H. P. Grice, "Searle on assertions and aberrations"

J. L. Austin was being recruited by the University of California at Berkeley in the late 1950s. He declined, saying, “I think I should be dead by then” and thereupon added, “Since you can’t get me, get Searle.” Searle became an assistant professor there in 1959, the same year he received his D.Phil. from Oxford. Aside from visiting appointments and leaves, he has spent his entire career at Berkeley, where he is Mills Professor of Philosophy. In addition to Austin, Searle had been a student of P. F. Strawson, Peter Geach (who directed his dissertation), and other distinguished Oxford philosophers during the heyday of ordinary-language philosophy. Adept at criticism, Searle is even more impressive as a constructive philosopher. Even in some of his early articles, which are ostensibly criticisms of others, his own positive theory is not far below the surface. Language In “Austin on Locutionary and Illocutionary Acts,” Searle shows how Austin’s original linguistic distinctions should be recast. The most important result is that paradigmatic cases of illocutionary acts should be understood as consisting of a force and a propositional content (see AUSTIN). Consider, for example, these sentences: I state that Jones will be at the party. I promise that Jones will be at the party. I question whether Jones will be at the party. It is obvious that all of these sentences have something in common. Each would be appropriately used to express the same content or proposition that Jones is at the party. (For simplicity’s sake, the temporal element will be ignored.) In these sentences, the propositional content is expressed with a “that” clause; but some sentences express their content with gerundive phrases or infinitives: I congratulate Jones for being at the party. I order Jones to be at the party. It is equally obvious that a standard use of each of the above sentences expresses their propositional content with a different “force,” the force of a statement, promise, quesBlackwell Companions to Philosophy: A Companion to Analytic Philosophy Edited by A. P. Martinich, David Sosa Copyright © Blackwell Publishers Ltd 2001 tion, congratulation, and order, respectively. So the structure of illocutionary acts can be represented as F(p). The p corresponds to the propositional content, and the F indicates the “force” attached to the proposition. In effect, Searle moved Austin’s rhetic acts from the category of locutionary act to a subpart of an illocutionary act. In “What is a Speech Act?” (1965) and Speech Acts (1969), Searle used promising to illustrate the appropriate form of analysis for illocutionary acts. With slight modification, Searle’s analysis was this: In uttering a sentence T, a speaker S promises an addressee H to do an action A if and only if 1 Normal input and output conditions apply. 2 S expresses the proposition that p in the utterance of T. 3 In expressing that p, S predicates a future act A of S. 4 H would prefer S’s doing A to S’s not doing A, and S believes H would prefer S’s doing A to S’s not doing A. 5 It is not obvious to both S and H that S will do A in the normal course of events. 6 S intends to do A. 7 S intends that the utterance of T will place him under an obligation to do A. 8 S intends [i -1] to produce in H the knowledge K that the utterance of T is to count as placing S under an obligation to do A. S intends to produce K by means of the recognition of i -1, and he intends i -1 to be recognized in virtue of (by means of ) S’s knowledge of the meaning of T. 9 The semantical rules of the dialect spoken by S and H are such that T correctly and sincerely uttered T if and only if conditions (1)–(8) obtain. Certain aspects of Searle’s analysis might be fine-tuned. A better analysandum is: “In uttering a sentence T, a speaker S explicitly and non-defectively promises an addressee H to do an action A.” Perhaps another preparatory condition is appropriate: “S is able to do A.” Notwithstanding these and other possible adjustments, the form of Searle’s analysis is powerful and easily adapted to analyze the full spectrum of illocutionary acts. Another merit of this form of analysis is that categories of conditions easily emerge from them. Most importantly, conditions (2) and (3) express requirements for the propositional content of the act. Conditions (4) and (5) express preparatory conditions. (6) expresses a sincerity condition. (7) expresses an “essential condition,” which conveys the aim or goal of the act. In Expression and Meaning, Searle developed a taxonomy of speech acts: assertives (for example, statements and asseverations); directives (for example, commands and suggestions); commissives (for example, promises and vows); expressives (for example, apologies and congratulations); and declarations and assertive declaratives (for example, declarations and verdicts, respectively). Unlike Austin’s taxonomy, which was founded on no principles and Zeno Vendler’s, which was founded on syntactic principles, Searle’s taxonomy is semantically based (and also closely related to the types of conditions already described). First of all, illocutionary acts are categorized on the basis of their point or purpose, as expressed by their essential conditions. Assertives aim at committing speakers to beliefs. Directives and commissives aim at committing someone to a course of action. Expressives aim at expressing a mental state, such as happiness JOHN R. SEARLE 435 or sadness. Declarations and assertive declarations aim at bringing about some fact about the world; when the Chairman of the Olympic Committee says, “I hereby open the Games,” the games are thereby opened. Another dimension of categorization is word/world fit. Thus, to assert “The door is open” is to aim at getting the words to fit or correspond to the way the world is. If the door is not open, the deficiency is with the words. In contrast, to command, “You will open the door,” is to aim at getting the world to fit the words. If the addressee does not open the door, the “deficiency” is with the world. Of course this worldly deficiency tends to make the addressee, not the world, culpable. Taking our lead from these two examples, we can say that assertives aim at having their words fit the way the world is, while directives and commissives aim at getting the world to fit the way the words say it is to be. Expressives do not have a direction of fit but presuppose some fact about the world. The use of “I congratulate you on winning” and “I apologize for stepping on your foot” presuppose a victory and an offense. A third dimension of categorization, related to sincerity conditions, is the psychological state expressed in the illocutionary act. Assertives and assertive declarations express the speakers’ beliefs. Directives express the speakers’ wants and desires. Commissives express the speakers’ intentions. Declarations do not express any psychological state; warranted by