Terra Firma

In the beginning then, there were not two substances, but a single organism. A natural classification of the organism's functions distinguishes the bodily from the mental. This distinction begins at the adjectival level, thinking being an activity of the organism that belongs in the latter category, and breathing being an activity of the same organism, an activity that belongs in the former category. Out of such adjectival forms naturally evolve nominalizations by a kind of linguistic concretion. One person is said to have a stronger mind than another; another a stronger body. Such linguistic forms do not commit us to new items in the inventory of the universe, human bodies and minds as separate items.

A naturalistic approach to mind is of course not secured by the rejection of substance dualism. One might still proffer a non-naturalistic account of mental events or activity, an account that sees mental events as sui generis events that occur in the physical organism. My aim has been to illustrate a tendency and to provide a point of departure.

B. Skepticism

We have been taught that the external world, other minds, past moments of time, all of these are problems, that their assumption requires philosophic justification, a convincing response to the skeptic. My naturalist, failing to have learned this lesson, begins in the external world. Indeed, to think of the world as external, at least in the way that we do, is already to concede too much to the power of modern philosophy. We begin in the world, constituted (this is our best shot at the moment) in the way that our colleagues in the sciences suppose. This world contains, importantly for us, other people whose "minds," for example, their intellectual and affective capacities, are no less evident than our own. That these things are so is not something that we come to philosophy to verify, that we think of as up for grabs, awaiting the results of our work.

To deny that such things are up for grabs is not to suggest that nothing is. People surely have looked and do look to philosophy for help with their beliefs. But there are beliefs and there are beliefs. What is required here is a sense of balance that, to slightly revise Russell, is required even in the most abstract studies. That my name is "Howard Wettstein," that I am married to Barbara, that honesty and integrity are important to human life, that language is one of the things that makes humans distinctive, none of these are, or ever have been, on the negotiating table for me, not to speak of my belief that there have been past moments or a physical world.

The naturalistic outlook, as I'm understanding it, thus makes it difficult for skepticism to get a foothold.⁽²⁾ It is important--although perhaps it goes without saying--that this is no part of the motivation for the naturalism, as if the outlook were another move in the philosopher's arsenal for defeating skepticism.

The naturalist's starting point, out in the world, also provides important perspective, a distinctive vantage point, on skepticism. Whatever else skepticism is, it is a perennial intellectual preoccupation, and so philosophy might have something to say about its power. Providing perspective on skepticism is, of course, not an undertaking for a few paragraphs. Again, I attempt only a point of departure.

Distinctively human is the ability to represent to ourselves the ways things are, and not only the ways things are but also the ways they might have been, and still other ways that are thinkable, whether or not they are possible. Philosophers, occupied with such things, are somewhat in the position of Leibniz's God, peering at the worlds. It's perhaps not altogether surprising that they are sooner or later drawn to wonder about which of the worlds is truly their own. It's difficult to tell. To put the point close to the terms of the "First Meditation" Cartesian skeptic, the appearances are compatible with our living in very different worlds than we ordinarily suppose. While real worry about such things, worry in one's gut, is thankfully rare, concern about the justification and epistemic status of our ordinary beliefs is almost inevitable.

Dwelling among the worlds thus occasions a kind of vertigo, a confusion about our own abilities. If we can discern the worlds, perhaps we should be able to discern our real place. We ought to be able to look down from the heights and see ourselves below, as on a kind of magical flight over one's home town. Or maybe such achievement is beyond our reach. Perhaps we can't intellectually defend our ordinary modes of thought.⁽³⁾ But then it will seem that the skeptic has won, that we do not really know what we thought we knew.

Alvin Plantinga once said, paraphrasing Hegel, that philosophy is "thinking about things." One would like to think that by focused and disciplined reflection one might arrive at quite a different place than one began, knowing better, as Socrates says, what one means and what one loves. This is not to suggest, however, that by thinking hard one might be able, for example, to demonstrate the reality of the past without begging important questions.

C. Philosophy of Religion

When an undergraduate gets bit by the bug of philosophy, it is often for its distinctive concern with fundamentals: fundamental concepts, fundamental human values, and institutions. Yet large and fundamental areas have received woefully inadequate attention from my own analytic tradition during the past 50 years or so.⁽⁴⁾ My favorite example is the philosophy of religion.⁽⁵⁾ This is a fertile area for naturalistic exploration.

Here again, the dominant kind of naturalism, in this case eliminativist, gives away too much to the opponent. The naturalist typically concedes that the supernatural is at the heart of religion. But theism--belief in a supernatural deity--however large its historical and personal role, is a theoretical position if there ever was one. Perhaps the theist should be likened to a mathematician who has strong views, say Platonistic views, about mathematical entities. Such views, even if demonstrably correct, would be in an important sense external to the mathematical work, per se. As Larry Wright commented to me, it is usually assumed that in the domain of religion the interpretation drives the institution, but perhaps the institution is primary. Perhaps religious practice and the religious life are not essentially tied to the prevailing theistic interpretation.

