Tuesday, February 18, 2014

Winch on social science

Wikipedia has a helpful page on Peter Winch. It quotes two passages from The Idea of a Social Science and its Relation to Philosophy which help to summarize what that book is about. These quotes include the following lines: 
to be clear about the nature of philosophy and to be clear about the nature of the social studies amount to the same thing. For any worthwhile study of society must be philosophical in character and any worthwhile philosophy must be concerned with the nature of human society
our language and our social relations are just two different sides of the same coin. To give an account of the meaning of a word is to describe how it is used; and to describe how it is used is to describe the social intercourse into which it enters. 
This sounds wrong, or at least not quite right, to me. Of course I must re-read the book before reaching any conclusions about it, but it seems worth trying to articulate some ideas to have in mind while re-reading it.

The first quotation obviously involves ideas about what is worthwhile, so Winch is making a value judgment rather than simply stating objective facts. The second quotation seems to involve a mistake, or at least something I disagree with that I don't consider to be a question of my having different values from Winch's. To give an account of the meaning of a word is to describe how it is used. This strikes me as partly true, but it ignores the part of Investigations 43 in bold here: "For a large class of cases--though not for all--in which we employ the word meaning it can be defined thus: the meaning of a word is its use in the language game." It (also) ignores essentially contested concepts. In short, Winch makes philosophy sound too much like a science.

But I mean only that in these brief passages he does that. Re-reading them in context might show them in a different light.


  1. Winch seems to be saying that social science and philosophy are investigating the same thing. But is he also saying they have the same point?

    Also, is there for Winch a connection between studying the social character of language and studying the use of a word? Is that implicit in what he says in those two quotations?

    1. I think the answer is yes in each case (with a very tentative yes because I need to re-read the book). With regard to whether social science and philosophy have the same point it looks as though he means not that they do have the same point but that they ought to. "[A]ny worthwhile study of society must be philosophical in character." That sounds as though it means that social science ought to be, or be replaced by, philosophy.

  2. Does that mean, for instance, that prediction of behavior should not be part of the social sciences, according to Winch? (Or is this part of philosophy?)

    1. I think the answer is No, but I really need to re-read the book. As I remember it he thinks that what social science aims at is understanding the meaning of human behavior, and that this is philosophy's job. But I'll have to look and see whether he says anything about prediction.

  3. check out: http://www.ashgate.com/isbn/9780754647768

  4. Hi -

    One way of putting the basic idea of On the Idea of a Social Science and Its Relation to Philosophy is that whereas natural science is in many respects a 'one-level' affair, such that if you understand the behavior of physical objects (the way they move, the way they respond to force and energy inputs, etc.) you have all the information you would need to understand or to construct a theory of such objects, social science is a 'two-level' affair, where in addition you need information about how the 'objects' (human beings) exhibiting that (physical) behavior understand themselves to be doing - their self-conception as it were - in order to understand the behavior itself. This is part of the reason Winch spends a considerable amount of time criticizing Pareto, Mill, etc. - he takes them to miss this and so to think that social scientific regularities can be dealt with in more or less the same way that natural scientific ones could be. Winch thought, though, that there was a difference, that the sense of people's physical behavior could only be made out against the background of what those people took themselves to be doing.

    There are thus some useful points of contact between this work of Winch's and e.g. Habermas' On the Logic of the Social Sciences - although by no means do the two books embrace the same theses or arguments, nonetheless both champion a position on which what might be called 'reflexivity' is essential to social scientific understanding.

    It should be noted that On the Idea of a Social Science and Its Relation to Philosophy is Winch's MA thesis and among the first (perhaps the first) substantial work he published. The essays collected in Ethics and Action are really useful for understanding the book, both in terms of the developments out of it that led to Winch's reputation in some circles as a principled 'relativist' and in terms of his own withdrawal or modification of some of the book's key claims, changes in his position which largely went in a direction congenial to the kind of later-Wittgensteinian concerns Dr, Richter raises in his blog post.