institutions, declarations do not need sincerity. A fourth dimension of categorization, related to propositional content conditions, concerns what illocutionary acts can be about. Assertives have virtually no restriction on propositional content. Directives and commissives must be about future actions. Expressives must be about present or past actions or conditions. Declarations can be virtually about anything, although there are limits. One cannot fry an egg by uttering, “I hereby fry this egg.” Other dimensions, such as the intensity of illocutionary point (suggesting versus insisting) and the way the act relates to the rest of the discourse (objecting versus replying), while informative, are not crucial to the basic taxonomy. As revealing as the taxonomy is, it will acquire even greater importance because of its connection with intentionality. So far, our discussion has focused on the nature of illocutionary force. This focus suggests the novelty of speech act theory, for philosophers had concentrated on propositions for two and a half centuries. Let’s now consider how Searle treats this hoary matter. Paradigmatic propositions consist of a reference and a predication. Traditionally, reference is considered the basic way in which words relate to the world. The spirit of that tradition is captured by Searle’s formulation of the axiom of existence namely, that everything referred to must exist. This invites the question of how the word or phrase used to refer gets hooked up to the world. The complete story requires a theory of intentionality (to be presented later) since reference depends on physical expressions such as words being backed up by inherently representational mental states. At this point, however, Searle is able to say that reference depends on “the axiom of identification,” namely, that a speaker must be able to identify the intended referent for his audience in a nontrivial way. The object is identified either by the descriptive content of the referring expression (“the first human to set foot on the moon”), by the referring expression plus the context (“this one here”), or by a combination of the two (“this red shoe here”) (Searle 1969: 80). A. P. MARTINICH 436 Traditionally, proper names have been the paradigmatic kind of referring expression; and the proper understanding of them has occupied philosophers for two millennia. To restrict our discussion to the last half-century, two theories have dominated: the descriptive theory and the causal theory. According to the causal theory, which Searle rejects, a proper name gets connected with its referent in virtue of a causal connection between that name, the speaker’s intention to have the addressee identify a particular object through her use of that name, and that particular object. In contrast, according to the descriptive theory, names refer to referents in virtue of their descriptive content. Taking off from Frege’s views, Searle maintains that proper names have both sense (Sinn) and reference (Bedeutung), and the reference is a function of the sense (see FREGE). Consider a name like “Aristotle,” which refers to Aristotle. That reference occurs in virtue of a certain descriptive content. Unlike general words like “red” or “human,” which seem to be obviously tied semantically to redness and humanity, respectively, proper names seem to have a looser but nonetheless indispensable sense. The meaning of “Aristotle” is “the logical sum [inclusive disjunction] of the properties commonly attributed to him” (Searle 1969: 173). Although Searle has never taken back any part of this view of the meaning of proper names, his focus seems to change in his later work. While insisting that some representational content must accompany the use of a proper name, he does not assert that the name’s meaning is that representational content. Further, like the causal theorist, Searle maintains that the representational content is causally related to the external world (Searle 1983: 238). I think that Searle ought to abandon the claim that proper names have Sinne. From the fact that every use of a proper name must be accompanied by some representational content, it does not follow that that content is the meaning of that name. It further does not follow that there is any stable content, shared by the people who use that name. Searle may be conflating Frege’s mode of presentation with the propositional content of an utterance (see Searle 1983: 249, 251). What is most important in Searle’s theory, it seems to me, is the role of intentionality in reference. He thinks that this commits him to descriptivism because he takes his main opponents, Kripke and Donnellan, to discount intentionality. I believe that Searle’s theory captures the intuitions of both descriptivism and the causal theory, without falling into the errors of either, and that he gives the wrong impression in claiming to be a descriptivist. Concerning the other part of a proposition, predication, Searle accepts part of the asymmetry thesis of Gottlob Frege and P. F. Strawson. While reference is the act of picking out or identifying an object for the purpose of classifying or categorizing it, predication is the act of assigning a property to the referent. A subject-predicate proposition is true if and only if the referent has the property ascribed to it. For Searle, the distinction between subject and predicate or reference and predication is primarily one of function, that is, of how something is operating and only secondarily one of ontology. As an ontological issue, there were two basic choices: nominalism (the denial that properties, or universals, exist) or realism (the assertion that universals exist as much as individual material objects do). Searle’s basic position is that there is no substantive issue here. Universals exist solely because predicates allow nominalization. Because of sentences like “Socrates is wise,” we can form sentences like “Wisdom is a virtue.” Consequently, although there are universals, they depend “merely on the meaning JOHN R. SEARLE 437 of words” (Searle 1969: 105). Predicating a universal means using “a predicate expression in the performance of a successful illocutionary act” (Searle 1969: 121; see also 124). According to the traditional taxonomy, Searle would count as a conceptualist, a nominalist of a liberal expression; words are predicated of the objects referred to (Searle 1969: 124) and universals exist. Consciousness Since a speech act is a kind of human action that requires a mental representation of the world, a complete theory of speech acts will be part of a theory of mind. For most philosophers the main problem in this area is “the mind–body problem.” What is the mind; what is the body; and how do they interact? For Searle, these questions are not problematic: The solution has been available to any educated person since serious work began on the brain nearly a century ago, and, in a sense, we all know it to be true. Here it is: Mental