apolliana: (Default)
2013-09-29 03:45 pm

Russell's Principle - What Do You Know When You Don't Know the Referent of a Demonstrative?

 photo IMG_1687.jpg

Russell's Principle: That in order to have a coherent thought, you must know what you are thinking about; in particular, you must have some kind of distinguishing knowledge of the referents of singular terms in your thought--i.e. a way of distinguishing them from other things.

Suppose you receive a postcard that says simply, 'I wish you were here,' with no pictorial representation of where 'here' is. Suppose, for this first case, that it's signed, so the only unknown is where 'here' is. Do you understand the sentence? My inclination is to say, yes and no. You understand it as a sentence containing a variable standing for 'the place where I [the writer] am,' wherever that happens to be. And the way you understand a sentence like that is certainly different from the way you understand a sentence with no undetermined referents. Why? Because you will invariably puzzle over where 'here' is. You have received a little mystery. You do, however, understand the kind of thing 'here' stands for: some place or other, where the writer is located at the time of writing.

John Perry's response is--extremely roughly--that you think of the referent of 'here' casually; this doesn't seem quite right to me. And that since you can, with the right resources, figure out what it is, then you know what its referent is. I agree that you can sometimes figure it out. And if there were no causality, and no event was connected to any other, this--and much else--would be rather difficult. But I think you can figure it out because you know that 'here' refers to the place the writer was at the time of writing. (For all I know, this is Perry's current position. I haven't read Reference and Reflexivity yet. But he does love to talk about ambiguous postcards as counterexamples to Russell's Principle.)

This effect is magnified if the sentence you're presented with contains more than one indexical or demonstrative with an unknown referent. On the wall above, 'It's you I love' is undetermined for both 'you' and 'I.' Supposing it actually was written by one person for another, we understand it to be one person telling another that they love them. And that is about all one understands. It's an empty schema. You can only imagine the circumstances that occasioned it.

It's also possible that it was written as an empty schema, in order to puzzle people, or to symbolize the act of telling someone that you love them; in this case there is no referent. It's art. This case is perhaps the furthest from any possibility of an ordinary understanding of the sentence. Not only is it an empty schema: it was written in order to be one. It's a representation of a type of utterance. It's not an utterance at all. (This case is interesting, because these sorts of thing are everywhere these days: mottos inscribed on things. They're not actual utterances--no one is saying them. And it's not clear who no one is saying them to. Yet we surround ourselves with them.)

All of this puts aside the more complex problem that even a sentence like 'wish you were here' might not have the meaning it superficially appears to have. It might be sarcastic: "Vesuvius in eruption. Wish you were here." There might be still more to the context of a spoken or written utterance than simply the referents of indexicals and demonstratives that we need to know in order to understand the utterance. If I turn on the television and see one character say to another, "It's you I love," having not seen what preceded that, though I will know its referents (sort of: this person and that person--though I may be intrigued to watch more to find out which particular people they are in the show's universe)--and I may be able to determine from its tone whether it's sincere, I won't really know its import, having not seen what preceded it. If I open a book I've never read to a random page and see "It's you I love" written there, then, though I may be able to see that it occurs in a conversation and thus is not a piece of conceptual art, I will be able to determine just from reading that sentence neither its sincerity, nor its import, nor the referents of 'you' and 'I.' I will, however, know the sorts of things 'you' and 'I' stand for--so long as they occur in ordinary speech. They stand for people (or, in this case, fictional people). I will have to read the rest of the story, however, to find out who.

I would like, then, to partially uphold Russell's Principle. I think the way one understands a sentence with indexicals and demonstratives whose referents one doesn't know is quite different from the way one understands a sentence with indexicals and demonstratives whose referents one does know. And I think that this ability to come to know said referents rests primarily on understanding the way such terms work--the sorts of things they refer to.
apolliana: (Default)
2013-09-14 10:37 pm

Evans and Strawson, "Language and Creativity" - Transcript

I have transcribed the discussion between Strawson and Evans in this television program, for the good of the many, but mostly of me. The transcript is downloadable at scribd here. There is much more here about generality, and whether language ought to be confining (no), and I will provisionally agree with Strawson about all of it. Except, that is, with his and Evans's sexist language.

(Also: Strawson looks uncannily like Patrick Cargill.)

Part 1

S (lecturing): And now to language, that marvelous liberating medium; the medium of thought, a degree of complexity, of reasoning of any but the simplest variety, of communication, of stored knowledge, and, if Wittgenstein is right, the source of all the bewitching perplexities of philosophy.

[Music, as of computers beeping.]

S: Language, we want to say, is the vehicle of thought, or words the clothing our thoughts put on when they make their appearance in public, or the outward and visible or audible manifestation of the inward and spiritual thing. Yet we also feel we must have managed the force of the idea, for at the level of any complexity the availability of any language, the vocabulary and syntax of our language, the availability of a sentence for expressing a thought is the possibility of thinking the thought. After a point, what we can’t say, we can’t think. So our language, or our languages, come to seem like an autonomous and highly structured realm, with all the thoughts we might think lying in it.

S: And yet again in a sense we are masters of this realm—the language, languages, our language; the limitless sentences and combinations of sentences of which we know in advance the significance, but of which we only ever use, or read, or see, or hear, a comparatively insignificant proportion of them. And even that comparatively insignificant proportion, which we so readily frame and utter in understanding what we say, or which we hear or read, understanding what we hear or read; even this is vastly numerous. So how is it that we have this vast and potentially limitless understanding?

[Formal logic and semantics book is shown. Cut to Evans talking about a farmhouse in Aberystwyth where he likes to go philosophize, walking around, natural surroundings. London and America cause tension. Doing philosophy is extremely hard. And when most at work appears to be asleep: an embarrassing occupational hazard for a philosopher.]

Interviewer: To what extent do you think that your investigation of language is an investigation of the organic structure of a living form?

E: Living… well, I think probably it is something which I don’t allow myself to forget that language changes, and is changing all the time.

I: Name two ways in which that influences you.

E: Well, there’s a pretty ordinary distinction between names, proper names (‘Evans’ and so on), and descriptions (‘the guy who’s got long hair’). There’s some moderately obvious distinction between those two forms of referential device. But I think it’s very, very important to the understanding of both of them to see that an expression can change from one category to the other over time. You’ve got to have your distinctions, your taxonomy and your theory, capable of allowing for that.

