Luke Morris

12/12/2004

PHL 341 – 20^th Century Analytic

Dr. Burke

John Searle’s Mental Realism and the Challenge to Skepticism[1]

            John Searle has been doing philosophy for nearly a half-century, and in that time he has produced important and influential work in many areas of the field, most notably in the philosophies of mind and language.  On reading his essays one finds numerous formulations and logical defenses of common-sense notions – a form of philosophy that echoes certain writings of G.E. Moore and the later Wittgenstein.  Though he is an analytic philosopher, Searle does not constrain his focus to minutiae of meaning in language, but willingly and ably tackles larger problems.  The central question that he has wrestled with throughout his career is: “How can we have a unified and theoretically satisfactory account of ourselves and of our relations to other people and to the natural world?” (1).  He has addressed smaller issues – the nature of speech acts, consciousness, intentionality, society, and rationality – as parts of this larger problem.  But his most influential (and controversial) work involves his conception of consciousness, and his creative arguments against alternative accounts of mind and meaning.  Searle posits a conception of mental realism with a physicalistic base, and uses this notion as a part of his attacks on behaviorism, functionalism, and different forms of skepticism.  After laying out his notion of ‘biological naturalism,’ he shows computer-type functionalism to be intuitively unstable, and demonstrates that the skeptical ‘indeterminacy’ arguments of Quine and Davidson cannot prove what their authors wish them to prove.

On Consciousness

            The notion of consciousness cannot be defined by enumerating necessary and sufficient conditions, but we still can say roughly what we mean by the word.  Searle, for one, uses the term ‘consciousness’ to designate our subjective states of awareness within our waking lives.[2]  As Searle sees it, this awareness is a biological phenomenon, similar to digestion and growth in living organisms.  What distinguishes consciousness from these other phenomena, though, is its subjectivity, the sense in which each person’s consciousness is private to that person.  Each person is related to his own experiences in a way that no one else is.  Subjectivity is “that feature of consciousness by way of which there is something that it’s like or something that it feels like to be in a certain conscious state” (8).  The term refers to the qualitative character of conscious states known as ‘qualia.’  Consciousness, then, is not to be confused with knowledge, attention, or self-consciousness.

            In his philosophy of mind Searle rejects the classic versions of dualism and reductive materialism, and solves the old ‘mind-body problem’ with a simple notion, stated thus:  “Conscious states are caused by lower level neurobiological processes in the brain and are themselves higher level features of the brain” (9).  The important thing to notice here is the causal relationship between consciousness and brain processes – “brain processes cause conscious states” – and the fact that consciousness is itself a feature of the brain.  The consciousness caused by brain processes is not a separate substance, but a higher level feature of the same system as the lower level neuronal elements.  ‘Higher level’ means that the feature is not reducible to particular neuronal events.  This is analogous to the liquidity of water or the transparency of glass – one cannot pick a water molecule out of a glass and say ‘this is wet,’ nor can one point to a synapse or neuron and say ‘this one is thinking of baseball.’  Thoughts, feelings, and emotions simply occur at a higher level than that of single neuronal events (10).

            Searle argues, for one, that there is no metaphysical or logical obstacle to claiming that there is a causal relationship between the brain and conscious states, while simultaneously saying that consciousness is just a feature of the brain.  Indeed, “we simply know as a matter of fact that brain processes cause conscious states” (10).  We do not know the details of this causal process yet, but just because we do not know how it works doesn’t mean we do not know that it works.  The problem of how neurobiological processes cause conscious phenomena is one for the biological sciences.  Still, we already know (as a matter of ‘common sense,’ so to speak) that it does so.  Searle’s solution to the mind-body problem, as he sees it, is actually a dissolution of it; we know ‘that,’ it is now up to science to tell us ‘how’ (11).

            Is Searle’s account of consciousness coherent?  It is rather simple, and it certainly seems plausible from a common-sense view at the outset.  One problem he might face is that, in deferring to science to tell us ‘how,’ he defers to it to tell us ‘that’ as well.  Just because there is no logical impossibility as regards a cause and its effect happening within the same physical system doesn’t mean that it actually does so.  If science is to act as arbiter, then it is also logically possible that biologists may be able to pin down aspects of conscious states to particular neuronal events, thereby reverting to reductionism.  Of course, Searle’s account seems more likely, since we can point to a definite distinction between a thought, which has logical structure and content, and the fact of a single neuron or group of neurons firing.  But does this not bring us back to dualism?  Searle might be perfectly willing to accept a sort of property-dualist account of consciousness, but the logical possibility seems just as open for a substantive dualism, in which the consciousness that the brain connects to really is a separate and irreducible substance from the physical matter.  This brings us back to the mind-body problem again.  Searle’s answer to these challenges, then, may go back to Ockham’s Razor – that his account is the simplest view that coheres with all of the available evidence.  On this issue he may have a point, but we will not spend any more time on this here. 