phenomena are caused by neurophysiological processes in the brain and are themselves features of the brain. (Searle 1992: 1; see also 1984: 14–15) Searle calls his view “biological naturalism,” and says, “Mental events and processes are as much part of our biological natural history as digestion, mitosis, meiosis, or enzyme secretion. . . . [Further,] intentional states stand in causal relations to the neurophysiological (as well as, of course, standing in causal relations to other Intentional states), and . . . that Intentional states are realized in the neurophysiology of the brain” (Searle 1983: 1, 15; see also p. 90 and 1984: 21–2). He is willing to admit that mental states causally supervene on brain states, but he is uneasy about the admission because supervenience was originally a logical relation and he worries that his causal form may be conflated with the logical one (Searle 1992: 124–6). Theories of supervenience are typically reductionistic. Is Searle’s? Yes, but in only one of the various senses of reduction. Theoretical reductionism, for example, tries to express or reduce all the laws of one theory T1 to those of another T2. Once this is accomplished, it is standard to claim that the entities referred to in T1 do not really exist and that only the entities in T2 do. So theoretical reductions, like most reductions, ultimately aim at some ontological reduction. But Searle’s does not. Rather, he explains that he espouses a form of causal reductionism because he holds that consciousness is causally reducible to the brain processes (Searle 1992: 116). This causal reduction, unlike most, does not have a corresponding ontological reduction because of the irreducible first-person ontology of consciousness. While a person’s pain is no doubt caused by a certain pattern of neuronal firings in the thalamus and other parts of the brain, a complete specification of this pattern would still leave out “essential features of the pain” (Searle 1992: 117), namely, how the pain feels. When heat was reduced to mean kinetic energy and colors were reduced to the refraction of photons, these two entities were redefined in order to eliminate the subjective element in the perception of them. But a similar kind of redefinition of consciousness is not possible because there is nothing to consciousness except the subjectivity (Searle 1992: 121–3). A. P. MARTINICH 438 Searle’s position is designed to avoid both materialism and dualism. Unlike a materialist, he affirms the existence of irreducible mental phenomena and denies that they are identical with brain states. Unlike a dualist, he asserts that mental properties are physical properties. Since many critics think that Searle is a property dualist malgré lui, something should be said about this. A property dualist holds (1) that there are only two kinds of properties, and (2) that all properties are mental or physical. Searle denies both (1) and (2). Either there is only one kind of properties, physical ones, or there are many kinds. And, if there are many kinds, none of them are mental or physical in the sense intended in (2). Searle thinks that dualism and materialism share the assumptions that give rise to the “mind–body problem.” But there is no such problem any more than that there is a “stomach-digestion problem” (Searle 1992: 15; see also 1984: 14). The serious issue in the philosophy of mind, in Searle’s opinion, is the nature and structure of consciousness. He dealt with its nature in The Rediscovery of the Mind (1992), and with its logical structure in Intentionality (1983). Searle came to appreciate how the structures of intentionality have to be understood as ways that consciousness exercises itself after he wrote Intentionality. So it is sensible for us to begin with his views about consciousness. Concerning the nature of consciousness, Searle has a negative and a positive project. The negative project is to show that current work on the nature of the mental, especially in cognitive science, is conceptually confused. The confusion does not just interfere with constructing a correct theory, but also motivates misconceived research strategies and covers over the fact that consciousness is the central phenomenon of the mind. His refutation centers on the “Chinese room” thought experiment. A person with no knowledge of Chinese is locked in a room. He is given a batch of Chinese writing (input); he has a rule book, written in his own language, that correlates the input with other Chinese writing and explains how to select or produce Chinese writing (output) that under certain conditions would be an appropriate follow-up to the input. To an observer outside the room and ignorant of the process in camera, it may appear that the person or mechanism inside the room knows Chinese. Of course, neither that person nor any conjunction of that person and anything else relevant to the outputting knows Chinese. He may not even know that he is dealing with Chinese or any language at all. He is performing totally formal operations, that is, the person treats the writing solely in virtue of physical shape (or other physical properties). The writing means nothing to the person. Since the person in the room is doing just what a computer does, Searle concludes that computers do not have minds nor cognitive states (Searle 1992: 45). Searle has variously described what the person in the room (and similarly any existing computer) lacks: a semantics, or an appropriate causal connection with the input and output, or understanding (see Searle 1992: 69). However, what seems to be most basic for Searle is the fact that neither the man in the room nor any existing computer has any understanding of Chinese. This is crucial because understanding requires consciousness and consciousness employs a unique kind of causation. This type, explained in Intentionality, may be described tentatively here as intentional causation (see Searle 1992: 107–9). The operation of intentional causation is important because some cognitive scientists have tried to circumvent the consequence of Searle’s scenario by constructing a “room” or robot that can perform complex JOHN R. SEARLE 439 operations. Suppose a robot is outfitted with television cameras, connected to levers and pulleys, powered by a motor that drives the robot on wheels to various locations to arrange and rearrange boxes or other objects. These scientists think that by increasing the complexity of the internal mechanisms and by having the robot be affected by and to affect its environment, they have undermined the Chinese room argument (Searle 1984: 34–5, 40–1). The scientists are mistaken in thinking that complexity or generic causality is the issue. Rather, it is the nature of the controlling mechanism (consciousness) and the type of causation (intentional) that are crucial for Searle. In Rediscovery of the Mind, he argues that the fact that cognitive scientists think that the brain is a digital computer and that