[Sun on leaves. Strawson stands on a low tree branch.]

E: Peter Strawson is now Professor of Metaphysics at Maudlin, but he only took up his chair there quite recently. Up until a few years ago, he was college tutor at University College, which is where I met him, just in the normal course of events. As an undergraduate there I was taught most of my philosophy by him. Most of his early stuff, his published stuff, as far as I know, was on logical theory—necessity, entailment, and of course the famous quarrel with Russell on the theory of descriptions. During the course of which he introduced and tried to refine, and over the years increasingly tried to refine the idea of a speaker’s referring to something. This idea moved very much to the center of the stage in his book Individuals, where he asked questions about how the world must be in order for a speaker to be able to refer to things like events, bodies, persons, sounds, nations, and so on. Since Individuals, he’s written on a variety—quite a wide variety—of different topics, but an experienced Strawson-watcher will be able to detect beneath the surface a concern with reference and with the distinction between particular and general.

E: I don’t think I came away from being taught by him with a whole set of philosophical doctrines, or dogmas. What he did manage to communicate to me was an idea that certain subjects are very interesting. As time has gone on, I’ve disagreed with him more and more about particular matters, for example, like the whole debate between him and Russell. But what has persisted is my conviction, shared with him, that the subject of reference, and the distinction between particular and general, for example, is a subject of enormous fascination. So if I gave a student one of Strawson’s things to read, say, his book Individuals, I wouldn’t give a damn if he thought it was all wrong. What I would mind is if he couldn’t see that a very wide range of interesting subjects were being discussed in a fruitful way.

E: Do you think there’s any sense in which philosophy differs from other disciplines in that its results should ultimately be intelligible to ordinary people?

S: Intelligible?

E: Intelligible to non-philosophers. Which is clearly not a constraint physicists operate under; although they can be intelligible, they don’t have to be.

S: Well, what I think is true is something like this: namely that philosophers must try to explain, and try to explain the relationships of, ideas which every man as a man has to operate with, has to use: ideas like that of knowledge, truth, identity, personal identity, perception—these are ideas with which everybody is bound to operate, being a man, and these are among—importantly among—the ideas which it’s the task of the philosopher to explain and explain the relationships among. Whether it follows from that that his results, and even his questions, except the most general questions of all, should be able to be intelligible to everybody, I’m not so clear.

E: I agree, I don’t suppose it would follow from that. But aren’t there some concepts that are as it were second-order—those of logical form, entailment and so on. Do you think these are of a different kind?

Part 2

S: Oh, they might well be, because of course there are philosophical problems that go under the title ‘philosophy of mathematics’ or ‘philosophy of physics,’ and if no one is required as a man to be interested in physics or mathematics, no one can be required as a man to be concerned with the philosophy of these subjects. I would go further and say concern with these ideas which are everybody’s property, which everybody willy-nilly operates with, is the fundamental concern of philosophy.

E: Ah, well what’s captured in that word ‘fundamental’?

S: Well, the tendency to ask philosophical questions or to think philosophically is possibly endemic in the species, that children ask philosophical questions, worry about philosophical issues. And of course when they do so, it is about these common notions that we all employ that they think. But I think I want to say that in another sense handling these notions is fundamental because I really think too that to explain the more sophisticated notions probably requires an adequate philosophy of the commoner notions.

E: Yes, then that gives a more solid sense to ‘fundamental.’ The sort of thing you have in mind for children would be what happens at the end of the world, or having an infinity…. Well, I don’t know—would infinity fall into your category of things that one has to operate with? In a sense it’s cross-cutting because he doesn’t have to operate with it, but in a sense it’s perplexing for every man.

S: As soon as he starts thinking of the position of things in space, or the series of events in time, questions of a very profound nature come up.

E: Well….yes. You talk about handling these things. Is it just that philosophy is concerned with laying bare, tracing paths between these central concepts: causality, person, body, space and time and so on? Is that really what it’s all about?

S: Well, I think it is what it’s fundamentally about. I think there are certain concepts or classes of concepts which are absolutely basic in our scheme of things—in everybody’s scheme of things. Notions like space, time, persistence, bodies, identification, reidentification, action—all these form, to my mind, a web with complex relations to each other. And one can’t hope to get really and finally clear without addressing a complex set of connections which bring all these in relation to each other.

E: Yes, well isn’t that a bit pessimistic in a way? Since the list is so long—and I don’t think you’ve exhausted it—it’s a bit like there being a building so large you can never take account its proportions all at once. Perhaps there’s something impossible about complete philosophical understanding…. What I suppose I’m skirting around is the question: is the critical role gone for the philosopher in the way you envisage it?

S: I can’t think of one.

E: So everything is alright as it is?

S: Let’s say provisionally, yes.

E: And not only do you believe that, but do you actually have an argument to think that it’s right?

S: I think there’s no privileged unique position from which we can organize and understand and if necessary criticize and throw out bits of the current commonly shared human conceptual scheme. There is no such position. On the other hand, of course, we can criticize not so much elements of the scheme as humans actually operate it, but typical philosophical misunderstandings of the way they actually operate it.

E: Yes. But it’s central to that way of looking at it that their operation and philosophical views about their operation are distinct. I mean, very much a rejection of what I said a moment ago, that ordinary people might take up positions on philosophical issues just by doing business.

S: Not by doing business, by carrying on their ordinary transactions with each other and the world; there they use, they deploy, they employ these ideas but they don’t, except in a trivial sense, espouse a philosophical view. They might be said to espouse a philosophical view in a trivial sense, in that as it were they observe, if I may be allowed the analogy, the correct grammar of those terms in using them. This of course is very far from saying that they have any notion of what that grammar is—any reflective and conscious notion of what that grammar is, and this is the role of the philosopher precisely to exhibit it to them.

E: I wonder…. I suppose it’s because I think…. I mean God provides us the precedent for someone who wants to say that human beings can engage in a great deal of talk which they think they understand…

E: Let me try again; how about the soul? ..when you try and lay bare the connections like you’ve done there are connections there, but they might be contradictory connections.

S: There is a distinction here between the case of the soul and the case of God in that you can make perfectly good sense of talking about human beings’ souls in a sense that fits in with the conceptual scheme although it may not be the sort of sense that is attached to it by religious believers.