We must next discover how Searle lays out some distinctive features of consciousness that science must study.  The most important distinguishing feature of consciousness, again, is its subjectivity, which is a phenomenon unlike any other in biology.  Science must seek an epistemically objective account of an ontologically subjective reality – that of subjective states of consciousness.  Another feature of consciousness is its unity.  Many conscious experiences occur simultaneously, within one unified experience.  This unity has two major aspects: first, “at any given instant all of our experiences are unified into a single conscious field,” and second, “the organization of our consciousness extends over more than simple instants” (12).  Scientific study has to determine how all of these experiences can fit together in a single system.

A third feature of consciousness, and one that Searle in the past has devoted a great deal of time to, is intentionality, the feature of many mental states that is directed at, or about, things or states of affairs in the world.  Intentionality subsumes such things as belief, desire, and fear, which all require a content, something existent (or non-existent) to be about.  Not all conscious states are intentional, of course; only those directed at something beyond themselves count as such.[3]  And we are not always conscious of our own intentionality – indeed, most of our intentional states at a given point in time are unconscious.  Still, “in order for an unconscious intentional state to be genuinely an intentional state it must be accessible in principle to consciousness” (12).  What does it mean to direct a conscious state at someone?  What does it mean to hold an unconscious intentional state?  How does one access and use intentionality?  These are some of the questions that a science of the mind must address.

While the features of subjectivity, unity, and intentionality are central, consciousness possesses other attributes that a scientific study must consider.  These features include, for one, the distinction between the center and the periphery.  Consciousness for a given entity exists as a field, within which one pays attention to some things (the center) and not others (the periphery).  One can at any moment, however, shift his attention to those things that had previously been on the periphery.  This brings us to another feature: the Gestalt structure of conscious experience.  “Within the field of consciousness,” Searle writes, “our experiences are characteristically structured in a way that goes beyond the structure of the actual stimulus” (13).  We have our conscious experiences in certain structures, and we have them as figures against backgrounds.  The brain will structure degenerate stimuli into certain forms,[4] and in the case of vision, for instance, we see these forms against a background, and that background against another, and so on, to the edge of visual consciousness.[5] 

We have here a somewhat Kantian picture of an ‘active mind,’ which comes even clearer when Searle describes the next feature: the aspect of familiarity.  For me to see the objects in front of me as a chair, a desk, a coffee cup, and so on, I must already possess those descriptive categories – I assimilate my experiences into a familiar set of categories like these.  Another feature that I bring to, or that comes with, each set of conscious states is a mood, an emotional flavor or tone, which pervades every conscious experience (14).  And these experiences, these states of consciousness, all fall within boundary conditions, a certain ‘situatedness.’  The details of the situation are not always a part of the content of conscious states, but the states themselves still come with the boundary conditions of, for instance, what year, season, time of day, and place they occur in (15).  The above features – subjectivity, unity, intentionality, center and periphery, gestalt structure, familiarity, mood, and boundary conditions – all play a part in conscious experiences, and they all must be considered if we are to have an adequate science of consciousness.

As Searle sees it, people make a number of common mistakes in talking about consciousness.  For one, “Instead of recognizing that consciousness is essentially a subjective, qualitative phenomenon, many people mistakenly suppose that its essence is that of a control mechanism or a certain kind of set of dispositions to behavior or a computer program” (15).  In other words, they treat consciousness as an objective, third-person phenomenon. The two most common mistakes made are that consciousness can be analyzed 1) behaviorally or 2) computationally.  The Turing test[6] disposes us to make these mistakes by supposing that, to be conscious, it is necessary and sufficient that a system possess the right computer program, giving the right outputs to certain inputs.  The problem with this view is similar to the old problem with behaviorism: “There is no logical connection, no necessary connection between inner, subjective, qualitative mental states and external, publicly observable behavior” (16).  In other words, the behavior that conscious states cause is logically distinct from the states themselves.  The computational model of consciousness is likewise distinct from the thing it models – the model itself is not conscious. 

Searle’s case against the computational model goes back to his classic “Chinese Room” argument, the conclusion of which is that

Computation is defined syntactically.  It is defined in terms of the manipulation of symbols.  But the syntax itself can never be sufficient for the sort of contents that characteristically go with conscious thought . . . Syntax by itself is not sufficient for semantic content.  (16) 

 

Today, however, Searle says that he conceded too much in claiming that the computational theory of mind was false.  Now he sees that the computational theory does not even reach the level of falsity, because it does not have a clear sense.  He shows this by first drawing a distinction in the features of reality between the intrinsic and the observer-relative.  Physical objects all have features that are intrinsic to reality, that can be studied by the natural sciences; but some objects, such as chairs, five-dollar bills, and ‘nice days,’ possess in addition certain features only in relation to observers or users.[7]  Computation is entirely observer-relative in this sense, since it is defined in terms of symbol manipulation, and “Something is a symbol only if it is used, treated or regarded as a symbol” (17).  There are no purely physical properties that symbols have that determine whether or not they are symbols.  Thus, “The question, ‘Is consciousness a computer program?’, lacks a clear sense” (17).  We can give a computational interpretation to anything, but nothing is intrinsically computational, and therefore consciousness cannot be.  “Computation exists only relative to some agent or observer who imposes a computational interpretation on some phenomenon” (17).  The notion of computation presupposes consciousness, and so cannot be used to explain it.