a digital computer can be constructed out of an infinity of materials, including paper or magnetic tapes, cogs and levers, “a hydraulic system through which water flows . . . an elaborate system of cats and mice and cheese . . . [and] pigeons trained to peck as a Turing machine” – all of these are actual examples from theorists – proves that the theory is bankrupt and irrelevant to the nature of the brain and mind (Searle 1992: 206). As he says, “we wanted to know how the brain works,” and it is no answer to say that the brain works like any digital computer might work, where “digital computer” is defined so broadly that “stomach, liver, heart, solar system, and the state of Kansas” count as digital computers (1992: 208, and 1984: 36). Another way of getting at the same general point is to say that essential to the operation of the brain is “intrinsic intentionality.” While this concept will be explicated later, it can be understood provisionally as meaning that the brain causes states that are inherently representational. In contrast, no current computer and no computer to be designed in the forseeable future has intrinsic intentionality. For the sake of simplicity, let’s say that the operation of every computer can be understood as involving the production of sequences of “1”s and “0”s. It is also common to think of “1” and “0” as syntactic (purely formal) entities, and one and zero as semantic entities, that is, the meanings of the syntactic entities. Even when “1” and “0” are not understood to be semantic entities, it is common to think that the computer itself takes those marks to refer to one and zero. Searle objects to this view. Computers understand nothing because they do not represent anything to themselves. Rather, the computer designers and the computer users interpret the marks, “1”s and “0”s, in some way that is useful to them, some computational way. But the interpretation and the understanding are in the designers and users only and not at all in the computer. This has the consequence that the computer “1”s and “0”s are not even intrinsically syntactic entities, since syntax and semantics are correlative ideas (Searle 1992: 209–10). Syntax is not in physics. It is no good to urge that computers can or do have physical states other than “1”s and “0”s. Searle’s point still holds: notions such as computation, algorithm, and program do not name intrinsic physical features of [computer] systems. Computational states are not discovered within the physics, they are assigned to the physics. (Searle 1992: 210) Searle’s scenario of the Chinese room is powerful because the only plausible locus of the intentional causation required for understanding Chinese is in the consciousness A. P. MARTINICH 440 JOHN R. SEARLE 441 of the person in the room, and it is obvious that that person is not conscious of and does not know Chinese. One consequence of Searle’s criticism of standard cognitive science is that an entire level of theory disappears; the idea of “an unconscious mental process” and hence the idea of their principles loses all justification (Searle 1992: 239–40). There are then only two legitimate objects of cognitive studies, the brain and consciousness. The study of the brain belongs to neurophysiology; the study of consciousness belongs to philosophy and a still misguided cognitive science. This brings us to Searle’s positive project about the mind, namely, to give a non-reductionistic account of consciousness that is consonant with neuroscience. What is ontologically essential to consciousness according to Searle is subjectivity: “the mental is essentially a first-person ontology” (Searle 1992: 70; see also pp. 77, 94– 5). It is this fact that makes the “first-person” understanding of it fundamental. Third-person access to consciousness via observation of behavior is inherently incomplete insofar as it has no access to the experience of consciousness. It is also derivative because while consciousness is essential to causing behavior, consciousness is logically independent of behavior (1992: 69). There can be consciousness with no behavior. Searle identifies a number of features of human consciousness. We shall divide them into seven categories. (1) Consciousness manifests itself “in a strictly limited number of modalities” (Searle 1992: 128). In addition to the traditional five – seeing, touching, tasting, smelling, and hearing – there is the sense of balance, bodily sensations, which includes “proprioception,” that is, the feeling of how one’s body and parts of one’s body is oriented, and the stream of thought. (2) Consciousness is unified with respect to both temporal continuity of impressions and the spatial unity of various impressions. Yesterday, today, and tomorrow are all part of the same temporal system; here, there, and the other place are part of the same spatial system; and the two form a spatiotemporal system. (3) Consciousness is a necessary condition for intentionality and typically is intentional, that is, directed at objects. All intentionality is aspectual. This is easiest to see in visual perception; things are always perceived from a point of view and as being things of a certain kind (1992: 133). A related aspect of intentionality is the fact that consciousness has a focus and this in turn gives rise to the difference between figure and ground in Gestalt psychology. Also, attention is directed to some contents of consciousness more than to others. The driver of a car may be paying more attention to his vacation plans than to his driving; yet both are simultaneously conscious. (4) Consciousness has a “subjective feeling.” There is a difference between human consciousness and what it is like to be a bat or a porpoise. (5) Although it is not a special feeling, there is an air of familiarity about the objects that a person is conscious of. Even the unfamiliar is familiar in the sense intended here. A person walks into an office building and expects it to have elevators; the elevators are found in a fairly predictable location; they are easy to operate; and the door opens to a floor, which, though never seen before, has enough familiarity about it that the appropriate room is discovered. There is a sense in which people have knowl- A. P. MARTINICH 442 edge of the world in a way that is more general than any particular bit of knowledge about it. The world is not strange and mysterious. This fact is highlighted by surrealist artists with their melting watches and ever-ascending-and-descending staircases. There is another kind of familiarity with the world: people know generally where they are and what time it is, in relation to many other places and times. Searle calls this general spatial and temporal familiarity with the world “situatedness” (1992: 141) (cf. CHOMSKY). Related to (5) is (6). (6) This is what Searle calls “overflow,” the feature that has some specific perception or belief connect to other beliefs seemingly without end in some elaborate, not