    Winch's thought continued to develop interestingly throughout his life. If you check out some of his best later works - e.g. "Waiting for the Vanishing Shed", his APA Presidential Address "The Expression of Belief", or his wonderful book on Simone Weil you will see a continuing deepening of his views over his entire career from the earlier works mentioned above.


    Sean Stidd

    1. Thanks very much, Sean (if I may). This is extremely helpful.

    2. Sure thing! Glad to see a post on this.

    3. Hi Sean, Duncan, Tommi,

      Sean invited me on over here to take a look. (He and I studied with Winch in the 90s in Illinois). Without using the "two-level" language (I don't know if that's an adequate account of the natural sciences), I would endorse Sean's view that his concern was to explore the different ways we think of or enact "understanding," and that this continued to be a concern in later works. Fundamentally the argument of ISS is against misconstruing the "logic of" human action as the same as the "logic of" physics. Using that phrase in the lovely mid-century British sense that is something like "how we make sense of, explain, and reason using the concepts of..."

      I re-read Charles Taylor's "Interpretation and the Sciences of Man" last year and I'd probably be hard-pressed to identify the difference between that Taylor and the Winch of ISS, though that may be an artefact of my own philosophical autobiography.

      As Sean and Tommi say, it was a young man's book, and Winch had decades of development in his thinking after that point. The preface to the second edition makes something like your second point, Duncan.

      Contemporary social scientists who take a page from Wittgenstein include Clifford Geertz from Anthropology and Davide Nicolini from Sociology or organizational theory.

      About your first point, Duncan, I nurse an anecdote that's interesting to me and perhaps pertinent to your observation. Late in my time at Illinois I asked Winch whether a Wittgensteinian perspective committed one aesthetically to prefer Tolstoy to a modernist writer like James Joyce (or...name a post-modern writer). I meant it as a reductio, but he embraced the consequence..."yes, perhaps it does."

      That's as reliable as human memory in general, which is not very reliable. Perhaps he just looked at me and said a non-committal but moderately intense "hmmm," as he often did, and it either meant he was taking seriously what you said, or it meant that he noted you had finally seen what he was saying, or it meant his thought had moved on to other, more important things, and he hadn't really heard you.

      He would quote the line reproduced in Culture & Value, "My ideal is a certain coolness. A temple providing a setting for the passions." This is from 1929, but it's worth asking whether this ideal is part of the ideal of philosophy as descriptive, leaving everything as it is. One question is whether this "setting for the passions" means purely scientifically descriptive, without value judgments. Another is whether neutral philosophical arguments force one to this descriptive philosophical method. I don't see Winch as thinking that the "description" philosophy engages in is neutral, or that there's a value-neutral argument that philosophy has to be descriptive in this way.

      It's an interesting question whether he ever thought that philosophy and the social sciences were or could be descriptive in the same sense. (Or whether one would, continuing the same line of thought.) The "description" of philosophy is always deeply shaped by the describer's interest in combatting metaphysics, dissolving a puzzle of thought. This is not sociology's preoccupation.

      I think that's something the Wikipedia entry on Winch fails to clarify. He was deeply and un-repentantly Wittgensteinian in his approach, but not a Wittgensteinian in the sense of an exegete or in the sense of continuing Wittgeinstein's train of thought himself (I mean the particular engagement with "philosophical logic"). He thought about his own philosophical problems, typically through reflecting on other philosophers (Rousseau, Plato, Spinoza) or through literature. And he challenged his students to do the same thing.

      - Lynette Reid

    4. Thanks very much, Lynette!

      This too is very helpful. Sorry it's taken me so long to respond.

    5. Late in my time at Illinois I asked Winch whether a Wittgensteinian perspective committed one aesthetically to prefer Tolstoy to a modernist writer like James Joyce (or...name a post-modern writer). I meant it as a reductio, but he embraced the consequence..."yes, perhaps it does."

      This is exciting. For the past year and a half or more, I have been working on a manuscript that keeps getting larger and larger, on some potentially problematic issues in Winch's and D. Z. Phillips's conception of philosophy (in relation to Wittgenstein's remark on the cool temple, for instance). This is extremely relevant for what I'm trying to do there. Thanks! I'll just try to keep in mind your disclaimer about the fallibility of memory.