 

Against Behaviorism and Functionalism

            I will now go over how Searle defends his account of mental realism by undermining behaviorist and functionalist accounts of consciousness.  For Searle, as we saw above, “thinking and other mental processes and events, like linguistic processes and events, are biologically based” (203), and are as real as digestion and other biological processes.  The behaviorist objection to mentalism arose from the logical positivists’ criteria of verification, since the only way to empirically verify statements about the mental, as the positivists saw it, is to observe behavior.  One objection to the behaviorist account, though, is the non-identity of sensations and behavior: to have a sensation of pain, for instance, is not the same thing as saying ‘ow.’ And again, one could demonstrate the behavior without actually having the sensation, and one could have a sensation of pain without showing it (204).  Another objection to behaviorism stems from the mentalistic element in human actions.  If a behaviorist seeks to analyze mental notions by an analysis of behavioral notions, and behavior is to include all human actions, this behavior must include intentional behavior, or mental processes, which leads him in a circle.  If, on the other hand, he wishes to constrain his analysis to bodily movements[8], independent of intentionality, then he can only explain bodily movements in terms of bodily movements, since a being could perform all bodily movements without the requisite intentions, beliefs, and desires that make up the mental.  Hence, behaviorism cannot describe what it wishes to describe – it collapses into describing movement in terms of movement, and the mental in terms of the mental (205).

Searle then brings back in the distinction between the intrinsic and the observer relative, this time as it relates to ascriptions of mental phenomena.  Intrinsic mental phenomena ascriptions involve one ascribing, to himself or others, thoughts, feelings, intentions, and so on.  Observer relative mental ascriptions, by contrast, do not ascribe intrinsic mental phenomena to the subject; the only mental phenomena are in the minds of the observers.  If I say, for instance, that I have a headache, I am ascribing intrinsic mental phenomena to myself; on the other hand, if I say that my calculator does simple algebra, I am not ascribing intrinsic mental phenomena to it.  There is only one type of mental phenomena, but there are two types of ascriptions of mental predicates (206). 

Proceeding to functionalism, Searle argues that it suffers the same problems that behaviorism does, and adds to them a confusion between observer relative and intrinsic mental ascription.  This is representative of “a recurring fallacy in analytic philosophy, the confusion between features of the language we use to describe a phenomenon and features of the phenomenon” (207).  The basic view of functionalism, for instance, follows an analogy to clocks and carburetors, which, though they may be differently constructed in their internal workings, nevertheless all function the same.  Thus, for the functionalist, “mental states are identified by their functions, and not by how those functions are realized in the brain” (208).  The problem with this view is that mental states cannot be functional states, since

[T]he mental states in question are intrinsic and functions are always observer relative.  The ascription of a function to a system or to an element of a system is always made relative to some goal, purpose, or objective; functions are never just causes, they are causes relative to a teleology.  (209) 

 

If we try to define mental states in terms of their functions, it is easy to see that something else could serve those same functions, even though it lacked the intrinsic features we want to define.  The mental level of description is intrinsic, not functional.

            ‘Carburetor’ functionalism soon fell by the wayside in favor of Turing Machine or ‘organizational’ functionalism, as advanced by Daniel Dennet and Hilary Putnam.  This view holds that mental states are a certain type of functional states: the logical states of a computer, which are intrinsic states at the computer program level of description.  According to the theory of mind proposed, mental processes are computational processes.[9]  The basic idea of organizational functionalism is that “psychological states are matters of functional organization” (210), and this is the notion that Searle attacks.  He first lays out Putnam’s version of the theory[10], which seeks to defend the hypothesis that “being in pain is a functional state of the organism,” and then argues that satisfying the conditions Putnam sets down for being in pain is fundamentally different from actually being in pain (211-12). 

            What Putnam’s argument amounts to, as Searle sees it, is:

Beings capable of being in pain are those and only those that instantiate a certain computer program; the computer program specifies a set of transition probabilities between input, output, and internal states, and the system will actually be in pain when the right input activates the system according to the program. (212) 

 

In other words, there are certain “machine tables” for pain, and specific inputs for pain, by which we can define functionalism.  The problem with this is that a system or organism could possibly satisfy Putnam’s conditions and not have any pains.  Searle presents two arguments for this conclusion.  The “Argument from Anesthesia” shows that one could be anesthetized to the point where he could not feel pain, yet he could still instantiate the right input/output program to, say, being punched in the nose, by ‘looking up’ the appropriate response (“Ouch!”) in the machine table and carrying out the instruction.  In other words, the instantiation of a given program for certain inputs could not be sufficient for pain, because “a human agent could in his own mental processes instantiate any machine table you like and could have any inputs you care to induce in him and still not feel the relevant sensations if his specific neurophysiological states are not appropriate” (213). 