Here, then, is an alternative picture.⁽⁶⁾ Imagine a community for which a certain narrative--say the Bible along with a tradition of interpretation--plays a central role. The narrative provides something of a history of the community, and perhaps a history of the world. It is in terms of this narrative that the community's rituals and ethical norms are understood. The narrative binds the community; it is treasured, and passed along from parents to children.

To this point, there is agreement amongst community members. Beyond this point, however, when it comes to matters of rather high level theory- specifically, the theoretical status of the narrative-there is much discussion and dissent. Some--the fundamentalists--take the theoretical status of the narrative to be like that of a weather report: it is precisely and literally true. Others--the orthodox practitioners--are with the fundamentalists on some fundamental matters, for example, on the existence of a supernatural creator. At the same time, they argue that there may be huge admixtures of mythology in the narrative. The story of creation, for example, is not one they take to be a serious historical record. Others--the naturalists--think of the narrative as allegorical and mythological at many key points. It's discussion of various historical events may or may not be correct; this is an open question. But it's talk of God, a being with supernatural powers, powers of creation, resurrection, miracles, and so on, is clearly mythological. At the same time, even the naturalists maintain that these myths are "our myths," that they play an enormous and enormously rich role in the life of the community, and constitute a crucial ingredient in the tradition's distinctive understanding of human flourishing.

It is not, as I see it, a philosopher's naturalism that inclines him to see no special value in, and no special place for, myth. Of course, what place that is needs exploration. Indeed that we have not seen the need for such exploration is part of my complaint. Consider a related topic: the role and value of ritual. Mencius, the Confucian virtue ethicist, maintained that a central category of virtue involves engagement with ritual. Some fundamental human things, certain forms of respect, for example, have no expression outside of ritual, according to Mencius. This is an exciting idea, suggestive with respect to our western religious and even secular rituals.⁽⁷⁾ That ritual, along with myth, are of no special interest to philosophy--nowadays a virtual dogma, and not only of empiricism--is not a dictate of naturalism.

The persistence of religion, not to speak of its power in human life, is often noted, and often given a deflationary explanation. Here as elsewhere with deflationary explanations, one may wonder whether there isn't something more positive and more important at work. Perhaps the power of religious institutions is a function of their touching something deep in what we are. Philosophy may have something to say on the matter.⁽⁸⁾

D. Philosophy of Language

Philosophical treatments of linguistic meaning and of--believe it or not--the soul sometimes bear a striking resemblance. Wittgenstein, in the Philosophical Investigations, speaks of meaning as "the soul of words,"⁽⁹⁾ and in The Blue Book addresses what we might call the vitality of language. He criticizes Frege's idea that

. . . the propositions of mathematics, if they were just complexes of dashes, would be dead and utterly uninteresting, whereas they obviously have a kind of life. And the same, of course, could be said of any proposition: Without a sense, or without the thought, a proposition would be an utterly dead and trivial thing. And further it seems clear that no adding of inorganic signs can make the proposition live. And the conclusion which one draws from this is that what must be added to the dead signs in order to make a live proposition is something immaterial, with properties different from all mere signs.⁽¹⁰⁾

Meanings, on the Fregean⁽¹¹⁾ approach with which Wittgenstein is unhappy, quicken the otherwise inert symbols, as souls are thought to vitalize bodies. Wittgenstein's express aim in The Blue Book is to bring the notion of meaning down to earth. Significance is not a function of the association of words with intrinsically alive entities, the meanings. The vitality of language is rather a function of what we do with the symbols, of our social practices.

But if we had to name anything which is the life of the sign, we should have to say that it was its use.⁽¹²⁾

This is not the place to explore Wittgenstein's tantalizing, if highly perplexing, ideas on the subject. It is enough simply to note that Wittgenstein's direction here seems of a piece with a naturalistic approach to "soul," one that sees our spiritual dimension in terms of natural capacities of the organism.

In other writings,⁽¹³⁾ I have explored what I see as a natural affinity between Wittgenstein's approach to meaning and the work of direct reference philosophers of language. A speaker's competence with a proper name is, for the latter approach, not a matter of the intellectual grasp of a meaning, as it was for Frege. Competence is rather a matter of the possession of a social instrument for making something a subject of discourse.

The really interesting and controversial heart of direct reference, as I want to develop it, is the denial of what I've called the cognitive fix requirement. This is the traditional idea that reference requires that the agent possess a strong cognitive relation to the referent. For Frege this meant that the agent needs to grasp a sense that has as reference the individual in question. For Russell, it was required that the agent stand to the referent in the extremely strong epistemic relation of direct acquaintance. Direct reference represents the denial of any such epistemic or cognitive requirement. One can speak about Aristotle because of the social power of his name, that is, because there is a socially available instrument for making him the subject of discourse. Direct reference thus shares something of Wittgenstein's naturalism about language.

II. WITTGENSTEIN'S NATURALISM ABOUT LANGUAGE

A. Language As A Refinement: In The Beginning Was The Deed

Wittgenstein's naturalism about language is highlighted in this passage from Culture and Value:

The origin and the primitive form of the language-game is a reaction: only from this can more complicated forms develop. Language, I want to say, is a refinement and in the beginning was the deed. (P.31)

That language begins as a refinement on the deed is perhaps uncontroversial. Who would deny that there was behavior prior to language, and that words evolve as a kind of refinement? Well, perhaps someone would find something objectionable here. But as a remark about origins, about pre-history, it does not really capture anything very distinctive, not to speak of distinctively naturalistic.