fully articulatable web: these east Texas trees are pines, like the pines of California, but not exactly; they flourish in wet areas not quite marsh, etc. Perhaps this feature is closely related to Searle’s concepts of the Network and the Background. (7) Clearly many states of consciousness are suffused with a mood (elation, depression, cheer) even though a mood “never constitutes the whole content of a conscious state” (1992: 140). Construed broadly enough, every state of consciousness has some mood or other. For most people, there is a permanent low level of pleasure connected with consciousness, and for some, there is a permanent low level of displeasure. (See also 1998: 73–80.) A theory of consciousness would not be complete without a theory of the unconscious. Searle’s main thesis is that “every unconscious intentional state is at least potentially conscious” (1992: 132; see also p. 152, and 1984: 43–4). In holding this, he is opposing the idea, standard among cognitive scientists, of a deep unconscious that can never be made conscious. They purport to subtract consciousness from conscious mental states and to find the “computational mind” as the remainder. This is an attempt to effect a separation of intentionality from consciousness, after which a mistaken account of intentionality as computation is presented. Searle’s proof of his thesis is roughly this: only mental states are intrinsically intentional in the sense described above. Unconscious mental states are intrinsically intentional. All intrinsically intentional states are aspectual. Unconscious mental states, as unconscious, exist only as neurophysiological events. Therefore, in order to be mental states at all, the unconscious ones must be capable of being brought into consciousness by the underlying neurophysiological events (1992: 155–9, 172). Searle’s theory of the unconscious permits a neat explanation for the existence of unconscious pains. For example, people with chronic back pain often awake from sleep with pain. It is plausible that the characteristic neurophysiological events that underlie the conscious pain are present during sleep, absent whatever additional events that bring the pain to consciousness. When these additional events get triggered, the pain becomes conscious and the sleeper awakes (1992: 164–7). Further, consciousness is transparent to itself. When David Hume looked into himself, in addition to not finding the “mind” he was looking for, he did not find consciousness either because it is always directed at other things and cannot be an object of direct study (Searle 1992: 97). That is why nineteenth-century “introspectionism” in psychology was doomed to be a failure and why it is natural to think that consciousness is not part of the physical world. The transparency of consciousness also explains why it makes no good sense to say that each person has “privileged access” to her own. That figure of speech requires one to think that a person enters a space separate from herself in which she alone can stand. But there is nothing in consciousness analogous to that space (1992: 97, 104–5, 170–1). Intentionality The cash value of Searle’s theory of the nature of consciousness depends upon his ability to explicate the structures of consciousness. The most salient structures have intentionality; that is, being directed at something in the sense of representing something (Searle 1983: 3, 11–12). While not all conscious states are intentional – undirected anxiety and nervousness are not directed at anything, nor, I think, is pure joy – the most important forms are. In order to emphasize that intending to do something is only one kind of intentionality, Searle capitalizes the latter term, but I shall not follow that convention. Because thoughts and feelings (prominent forms of intentionality) are expressed in language, Searle thinks that much of the structure of intentionality can be read off from the structure of speech acts. So, just as the basic form of an illocutionary act is F(p), the basic form of an intentional State is S(r), a psychological mode, such as believing, hoping, or loving, and a representational content. Like speech acts, the content is often a proposition, as in believing that George Washington was the first president. But it can also be an object as when Jones loves Smith. In this latter case, a distinction needs to be drawn between Jones’s representational content, and the real person Smith, which is the intentional object of that content. Four features shared by illocutionary acts and intentional phenomena are especially revealing. (1) Analogous to the force and propositional content of speech acts, intentional states have a psychological mode and a content. “I state that you broke the vase” and “I order you to wash this floor” are structurally identical with “I believe that you broke the vase” and “I want you to wash this floor,” respectively. (2) Direction of fit: like statements, beliefs have a word-to-world direction of fit, and like orders, desires and intentions have a world-to-word direction of fit. Like apologies, sorrow has no direction of fit but presupposes something about the world, an injury caused by the agent. (3) For illocutionary acts that have a sincerity condition, the expressed propositional content purports to be the content of an intentional state of the speaker. For example, stating that snow is white is representing that one believes that snow is white. And promising to go to the party is representing that one intends to go to the party. Further, the relevant intentional state is the sincerity condition of the illocutionary act. Both the sincerity condition and the intentional state of asserting that snow is white is the belief that snow is white. (4) Both illocutionary acts and intentional states have conditions of satisfaction. A statement is satisfied if and only if it is true; an order is satisfied if and only if it is obeyed; a promise is satisfied if and only if it is kept. Correspondingly, a belief is satisfied if and only if it corresponds to the way things are; a desire is satisfied if and JOHN R. SEARLE 443 only if it is fulfilled (idiomatically, desires are said to be “satisfied”), and an intention is satisfied if and only if it is carried out. So far, nothing has been said about a traditional part of the concept of intentionality, namely, that what is intentional is always directed at an object. While Searle holds that some states have intentional objects, he denies that all of them do, because he does not hold the standard view about what an intentional object is. For him, it is always an existent object. The intentional object of believing that Jones is happy is Jones, the very person herself. The intentional object seems analogous to the referent of a reference. Both referents and intentional objects are idiomatically spoken of as being “in” mind, although neither, of course, could literally be spatially in the mind. And the referent