            Searle’s second response to Putnam, the “Argument from Biology,” is a bit more complex.  Putnam’s two claims are, first, that for a human to be in pain is to instantiate a certain type of program with the right inputs and outputs, and second, that any system (i.e. a robot) that was functionally equivalent to a human could instantiate such a program, and would be in pain when it did.  Searle’s counter-argument begins with the shared assumption that humans (and some animals) “have nervous systems that are causally sufficient to enable them to feel pains and have other sorts of mental states” (215).  Any system that could have such states, then, must also possess the relevant causal powers [is this a logical leap?].   And as seen from the previous argument, instantiating a program could not be sufficient for mental states; thus Putnam’s first claim does not hold.  Therefore, a program with input-output relations could not by itself give a system the relevant causal powers required for mental states.  Humans, though, have a causal capacity to produce mental states that does not consist merely of such a program.  Hence, robots and other such systems would need more than a program with certain input-output relations in order to have mental states – they would have to have causal powers equivalent to those of a brain.  This conclusion negates Putnam’s second claim (216).

            The Argument from Biology is strong, but it has a possible flaw in that its second step does not seem to follow from the first:  Just because humans and animals have nervous systems with the causal capacity to allow them mental states does not seem to necessitate that any system capable of mental states must have equivalent causal powers.  If this step is valid – and it seems that more detail is needed to show that it is – then the argument holds.  To show that the step is not valid, though, the functionalist might need to show that it is possible for a system without such causal powers to have mental states; in which case, they would need to turn the ‘program’ model in an entirely different direction than trying to show sufficient causality.  But this is a project for another day. 

            Searle then turns from Putnam to Daniel Dennett, who even more explicitly opposes the reality of mental entities.  Dennett believes that mental phenomena such as pains are not ‘good theoretical entities,’ but Searle is unclear whether by this term he means that they are not scientifically studied or that they do not exist “as ordinarily understood.”  Dennett wants to hold the view that there are no such things as pains, though we do in fact feel pain.  The argument that there could be no such thing as pain is as follows: the intuitions we have of pain do not form a consistent set; therefore, there can be no true theory of pain for a computer to instantiate, and none for a human or animal to instantiate, which leads us to the ‘outrageous conclusion’ that no one feels pain.  As Searle points out, this argument is an exercise in flawed logic: “From the fact that the set of intuitions we have about a class of objects is inconsistent, and therefore the class is such that none of the objects could satisfy all of the intuitions, it simply does not follow that no such objects exist” (217). 

            Dennett’s positive conception of the mental fares little better.  He describes a flow chart for a system that has consciousness, and posits there is something it is like to be a being that instantiates the chart.  He holds that consciousness involves a sort of functional organization.  He then tells us that we are realizations of the flow chart, that this is what gives us something that it is like to be us, and he challenges us to give grounds for denying it.  For Searle, though, the response is simple: “Having conscious states such as bodily sensations and visual experiences is one thing, and instantiating the flow chart is something quite different” (218).  The flow chart itself is independent of the forms of its realization – all sorts of substances could instantiate it and still not have conscious states.  A modification of the argument from anesthesia also applies to Dennett: a person with his eyes closed could go through all the steps in the flow chart that are specific to vision, using other methods.  He will have conscious states, but they will not be the correct states for visual experience (219).

            Another argument Searle makes to Dennett’s conception of consciousness is that it confuses intrinsic mental states with the causal features that require observer relative ascriptions of mental states.  This shift becomes a matter of principle in Dennett’s notion of ‘intentional systems’ and ‘the intentional stance.’  Anything that instantiates certain sorts of programs, whether human or computer, will be an ‘intentional system,’ which can adopt the ‘intentional stance’ – predicting behavior by ascribing the possession of certain information to the system, and supposing the system to be directed by certain goals, then figuring out the most reasonable action to take on that basis.  After that, it is “a small step to describe this information and goals as beliefs and desires” (219).  By Dennett’s definition, though, intentional systems do not really have beliefs and desires – the choice to ascribe beliefs and desires to systems is merely a pragmatic one.  But this leaves open the question whether an ascription of mental states is one of information and goals, or beliefs and desires.  As Searle states,

Even if we have not defined intentional systems in such a way that we say that they really have beliefs and desires, there will still be a difference between those intentional systems that really have beliefs and desires and those that do not.  For example, in the case of those that do, the ascriptions have entirely different interpretations from the case of those that do not.  (219)

 

Dennett thinks that for both the computer and the human being, we adopt the intentional stance because we find it the most useful.  But, Searle argues, there is a big difference between my attributing to a chess-playing computer a desire to put me in check and my saying I have a desire to drink a cold beer.  In the computer case, there are no intrinsic mental phenomena – I am merely using a descriptive shorthand for practical purposes.  But in my own case, “I am stating facts about intrinsic mental phenomena” (220), regardless of whether anyone takes the ‘intentional stance’ towards me.  To put it another way, “whether a system really has beliefs and desires is quite independent of whether or not we find it useful to make observer relative ascriptions of beliefs and desires” (220).[11]

            What Dennett’s project (and functionalism in general) amounts to is replacing the question of analyzing the intrinsic features of mental states with the question of under what conditions we find it useful to speak as if an entity had mental states.  This technique is, in essence, a self-deceptive way of changing the subject.  Searle’s objections to Dennett, then, as we have seen, are that his argument against the existence of pains is invalid; that “the cognitive theory of consciousness is subject to counterexamples;” and that the ‘intentional state’ idea does not eliminate the distinction between observer relative attributions and attributions of intrinsic mental phenomena (221). 