A suggestion of something more interesting emerges when we place the quotation in the context of Wittgenstein's work. The suggestion is that language, as a refinement, never quite forgets its roots, that its ancestry remains present in later forms. I want to explore two of Wittgenstein's examples, one immediately on the topic of conceptual precision, and one later in connection with Wittgenstein's ideas about pain vocabulary. In each case, what Wittgenstein does with the refinement idea makes for a stark contrast with traditional philosophical views about language, for which Wittgenstein often takes Frege to be the leading spokesman.

To begin with conceptual refinement, Wittgenstein, in Philosophical Investigations §71, criticizes Frege's idea that concepts, like geometrical areas, require absolutely sharp boundaries. To place Frege's view in a larger context, Frege¹¹ sees (what he calls) concepts, as well as (what he calls) senses (these are what we usually call concepts) as perfectly well-defined, changeless, eternally existing. Such abstract objects occupy a "third realm," a realm of things that are neither physical nor mental. Most important for our purposes are Frege's senses, the meanings, roughly, of linguistic expressions. Pieces of language are, as it were, the worldly (and so imperfect) incarnation of the well-defined, changeless, eternally existing senses, and Frege points out ways in which our language falls short of logical perfection.

Wittgenstein, like Frege, emphasizes the "worldliness," that is, the this-worldliness, of linguistic symbols. But for Wittgenstein, this emphasis is not by way of contrast with other-worldly perfection. The point is rather to underscore the connectedness of the symbols, and of the practices in which they are embedded, with the human goings-on from which the symbols and practices evolve.

In the beginning, we might say, were prelinguistic brutes, bumping into each other and grunting. How such grunting evolves into articulated speech is a difficult question, but that it does is clear. Indeed, over time the grunting matures into something quite sophisticated. Don't think of this emerging conceptual distillation as reflecting the absolute refinement of other-worldly concepts. Articulation is a gradual and relative thing. Concepts are articulated as they need to be for purposes at hand. And purposes at hand, even in the most abstract studies, never require concepts with absolutely sharp boundaries. Language, even at its most rarefied, reflects its beginnings in the deed.

The global difference of orientation at issue here--Frege's worldliness vs. Wittgenstein's--leaves its traces throughout the philosophy of language and the philosophy of mind. One example, briefly noted in the last section, concerns linguistic competence. Frege saw competence as requiring the speaker's intellectual grasp of a (third realm) sense. On more naturalistic approaches, Wittgenstein's included, competence with language is characteristically understood in terms of knowing how, in terms of the speaker's getting the hang of the practice with the expression.

B. Two Grades Of Worldly Involvement

I will return below to the theme of language as a refinement on the prelinguistic. I turn now to another aspect of Wittgenstein's naturalism about language. I begin with a brief discussion of an interpretation of Wittgenstein that distances him from my sort of naturalism.

Wittgenstein's emphasis on the use of language, on language as social practice, and perhaps especially his obscure remarks to the effect that meaning is just use, has encouraged some to suppose that significance can be understood in abstraction from what words are about, as if the real world of things under discussion had nothing to do with meaning. Whatever there is to meaning can be discerned by a study that is restricted to the actions, social interactions, and mental goings on of speakers, in abstraction from the realm of things "out there" about which they might be supposed to be speaking. A cousin of this view, on the topic not of meaning but of reference, has it that Wittgenstein is an "anti-realist" about reference.

Such an "idealist" reading of Wittgenstein seems to me to misfire rather radically. To begin with reference, while Wittgenstein surely does remark critically on philosophers' use of the notion, he seems to me about as serious and naive concerning the application of words to the real world, about the fact that words pick out things, as are, say, contemporary direct reference advocates. Indeed, he writes, "it will often prove useful in philosophy to say to ourselves: naming something is like attaching a label to a thing."⁽¹⁴⁾

Let's turn from reference to the more general topic of significance. I began with the "idealist" reading of Wittgenstein in order to focus attention on the question of the contribution of the natural world to meaning. Wittgenstein, as I understand him, maintained that the natural world is deeply embedded in the meaning of our words. I need to do better, though, by way of saying what's at stake here. What is it for the world to be embedded in the significance of words? I begin with Frege's view, one that de-emphasizes the contribution of the world to meaning, but not in the service of any kind of idealism.

For Frege, linguistic significance arises when a mind grasps a sense and associates it with a linguistic expression. What happened to the world of references in this picture? Where is its contribution to meaning? For Frege, any such contribution is indirect. Senses, of course, have references--at least when the universe cooperates. Stretching usage a bit, we might say that a sense means its reference.⁽¹⁵⁾ So it's not that the world is irrelevant, that it fails to show up in the picture. But what makes an expression, for example 'Gabriel', significant is not its relation to a particular piece of the world, in this case to my dog. Indeed the expression could have just the significance it has even if Gabe did not exist.