is no more a part of the proposition expressed than the intentional object is part of the representational content of a mental state or event. Given the analogy between intentional objects and referents, one might think that it forms the basis for a fifth shared feature. But Searle does not go this route; and in general he mentions intentional objects only to put them aside. In addition to these four parallels between illocutionary acts and intentional states, there is an important disanalogy. Intentional states are intrinsically intentional, while illocutionary acts are not. Illocutionary acts have to be realized in utterances (sounds, marks, or gestures) that are intrinsically physical and only derive their intentionality from the fact that humans intend them to represent states that are intrinsically intentional (Searle 1983: 27; 1992: 78–80). In other words, when a speaker utters a sentence, he (i) intends that sentence to have conditions of satisfaction that are given by (ii) the intentional state expressed. The parenthetical roman numerals indicate what Searle calls a “double level of Intentionality in the speech act,” namely, (i) the intention of giving the utterance conditions of satisfaction that are (ii) the intentional states. The condition of satisfaction of “snow is white” is that snow is white. To ask for the meaning of a sentence is to ask for the intentional state that is the condition of its satisfaction (1983: 28, 164). What determines that some intentional state or event is one with some particular content rather than some other? In Intentionality, Searle’s answer was that an essential element is its place within a system of intentional states. For example, the intentional state of running for the presidency of the United States is possible only as part of a Network of other intentional states involving such beliefs as that there is a United States, constituted in the eighteenth century, separate and independent of every other country, bounded in large part by Canada on the north and Mexico in the south, and so on. So Searle endorses a kind of holism. Further, the Network of intentional states is possible only against a background of nonrepresentational “mental capacities” (Searle 1983: 21). The background is generalized know-how. It is a worldly competence that makes possible particular kinds of know-how, such as knowing how to open a door or knowing how to write a letter. But, because particular know-hows can be explicated as essentially involving propositional contents, such as that something is the case or that something is to be done, these instances of know-how are not part of the background itself. Opening a door involves representations, but “the ability to recognize the door and the ability to open it are not themselves further representations” (1983: 143). It is difficult to be explicit and precise A. P. MARTINICH 444 about the background just because background capacities are not propositional and explicit while explanations are. Searle may be discussing the same phenomenon as Martin Heidegger when he talked about human beings as beings-in-the-world with things being ready-at-hand and as Ludwig Wittgenstein when he talked about certain things being so basic that they “stand fast” and hence are objects neither of certainty nor doubt. In this spirit, Searle claims that metaphysical realism is not a “hypothesis, belief, or philosophical thesis . . . but the precondition” of any of these things (1983: 158–9). Consciousness is not a hypothesis either (1992: 79; see also 1995: 178, 195). Searle’s views about the Network and the Background changed in The Rediscovery of the Mind although the core remained. “Intentional states do not function autonomously,” he says. Their conditions of satisfaction depend upon a set of Background capacities. Some of these capacities are capable of generating other conscious states. Of course these newly generated conscious states are just like the first ones mentioned; they do not function autonomously and they depend upon a set of Background capacities. The Background role is so important that the “same type of intentional content can determine different conditions of satisfaction” when the Background capacities are significantly different (1992: 190). Consciousness then consists of an occurrent representational content (item (a)), which depends upon a neurophysiological base with “the capacity to generate a lot of other conscious thoughts.” This capacity (item (b)) itself is part of a neurophysiological system that is nonrepresentational but necessary for both the representational content and the individual consciousness. Item (a) replaces Searle’s earlier idea of a Network; item (b) replaces the earlier idea of the Background (1992: 190–1; see also 1995: 181–2, 184–9). Having laid out the general structure of intentionality, we can look at Searle’s treatment of the “primary forms of Intentionality, perception and action.” Concerning perception, he is a realist; people see cars, tables, trees, and so on. They do not see their perceptual experiences; they have them (1983: 36, 38). Experiences, like beliefs and desires, are intentional, being directed at objects and having conditions of satisfaction. The experience of perceiving a yellow car is directed at the car and is satisfied only if there is a yellow car where it is perceived to be. While the car is yellow and car-shaped, the experience is neither. Further, the car itself has to cause the visual experience. How should this information be represented? Searle proposes the following: “I have a visual experience (that there is a yellow station wagon there and that there is a yellow station wagon that is causing this visual experience)” (1983: 48). The crucial point about this analysis of the conditions of satisfaction of visual perception and related phenomena is the element of self-referentiality: the visual experience that is to be satisfied is mentioned in the conditions of satisfaction. This does not mean that the self-referential causal relation is itself seen (Searle 1983: 48–9). Searle has some sympathy with the suggestion that a clearer way to represent the conditions of satisfaction is this: VEc (that there is a yellow car), where the subscript “c” indicates the causal self-referentiality. The visual experience is actual seeing only if the appropriate object in the nonmental world, in this case, the right yellow station wagon, is causing the experience. This allows Searle to say that on his account JOHN R. SEARLE 445 perception is an Intentional and causal transaction between mind and the world. The direction of fit is mind-to-world, the direction of causation is world-to-mind; and they are not independent, for fit is achieved only if the fit is caused by the other term of the relation of fitting, namely, the state of affairs perceived. (Searle 1983: 49; see also pp. 61–2) Does Searle’s account solve a familiar philosophical puzzle? Suppose that two identical