E: Ah, but the critical question is whether it is the sense attached to it by ordinary people.

S: See, the word ‘soul’ is one among many, isn’t it—mind personality temperament, thoughts, and so on.

E: I’m particularly thinking of the soul as in the possibility of an afterlife.

S: Well so long as it is connected to a set of transcendental beliefs, it suffers from the general deficiency of other such beliefs. So long as it’s integrated with one’s beliefs about human beings anyway, then we can rescue it and make sense of it.

S (lecturing): Surely we don’t learn the meaning of every new sentence independently. The whole point of the fact of the infinite potential is that we don’t need to do this. So out of some limited means or material is generated this potentially limitless understanding. And here we have a problem which engages many thinkers in the philosophy of language today, and is likely to continue to do so for some time to come. It seems it must be solved by crediting us with implicit mastery of a structure of general rules or principles of combination of linguistic elements; a grasp of a grammar or syntax, in fact. And that this together with a grasp of a finite vocabulary of elements which do have to be learned independently and individually, these together contain in themselves and explain the possibility of this limitless understanding. But where shall we look for the abstract underlying structure? Here the logician or semantic theorist inspired by logic may modestly raise his hand. The central consideration, he points out, in understanding the significance of sentences, is a grasp of their truth conditions. To understand a sentence is to know what--

Part 3

S: --expresses, or is capable if given contextual conditions of expressing, and to know this is just to know what we would be believing if we took that thought to be true. Now suppose we understand the basic notion of a true predication—of truly applying a concept in an individual case. The notion schematized in logic in the form of the atomic proposition; you can write these forms as you know and as I’ve done on the board as Fx, Fx,y etc., where the xs and ys are placeholders for the names of individuals and the Fs are placeholders for predicates. Now of course for this structure to have content we also must learn the sense of individual predicates; the difference for example between the truth conditions for predicating ‘is blue’ and those for predicating, say, ‘is square’ or ‘is a fracas.’ And this will be so for any structural explanation. Formal logic on this view supplies the necessary structural key for an adequate semantic theory for any natural language. Now the semantic theorist who thus takes a stand on formal logic is aware that adjustments are necessary before this key will turn. To illustrate the point with a very simple case: take the true attributive adjectives ‘good’ and ‘bad,’ and an indefinite list of nouns, such as ‘hockey-player’, ‘draughtsman’, ‘king’, ‘husband’ and so on. It seems clear that in mastering the semantic force of the 2 adjectives, we also master a very simple rule of combination such that given that we also know the meaning of the relevant nouns, plus the notions of predication and conjunction, we can put the right semantic interpretation on an indefinitely large class of sentences, of which are instances sentences (1) and (2) on the blackboard.

(1) Charles is a good husband and a bad king.
(2) John is a good hockey player and a bad lecturer.

But now if we turn in a naïve spirit to formal logic to help us with the elucidation of the principles of our semantic understanding here, we are immediately in a dilemma. We can’t treat these sentences in the spirit in which we can treat, say, our sentence number (3).

(3) John is a 38 year-old hockey player and a long-haired lecturer.

With this one, number 3, we can just render it perhaps as a conjunction of four simple predications which can be redistributed in any order without affecting the truth-value of what’s been said. We can treat it to the effect as ‘John is 38 years old and John is hockey player and John is long-haired, and John is a lecturer,’ and so on. But to apply the same treatment to sentence (1) and (2) would be semantically disastrous. It would lead for example to the result that sentence (1) is equivalent as regards truth-conditions to sentence (4).

(4) Charles is a good king and a bad husband.

So in general no simple maneuver stands the slightest chance of displaying as subject to the structural rules of logic all those structural features which are at work in generating sentence meanings out of sentence elements. So the semantic theorist who follows this line, who takes a stand on logic, is bound to work hard at recasting whole classes of ordinary sentences in such a way, as he would put it, as to reveal their true logical form. And studying them in their restructured, recast form, he hopes, we should be able to see just how just the same logical principles are really at work in determining their meanings as we see at work in those well-behaved sentences whose meaning is simply mirrored in their surface appearances. And the idiosyncratic semantics and syntax of particular natural languages will on this view be complete when we have framed the clinically applicable rules which will enable us to transform the superficial structures of all sentences of a language into their deep or true logical form.

Now this is one program, and a program to be treated with great respect. But I don’t think we can say that it has an exclusive claim to be treated as the right way to set about the problem.

E: In this lecture we just heard, you characterize a way of doing semantics for a natural language, such as English, the formal logician’s way of doing it; and you hint that this isn’t the way you yourself think that the subject ought to be pursued, though you treat it with a considerable amount of respect. Can you perhaps explain what you think wrong with doing it that way?

S: Well, in general I think it’s unrealistically restrictive. Let’s take an example of the sort that’s been discussed recently. Let’s take, ‘John kissed Mary in the garden at midnight.’ Now this is typical of course of a whole host of sentences in which we say that somebody acted in some way, some time, some when. And part of our understanding of these sentences is knowing that for example, ‘John kissed Mary in the garden at midnight’ can be true only if John kissed Mary period is true. A grasp of this logical consequence, of the inference of John kissed Mary from John kissed Mary in the garden at midnight is part of our understanding of this class of sentences. This is true of the whole lot of sentences of this class. Now it might look straight off as if our understanding of this, each of these sentences, was to be explained by our having hold of a simple rule. After all, it’s in the nature of action that when people act they act somewhere, sometime. So it seems that we can, given an action predicate like ‘kissed,’ always modify that predicate by combining it with a locality specifying expression like ‘in the garden,’ time specifying expression like ‘at midnight’ to form a modified predicate that will be true, which will hold true of just those agents of which the unmodified predicate in ‘John kissed Mary’ is true. I call this form of combination adverbial modification. Here we seem to have hold of a perfectly general and perspicuous principle, our grasp of which is sufficient to account for our grasp of this general logical feature of these sentences. But if you accept the approach that I was critical of, then you can’t accept this; and the reason why you can’t is that the predicate calculus, that is to say the form of logic, of current logical theory which we are talking about and which is regarded as the framework for semantic explanation by these theorists—the predicate calculus simply doesn’t allow for this style of predicate modification. It isn’t catered for in the forms of the predicate calculus. So. what are the theorists of this style to do? Well, what they do is to say that we understand the logical consequence here, we understand this type of sentence, because we really understand it as of a different form from the form it superficially appears to have. And to render this different form in something like ordinary English, one might say that a sentence like ‘John kissed Mary in the garden at midnight’ really has a form like this: ‘There is a kissing which was by John of Mary, and which occurred in the garden, and which occurred at midnight.’