Consider, in connection with Frege's setting the world at a distance from meaning, Frege's notion of a thought, what we nowadays call a proposition. Frege took thoughts to be (roughly) constellations of senses. Frege's approach, his idea that the constituents of propositions are purely conceptual, is perhaps the most intuitive way to construe propositions. To approach the subject this way is, however, to set the world of references at some distance not only from meaning, but from the contents of thought. The things we think about, real-world references, become outsiders to the contents of our thought.

One way to emphasize this alienation of the world of references from thought is to focus upon names that fail to refer and on thoughts expressed by sentences that contain such "empty" names. "Odysseus was set ashore at Ithaca while sound asleep," Frege tells us, formulates a complete thought, albeit one that lacks a truth-value, since 'Odysseus' fails to refer. Let the whole world disappear. At least some thoughts would lose their truth-values; all would remain unexpressed. Still, the internal integrity of the thoughts would be unaffected.

Let's turn to direct reference for a view that gives considerably more weight to the world's contribution to significance. The direct reference idea--it's a bit like the story of Adam naming the creatures--is that the world presents us with items to which we may assign names, thought of as tags or labels. Proper names, on this view, "directly" designate the things to which they have been assigned. They designate their bearers, that is, without the mediation of ideas, senses, or concepts. A name is significant in that it stands for this particular thing. The significance of a name comes, so to speak, not from above, from the realm of senses, but from below, from the worldly realm of references. If one did not mind stealing cute turns of phrase, one might say that the direct reference approach represents a further grade of worldly involvement.⁽¹⁶⁾ The world, as it were, gets into the meaning or significance of one's words.

What then becomes of propositions, which Frege understood as constellations of senses? Direct reference people who posit propositions give the notion a Russellian twist. For Russell, the proposition expressed by "John is happy" is constituted not by senses but by references, that is, by John himself⁽¹⁷⁾ and happiness, a universal. Such propositions, championed by David Kaplan who dubbed them "singular propositions,"⁽¹⁸⁾ have been found by some to be unproblematic and by others to be completely unintelligible. (Such is the way of the life we have chosen.)

Whatever we do about singular propositions in the end,⁽¹⁹⁾ the fundamental direct reference idea is that an assertion about John does not involve a conceptual representation of John, but rather John himself. Talk of "involvement" is fudgy, even if heuristically useful. So it would be better to put the point otherwise. We might say that on Frege's view assertion requires that the mind apprehend, that it be directed upon, a thought content, in this case the proposition that John is happy. A constituent of this proposition is the sense of "John," and the mind, in apprehending this proposition, apprehends the sense of "John."⁽²⁰⁾ The direct reference idea can now be put in this way: In thinking a thought about John, or in making such an assertion, the mind is directed not upon any such conceptual representation of John, but upon John himself. The world gets into the significance of one's words, even--given singular propositions--into the contents of one's thought.

One reflection of this greater degree of worldly involvement is the severe problem non-denoting names present for direct reference. If the significance of a name is purely a matter of its reference, then either 'Odysseus' really fails to be significant or someone had better say something very smart, very fast. To editorialize just a bit: This is a fundamental problem for direct reference that should not be avoided or finessed. Given all the powerful considerations direct reference people have urged against Frege, and even given that one may (or, of course, one may not) feel that greater worldly involvement is a good thing, even so, one ought to hesitate over this one.⁽²¹⁾

So far we have considered two grades of worldly involvement. Notice the inverse proportion: as the world's contribution to significance becomes greater, less is required of intellect. Frege's view is relatively undemanding of the world, relatively demanding of us. Both the introduction of a name and its subsequent use have a substantial cognitive prerequisite: the speaker must be in a position to conceptually single out the referent. Frege, you might say, presents a highly intellectualized picture of language.⁽²²⁾

It is just this highly intellectualized conception that Wittgenstein criticizes time and again. And direct reference philosophers, joining with Wittgenstein, emphasize that the use of a name requires no such cognitive achievement. This is not to deny that mastering the practice of proper names represents a cognitive achievement, even a substantial one; Gabe, as smart as he is, can't do it. But this is cognitive in a different sense, mastery of a highly sophisticated practice, knowing how.

Once the practice is in hand, moreover, the use of a particular name demands very little of intellect. A student wanders into a classroom where Aristotle is under discussion, hears the name, picks up a few facts about Aristotle (and perhaps some misinformation), and is off. The student, even though he may lack anything close to an accurate, uniquely identifying characterization of Aristotle is now able to say and think things about Aristotle. It's like magic.

When we turn to name introduction, as opposed to subsequent use, direct reference provides more role for intellect, the role of assigning names. This is important, and will contrast with the more extreme Wittgensteinian idea below. But assigning names is still a modest exercise of intellect relative to what Frege had in mind.

C. Wittgenstein On Pain: A Third Grade Of Worldly Involvement

I want to explore the possibility of a deeper level of worldly involvement. Wittgenstein, as we will see, apportions even more responsibility to the world, even less to intellect.

How do words refer to sensations?--There doesn't seem to be any problem here; don't we talk about sensations every day, and give them names? But how is the connexion between the name and the thing named set up? This question is the same as⁽²³⁾: how does a human being learn the meaning of the names of sensations?--of the word "pain" for example. Here is one possibility: words are connected with the primitive, the natural, expressions of the sensation and used in their place. A child has hurt himself and he cries; and then adults talk to him and teach him exclamations and, later, sentences. They teach the child new pain-behaviour.