      And it's very nice indeed to hear from you. For a long time I have heard nice things about you (e.g. from David Stern), but I have never heard from you. I only know – and appreciate – your paper "Wittgenstein's Ladder: The Tractatus and Nonsense".

  5. "For a large class of cases--though not for all--in which we employ the word meaning it can be defined thus: the meaning of a word is its use in the language game."

    I take this to be saying that some uses of "meaning" don't ask linguistic questions. Such as, "what is the meaning of life"? Here we do not want to know the meaning of the word "life." Instead, this is an existential question. So, for all cases where meaning does pose linguistic questions, meaning is use.

    1. Perhaps something like this in some way. But the example is unworkable, as the German word Wittgenstein uses is Bedeutung – and not Sinn, which would be the word for existential meaning. "The meaning of life" is der Sinn des Lebens in German. One can speak of die Bedeutung des Lebens, but it would then mean 'the goal of life' or 'the import of life'.

      Frege's "Über Sinn und Bedeutung" is usually translated into English as "On Sense and Reference". Using this translation, what Wittgenstein speaks of in this passage is reference exclusively.

    2. I read some importance into who the "we" are—"in which *we* employ the word meaning," I read it as we people who are puzzling over philosophical and logical questions. We who are looking to that concept (Bedeutung) to do philosophical work should see that in most cases we can't really get more out of that concept than we can get out of reflection on how the word is used in the language game. (And what we get out of that may be more "dissolution" of our questions than "solution.")

      I think he would readily say that in general, there is no one thing that is linguistic description or use.

      And that, in a given artificial or language, meaning may be given by stipulation, by a table of concordance, or what have you, rather than by a description of use--but then even in such cases, if you get philosophical about it, you very quickly have to start describing how that table is used or what follows or doesn't from that stipulation in (what speakers of that language make sense of in) practice.

    3. ... I agree that use involves different behaviors (e.g., stipulating). But the passage appears to say that there is a small circumstance where the word "meaning" does not call for looking at usages. This can't be a class of words -- for that would put forth a bona fide linguistic thesis for which we would have seen extensive remarks (Wittgenstein's exceptions). Not to mention his disavowal of theses. Instead of going down this road, it seems more rational to seek an answer that is utterly consistent with the Investigations: namely, that some uses of "meaning" are not like the ordinary ones. As in, "the meaning of life," which asks why I exist. So, absent this kind of inquiry, the usual way that we inquire about meaning involves looking at its use in language games.

    4. Certainly, if he were to turn the conversation to "meaning of" in the sense of "significance of" something (ranging from this painting to life), he would not maintain that that is its use in language.

      But about this passage, I guess I'm not convinced that this is here point here. First, he says "the meaning of a word," which strongly suggests he is not referring to a broader class of applications of "meaning" to phenomena like life. Second, I agree with the sinn and bedeutung point that Tommi makes—not that I think that in general Wittgenstein uses those terms technically after the Tractatus, but in this specific passage he's recalling to our minds the technical distinction of his earlier conversations with Frege & Russell. Third, what I said about who the colleagues and interlocutors are at this point in the discussion. So, while I wouldn't adopt the "linguistic uses" language, I would say that he's saying that for most but not all instances in which we are puzzling philosophically about language and meaning, the meaning of a word is its use in the language game. In a way this is stipulative, in the process of developing this instrument of his new philosophical art, the description of a language game.

    5. I think I once heard Bill Child saying that in the qualification in §43 Wittgenstein has in mind words employed in secondary senses. It is the most promising suggestion I have so far heard. It creates a problem, however—at least an apparent one: it would seem to imply that words employed in secondary senses are not used (at least in some sense of ‘used’).

      Come to think of it, there would seem to be a similar problem with any other suggestion—a problem of disconnecting the meaning of a word from the way it is used.

      I personally think that Sean’s problem—having a reading of §43—is a pressing one.