E: Yeah, but the logician doesn’t have to claim that he’s reporting on a stage of mental activity in providing his regimented sentences. What he really wants to say of them is that they represent, in a clearer, more perspicuous form, a form in which the possibilities of inference, of entailment and so on are more clearly realized, and in another way more clearly related to other forms of inference—they represent these better than the ordinary English sentences themselves.

S: Well, of course, by talking of a process I didn’t mean to saddle him with the suggestion that we actually went through this process; only that it was so to speak available to us. Even if this is made quite clear, why is it that we go so remote, so far from the explicit forms that we really understand them to have and instead as it were mask them as having the form that they, on his view, don’t?

E: Well, he might point out the way language has been built up. It wasn’t built up like a French city—or, for that matter, an American city—with the streets laid out before it’s built up. Like a very ancient city there’s a long history, and bits are added on here and there…

Part 4

E: (continued from Part 3)… and fashioned to fit into the sometimes very un-ideal plot that’s already there. I suppose what’s interesting is not merely the negative aspects, the criticism, but the extent to which you have an alternative to the formal semantics given by the formal logicians.

S: I think what we have to bear in mind is what the object of the whole exercise is. The object, isn’t it, is to explain the ordinary speaker’s mastery of a limitless range of sentences; grasp of a limitless range of sentences and their logical relations. Now, it’s …how do we do this? Answer: because we credit him with mastery of a limited set of rules of combination or styles of combination and elements.

E: Well to give—yes: let me think again through this by suggesting another example and seeing what you would say of it again if I can see the blank wall that I saw gazing me in the face again. The old problem about ‘good’ and ‘bad’ and all that sort of thing—; here’s the problem: you get--

S: I discussed that in my lecture. In other words, what solution is to be offered to the problem of attributive adjectives of an evaluative sort? You could also admit this category too; admit the category of evaluative or appraisal words, point out with respect to them that the criteria for their application depend upon the type of thing you’re applying them to. If you take these two facts together, you’ll see that you won’t be able to know the truth conditions for calling an x a ‘good x’ unless you know what sort of thing an x is, or what the value of x is here.

E: But with those—I mean, taking ‘good knife’ for example, which in a sense raised a good number of the problems which ‘good king’ raised—would you suppose that to be some sort of algorithm for getting the meaning of the compound, ‘good knife’ from that of the element ‘knife,’ together with what you said about ‘good’?

S: Suppose you took ‘king’ as a sort of functional term like ‘knife’ so that in a sense you haven’t grasped the concept ‘knife’ unless you know what properties it’s required, by those who require knives, to have. You might say ‘king’ was like this—he kings it well; he’s just, and so forth. But of course what complicates the matter with these evaluative predicates applied to social phenomena is that your standards may change; I mean elements of evaluation may change over time. So it isn’t quite a straightforward semantic issue, in the way that perhaps the combination of an evaluative word with a purely instrumental object like a knife might be.

E: At least for that class of cases it wouldn’t be true that you’d need under the entry for ‘knife’ in the dictionary some statement as to what properties you’d need to have to be a good knife.

S: Well, let me ask you—do you think you understand ‘good oscillograph’?

E: I’m going to have to reveal my ignorance.

S: But surely you want to say, in a way, you understand it perfectly well. There isn’t a clear yes or no answer here, is there?

E: Well, when it becomes a yes or no answer, when it becomes interesting I suppose is to say that it isn’t just a trivial matter what would count as understanding; when it really becomes interesting is whether we say this is part of the province of explanation for semantics. That’s really what I’m getting at. What is an oscillograph, by the way?

S: I haven’t the vaguest, myself.

E: The reason I’m worried about that is this sort of point. In some sense I want to say that in coupling a term like ‘cat’ with a term like ‘run,’ in order to understand it you’ve got to know what state of affairs is being asserted to obtain, and—I can’t put this very clearly, but do you see what I’m getting at?

S: Suppose you took the expression, not ‘cat runs,’ but ‘water runs.’

E: Or ‘nose runs’!

S: If you wish. If you think that the primary application of the word ‘runs’ is to some creature with legs and they twinkle and that’s the matter of the thing running, or whatever their particular movement looks like. But obviously—or at any rate, plausibly—when you make the jump from ‘[an animal of any species] runs’ to ‘water runs,’ there is a real inventiveness here, a real extension of your vocabulary, which is in a sense rather mysterious; because it isn’t already covered by the semantics of ‘runs,’ and yet it’s graspable—it’s humanly graspable.

E: And the critical question is, is this something that semantic theory should try and explain, or can it explain? That’s really what I was getting at, you see.

S: But what does seem to me important and interesting is that there really aren’t any rules here—

E: Well, in a sense there can’t be!

S: Whereas in the sort of semantics we’ve been talking about the whole idea is to explain human capacity by reducing the styles of combination we have to general, graspable rules. It seems to me in this sort of example, which is perhaps a relatively trivial extension of something much more important in human thought, rules are not to be had.

E: Of course what happens when rules are to be had, that is when the metaphor becomes dead and conventional, it ceases to draw upon the imaginary capacity we’re talking about. But you see what’s interesting—well, one of the things; I made the case rather difficult for myself by starting with ‘man’ and ‘cat’ because in a sense I want to say you can’t draw the line where this extension stops.

S: Surely, when the child learns, he doesn’t learn an abstract specification for running, like ‘legs moving relatively fast in relation to each other.’ He learns in relation to a dog, perhaps, or himself; and he does make this extension to cat, and even in that, you might say, there’s an element of imaginative creativity or something of the kind. That ties in with something I said in my lecture which seemed to arouse some sort of response: I said that what you can’t say, you can’t think. And there's something not quite right about this because of course we as it were constantly enlarge the boundaries of our language by these creative or imaginative sets. It’s still in a way true, but it doesn’t mean that we’re bound within the language as it is in any particular stage.