"So are you saying that the word 'pain' really means crying?" -- on the contrary: the verbal expression of pain replaces crying and does not describe it. (PI, §244.)

I don't know if Wittgenstein is correct about pain language. Nor will I explore that question here. The mere possibility will keep us busy. Moreover, I don't know how far to extend his idea. I bet he did not intend his remarks to apply to our entire linguistic repertoire for talking about pain,⁽²⁴⁾ not to speak of other sensations. On the face of it, his idea does not seem to apply to third person reports, to statements to the effect that someone else is in pain. My remark about your pain certainly doesn't seem like pain behavior on my part. Perhaps Wittgenstein's idea has merit only with regard to remarks like "It hurts," or "My foot hurts," perhaps to all first person (present tense?) remarks that attribute pain. Or perhaps there is a way to extend it to even third person cases,⁽²⁵⁾ or to other cases in addition to pain.

Wittgenstein's pain-behavior idea seems to touch a number of bases. First, it is suggestive about the functioning of at least some pain locutions. Second, it provides a model for how we learn pain vocabulary. Finally--thinking about the Culture and Value remark about language as a refinement on more primitive things--the pain-behavior idea is suggestive as to the birth of such vocabulary, how it may have entered our linguistic practice.

It is characteristic of Wittgenstein to focus our attention on the second and third of these, and this is very unlike what many of us do if left to our own devices. It seems to me salutary to consider such questions, perhaps especially fitting in the context of naturalistic exploration. Late twentieth century discussions in the philosophy of mathematics have emphasized that there must be a fit between an account of things known, an account, for example, of the nature of the numbers, and an account of our coming to know such things. It would be an embarrassment if the things allegedly known turned out to be unknowable by creatures with our cognitive capacities. A parallel here is that a theory about the meaning or function of an expression has to square itself with how we learn language, and how expressions with such functions come to enter the practice.

This is of course not to suggest that one cannot profitably work on questions of function without exploring questions of language learning or of origin. But it is to recommend reflection on the learning process and the evolution of our practices. Such reflection may serve as a kind of corrective to high flying accounts of linguistic function. More positively, plain facts or even speculations about language learning, and about origins, may be suggestive about the character of what we learn when we learn language. And a conception of linguistic function that coheres with a plausible story or stories from these other domains is that much stronger.

In my earlier discussion of Wittgenstein's language-as-refinement idea, I emphasized Wittgenstein's suggestion that elements of our practice never forget their roots, that ancestry remains present in later forms. This doesn't really predict anything in particular. But it does counsel that we keep an eye out. One example was the question of concepts without sharp boundaries discussed above. Another comes to the fore here.

Some pain locutions, let's assume, really did begin as refinements on primitive pain behavior. To start small, groans evolve into something more like articulated words; "Ow!" is born. Let's take another step backward, beyond such linguistic or quasi-linguistic forms to groaning itself or wincing. We can distinguish two stages at this pre-linguistic level.

Stage one: One groans in immediate response to pain.

Stage two: One becomes attuned to the fact that, say, one's mother's care and attention are forthcoming when the groans are evident. So one learns to modulate one's groans for maximum advantage.⁽²⁶⁾

Here, already, something of enormous significance begins to emerge, a birth of meaning. At stage one, the groans express pain, but not in the sense that they express pain at stage two, the modulation stage. Stage two groaning is behavior with a teleological aspect, aimed at eliciting response, even if none of this is formulable by the agent. Stage two groaning thus approaches linguistic expression. We might mark the distinction by saying that at the first stage, the pain expresses itself in the behavior, while at the second the person expresses the pain.

To continue the fantasy, we can imagine that if it's in the interest of the creature to advertise his pains to others, then more articulate expressions eventually emerge. Perhaps "Ow!" emerges first, as an "expression" of pain in both senses just distinguished; it's the noise one learns to make as an immediate response when one hurts, and it's the noise one learns to make to let others in on one's hurts. Later, various refinements on, or substitutes for, "Ow!" emerge, for example, different noises for different kinds or intensities of pain. Still, the expressions remain primitive. They function, that is, only at stage two: the speaker makes these sounds to express his pain, but not, as we use pain expressions, to speak about his pains. Accordingly, his utterances do not count as assertions, as true or false. (Which is not to say that they cannot mislead, or even intentionally mislead.)

The way is now prepared for a major new development in meaning's evolution. It now seems like a small step--one that could take no more than a few eons-- to pain vocabulary as functioning in a referential fashion.⁽²⁷⁾ Here's how it might go: A painful experience or episode, once expressed by means of a stage two utterance, is in the public domain. Indeed, part of the speaker's motivation for utterance may have been to let others know. And it would certainly be nice if the others (and the speaker himself later on) had a way to get at the episode in thought and speech. All it takes is an ingenious wordsmith to extend a piece of pain vocabulary from it's purely expressive function to a referential one. Indeed, we witness, and participate in, such linguistic developments with our own children. The term, "owie" is often used by parents as a noun, in reference to a child's pain expressed by "Ow!"