    6. Okay, I hadn't got so far as to see precisely what construal Sean W was putting forward—I guess it's that Wittgenstein could have more clearly said: "for a large class of cases in which we employ the word 'meaning,' namely, those pertaining to language, the meaning of a word is its use in language.' Our ears apparently respond differently to the question of whether that's a reasonable construal...doesn't seem so to me but I can see that it might seem so to others.

      A few questions this raises for me...

      1) Is saying that 'linguistic meaning is use' any less thesis-like than anything I was saying? I don't see how. We're both talking about situations in which we use the word 'meaning.' Sean's take is broader: situations in which we are talking about the meaning of a word are situations in which it is correct to say meaning is use. Mine is narrower: in situations where we are addressing philosophers concerned with certain philosophical puzzles--the meaning of words in the absence of referents--and we are doing so using my newly proposed heuristic device of 'language games,' we can say that the meaning of a word is its use in the language game. To my mind, stipulating a definition of "linguistic meaning" in terms of language games—commonplace today but introduced for the first time in these passages—would be much more thesis-like. 'I, LW, have discovered these things called language games, and it's use in these that gives words meaning, contrary to what Frege or Russell or the author of the TLP thought.'

      2) Are there counter-examples on the table to what Sean W suggests? Are we arguing over a null class? What about this: does Wittgenstein give us an example right in the same remark? If I say ' what does 'chartreuse' mean,?' and you point at a swatch (or "what is a trivet?" and you point at a trivet), certainly this explanation would not take without much mastery of use and my mastery of the meaning of this word would be displayed in how I went on to use the word—nonetheless, you'd be doing philosophy on me if you answered that question with a description of its use in the language game. Your explanation of the meaning and my ability to grasp it presuppose the use and its mastery, but couldn't/wouldn't be replaced with a detailed description of that use in the language game. (We can debate that couldn't/wouldn't. The demonstrative explanation is perfectly all right as it is. Which is against the backdrop of use but not shorthand or a sloppy way of indicating that backdrop.)

      3) What does 43 and any disagreement we might be having about 43 have to do with the context in which it appears? What is the move in the conversation?
      to be continued...

    7. He's diagnosing the source of his TLP, Russell's, and Frege's various views around names without reference. Contrary to all those folks who needed to track the reference of names of things that don't exist to something (through logical analysis to simple names for W, through theory of descriptions to this for R, through whatever account you accept of F on sense and reference for non-referring names and the implication for sense/reference of the sentences in which they appear), Wittgenstein says, look, in your language game you could have all sorts of ways of responding to what we might call 'names of things that don't exist.' It could be that you stop using the name; it could be that you keep a list of lost tools; it could be a joke. Any of those things could be true, although when you ask what the word means, someone might well point to its bearer, and that would be a correct explanation of what it means. For that context in which "meaning of the word" is used.

      If I continue to trace where this difference in reading 43 might lead, it leads me to my whole quirky Ishiguro-like view of the meaning of names in the TLP, and therefore what exactly he's diagnosing here, and ladders and nonsense and so on, and that gets us even farther from Duncan's initial point in raising the passage, proving that philosophers cannot appeal to a single sentence of Wittgenstein without many Wittgenstein exegetes stopping by to disagree about what Wittgenstein really meant.

      Back to Winch and ISS, 1) is there one thing that is a description of the use of a word in language, and 2) does 'meaning' always refer to that, and 3) is that kind of description of use of language the same thing in philosophy à la Winch-genstein and in Sociology? I think Duncan was saying no to 2 and setting it alongside Winch in the ISS, and I think Winch would have said later 'yes, Duncan, you have a point,' because he would have later read 43 something like what I'm laying out. (I'm laying out a reading of this passage based on a style of reading the PI that I learned from him—not the other way around.)

    8. Lereid: he doesn't say "meaning of a word" until he is done dismissing the irrelevant senses of meaning. In other words, it's straight forward. He says: excluding irrelevant usages, meaning is use. That's pretty much the structure of it. The difficulty with the other view is that we don't have any remarks about those genuine times when "puzzling about language and meaning" is (what?) more orthodox. That is to say, in all the writings going on at this time, we are not treated with "Wittgenstein's exceptions." Also, I didn't read Tommi as disagreeing here. Even under his view, the German word has a different meaning when asking about the goal of life than it does when asking about the meaning of a word.