E: No, and thank God for that! I mean, what an awful and nightmarish possibility it would be if that were true. And interestingly enough there’s another fact—

S: Actually, how misused the word ‘creativity’ is by those linguists who refer precisely to our mastery of an indefinitely large number of sentences by virtue of our mastery of a certain restricted set of rules—and say this is creativity. It isn’t creativity at all!

E: Exactly! Creativity precisely comes in where the rules give out!

S: Or where you stretch them.

E: Well, I mean, actually we don’t want to fiddle over the word ‘creativity’; in a sense they can have creativity because in a sense we know what they’re talking about; one can understand new sentences, but—

S: Why give them a good word? Now, if we ask ‘what is fundamental to thought at large?’ not thinking of the subject matter on one hand, or logic on the other, I think we’re bound to say that the most fundamental feature of all is a certain dichotomy, or duality perhaps better, of on the one hand a notion of the general—the universal, the repeatable in multiple cases—and on the other hand, the individual case, the particular case. And obviously this dichotomy is represented in logic, is represented in the basic notion of predication, of saying of an individual case that it’s of such and such a character. In logic this notion has an entirely general, subject-free character. Now might we not see the unlimited generality of logic—the notion that it’s totally indifferent to what you’re talking about, be it numbers, or objects of an ordinary kind, or not—as arising out of an extension, a kind of imaginative extension, from this basic case on the strength of certain analogies, not of a subject matter kind, but of a logical kind; the extension from this basic case to other more rarified cases. So that we can take numbers as individuals and talk of the properties of numbers; so that we can take, if you like, actions as individuals and talk of the properties of actions; so that we can take propositions as individuals and talk of the properties of propositions, and so forth.

E: And the examples that come to mind do have such a character, don’t they?

S: So that some sort of exercise of as it were human imagination is involved in making this step—

E: So—

Part 5

S: in the sense that we have an analogical extension of forms, this time, rather than semantic contents, beyond their primary application.

E: Well, one day—as one likes to dramatize these things—something of this form was used to express a proposition of a rather different kind. And in order for the hearers of this new proposition involving making new uses for old means, they had to be imaginative in some of the ways we’ve been talking about to understand what was going on.

Focusing on that move, isn’t there a danger that he could only understand what was going on by incorporating error into his thought—somehow thinking of the subject too much like the subject of the other kind, let’s suppose the sentence is something like ‘his proposal was defeated’?

S: I don’t think that there has to be any sort of error made by the actual people who talk in this way. What I do think, and I think this is a unique thing about the history of philosophy, is that the constant charge of reification—of treating what aren’t things as if they’re things—is justified insofar as it is justified by philosophers—not ordinary men, but philosophers—taking these derivative forms too seriously, and at the same time being haunted by the original model. If the basic entities, the basic individuals, are spatiotemporal objects or spatial objects, then if you take the derived grammatical forms too seriously, then you may suppose that properties, propositions, all sorts of things, have a quasi-substantial character, which they don’t have. But I don’t think, so to speak, that this basic error which philosophers can make is implicit in the language; I think it’s a philosophical error.

E: So it would be as wrong in these kind of cases to suppose that the ordinary man is mislead as it would if you took some of his metaphors for talking about his mind, like ‘in the back of my mind,’ and all that sort of thing—

S: Right. There of course as soon as he starts to reflect—and as I said, the tendency to reflect philosophically is endemic in the species; good!—as soon as he starts to reflect, maybe he’s liable to fall into these traps. As Wittgenstein pointed out, we are mislead by surface grammatical analogies, by metaphors, to bring the two points together as you did. But we’re mislead not when we use these things in ordinary exchanges, but when we start to reflect.

E: Do you think that line can really be drawn so clearly, actually?

S: Yes! Yes, I do. What’s wrong with the philosophers of the critical kind, is that they over-correct, as it were. Because they see that there is a theoretical risk here they—

E: They try to provide translations of them.

S: Yes, they try to provide translations of them—

E: And you can’t.

S: So in a way you can see the whole new move in theoretical semantics as a very old thing, fished up again.

E: I suppose it would follow from our speaking of this in the ways that we have—imaginative, creative—that those translation procedures just aren’t going to be available in general.

S: So they’re caught in a terrible dilemma. Terrible for them, but amusing for us.

E: One of the ways it’s efficacious—now, I’m not sure this is what you had in mind, but maybe by seeing how little this gets your thought you can tell me the rest of it—is that of course it enables us to use—I mean, we’re talking about the extension of a grammatical form—it enables us to use, coupled with the extension of the new term, in subject position, some term like ‘his proposal,’ ‘his defeat,’ and so on, a whole lot of already existing predicates could now change.

S: Good, this is excellent. So this is where the exercise of imagination as you put it, which is involved in the adaptation of a grammatical form for a new purpose, goes along with and is married to that extension of particular predicates from their original application to their new application. I mean a defeat is something that happens to somebody in a battle.

E: Not in a proposal.

S: Not in a proposal.

E: I wonder whether we couldn’t—I don’t know, maybe that’s too bold.

S: So this is extraordinarily interesting; the notion that there’s a union here of these two kinds of extension—the semantic extension, which you earlier illustrated by talking about water running on the one hand, and the adaptation of the grammatico-logical form to receive new inhabitants. There is a continuity here, isn’t there, between things which we are accustomed to regard as discrepant or distinct, but which linguists or theorists of language, who are concerned after all with all uses of language, should be able to cope with. We spoke earlier of those as we called them of those imaginative extensions of the uses of ordinary predicates, ordinary words; and we considered the example of extending the word ‘runs’ from cats or animals to water, and I think rightly called this an imaginative step. And couldn’t it be said that it’s just the very same sort of imagination at work in characteristically poetic uses of language, in figures and tropes such as poets use. I mean, to see water as running is as it were the same sort of thing as, to take a childish example, to see a cloud as a camel. Or, to take a sophisticated example, to see the water, the river, as a strong brown god, or something like that, as in Eliot. But in general what one would like to stress is the kind of continuity here in use of language, and in thinking in general, between the poets view of the world and the ordinary person’s vision of the world.

E: Yes, so poetry is contained as a possibility right in the germ of all language.

S: Yes, and otherwise it would be utterly mysterious how we could understand the poet’s tongue if this were not so, if the poet’s tongue wasn’t after all our own tongue in a rarefied, refined, and intensified form.