Talk of an ingenious wordsmith extending linguistic function makes it sound like a matter of conscious decision, the product of deliberation. And there may be moments of deliberation involved. But the development we are envisaging is no doubt gradual and less conscious than the wordsmith image suggests. We can imagine intermediate cases: before the epoch of straightforwardly referential pain talk, the primitive expressions may find their way into sentence frames. "There is pain in my leg," "My foot hurts," "I am in pain" may emerge, where the embedding sentence frames, for example, "There is x in y," have other uses, some of which are farther along the linguistic evolutionary ladder. What effect does the presence of the pain locution in such frames have on the functioning of the pain locution? Does it still function primitively, so that the resulting utterance fails to count as an assertion? Or does the presence of the frame determine that the speaker has unknowingly arrived at referential pain talk? About such intermediate cases, it is unclear what we are to say. Perhaps these questions, in such a context, admit of no determinate answers.

So "Pain!" as a utterance of a primitive, pre-referential expression may yield to "pain" as a noun, as a referential device, as a device for attribution. You can think of attribution here as description, as long as you don't read description in the traditional fashion. To say that the term, 'pain' becomes available as a device for attribution is not to say that there becomes associated with it a characterization of pain, or even that it now designates a property shared by all instances of pain. It is rather to say that the term becomes our device for collecting the rough assemblage of hurts originally expressed by its use. Given our crucial ability to extend usage in ways that seem to us and our fellows entirely natural, the term becomes available for pain ascription generally.

The foregoing story about origins is suggestive about language learning as well as about actual function. As for the latter, we need to go slow. Let's grant such a story about origins, and even that something of the primitive function is preserved, so that first-person pain utterances are expressions--perhaps in both senses--of pain. Still, it does not follow that first-person utterances are not really ascriptions, that they are mere expressions of pain, mere "avowals," that they fail to possess truth-values.

Pure expression, mere "avowal," may have been the correct characterization of the practice at the primitive stage, but we are no longer there. We now have a general term in the public domain that we use to characterize others' experiences. Is there something that prevents us from characterizing our own? Why insist that our first-person utterances are mere anythings? Such insistence seems unmotivated, and in any case conflicts with so much that we ordinarily say and think about people, for example, that they can lie about their own pain, or tell the truth. If Wittgenstein is correct about the core idea, perhaps the lesson should be that speech that constitutes "expression" can also represent, or purport to represent, the way things are.

In exploring Wittgenstein's ideas about pain ascription, we have been exploring another way in which language constitutes a refinement on the prelinguistic. I introduced Wittgenstein's pain-behavior remarks with the promise that there was here another kind of naturalistic theme, a third grade of worldly involvement. I turn to that promise.

If there is indeed another level of worldly involvement in meaning, and if primitive pain locutions are the ones that exhibit the phenomenon, then it would seem to be important that they are genuine bearers of linguistic meaning. Is this so? Should we count the primitive, pre-referential pain locutions as meaningful linguistic expressions? Clearly such locutions do some work; they have a function. The question is one of continuity: How similar in function are the primitive expressions to paradigm linguistic expressions? I find the continuities impressive. The primitive expressions, after all, are repeatable expressions that have a discernible function in communication. When we turn to the somewhat later development considered above in which the still primitive pain vocabulary is embedded in sentence frames, the case gets stronger for linguistic significance. In such examples, the pain locutions begin to function in more deeply systematic ways, combining with other expressions in a variety of constructions. Since it is unclear, even indeterminate, whether in this intermediate case the pain locutions have a referential function, one might take it to be unclear, or indeterminate, whether they have linguistic meaning. But referential function as a criterion of linguistic significance is surely too strong. It would rule out clearly meaningful words like our "no." I will provisionally assume that the primitive pain vocabulary bears linguistic significance, and return below to this assumption.

So far we have discussed two models for how expressions come to bear meaning. In ascending order of worldly involvement, and descending order of the contribution of intellect: Frege's, that an agent associates an expression with an intellectually apprehended sense, and the direct reference idea that an agent assigns an expression to something provided by nature. The model we are now considering seems very different. Like direct reference, the significance of the primitive "pain" pertains to association, in a very general sense, with the worldly realm; certainly no senses are involved. But "assignment" of words to aspects of the worldly realm is the wrong idea. Here's how it works. Early on, the agent learns to control himself, not to writhe in pain in ways that may hurt others, not to scream offensively. At some point vocables emerge as even more acceptable expressions of pain. After some training, the agent learns to make these noises rather than engage in cruder forms of pain behavior. (This is stage one.) When he begins to make these noises in order to elicit a response--that is, when he begins to utter what we are assuming to be meaningful vocables--he is making use of locutions whose associations with the world have already been fixed at an earlier stage, at a stage at which the vocables failed to bear linguistic meaning.

With respect to world/intellect inverse proportion, Wittgenstein's primitive pain expressions provide a limiting case. Notice how little the mind has to do: no associating words with concepts, not even assigning words to things. The primitive symbol is born with meaning; it comes to us interpreted.⁽²⁸⁾ This represents a new and third grade of worldly involvement in meaning, a new idea.