            The problem with all of the definitions of functionalism is that “they do not tell us the intrinsic properties of mental states, only their functional role” (221).  The functionalist thesis, though, is that mental states have no intrinsic properties besides their functional role.  Returning to the intrinsic-observer relative distinction, the principle question was whether the acknowledged causal properties of mental states are sufficient to define their intrinsic mental properties.  Searle’s answer is a definite no.  The intrinsic features of beliefs, pains, desires, and other mental states, are intrinsic.  Functionalism, like behaviorism, tries to get us to take a third person observer relative stance in agreeing that mental states stand in causal relations.  But Searle’s arguments above are designed to show that “something could have all the right causal relations and still not have the right mental properties” (222).

            In Searle’s final analysis, functionalism, like behaviorism, requires diagnosis[12] more than refutation.  Why do people accept it?  Searle picks out three main reasons for doing so: the first is verificationism – the remnants of the positivist idea live on in functionalism.  Functionalists, as we have seen take a third person point of view, which leads to confusion between intrinsic and observer relative ascriptions; they overlook the difference between what it is to have mental states and what the functional role of the ascription of mental states is.  The cure to this particular ailment is to insist on a first-person point of view.  “It reveals a fundamental confusion to suppose that we can get clear about mental states entirely by examining the functioning of third-person attributions of the mental” (223). 

A second motivation behind adopting the functionalist notion has been the rise of ‘cognitive science,’ which gives us the illusion that the computer has given us remarkable insight into the operation of the mind, and that philosophical problems about the mental can now become technical questions in automata theory, which can be solved by the cooperation of philosophy and cognitive science.  According to Searle, “the fallacy is in moving from the true premise that the brain is a computer, in the sense that it instantiates computer programs, to the false conclusion that all the brain does which is relevant to the production of the mental is to instantiate computer programs” (224).  This conclusion is false because “no program by itself guarantees the causal powers that are specific to the biochemistry of the brain, and because a system such as a human agent could instantiate the program and still not have mental states” (224). 

The third factor that drives philosophers to embrace functionalism is the fear of Cartesianism.  Many people worry that admitting the existence of the mental at face-value would require postulating, one, a class of mysterious entities in another metaphysical realm; two, the substantive ‘self’ or the soul; and three, even worse, a ‘mind-body problem.’  Neither behaviorism nor functionalism was independently motivated, by one examining his own pains or other mental states.  They were rather proposed as solutions to other philosophical issues, such as the problem of other minds or the ‘mind-body’ problem.  Searle posits, though, that if thinking and perceiving are simply natural, biological phenomena, the problem goes away – we have no reason to suppose it if mental phenomena are both caused by and realized within the structure of the brain (224).  We can therefore drop such terms as dualism, ‘monism,’ and ‘physicalism,’ since as Searle says, “the antidualist jargon only has a point if we accept the dualist categories” (225), such as kinds of substance.  Searle’s conception of ‘biological naturalism’ removes the philosophical picture that motivated behaviorism and functionalism in the first place.

 

On the Indeterminacy Thesis

            W.V. Quine is one of the most influential philosophers of language in the twentieth century, and is particularly famous for his version of meaning-skepticism,[13] as laid out in his thesis of the indeterminacy of translation.  In Quine’s view, the objective reality of meaning “is simply a matter of being disposed to produce utterances in response to external stimuli” (226).  He denies that between the stimulus and the verbal response there are any mental entities; there is no consciousness, intentionality, thoughts, or internal ‘meanings’ connecting the stimulus and the response.  All we have for meaning are the patterns of stimulus and learned response – a computer that could produce the right sounds in response to the right stimuli would have ‘mastered’ a language.  Whether mental entities exist is unimportant, since, for Quine, inner mental states and processes are useless and irrelevant in an empirical theory of language (227).

            Quine’s thesis, as Searle sees it, is a type of linguistic behaviorism.[14]  But Quine himself has offered a reductio ad absurdum of the basic premises of extreme linguistic behaviorism, and shown that, “If behaviorism were true, then certain distinctions known independently to be valid would be lost” (227).  Quine’s argument runs as follows:  when a person says a word or phrase, say ‘rabbit,’ there is a distinction between his meaning by the expression rabbit, rabbit stage, or undetached rabbit part.  But if we apply behaviorism’s assumptions to an unknown alien language, we could not make such distinctions as plain facts of the matter about the language.  “So, if all there were to meanings were patterns of stimulus and response, then it would be impossible to discriminate meanings, which are in fact discriminable” (228).  The conclusion to Quine’s argument from radical translation is that, if we assume linguistic behaviorism, no behavioral evidence gives us anything by which to choose between two mutually inconsistent translation manuals of an alien language, both of which fit the evidence available. 