      Reshef: I'm not convinced by the secondary sense thesis, not only for reasons you raise. When he spoke (locally I think) about "secondary sense," he said it was not metaphorical, which meant it functioned like primary ones. And he also made comments about how it was meaningful to call the vowel "e" yellow and to speak of Tuesday being "fat." Point being: the sense in which "e" is yellow is surely a function of its use (context).

      Anyway, I appreciate all of the responses so far: good topic to discuss.

    9. Sorry. I responded to the very first comments, without seeing the others that came next. They came when I was typing!

    10. This discussion is indeed at a bit of a distance from Duncan’s original concern, I’m sure he wouldn’t mind.

      Anyway, let me try to sort our positions out. I got a little confused. Tell me if I got things right:

      There are two separate questions here about §43:

      1) Does Wittgenstein leave room for talking of the meaning of other things, beside that of words?

      2) Does he leave room for the possibility that the meaning of a word is not its use in the language game?

      It seems to me that we mostly agree that the answer to the first question is yes. – Do we? If so, is there still a disagreement?

      Regarding the second question, Sean’s answer seems to be no. (“for all cases where meaning does pose linguistic questions, meaning is use). Lynette’s answer is yes. (“Your explanation of the meaning and my ability to grasp it presuppose the use and its mastery, but couldn't/wouldn't be replaced with a detailed description of that use in the language game”). My own secondary meaning answer is also a kind of yes.

      Have I got things right?

      If this is right, then it seems to me that one thing we need (if we are to come to an agreement, and if our disagreement is not merely “semantic”) is a better notion of ‘use,’ and of how use is given in an explanation of meaning. (We might put my own suggestion aside for a while, because it would also require an account of secondary meanings and uses, and metaphors.)

    11. Thanks, everyone, for the comments. And of course I don't mind that the discussion has taken on something of a life of its own.

      PI 41 and 42 are very interesting. Wittgenstein talks about what would happen if the tool breaks for which some word is the name. Is the name then meaningless? The names in the language game in question are commands ("slab!" etc.) so they both refer to objects and call for a response. If the appropriate response is impossible, then the would-be responder might be at a loss. But there might also be a convention about what to do in such circumstances, so that he is not at a loss. The same goes if a new word is introduced. There could be a convention of doing this and responding in a certain way (as a game, say). (Has anyone compared this with Heidegger on broken hammers or Nietzsche on the death of God? I wouldn't be surprised.)

      Then here's PI 43:

      For a large class of cases - though not for all - in which we employ the word 'meaning' it can be defined thus:the meaning of a word is its use in the language.
      And the meaning of a name is sometimes explained by pointing to its bearer.

      The answer to Reshef's question 1 seems to be Yes, but it's by no means clear that Wittgenstein is trying to say anything here about the meaning of anything but words. The answer to question 2 also seems to be Yes. He doesn't explicitly rule other possibilities out.

      It's a funny passage though. Section 40 warns against thinking that the bearer of a name is its meaning (because then the word would become meaningless when the object it names is destroyed). So presumably 43 is not saying that meanings ever are objects in that way, even though we sometimes explain/define a word's meaning by pointing to its bearer.

      Section 41 refers to "N" as a name, a sign, and a command. This blurring of distinctions seems worth noting, although so too does the fact that what is under discussion is a particular language game, not (necessarily) all uses of language.

      When Wittgenstein talks about 'use' in these passages he seems to have in mind conventional use, but he also brings up the possibility of a word's being introduced or admitted into the language, and different ways in which a word might become meaningless. And, it seems to me, he implies that 'meaningless' might have different senses: there are games in which we conventionally shake our heads at reference-less words and then games in which we are on our own if someone utters a reference-less word. Both might be called meaningless, it seems to me, although Wittgenstein treats the former as having a meaning (because these words have a use, a role, in the language game).

