construction ofsocialreality
connecting me to various actual examples of water in the (Searle 1995).
world. This view is called 'externalism',and it is usually Specifically,I believe that in our study of political and
opposed to the traditional view, called 'internalism'. social reality,we need a set of concepts that will enable us
Externalism has led to an extensive research project of to describe political and social reality,so to speak from
tryingto describe the nature of the causal relations that the 'middle distance'. The problem that we have in
give rise to meaning. The problem with this research attemptingto cope with social realityis that our concepts
project is that nobody has ever been able to explain, with are either immensely abstract, as in traditional political
any plausibility whatever, the nature of these causal philosophy, for example the concepts of the social
chains. The idea that meanings are somethingexternal to contract or the class struggle; or they tend to be essen-
the mind is widely accepted, but no one has ever been tially journalistic, dealing with day-to-day questions of
able to give a coherent account of meaning in these policy and power relations. Thus we are quite sophisti-
terms. cated in abstract theories of justice, and with developing
My prediction is that no one will ever be able to give criteriaforassessing thejustice or injusticeof institutions.
a satisfactory account of meanings as something Much of the progressin this area is owed toJohn Rawls,
external to the head, because such external phenomena who, as I mentioned,revolutionizedthe studyof political
could not function to relate language to the world in philosophy with his classic work A theory ofjustice (Rawls
the way that meanings do relate words and reality. 1972). But when it comes to political science, the cate-
What we require in order to resolve the dispute gories traditionallydo not rise much above the level of
between internalistsand externalistsis a more sophisti- journalism. Therefore,if,forexample, you read a work in
cated notion of how the mental contents in speakers' political science as recent as 20 years old, you will find
heads serve to relate language in particular, and human that much of the discussion is out of date.
agents in general, to the real world of objects and states What we need, I believe, is to develop a set of cate-
of affairs. gories that would enable us to appraise social reality in a
way which would be more abstract than that of day-to-
(d) The philosophy of society day political journalism, but at the same time would
It is characteristicof the historyof philosophythat new enable us to ask and answer specific questions about
branches of the subject are created in response to intellec- specific political realities and institutionsin a way that
tual developmentsboth inside and outside of philosophy. traditional political philosophy was unable to do. Thus,
Thus, for example, in the early part of the 20th century for example, I think the leading political event of the
the philosophy of language in the sense in which we now 20th centurywas the failureof ideologies such as those of
use that expression, was created largely in response to Fascism and communism, and in particular the failure of
developments in mathematical logic and work on the socialism in its differentand various forms.The inter-
foundations of mathematics. A similar evolution has estingthing fromthe point of view of the presentanalysis
occurred in the philosophy of mind. I would like to is that we lack the categories in which to pose and answer
propose that in the 21st century we will feel a pressing questions dealing with the failure of socialism. If by
need for,and should certainlytry to develop, what I will 'socialism' we mean state ownership and control of the
call a philosophy of society. It is characteristic of the basic means of production,then the failureof socialism so
social parts of philosophy that we tend to construe social definedis the single most importantsocial developmentof
philosophyas eithera branch of political philosophy,thus the 20th century.It is an amazing fact that that develop-
the expression'social and political philosophy',or we tend ment remains unanalysed and is seldom discussed by the
to construe social philosophyas a studyof the philosophy political and social philosophersof our time.
of the social sciences, just as the philosophy of natural When I talk of the failure of socialism, I am referring
sciences is a branch of the philosophy of science. I am not only to the failure of Marxist socialism, but the
proposing that we should have a social philosophy,which failure of democratic socialism as it existed in the coun-
stands to social sciences in the same way that the phil- tries of Western Europe. The socialist parties of those
osophy of mind stands to psychology and cognitive countries continued to use the vocabulary of socialism,
science, or the philosophy of language stands to linguis- but the belief in the basic mechanism of socialist change,
tics. It would deal with more general frameworkques- namely the public ownership and control of the means of
tions. In particular, I think we need much more work on production, apparently has been quietly abandoned.
questions of the ontology of social reality. How is it What is the correct philosophical analysis of this entire
possible that human beings, through their social inter- phenomenon?
actions can create an objective social reality of money, A similar sort of question would be the appraisal of
property,marriage, government,wars, games, etc., when national institutions.So, for example, for most political
such entitiesin some sense exist only by virtue of a collec- scientistsit would be very difficultto attempt to analyse
Phil.Trans.
R. Soc.Lond.B (1999)
2078 J. R. Searle Thefuture
ofphilosophy
the backwardness,corruptionand general dreadfulnessof but my impression is that, more than any other single
the political institutionsof several contemporary nation factor,Rawls's work not only revived political philosophy
states.Most political scientists,given theircommitmentto but made substantiveethics seem possible as well.