            For Quine, the unit of analysis in translation is whole sentences, and in translation they can only be empirically verified if they are associated directly with stimulus conditions, via ‘observation sentences.’  For instance, ‘gavagai’ (in the foreign language), ‘rabbit,’ and ‘rabbit stage’ all have the same determinate stimulus meaning – they have ‘stimulus synonymy.’  Indeterminacy of translation comes in when we state analytic hypotheses on the meanings of particular words and expressions (228).  Searle argues, though, that sentences that have the same stimulus meaning do not have the same meaning as we use the term; it is a matter of objective reality that ‘rabbit’ and ‘undetached rabbit part’ do not mean the same thing.  In other words, Quine is trying to hold two incompatible positions:

1.     The thesis of behaviorism:  The objective reality of meaning consists entirely of correlations between external stimuli and dispositions to verbal behavior.

2.     In a given case of speech behavior, there can be a plain fact of the matter about whether a natural speaker meant, for example, rabbit, as opposed to rabbit stage, or undetached rabbit part, by the utterance of an expression.  (229)   

 

If we accept (2), (1) must be false.  Quine’s argument has thereby refuted extreme linguistic behaviorism.

            Turning to a first-person scenario, if behaviorism were correct, it would have to be so for English as well as for any unknown foreign language.  But when I see a rabbit, and say “rabbit,” I mean rabbit, not rabbit stage or rabbit part.  So a theory that tells me that there is no difference between my meaning rabbit and my meaning undetached rabbit part must be mistaken.  As Searle says, “In all discussions in the philosophy of language and the philosophy of mind, it is absolutely essential at some point to remind oneself of the first-person case” (229).

            Quine, however, wants to hold on to both (1) linguistic behaviorism, and (2) the distinction between different translations.  He posits three theses:

A.    The Indeterminacy of Translation

B.    The Inscrutability of Reference

C.    The Relativity of Ontology

 

As regards A, Quine claims that “Where questions of translation and, therefore, of meaning are concerned, there is no such thing as getting it right or wrong . . . there is no fact of the matter to be right or wrong about” (230-31).  B then follows from A; if there is no fact of the matter about whether one means rabbit, then there is no fact of the matter about whether he is referring to a rabbit.  Thesis A says, first, that there is no meaning beyond the correlation of external stimulus and verbal response, and further that there are an indefinite number of equally valid but inconsistent ways of correlating stimulus and response in translation.  Quine is basically claiming that indeterminacy is underdetermination at one remove,[15] and thus that there is no fact of the matter about meaning (or semantics). 

            If this is the conclusion that Quine wishes to show, though, he faces a dead end, since to get his argument going he assumes from the start that there are no objectively real meanings in any psychological sense.  If we assume from the beginning that intentionalistic or mentalistic meanings are possible, on the other hand, Quine’s argument fails, because underdetermination at one remove is still underdetermination – it does not show that there is no fact of the matter.  Just because a body of evidence is underdetermined by an indefinite number of explanatory theories does not mean that there is no fact of the matter as to which theory is right.  The real issue in Quine’s argument is not indeterminacy, then, but extreme linguistic behaviorism.  The argument does not and cannot refute mentalistic or intentionalistic theories of meaning, since it must assume the nonexistence of intentionalistic meanings in order to begin the argument for indeterminacy (232).  Once we stop begging the question against mentalism, we see that the underdetermination of theory by evidence in psychological meanings applies in addition to the underdetermination of theories on physical particles or observable behavior.  As Searle puts it, “There is nothing special about meaning and nothing to show that where meaning is concerned there is no fact of the matter” (233). 

            That said, Quine seems to recognize the absurdity of the indeterminacy argument when applied to the first-person case.  After all,

If the argument is valid, then it must have the result that there isn’t any difference for me between meaning rabbit or rabbit stage, and that has the further result that there isn’t any difference for me between referring to a rabbit and referring to a rabbit stage, and there isn’t any difference for me between something’s being a rabbit and its being a rabbit stage.  (234)

 

This is an obvious consequence of assuming that there is no meaning beyond behaviorist meaning.  But we can only understand Quine’s indeterminacy (‘gavagai’) argument because we know the difference, for us, between meaning rabbit and meaning rabbit stage.  Thus, “If the indeterminacy thesis were really true, we would not even be able to understand its formulation” (234).  Quine’s notion of meaning as stimulus meaning generates a blatant absurdity.

            Quine’s response to the above objection is that it only makes sense to speak of an expression’s reference relative to a background language, and doubly relative in the case of translation.[16]  Searle replies that this defense does not solve the problem, but merely repeats it – it still leaves us with no empirical difference of meanings within our own ‘background’ language.  “The absurdity is that, if I assume my idiolect is a fixed set of dispositions to verbal behavior, then any translation of one word into itself or another of my idiolect is absolutely arbitrary and without empirical content” (235).  In other words, I cannot justify the claim that I mean rabbit when I say “rabbit.”  Quine cannot have it both ways, first holding that there is no fact of the matter about meaning, and then saying that we should take words in our mother tongue at their face-value, in excess of their empirical content.  Indeed, any problems we find in translating from one language to another we will find in one language alone (236).  Even with an account of relativised meaning, we have the same absurdities: two people who speak the same mother tongue can translate a word from another language in two different ways, relative to them, and still both be wrong.  Obviously, “there is more to meaning that just the relations that a word has to the language of which it is a part; otherwise the question of translation could never arise in the first place” (238).  A word means what it does relative to the language of which it is a part, but “the very relativity of the possession of meaning the nonrelativity of the meaning possessed” (238).