      It is not as if there are words that have a role and words that do not and that's that. We might have various rules or conventions for dealing with words that have no conventional use yet or any longer. (And, in fact, it seems that we do have different ways of responding to unfamiliar words.)

      Reading this over again it all seems very disjointed. I hope it makes sense.

    12. "The answer to question 2 also seems to be Yes. He doesn't explicitly rule other possibilities out."

      Can you say more what you think those other possibilities might be?

      So far, we've been having troubles identifying, or agreeing, on even one!

    13. Perhaps there aren't any, but if there are then he allows for them. That is, all I meant was that he doesn't insist that there are no other possibilities.

      Having said that, it does seem worth thinking about what these might be (if anything). Two possibilities to consider are that 'meaning' might not always be used to refer to meaning, as when I ask what the meaning of a name is and you point to its bearer (and I am satisfied), and the use of a word in a secondary sense. In the former case we seem to have something annoying or cheating going on. When I ask for the meaning I am asking what object bears the name (so I don't mean 'use' when I say 'meaning'), but by pointing to the name's bearer, given the background, you are telling me all I need in order to be able to use the word, so you are explaining its use and that is what I want. Still, in cases like this the object itself plays a role in the use of the word (at least in some cases it does), so it isn't irrelevant to the meaning of the word. It's not as if you have the use of words over here and the rest of the world over there, so that one might be constructed out of the other or questions arise about how the two relate. There are not two wholly separable things here.

      In explaining what he means about the use of words in a secondary sense, doesn't Wittgenstein say that these are cases where we want the normal meaning but not the normal use of the word in question? If so then here he does not seem to mean 'use' when he says 'meaning.' I don't know whether he just means tone, association, or color either. If I say that "Wednesday is blue" then I don't mean that Wednesday is like the color blue. I want to say that it is blue. That's what I mean. Am I using the word in that way? Well, I'm trying to. There is no convention of using it this way, but I suppose I'm hoping that my attempted use of words will be successful, will be a use. If I fail then there is no question of what I meant, only of what I was trying to say. So maybe meaning is use in this case too, although 'use' here does not mean conventional use, which is what "its use in the language" suggests to me. Nor is it use in the cause-and-effect way that I might use a word to frighten an animal (by shouting the word at the animal, say). So it's use in a slightly mysterious sense.

      Another case to consider is that of 'is' in section 561 and following. Wittgenstein wants to say that the use of 'is' as the copula and as the sign of equality amounts to the word's having two different meanings, not one complex use. I think we could again say that meaning really is use in this case, but he wants to say that the two uses are essentially different, and makes this point by saying that the word has two different meanings.

    14. ... I just think the easiest thing to do here is ask, when, in later Wittgenstein, is meaning not use. I can't think of anything. That's the big thing with 43. People read it as though it is a law statement and not consistent with everything else. (And in fact, they seem to even misread it as a law-statement)

      I also think that what 43 says about meaning being use is not meant to be local to the names discussion, because there is nothing in the names discussion which differs on this point from what comes right around the corner (such as, family resemblance). Think about it. He just gets done showing that even names have more than one sense. This allows a person to say "Federer looked like 'Federer' today on the court," and mean something different with each use of Federer. And so, I think 43 is telling us why this is so. And the reason why it is so happens to be the same thing that operates throughout the Investigations when discussing subjects other than names.

    15. I can't think of a time when meaning is not use, but I think 'use' has different uses (conventional use, unique application on one occasion, etc.). And sometimes we use the word 'meaning' to mean something other than use. So basically I think I agree with you: meaning is use but this should not be taken as the statement of a law.

    16. With regard to the first case Duncan mentions--that of names--I admit that I don’t see it yet. I mean, if we are supposed to be looking for cases in which the meaning of a word is detached from its use, then this does not seem to be such a case. It seems to me that the only reason for thinking of this case in this context is that the use here is given not by a specification of a rule but by pointing to an object. – But so what? I mean, I don’t think it is unimportant that ‘giving the use of a word’ takes different forms in different cases, but I don’t yet see how these differences make a difference regarding the question we are asking: “Can meaning be detached from use?”