twins have type-identical visual experiences while looking at two different but type-identical station wagons. What makes one twin’s perception a perception of car A and the other twin’s perception a perception of car B? Searle says that there must be something in the representational content itself that specifies the proper car as a condition of satisfaction. The specification is achieved because “each experience is selfreferential” (Searle 1983: 50). The visual experience of one twin includes as a condition of satisfaction the fact that that very experience is being caused by a yellow station wagon. In other words, perceptions as a mode of consciousness are essentially firstperson phenomena. In contrast, a causal theory of perception according to Searle takes a third-person perspective: car A causes one perception and car B causes the other. The theory is inadequate, however, as Searle observes, because it has no account of how the perceiver’s intentionality matters to the perception (1983: 64). While Searle is right in his criticism, he does not seem to concede as much to the causal theorist as he should. The fact that any perceiver has her own individual visual experience with a self-referential content depends on the fact that some individual nonintentional object is causing that unique experience. That is, the object of perception helps individuate the visual experience. Searle’s opposition to the causal theory inclines him to assert that human beings are “brains in vats,” the vats consisting of their skulls (Searle 1983: 230). Although he may have primarily intended the then fashionable phrase “brains in vats” metaphorically, it is hard to see how this position is consistent with central features of his theory. A brain in a vat does not receive its intentional contents in the right kind of way; its prior intentions do not cause actions because the brain is not hooked up to the appropriate biological organs, and so a brain in a vat does not perform any non-mental physical action. A different, and, I think, ineffective objection to Searle’s theory of perception is that it leads to a familiar form of skepticism. (1) Since the car causes the visual experience, the visual experience is the basis for believing that the car is seen; and (2) since the perceiver infers from the perceptual experience that the car exists, it might be the case that the experience exists without the car existing. Searle rejects both parts of this line of reasoning. The perceptual experience is not evidence for the belief that the car is seen; and the perceiver does not infer that the car exists. One simply sees the car: “The knowledge that the car caused my visual experience derives from the knowledge that I see the car, and not conversely. . . . [W]e perceive only one thing and in so doing have a perceptual experience” (1983: 73, 74). Let’s now consider Searle’s theory of action. Intentional actions are conditions of satisfaction of intentions to act; but not every intentional action is preceded by a prior intention to act. People often act without planning (1983: 84–5, 107). That is, “prior intentions” should not be confused with “intentions in action.” Prior intentions are like plans; they temporally precede an action. They may be expressed by sentence-forms like, “I intend to A” or “I will A.” In contrast, an intention in action is part of the warp and A. P. MARTINICH 446 woof of an action. All actions are intentional. So-called unintentional actions, such as Oedipus marrying his mother, is related to something intentional, namely, Oedipus marrying Jocasta. Breathing, snoring, and sneezing are bodily movements but not actions, because they are not intentional. Like perception (and memory), both “prior intentions and intentions in action are causally self-referential” (Searle 1983: 85). Intrinsic to their conditions of satisfaction is a causal relationship between the intentional state and the thing done. To raise one’s arm is to have one’s intention of raising it cause it to go up (p. 86). Searle says that “the prior intention causes the intention in action” (p. 94). This is true whenever there is a prior intention to do something immediately; such intentions cause at least attempts or efforts. To try to do something is to do something. However, prior intentions to do some act A are not sufficient to cause the person to do A. There is a gap between the intention and even the decision to act and the acting itself. Perception and action are nicely contrasted by two facts: (1) While perception has a mind-to-world direction of fit, action has a world-to-mind direction of fit. (2) While an object causes a perceptual experience, an experience of acting causes an event. But perception and action are the same insofar as both present, rather than represent, their experiences, in contrast with, say, memory or imagination, which do re-present things. Searle’s researches into the logical structure of intentionality caused him to rework the analysis of meaning that he gave in Speech Acts, where his main goal had been to give an analysis that would not include intending to cause an effect, as H. P. Grice had in his. In Intentionality, Searle claims that the essence of meaning is representing; in effect, it is imposing conditions of satisfaction on something that is not inherently representational. For example, in raising his arm, a person means that the army is retreating if and only if his intention to raise his arm causes his arm to go up and his arm raising has as a condition of satisfaction that the army is retreating. Obviously, the fact that the arm raising has a condition of satisfaction is due to the mind imposing that condition on it. All meaningful gestures or utterances have an intentionality that is derived from mental states that are inherently intentional. Although representing is the heart of meaning, according to Searle, since meaning is standardly used to communicate, a complete account has to say how communication occurs. What needs to be added to meaning as representation is an intention to get the audience to recognize what the condition of satisfaction of the person’s gesture or utterance is and to have the audience recognize it in virtue of the gesture or utterance itself (Searle 1983: 168). One merit of this analysis is that it gives a precise and informative answer to the question “What is the difference between saying something and meaning it versus saying something and not meaning it?” The answer is that the former has conditions of satisfaction and the latter does not. When “Es regnet” is said and meant, a condition of satisfaction is that it is raining, whereas when it is said merely in the course of practicing German, the weather is irrelevant. Social reality At the end of The Rediscovery of the Mind, Searle offers some guidelines for the proper study of mind; the last of these, the last sentence of the book, is: “we need to rediscover JOHN R. SEARLE 447 the social character