scientificobjectivity,and the limited categories at their
disposal, cannot even attempt to describe how dreadful (f) The philosophy of science
many countries are. Many countries have apparently In the 20th century,not surprisingly,the philosophy
desirable political institutions,such as a writtenconstitu- of science shared the epistemic obsession with the rest of
tion, political parties, free elections,etc., and yet the way philosophy. The chief questions in the philosophy of
they operate is inherentlycorrupt.We can discuss these science, at least forthe firsthalf of the century,had to do
institutionsat a very abstract level, and Rawls and others with the nature of scientificverification,and much effort
have provided us with the tools to do so. But I would like was devoted to overcoming various sceptical paradoxes,
an expanded social philosophy which would provide us such as the traditional problem of induction.Throughout
with the tools foranalysing social institutionsas theyexist most of the 20th century the philosophy of science was
in real societies, in a way that would enable us to make conditioned by the belief in the distinction between
comparativejudgements between differentcountries and analytic and syntheticpropositions.The standard concep-
larger societieswithout,at the same time, risingto such a tion of the philosophyof science was that scientistsaimed
level of abstraction that we cannot make specific value to get syntheticcontingenttruthsin the formof universal
judgements about specific institutional structures.The scientificlaws. These laws statedvery general truthsabout
work of the economist-philosopherArmatya Sen is a step the nature of reality,and the chiefissue in the philosophy
in this direction. of science had to do with the nature of their testingand
verification.The prevailing orthodoxy,as it developed in
(e) Ethics and practical reason the middle decades of the century, was that science
For much of the 20th centurythe subject of ethics was proceeded by somethingcalled the 'hypothetico-deductive
dominated by a version of the same scepticism that has method'. The scientistsformed the hypothesis, deduced
affectedother branches of philosophy for several centu- logical consequences fromit, and then testedthose conse-
ries. Just as the philosophy of language was damaged by quences in the formof experiments.This conception was
the urge to treat the users of language as essentially articulated, I think more or less independently,by Karl
researchers engaged in an epistemic task of trying to Popper and Carl Gustav Hempel.
figureout what a speaker of a language means, so ethics Those practising scientistswho took an interestin the
was obsessed by the question of objectivity.The principal philosophy of science at all, tended, I think, to admire
issue in ethics was about whether or not there could be Popper's views, but much of their admiration was based
epistemic objectivity in ethics. The traditional view in on a misunderstanding.What I think they admired in
analytic philosophy was that ethical objectivity was Popper was the idea that science proceeds by acts of
impossible, that you could not in Hume's phrase derive originalityand imagination. The scientisthas to form a
an 'ought' from an 'is', and consequently ethical state- hypothesison the basis of his own imagination and guess-
mentscould not literallybe eithertrue or false, but func- work. There is no 'scientific method' for arriving at
tioned only to express feelings or to try to influence hypotheses.The procedure of the scientistis then to test
behaviour, etc. The way out of the sterility of these the hypothesis by performing experiments and reject
debates is not, I think, to try to show that ethical state- those hypothesesthat have been refuted.
ments are true or false in a way that, for example, scien- Most scientists do not, I think, realize how anti-
tificstatementsare true or false, because there are clearly scientific Popper's views actually are. On Popper's
important differencesbetween the two. The way out of conception of science and the activityof scientists,science
the impasse, I believe, is to see that ethics is really a is not an accumulation of truths about nature, and the
branch of a much more interestingsubject of practical scientistdoes not arrive at truthsabout nature, rather,all
reason and rationality.What is the nature of rationalityin that we have in the sciences are a series of so far un-
general and what is it to act rationallyon a reason foran refutedhypotheses. But the idea that the scientist aims
action? This, I believe, is a more fruitfulapproach than aftertruth,and that in various sciences we actually have
the traditionalapproach of worryingabout the objectivity an accumulation of truths, which I think is the pre-
of ethical statements. suppositionof most actual scientificresearch,is not some-
Something like the study of rationality,as a successor thingthat is consistentwith Popper's conception.
to ethics as it was traditionallyconstrued, seems to be The comfortableorthodoxyof science as an accumula-
already happening. Currentlythere are, for example, a tion of truths,or even as a gradual progressionthrough
number of attemptsto revive Kant's doctrine of the cate- the accumulation of so far unrefuted hypotheses, was
gorical imperative. Kant thought that the nature of challenged by the publication of Thomas Kuhn's Structure
rationality itself set certain formal constraints on what revolutions
of scientific in 1962. It is puzzling that Kuhn's
could count as an ethically acceptable reason for an book should have had the dramatic effectthat it did,
action. I do not believe these effortswill succeed, but because it is not strictlyspeaking about the philosophyof
much more interestingthan their success or failureis the science, but about the history of science. Kuhn argues
fact that ethics as a substantivebranch of philosophy- that if you look at the actual history of science, you
freedfromits epistemic obsession to finda formof objec- discover that it is not a gradual progressiveaccumulation
tivity and the inevitable scepticism when the quest for of knowledge about the world, but that science is subject
objectivity fails seems now to have become possible to periodic massive revolutions,where entire world views
again. I am not sure what the reasons forthe change are, are overthrownwhen an existingparadigm is overthrown
R. Soc.Lond.B (1999)
Phil.Trans.
Thefuture J. R. Searle
ofphilosophy 2079
R. Soc.Lond.B (1999)
Phil.Trans.
2080 J. R. Searle Thefuture
ofphilosophy
space. Analogously, I believe, we should have the same greater theoretical understandingand more constructive
attitude towards the paradoxes about the possibility of theoreticalaccounts than we have had at any time in the
knowledge that were advanced by sceptical philosophers. past historyof the subject.
That is, these are interestingpuzzles, and they provide
good five-finger exercisesfortrainingyoung philosophers,
but we should not suppose that the possibilityof knowl- REFERENCES
edge and understanding rests on our firstbeing able to
Kuhn, K. 1962 Structure of scientific Universityof
revolutions.
refute Hume's scepticism. I cannot, of course, predict
ChicagoPress.
what is going to happen in the 21st century,but I can Rawls, J. 1972 A theoryofjustice.Cambridge,MA: Harvard
express the hope, and I thinkat this stage in our intellec- University Press.
tual historyit is a well-foundedhope, that with the aban- Searle,J. R. 1980 Minds,brainsand programs.Behav.BrainSci.
donment of the epistemic bias in the philosophy of 3, 417-457.
language, the philosophyof mind, ethics, political philo- Searle,J. R. 1995 Theconstruction NewYork:The
ofsocialreality.
sophy and the philosophy of science, we may achieve FreePress.
R. Soc.Lond.B (1999)
Phil.Trans.