E: Part of the thing that might be accounted for by looking at it in this way—reversing back now to what we said about the meanings of the composite is based by a rule on the meanings of the two parts—the characteristic ambiguity of poetry, because it isn’t yielded; I mean, you couple two terms, ‘water’ and ‘run’ perhaps (I can’t think of one that has a very obvious ambiguity at the moment)—the output isn’t yielded automatically by the semantic content that you have already. And that’s why you have this ambiguity—

S: And one’s inclined to say that no output is yielded automatically, or almost none. This refers to a point that you made, and that after all Hume made too when he described the imagination, and attributing to this faculty the key role in the application of words at large, he described this as a magical faculty in the soul, which is properly called a genius—he used these highly elevated forms of words to talk about the ordinary person’s capacity to apply ordinary general words over a range of similar, and also quite dissimilar cases. And it as you might say is an extension of just the same faculty which you might find in poetic use of language.

E: And do you think the mystery he found in it was the extent to which it isn’t rule-determined?

S: Right, I think that’s so. That in a sense the rules escape us. And this in a way is equivalent to saying that we can’t reduce it to a rule, or not to a mechanical rule; and yet it’s a faculty we must all be possessed of for the most, in a way, prosaic exercises of the mind and uses of language as for the most obviously and conventionally poetic.

E: What comes to mind is that perhaps a lot of the emphasis placed on rules as an explanatory concept in our understanding of language has been actually been misplaced.

S: And of course it’s not only the matter of the extension of the trope in the individual word, the figurative use of the individual word, but—I suspect I haven’t been into this at all—there is a similar imaginative use of grammar in poetic language. But this is a subject of investigation; I have no views about it really, except that I suspect it might be so.

[Part 6. Walking—Credits.]

E: Do you remember that footnote in Austin’s whatsit, a lovely thing that use to puzzle me, let’s see if I can remember it—why is beer, or water, in a toy beer bottle not toy beer, but pretend beer?

S: Aha.

E: So if I was being really belligerent on some occasion I would say,

S: You’ve really complicated the setting because the whole phrase ‘toy beer’ is odd, like ‘toy wine,’ ‘toy wool’—

E: But it wasn’t real beer, you see, it was water! That was the point!

S: Yes, yes, yes. The point still holds.
apolliana: (Default)
2013-06-09 03:20 am

The Generality Constraint's Strawsonian Origins

I was reading along in Individuals, when I came across as passage that stated the way I'd been thinking of the Generality Constraint. I thought I was perhaps seeing what I'd been primed to see, as this passage was a footnote to an argument against what he calls the "no-ownership" or "no subject" view of the self (the view that there is no subject of mental states--that there seems to be is only an illusion); but it is indeed the passage Evans cites when he introduces the Generality Constraint.

Strawson says, "When I say that the no-ownership theorist's account fails through not reckoning with all the facts, I have in mind a very simple, but in this question a very central, thought: viz. that it is a necessary condition of one's ascribing state of consciousness, experiences, to oneself, in the way one does, that one should also ascribe them, or be prepared to ascribe them, to others who are not oneself" (99).

The footnote:

"I can imagine an objection to the unqualified form of this statement, an objection which might be put as follows. Surely the idea of a uniquely applicable predicate, i.e. a predicate which belongs to only one individual, is not absurd. And, if it is not, then surely the most that can be claimed is that a necessary condition of one's ascribing predicates to a certain class of individuals, i.e., oneself, is that one should be prepared, or ready, on appropriate occasions, to ascribe them to other individuals, and hence that one should have a conception of what those appropriate occasions for ascribing them would be; but not, necessarily, that one should actually do so on an occasion.

"The shortest way with this objection is to admit it, or at least refrain from disputing it; for the lesser claim is all that the argument strictly requires, thought it is slightly simpler to conduct it in terms of the larger claim. But it is well to point out further that we are not speaking of a single predicate, or merely of some group of other predicates, but of the whole enormous class of predicates such that the applicability of those predicates or their negations defines a major logical type or category of individuals. To insist, at this level, on the distinction between the lesser and the larger claim is to carry the distinction over from a level at which it is clearly correct to a level at which it may well appear idle and possible senseless.

"The main point here is a purely logical one: the idea of a predicate is correlative with that of a range of distinguishable individuals of which the predicate can be significantly, though not necessarily truly, affirmed (99)."

In the context of the section of the chapter in question, Strawson is talking about mental predicates, e.g., 'is in pain." So it is clear that he doesn't think it should make sense to say that, e.g., a rock is in pain. He is concerned with explaining why mental predicates are properly applied to persons, rather than to unminded bodies or disembodied minds. I'm happy to accept that if I can sensibly self-ascribe a mental state, it will be logically possible to imagine ascribing it to other subjects to which mental predicates may be ascribed.

Strawson's argument against the no-ownership view is that if it makes sense to ascribe a mental state to oneself, it makes sense to ascribe them to others, in principle. And the no-ownership theorist is committed to ascribing mental predicates to himself insofar as he talks of "my experiences" as things that can't properly be ascribed to a self.
apolliana: (Default)
2013-06-06 03:08 pm

A transcendental argument for Evans's location requirement for successful reference using 'I'

I'd been thinking that one argument Evans might but doesn't make for his requirement that to successfully refer to yourself using 'I' you must possess the ability to locate yourself in space is the following: without space, there would be no other selves and thus no need to distinguish between yourself and other entities.

This of course is implicit in what he says about Strawson's "Sounds" chapter, where Strawson himself concludes that although we can achieve an analogue of spatial distance in the sound-world via the master-sound, we can't as easily make sense of distinguishing between oneself--Hero, hearing the sounds--and other entities. (And very possibly one reason it makes more sense to do this in a spatial world is that we can close our eyes and thus separate ourselves from the things we see.)

I had been thinking this, without knowing that Strawson himself suggested such a thing, at the opening of the subsequent chapter, "Persons":

"We drew a picture of a purely auditory experience, and elaborated it to a point at which it seemed that the being whose experience it was--if any such being were possible at all--might recognize sound-universals and reidentify sound-particulars and in general form for himself an idea of his auditory world; but still, it seemed, he would have no place for the idea of himself as the subject of this experience, would make no distinction between a special item in his world, namely himself, and the other items in it. Would it not seem utterly strange to suggest that he might distinguish himself as one item among others in his auditory world, that is, as a sound or sequence of sounds? For how could such a thing--a sound--be also what had all those experiences? Yet to have the idea of himself, must he not have the idea of the subject of experiences, of that which has them? So it might begin to look impossible that he should have the idea of himself--or at any rate the right idea. For to have the idea at all, it seems that it must be an idea of some particular thing of which he has experience, and which is set over against or contrasted with other things of which he has experience, but which are not himself. But if it is just an item within his experience of which he has this idea, how can it be the idea of that which has all of his experiences? (Individuals, 88-9).