of the mind” (1992: 248). This was an advertisement for his next book, The Construction of Social Reality (1995). The title is a direct challenge to a diametrically opposed idea, the social construction of reality. This popular alternative, which maintains that the world is a construct of the human mind, has no appeal for Searle, who says it is a “preposterous” view that rests on “an array of weak or even nonexistent arguments” (1995: 160). He is a realist about the real world. The natural world consists of two basic kinds of things: non-mental and mental. The non-mental things at a relatively fundamental level are atoms and at the most fundamental level are space-time points, according to current science. The mental things evolved out of the non-mental after billions of years. The non-mental things are ontologically objective; they exist independently of minds. The mental things are ontologically subjective; they depend for their existence on minds. For example, a pain is a part of the natural world, but subjective. The social world, which includes money, government, and marriage, arises out of mental reality, largely because the mind can represent things as other things. To take the most powerful instrument of representation, language uses sounds, marks, or gestures to represent other things; and this is possible because people are willing and able to take them as representing other things. In short, the social world is observer- and user-dependent. Money is money because people take pieces of paper or metal to have exchange value. Citizens are citizens because people treat them as having certain rights and responsibilities. In contrast, the natural world is observer-independent. A rose is a rose is a rose, whether anyone views it or not. Given this ontology as background, Searle uses three elements to explain social reality. First is the idea that people can impose functions on objects that do not have that function beforehand. Never “intrinsic” to the thing itself, functions are always observer-dependent and introduce a normative dimension (Searle 1995: 19). Originally, some object became a hammer when it was used to hammer and not before; and, in becoming a hammer, it became possible to judge good and bad hammers, depending upon how well they functioned. The second element is collective intentionality. A lineman blocks an opposing linebacker only as part of his team’s play, and a musician plays first violin only as part of the orchestra’s symphony. Football players play as much in concert as musicians do (Searle 1995: 22). Collective intentionality is not reducible to individual intentionality. In doing something together, each participant has her own individual intentions, but these derive from the collective intentionality of the group (pp. 24–5). The third element is the distinction, introduced by Searle in Speech Acts, between regulative and constitutive rules. Regulative rules direct or control pre-existing behavior, as the rules of etiquette control how people should eat. They typically take the form of imperatives: Use a napkin, not your sleeve, to wipe your mouth; eat your peas with a fork, not your knife. Constitutive rules create new forms of action. The rules of football create the game of football. More importantly, the basic rules of a government, formulated in constitutions, create governments. When one government falls and another arises, a new constitution is formulated. These constitutive rules often take the form of indicatives: “The Supreme Court is the highest court of the judicial branch of the United States.” Such sentences may seem to be statements of fact but they are more properly seen as declarations (Searle 1995: 55, 74). Searle believes that the deep form of conA. P. MARTINICH 448 stitutive rules is “X counts as Y (in context C),” for example, “A person born on American soil counts as an American citizen” (p. 28). In keeping with the theory of Speech Acts, one might suggest that a deeper form is, “We declare that X is Y (in context C)” (cf. 1995: 104–11). The suggestion is tempered by the fact that so many institutional acts evolve slowly, haltingly, and unreflectively. The “X counts as Y” formula can be iterated. Something that once occurred as a Y term, for example, “citizen,” can occupy the place of the X term in another formula. “A citizen counts as the president when duly elected, etc.” Roughly, the more complex the society, the more numerous and iterated the “counts-as” formulae (Searle 1995: 80). When constitutive rules are enacted, institutional facts are created. These facts are self-referential: something is a five dollar bill because it is accepted as a five dollar bill. At a deeper level, a level that exploits the three elements of function, collective intentionality, and constitutive rules, the logical structure of institutional facts is this: We collectively accept (S is enabled/required (S does A)). Applying this structure to a five dollar bill, call it X, we get, “We accept (S, the bearer of X, is enabled (S buys with X up to the value of five dollars) )” (Searle 1995: 97–8, 104–12). Contrary to appearances, this analysis is not viciously circular, because “five dollar bill,” and, more generally, “money,” occupy only a couple of nodes in “a whole network of practices, the practices of owning, buying, selling, earning, paying for services, paying off debts, etc.” (p. 52). The range of institutional facts is enormous, from “wives to warfare, and from cocktail parties to Congress” (p. 96). They, and the powers that go with them, come into existence when people accept them as facts and continue to exist as long as people accept them as facts. Conclusion The claim that Searle counts as a philosopher of the first rank turns on this point: he uses a small number of interlocking elements to explain a broad spectrum of reality in an illuminating way. The most important elements are the ideas of representation, direction of fit, self-referential intentional causation, and the distinction between constitutive and regulative rules. The spectrum includes the nature of language, mind, and the social world, all presented within a naturalistic but not materialist world-view. Bibliography Works by Searle 1969: Speech Acts, Cambridge: Cambridge University Press. 1979: Expression and Meaning, Cambridge: Cambridge University Press. 1983: Intentionality: An Essay in the Philosophy of Mind, Cambridge: Cambridge University Press. 1984: Minds, Brains and Science, Cambridge, MA: Harvard University Press. 1985 (with Vanderveken, Daniel): Foundations of Illocutionary Logic, Cambridge: Cambridge University Press. JOHN R. SEARLE 449 1992: The Rediscovery of the Mind, Cambridge: MIT Press. 1995: The Construction of Social Reality, New York: Free Press. 1998: Mind, Language and Society, New York: Basic Books. Work by other authors LePore, Ernest and Gulick, Robert van (1991) John Searle and His Critics, Oxford: Blackwell Publishers. A. P. MARTINICH 450 451 37