            Searle’s argument against Quine breaks down thus:  first, we grant a distinction between meaningful and meaningless words, like ‘rabbit’ and ‘flurg’ in English. We can see that there is a feature of ‘rabbit’ in English which ‘flurg’ lacks – its meaning.  Yet “from the fact that ‘rabbit’ has the particular feature it has relative to English, it does not follow that the feature, its meaning, can exist only relative to English” (239).  The question of translating ‘rabbit’ into another language is the question of whether that language has an expression with the same feature (meaning).  “Quine’s argument is a reductio ad absurdum because it shows that the totality of dispositions to speech behavior is unable to account for distinctions concerning the feature, meaning, which we know independently to exist” (239); these distinctions include that between the meanings of ‘rabbit’ and ‘rabbit stage’.  Searle concludes, then, that the feature of meaning in ‘rabbit’ exists not only relative to English, because the reductio is about the feature itself, which is not relative.  Thus, as pain behavior is not sufficient to account for pain, so, “on Quine’s argument, dispositions to verbal behavior are not sufficient to account for meanings, because one might exhibit behavior appropriate for a certain meaning, but that still might not be what one meant” (240).  Quine’s argument certainly proves something, but it cannot prove what its author set out to prove.

            After his analysis of Quine, Searle shifts his aim to the indeterminacy theory of Donald Davidson.  Davidson, though not a behaviorist, does agree with Quine on the indeterminacy of meaning and translation, and Searle claims that Davidson does hold a modified version of Quine’s empirical theory of language.  For Davidson, though, the first-person case has no special status, since semantic facts about meanings must be equally accessible to all the language participants involved.  As a basis for a theory of meaning, Davidson takes the intentionalistic notion of ‘holding a sentence true.’  For the indeterminacy thesis, “The basic idea is that there will be different ways of matching up objects with words, any number of which could equally well figure in a truth theory that explained why a speaker held a sentence true” (241).

            Davidson runs an argument of ‘radical translation,’ similar to Quine’s:

1.     The unit of empirical analysis in radical interpretation is the sentence (as opposed to subsentential elements).

2.     The only empirical evidence for radical interpretation is the fact that speakers ‘hold true’ certain sentences in certain situations.

3.     There are alternative ways of matching words with objects which are inconsistent, but any number of which could equally well explain why a speaker held a sentence true.  (241)

 

These three steps, however, do not entail indeterminacy about what a speaker actually means and refers to.  For this Davidson needs an added premise:

4.     All semantic facts must be publicly available to both speaker and hearer.  If the interpreter cannot make a distinction on the basis of public, empirical evidence, then there is no distinction to be made.  (242)

 

All facts about meaning must be public for inscrutability to follow from the underdetermination thesis.  In our own case, we have determinate senses with determinate references for given expressions; this is what we are looking for in others.  Why should another’s senses and references not be just as determinate as ours?  Again, underdetermination does not demonstrate inscrutability (243).

            Davidson, though, objects to the first-person case.  He holds that the semantic features of language are public, and therefore a speaker cannot intend his words to have a unique reference, because the words cannot convey that reference to others.  Indeed, there is no such thing as a unique reference, since the public features of language are subject to indeterminacy.  And Davidson adds that we must know this in some way, so we cannot even intend to refer to, say, a rabbit instead of a rabbit part.  In Quine’s view, as we saw, indeterminacy follows from underdetermination only if we deny mentalism from the start; likewise, for Davidson, “indeterminacy follows only if we assume from the start that different semantic facts must necessarily produce different ‘publicly observable’ consequences” (244).  But Searle argues that we know independently that this conclusion is false.

            We know that Davidson’s conclusion is false because, in our own case, we know that by ‘rabbit’ we mean rabbit, not rabbit stage.[17]  To understand someone else’s words, then, I must look for in him what I already have for myself.  As Searle writes, “I know what I mean.  Furthermore, if another person understands me, he will know what I mean, and this goes far beyond just knowing under what conditions I hold what sentences true” (244).  In other words, knowing another person’s ‘hold true’ conditions is not enough to fully understand him; indeed, knowing your own ‘hold true’ conditions is not enough to fully understand yourself.  Davidson claims that meaning can only consist in what an interpreter can figure out form the totality of relevant evidence.  What counts as ‘figuring out?’  Language is a public matter, and we generally know what someone means if we know what he says, and under what conditions he says it.  But our certainty does not come from his meaning being a sum of the publicly available evidence; we know what the person means, rather, because we know how to interpret the evidence.  Really,

[T]he fact that the interpretation of the speech of another is subject to the same sort of underdetermination as any other claim about other minds does not show either that there is any indeterminacy or that we cannot, in general, figure out exactly what other people mean from what they say.  (245-46)

 

The conclusion to Davidson’s argument, like that of Quine’s, should be construed as a reductio ad absurdum of its premises.  When we plug in the first-person case, Davidson’s claim that I cannot intend ‘rabbit’ to mean rabbit seems entirely implausible.  I know what I mean, there is a fact of the matter, and therefore I can reasonably infer this of others as well (246). 