This is an argument to the conclusion that the no-space world is solipsistic. If there is no space, there will be no use for 'I.' We might also infer that if there is a use for 'I', there is space. It would seem to be a further step that one be able to locate oneself within that space.

Evans is also concerned to explain how our indexicals and demonstratives link up with the objective world. And thinking about ourselves as positioned in the objective world, he thinks, requires knowing something--which, I believe, is actually quite minimal--about where in that world one is. I take this as roughly meaning that one has the capacity to use a map, even if one doesn't currently have the knowledge it would bestow upon one. It is the requirement that one be able in principle to link up one's egocentric space with objective space. This might have the consequence that small children who do not yet possess this ability do not refer to themselves in quite as rich a way as adults do (if indeed we wish to describe it as successful reference), but that it a consequence that seems correct.
apolliana: (Default)
2013-06-04 01:12 am

Conclusions from the World of Sounds

"if Hero is to think of unperceived particulars existing simultaneously with, and in relation to, perceived particulars, he must have simultaneous spatial concepts, and not those that turn essentially upon change" (Evans, "Things Without the Mind," Collected Essays, 290).

"...It is not the opposite of Capgras syndrome, but a natural consequence of the same cause: a loss of the sense of a unique whole. Such 'delusional misidentification' applies not only to people, but to objects: another patient of mine began a vendetta against someone who, she believed, had entered her bedroom and changed all her clothes for copies of a slightly inferior quality. It can even apply to places: one individual held that there were eight 'imposter' cities, duplicating his own, and said he had spent the last eight years wandering between them, without finding the real one. There were also eight duplicates of his wife and children, each duplicate living in a separate duplicate city with a double of the patient" (McGilchrist, The Master and His Emissary, 54).


1. Objects in the World of Sounds

In "Things without the Mind," Evans examines Strawson's famous exploration in Individuals of the possibility of a world wherein the only sensible qualities are sounds. This is an indirect way of testing whether we need space, as well as time, in order to have a conceptual scheme recognizable as our own--a conceptual scheme in which "it is logically possible that such [auditory] items should exist whether or not they were being observed, and hence should continue to exist through an interval during which they were not being observed" (Strawson, Individuals, 72) . (Of course, in reality, sounds do require space in order to travel; but if one lost vision and touch, the world one would thereby inhabit might seem purely auditory and spaceless.) Both Evans and Strawson think a thought experiment of this kind shows that space, or something that performs the role of space in allowing for simultaneous instances of the same universal, is in fact necessary for a conceptual scheme like our own, though Evans thinks Strawson makes his argument--to the extent that it is an argument, and not simply an exploration--weaker than it could be through the use of a sound against which to locate all other sounds, the "master-sound." The master-sound, Evans thinks, is too obviously weak, as it does not allow for a distinction between qualitative and numerical identity, in the following way.

Imagine that you are the person Evans calls Hero (the 'hero' of our tale about the sound-world), who inhabits the purely auditory world. How do you tell whether the sound you hear at time t1 is the same sound you heard at t0? There is, in the background, a single continuing master-sound, which changes in pitch; other sounds can be 'located at' different pitch levels of the master-sound, with which they correlate. The pitch-level of the master-sound, and your memory of the other sounds and the pitch-levels of the master-sound with which they correlated, is your only 'landmark.' There is, Evans thinks, no way of deciding whether a sound S2 is a new instance of a previous sound, or another sound altogether. Is it "one of those" again, a token of a certain type, or a distinct entity? As sounds occur in serial order, Hero can never directly compare two sounds to say whether or not they are of the same type. Further, he can never say they are completely identical. They aren't; they occur at different times. The patient with right-hemisphere brain damage in my second epigraph seems to suffer from the inability to consider identical two 'instances' of a person, or a place; City at t1 and City at t2 are, for him, different cities. Inhabitants of the normal world do not usually suffer from this problem, however, as we have the ability to think about objects continuing to exist, and to be the same individuals, though some of their observable properties may change. (McGilchrist believes that this dysfunction points to a role for the right hemisphere of the brain in recognizing and reidentifying individuals.)

The largest part of Evans's essay is taken up with an attempt to say what is necessary for a conceptual scheme that allows for simultaneous instances of the same universal. Someone who lives in a purely serial, purely auditory world, he thinks, will not develop a theory that allows for primary, non-relational properties of objects. That is, the auditory world can only tell you about what exists insofar as you hear it; you are a Berkeley-an ear--the world is only what you can hear of it. Both Strawson and Evans gesture at ways one might draw conclusions about where sounds might be when one isn't hearing them (one sound drowning another out, for example), but neither think they're very compelling. Sounds don't really seem to demand such a theory. The spatial world--at least insofar as we see it--does. Evans does not make explicit why he thinks this is so, but I think the implied answer is, distance--particularly visual distance. If something can get farther away, yet remain itself, surely it can go out of view and remain itself. But while this feature of our visual world seems to prompt the development of a naive physical theory in which unperceived objects can continue to exist, Evans seems not to want to tie concepts of primary properties of objects to any one sensory modality. For this reason the congenitally blind can have such concepts; if primary properties are not limited to connection with a single sensory modality, it shouldn't matter whether one modality is lost: "it is no more possible to have a purely sensory concept of hardness than it is to have a purely kinaesthetic conception of what it is for one's legs to be crossed, or to have a purely muscular conception of the motion of one's body, or to master the concept of electricity solely by learning to recognize electric shocks" (270).

An inhabitant of the auditory world would have no reason to believe sound 'objects' exist when no one is around to hear them, since he has no concepts of 'primary' sound properties--properties not inherently dependent on their being heard. This, roughly, is Evans's conclusion.