      Regarding the use of a word in a secondary sense: Duncan, where does Wittgenstein say that in such cases we want the normal meaning but not the normal use of a word? – This is what I would like to say. He comes close to saying that in the ‘Wednesday is fat’ passage. He also says something that is in a way close to that about aspect-seeing in RPP1 §411: “[…] I take it as the typical game of ‘seeing something as something,’ when someone says ‘Now I see it as this, now as that.’ When, that is, he is acquainted with different aspects, and that independently of his making any application of what he sees. // So I should like to say this: I do not see any application of the picture as a sign that the picture is seen this way or that.” Here the application (the use?) is not taken as a criterion for the experience.

      But I think Duncan is right, there is something mysterious here. We need to talk about “use” in such cases. Wittgenstein himself does that. But at the same time, the word "use" here means, implies, something different. I take this to mean that we have a task here of describing exactly in what way a use of a term in a secondary sense is not a use, and in what way it is. Perhaps, the thing to say is that it is a secondary sense of “use.”

      Regarding the third case Duncan mentions: Wittgenstein says in §561 about the word ‘is’ that he should not care to say that its meaning is its use. And I guess this is why Duncan mentions this case in this context. But it seems to me that Wittgenstein says that only because saying that the meaning of the word ‘is’ is its use might be taken to imply that the word has a single use. His concern here seems to be with separating uses—with the notion of “use” itself. It seems to me that he is saying, among other things, that not any specification of rules for the application of a term would indeed be considered as a specification of use. I cannot, for instance, specify the use of a new word ‘bnox’ by saying that is use is partly like that of the word ‘tea’ and partly like that of the word ‘funny’–at least that would be very odd. One would not get a sense here that this specification has any point.

    17. Thanks, Reshef.

      I don't have much to add, but to answer your question: I was thinking of the 'Wednesday is fat' passage:

      Now have "fat" and "lean" some different meaning here from their usual one? -- They have a different use. -- So ought I really to have used different words? Certainly not. -- I want to use these words (with their familiar meanings) here.

  6. Two considerations:

    1) Social science has moved on since the 1950s, so Winch's criticisms, regardless of how relevant they were in their contemporaneous context, are no longer so. Reading a recent randomly picked issue of a cutting-edge journal in, say, sociology, it keeps getting harder and harder to recognise Winch's image of social science in it – and conversely, the further we rewind back to the 1950s, the easier it gets. What Winch speaks of a "the Idea of a Social Science", with a definite article, is an idea of a social science, but by no means the only one. (By far the best thing I know on this is Michael Lynch's "A new disease of the intellect?", esp. pp. 143–146.)

    The book is indeed by a very young philosopher who continued to develop later, and kept on developing in some ways to the end of his life. Personally I view Winch as one of the 5 or 10 most important philosophers of the 20th century, but I hardly ever return to The Idea of a Social Science, because the world it speaks about is really neither my own world nor that of Winch himself in his later career.

    2) It's unclear whether Winch's making (what are obvious) value judgements is compatible with his official self-image as a Wittgensteinian philosopher who views the use of philosophy to underwrite value judgements as a non-starter. In "Moral Integrity", his 1968 inaugural lecture at London, Winch wrote that "philosophy can no more show a man what he should attach importance to than geometry can show a man where he should stand". But to say that "any worthwhile study of society must be philosophical in character" is precisely to say where we should stand, or at least where we should not stand.

    We had an exchange on all this (and related issues), here on the blog two and a half years ago, which you've probably forgotten, but which I keep coming back to all the time.

    1. Thanks, Tommi.

      I had not forgotten that we talked about this before, but I didn't remember all the details. Thanks for reminding me.

      I think I agree with everything you say here, although the problem of being a Wittgensteinian philosopher and making value judgements is a tricky one. I don't mean that you're wrong, just that, as you say, it's unclear. Philosophy can't make value judgements, but philosophers can (albeit perhaps not qua philosophers).