Let's return to my provisional assumption that the primitive expressions bear linguistic significance. If someone has strong intuitions that this is not so-- perhaps on the grounds that until the arrival of the referential function there is no real language and/or no real linguistic meaning--I will not argue. Still, we now see continuities where we saw discrete forms. Even if the primitive "pain" does not possess specifically linguistic meaning, "Pain!" is surely a significant noise. The primitive "pain" is a precursor to our word "pain," and what we might call the communicative significance of the former is a precursor to genuinely linguistic significance. Even if there is no third grade of worldly involvement in linguistic meaning, there is such a grade of worldly involvement in pre-linguistic significance.

There is more that is new. Wittgenstein is no fan of intermediate entities as explanatory posits, but he sees great importance in "finding and inventing intermediate cases." (PI, §122) What Wittgenstein has done in the pain case is, as I see it, to find or invent a crucial intermediate case, and I have tried to assist in such efforts above. Indeed, on any view that sees great significance in the fact that language is a refinement on the pre-linguistic, the transition from the pre-linguistic to the linguistic will be crucial. Do we have any sense of how such a transition might go? Wittgenstein's discussion provides insight on this question.

Grice distinguishes natural from non-natural meaning, meaningⁿ from meaningⁿⁿ. That smoke means fire exemplifies the former; that "smoke" stands for smoke, the latter. Natural meaning is the original; non-natural meaning, of which linguistic meaning is one kind, presumably evolves from the more primitive natural meaning. Wittgenstein's intermediate case--pre-referential "pain"--can help us understand this development.

Notice that Wittgenstein's primitive "pain" furnishes an example of Grice's natural meaning. The stage one utterance, "Pain!," no less than a groan or grimace, is correlated with pain, and can serve as an indication that pain is being experienced. So "Pain!" meansⁿ pain. That stage one "Pain!" meansⁿ pain figures in its coming to be used by a pained person to communicate his pain--we are now at stage two. The stage two utterance, with its non-natural meaning--I've allowed that this may be only "communicative" significance--evolves from stage one pain-expression, and gives rise, given human ingenuity, to later stages, to stages at which the expressions are paradigms of linguistic significance. Wittgenstein's intermediate case provides a picture of one way in which non-natural meaning can arise out of natural meaning.

The sort of naturalism that I advocate, one that rejects reductionist and eliminativist strategies and that extends as far as religion, has not had much play in recent times. My aim in this paper has been first to create a feel for such an approach, and then to explore several themes in Wittgenstein's later work that seem to me congenial.

While it is well known that Wittgenstein emphasizes the deeply social character of language and meaning, the naturalistic character of Wittgenstein's conception is less widely acknowledged. My focus here has been the contrast between Wittgenstein's view that emphasizes the natural basis for language and the continuities between later forms and the prelinguistic, and one that sees language as the embodiment of the eternal structure of thought, as Tyler Burge once put it.⁽²⁹⁾

1. < From The Realm of Mind (New York, 1926), pp. 2-4.

2. The view expressed here is very far from the naturalism associated with Hume, according to which there are philosophically convincing skeptical concerns but such concerns are to be forgotten when one leaves one's study. The naturalistic outlook makes skepticism unattractive even in the study, or so I'm arguing.

3. < Cf. Pears's "Wittgenstein's Naturalism," p. 3

4. Such complaints are often heard from those who have no love for analytic philosophy or appreciation for what I see as its genuine and substantial accomplishments. To say that we have been too narrow in our vision is not to denigrate the accomplishments. In my undergraduate days, I studied a number of American philosophers, naturalists like Santayana and Woodbridge, who were genuinely visionary, but who lacked the analytic virtues so emphasized in our times. Nowadays, vision is almost dirty word. As Hilary Putnam says, "The besetting sin of philosophers seems to be throwing the baby out with the bath water. From the beginning, each "new wave" of philosophers has simply ignored the insight of the previous wave in the course of advancing it's own." ("Sense, Nonsense, and the Senses: An Inquiry into the Powers of the Human Mind," The Journal of Philosophy [Vol. XCI, September 1994], p. 445.

5. In recent years there has been considerable attention paid to religion by analytic philosophers. Indeed some of the age-old battles between theists and atheists have never disappeared, and theists have become energized in recent years by the application of analytic techniques to the justification of theism. What has been absent, however, although present earlier in our century, is the general perception that it is in the domain of philosophy to explore the significance and power of religious ideas and institutions.

6. My remarks here are inspired by reflection upon my own religious tradition, Judaism, which is not to say that these remarks are uncontroversial even in that context; quite the contrary. Santayana's thinking about religion, perhaps inspired by his reflection on his own Catholic tradition, seems to me congenial. See especially, Chapter III, "Reason in Religion," in The Life of Reason, (New York, 1955). Insofar as the emphasis is on community and ritualized forms, this picture is less straightforwardly applicable to many strains of Protestantism, but there may be reverberations there too. The question of the applicability of this sort of picture to the other great religions is even further beyond my competence.