2. Objects in and of the Mind

What conclusions from the sound-world can we draw about the way we identify and reidentify our own mental states? It seems, few, as we are not inhabitants of such a world; we do have access to things we more properly call 'objects.' The problems that plague Hero--not being able to distinguish numerical from qualitative identity, not knowing whether or how to divide a process into distinct events--still arise insofar as we aim to identify particulars to which others have only indirect access, and which we know mainly in serial order. But I don't think we need be so deeply troubled as Hero might be (or as we ought to be on his behalf), for this is not the whole of our world.

For bodily sensations, emotions, and thoughts, minus any references they make to the outside world, one will have to decide, based upon one's purposes, whether to consider Sensation1 and Sensation 2 instances of a single type. Since they occur at different times, they will never be thoroughly identical. But it is much less problematic for normal humans not trapped in a purely auditory universe, as we can simply borrow and adapt our ordinary ways of handling instances of universals, to suit our purpose. There is no reason we cannot refer demonstratively to our mental states (this pain, that sadness).

Keeping track of them makes slightly less sense, however. One can make successive reports tracking the continuation, increase or decrease of some sensation (e.g., anxiety), but--absent rigorous Buddhist training--it's hard to see what it could mean to 'keep one's eye on it' in the same way one keeps one's eyes on an object in one's visual field. (However, the 'keeping track of' function is not so literal with indexicals: one doesn't keep one's eye on the day with 'today' in anything like a literal sense.) If I keep my eye on my current mental state and note that the anxiety has lessened, it seems what I'm keeping my eye on is not my anxiety, but my current mental state. I can't follow the anxiety offstage and see where it's gone. (Sensations are not like rabbits.) Thus the 'reidentification' of mental states can never be literal, as it can be with rabbits and other space-occupying entities. I might identify another instance of a state I've experienced before, and I may say--"it's that otherworldly feeling again," but it is not literally, numerically, the same feeling. So talk of reidentification will only be of state-types. The only thing one can continuously monitor is one's current mental state, and that is too vague to admit of reidentification.

Sensations, thoughts, emotions and sounds are not--obviously--the only particulars in our universe, if indeed we wish to call them particulars at all. They are dependent upon the persons in whom they are located, and we refer to them as particulars, rather than as instances of types, very rarely. In what sense, then, can we refer to them on an Evansian view? In Varieties of Reference, mental states are treated in the chapter on self-reference, so I believe the idea would be that we refer to mental states as states of persons. This will say enough to position a mental state referred to non-demonstratively in the objective world. It will also say enough to refer to someone else's mental state demonstratively ('that otherworldly feeling you keep having'); you need to know it's mine, and that I've had it on certain past occasions. For demonstrative reference to one's own mental states, it is less clear whether to use the requirements on the use of the indexical 'I' to refer to oneself, or the requirements on using 'this' and 'that' to refer to objects. In any case, the stricter of the two sets, the requirements for 'I,' would probably be met, so long as one possessed the ability to locate oneself in some way in space; the information- and action-links are clearly fulfilled if in fact the states to which you're demonstratively referring are your own mental states, barring complicated multi-brain cross-wiring cases.

Objectively speaking, mental states will be located at the persons whose states they are. This requirement seems reasonable enough, as we can't conceive of mental states occurring outside of minds. We will not be able to say, "If you were to go to position x [Bob's mind], you would experience sadness," however, as they are not events everyone can view in the same way; though perhaps we can say that if you go to Bob, you can discover sadness there."

apolliana: (Mum)
2012-11-10 03:01 pm

The Nonconceptual Content of Dreams

That the content of dreams might be nonconceptual is suggested by the regularity with which I remember them, but cannot describe them, even to myself. I find myself saying things like,

"It was a tree, but also a house; and you were there, but you were also Bill Clinton, and then you disappeared; and we were in a theme park, then airport, then country house in Italy, and everyone was naked and on drugs."

And invariably the description will seem to miss the experience. This happens with descriptions of ordinary experiences, too; but dreams even moreso because of the apparent inability of the scenery and characters in dreams of fitting into conceptual categories. There's no question that "it was a tree, but also a house" will fail to miss what it was like. The strange superimposition of things in dreams takes care of this. Places and people are often composites; and the way in which they are composited is difficult to describe. It's possible that some cognitive attitudes come in here: "knowing" the identity of someone in a dream, while in fact they look like someone else.

But perhaps we might want to say there are concepts in play, but they don't fit things properly because the relevant things are ambiguous and mutable. But such are my paradigm cases of nonconceptual content: the object in the dark room that looks suddenly too strange for recognition. In that case there is something you're looking at, but it does not immediately present itself to you as falling into a category smaller than 'something.' The content of the experience, phenomenologically, is an indeterminate object. In dreams it's a raging river of them.

But is "an indeterminate object" nonconceptual? Object individuation is partly involved; a thing can be discerned. Can it be distinguished from all other things? Only partly, as the perceiver doesn't know what it is. Is a thought about it Russellian? Yes; where there is no thing, there is no thought, regardless of the indeterminacy of the object. However, when the thing is perceived in proper lighting, the previous thoughts about it as indeterminate will lose their force.

Very often I decide the content of one of my dreams cannot be the content of a thought. I give up on describing it. I decide whatever it was doesn't suit the content of a thought; whatever the thought purports to be about is not the sort of thing to sustain the thought. There is no thought there. Perhaps the combination of fluid ambiguity and being past-tense (not something I'm presently looking at) renders the content of dreams less thought-apt than even the vague object in the dark. Perhaps this is because demonstratives don't work in dreams (or in past-tense accounts of them). We cannot use 'this', 'that', 'I', and 'here' where none of those work as they normally do. Demonstratives themselves are undermined in a world that cannot be still. Even if the properties that won't be still are conceptualizable properties, demonstratives--which do not seem to rely on concepts--rely at least on their stability, or on at least a backdrop of stability, so that if the object itself is not stable at least the environment is.

Do dreams have nonconceptual content? They have content, which is not apt to be the content of thoughts. But it is less apt to be the content of thoughts than the content of indeterminate experiences, as demonstratives don't quite work. And it is less apt to be the content of thoughts than non-indeterminate experiences, because the qualities of things that go along with their conceptualizable properties aren't stable. But the sense in which the contents of dreams cannot be contained in the contents of thoughts partly parallels the sense in which the same is true of non-ambiguous experience, of which the detail, the experience, cannot be conveyed. But in dreams there's nothing else. So many a dream-thought turns out not to have been a thought at all.