            Both Davidson and Quine use the model of ‘radical translation,’ which forces us to adopt the third-person viewpoint, and then asks how we would know some other person’s meaning through his words.  This way of framing the question causes confusion between the epistemic and the semantic, between how we know and what it is that we know when we know.  Yet the linguistically relevant facts must be the same for me as for another, and vice-versa, since “what I have when I understand him is exactly what he has when he understands me” (247).  And I already understand myself. 

            The thought experiment of radical interpretation is useful in understanding communication, but we can frame the problem in common-sense, mentalistic terms; we don’t need Quine’s behavioristic or Davidson’s ‘empirical’ constraints added into the mix.  Quine, for one, holds that all our talk of things is “a conceptual apparatus that helps us to foresee and control the triggering of our sensory receptors in the light of previous triggering of our sensory receptors” (qtd. 247-48).  Searle sees Quine as declaring a preference for a certain level of description.  However, “the level of semantics that we need to analyze also involves a level of intentionality” (248), and this level is implicit in Quine’s use of ‘foresee’ and ‘control,’ which convey intentionalistic notions.  It is Quine’s dogmatic rejection of mentalism in analysis that leads to his demand for a third-person point of view.  But “Once you grant that a fundamental unit of analysis is intentionality, then it seems you are forced to accept the first-person point of view as in some sense epistemically different from the point of view of the third-person observer” (248).

            The thesis that language is a matter of stimulations of nerve endings is a metaphysical preference unwarranted by the facts.  “The crucial fact in question is that performing speech acts – and meaning things by utterances – goes on at a level of intrinsic first-person intentionality” (249).  We cannot simply disregard the first-person, as Davidson seems inclined to do, since to do so would be to negate oneself as a subjective, speech-act performing agent.  Davidson asserts that “what is empirical must be equally and publicly accessible to any competent observer” (249).  But why must it be so?  The fact is that Davidson’s constraint on empirical facts is far too stringent, since “what he means is: What can’t be conclusively settled on third-person objective tests cannot be an actual feature of language as far as semantics is concerned” (249).  Another meaning of empirical, however, is ‘actual or factual,’ and the argument against the first-person case only works if we equate actuality with third-person testability.  For, “once we grant that there is a distinction between the public evidence available about what a person means and the claim that he means such and such . . . there is no argument left for inscrutability” (249).

            Searle’s main point is that, “when we understand someone else or ourselves, what we require – among other things – is a knowledge of intentional contents” (250).  This knowledge is not that of matching public behavior with stimuli, nor is it that of matching utterances with conditions in the world.  We can only come to this knowledge through an account of our first-person case, which a true model of the understanding of language requires. 

            Is Searle’s argument against meaning-skepticism sound?  It certainly seems strong, in that it leaves the skeptic with little room to play with.  A challenger might respond to the objection from the first-person with a question like, “How do you know that by ‘rabbit’ you meant rabbit and not undetached rabbit part?”  If he does, I might have nothing to respond with, other than that rabbit is what I mean, and a rabbit is what I refer to when I point at a rabbit and say ‘rabbit.’  The challenge to the skeptic is to come up with a good reason for me to doubt my own meaning in such an instance.[18]  Unless the skeptic can find such a reason, we are justified in holding to a realistic view of mental phenomena.  Whether this argument against indeterminacy supports Searle’s own account of consciousness is another matter, but if his arguments against competing notions hold he does seem have a leg up on other physicalist and empiricist theories of mind.

 

Conclusion

            John Searle’s ‘biological naturalist’ idea of consciousness – as a higher level feature of the brain caused by lower level neuronal processes in the brain – is a thought-provoking version of mental realism that avoids the traps of reductionism and functionalism, without collapsing into substance dualism.  Whether the theory is completely coherent is, again, a matter for another day.  Searle’s defense of the mental against behaviorism and functionalism is worthy of merit, in that it makes a strong case for consciousness being something more than a mere computer program, and points out that the causal features of the brain must logically involve more than a set of structured inputs and outputs.  And lastly, Searle’s demonstration of how the indeterminacy thesis collapses into reductio ad absurdum serves as a strong case against skepticism and a firm support for the reality of intentional states.

 

 

 

 

 

 

 

 

Works Cited:

Searle, John R.  “The Problem of Consciousness.”  Consciousness and Language, p. 7-17.  Cambridge University Press, New York 2002

-        “Analytic Philosophy and Mental Phenomena.”  Consciousness and Language, p. 203-225.  Cambridge University Press, New York 2002

-        “Indeterminacy, Empiricism, and the First Person.”  Consciousness and Language, p. 226-250.  Cambridge University Press, New York 2002