7. A particularly powerful example of Mencius's point was recently powerfully described to me by David Shulman, the first government AIDS attorney. Shulman described participating in the production of the AIDS quilt. The activity and the product were expressions of a powerful sense of communal support and solidarity. This is a dramatic example, but once one starts to think of it, there are many everyday sort of examples.

8. Needless to say, I am here only scratching the surface. What, for example, does one make of prayer, and of religious belief, on such a non-theistic view? I believe there is much to be said for this approach to religion, and I will try to make good on this elsewhere. See my "Awe and the Religious Life: A Naturalistic Approach," presented at the Pacific APA, March 1995.

9. Philosophical Investigations, §530.

10. The Blue Book, p. 4.

11. Frege's views have been vigorously criticized and defended throughout this century, defended sometimes on grounds that the critics are attacking views that Frege never held. What is most important especially to the contemporary anti-Frege literature, however, is the correctness of these allegedly Fregean ideas, whether Frege's or not. Perhaps the designation "Fregean" (and related forms), in this paper and throughout this critical literature should be read as naming an important constellation of views that many have found in Frege's writings. This "Fregean" outlook is important because it represents a natural take on the questions it addresses and is, in the best sense, unsophisticated. It represents, to be facetious, what Frege might have thought before he became sophisticated.

12. P.4.

13. See my forthexisting book, The Magic Prism: An Essay in the Philosophy of Language, as well as a number of previously published papers, reprinted in Has Semantics Rested On A Mistake?, and Other Essays (Stanford University Press, 1991). See esp. "Frege-Russell Semantics" (Chapter 8), "Cognitive Significance Without Cognitive Content" (Chapter 10), and "Turning the Tables on Frege" (Chapter 11).

14. PI, §15.

15. If a sense means it's reference, then the contribution of the world to the significance of an expression consists in the world's feeding directly into the significance of a posited intermediary. This seems related, or just the other side of the coin, to these ideas: thought about an object consists, for Frege, in the direct apprehension of an intermediary, and linguistic reference to an object consists in the expression of an intermediary.

16. In addition to the primary theft from Quine, I've been borrowing this use of "worldly" from Joseph Almog.

17. I here take Russell's lead and omit consideration of his doctrine about logically proper names (and the related idea that direct acquaintance is a necessary condition for name reference). Russell often gives examples of his view which abstract from these features, and contemporary singular propositions theorists, not taken with Russell's epistemic views, have largely ignored these matters.

18. See Kaplan's classic monograph, "Demonstratives," in Themes From Kaplan, ed. J. Almog, J. Perry, and H. Wettstein, Oxford University Press, 1989, pp. 481-566. The focus of direct reference advocates has been upon the subject position, and the contribution of a subject term to the proposition. So a commitment to singular propositions has come to mean a commitment to objects as constituents of propositions, but has not meant anything very clear about the predicate constituent of the proposition.

19. My own view is that direct reference would do well to eschew propositions. See "Cognitive Significance Without Cognitive Content."

20. Perhaps grasping a thought does not entail grasping its constituents. The usual picture of Frege's conception, however, is that in grasping a proposition one grasps its constituent senses.

21. I have addressed this problem in an unpublished talk, "That Name Means Nothing to Me." My contention is that insufficient attention has been given by direct reference to the radically unFregean social conception of language that, I argue, underlies the direct reference view. The notorious problems of "cognitive significance," including the problem of non-denoting names, look very different from the more social perspective.

22. Russell's view, while it does not require the association, or even the possession, of an individuating concept, does require direct acquaintance, an exceedingly strong cognitive relation between speaker and referent.

23. The two questions Wittgenstein formulates seem very different. Andrew Hsu suggests that we read "This question is the same as . . . ." as something like "We might more profitably ask . . . ."

24. The "about" in "talking about pain" is to be read noncommittally. Perhaps such "talking about" is mere non-symbolic pain behavior.

25. See the passage in Zettel, §§540-542, mentioned by Pears, in which Wittgenstein may be suggesting that our natural reaction to the pain of others may ground our third person ascriptions, so that such "ascriptions" may turn out to be their own kind of behavior surrogates.

26. Cf. G. H. Mead's views on the evolution of language. See, e.g. "Mead on Mind, Self, and Society," in I. Scheffler's Four Pragmatists (London, 1974).

27. Reference here is to be taken broadly. When we hear "reference," we often think of "singular reference." But "table," "evolution," "inertia" are all nouns, and in some sense referential devices. No account of the nature of pain is presupposed here; only that "pain" and other pain locutions are in the broad sense referential.

28. Cf. Pears: "We do not face the intellectual task of identifying a sensation as painful, because nature has done most of the work for us. All that we have to do is to substitute the word 'pain' for the natural reaction, and the contribution required from out intellects is minimal."

29. The present paper derives from remarks I offered on David Pears' "Wittgenstein's Naturalism," (this volume) at the UCSD conference on Wittgenstein organized by Avrum Stroll. I am grateful for many illuminating discussions on these matters with Larry Wright, and for comments upon earlier drafts to Arthur Collins, Eros Corazza, Andrew Eschleman, John Fischer, Carl Hoefer, Paul Hoffman, Andrew Hsu, and Genoveva Marti.