since handing in my paper for 'metaphor and thought' today, i've been able to finally formulate my own metaphor for the great divide in the philosophy of language that's been bouncing around in my head since i watched that 'elegant universe' documentary.
first there was the newtonian theory of physics, which was fairly over-simplified and deterministic. then (to over-simplify a bit myself), along came albert einstein and came up with a much better theory that explained a lot more phenomena and required a major shift in everyone's thinking about the universe. einstein was pretty smart, but toward the end of his life when people started to talk about quantum mechanics, he couldn't handle it. 'gott würfelt nicht,' he said, meaning, 'god doesn't play dice, the laws of nature aren't based on probabilities.'
this is the same position analytic philosophers are in. (i think it's a nice metaphor for them too... they get to be einstein!). they cannot believe there is not a truth 'out there' in the universe to be discovered (i think a lot of philosophers might consider themselves atheists, but i'm just saying... gott might würfle). Their theories work a lot better than older classical models of language, but just try telling a philosopher that reference isn't deterministic, that language doesn't have to have a strict dependence on the world, that the odds determining how a word is used depend on the odds that another word was used two weeks ago and so on..., that vagueness isn't a problem that needs to be solved. you'll get a funny look and a bad grade.
now imagine a contemporary university physics professor who not only doesn't teach quantum physics, but doesn't believe in it. if you told them that light is both a particle and a wave (a very post-structuralist move, by the way!), they would give you a funny look and a bad grade, but then they also might get fired for being 50 years behind the times (unless they had tenure, in which case they might just be ridiculed in the literature and discussed angrily in the cafeteria and department meetings... i don't know how these things work.)
granted, from my understanding, there are more outlandish theories out there (like string theory) that aren't widely accepted or taught in most physics courses. to continue the analogy, i'm not saying philosophy classes should start teaching all the craziest semiotics they can get their hands on. i'm also not saying there's no value to studying analytic philosophy. it's smart, it's interesting, and it maps out a lot of the territory of philosophy of language and shows where the problems are. i'm just saying that philosophy departments and classes exclude out of hand, for no apparent reason other than the fact that it would force them to rethink 100 years of their own research, a huge branch of thought about the same questions they claim to be asking.
some quest for knowledge, huh?
Subscribe to:
Post Comments (Atom)
19 comments:
You should try and explain to "einsteinian" semanticists that meanings have fuzzy boundaries and are basically strange attractors (to stay with the physics analogy). That should get you a good blank stare...
Hold up.
First, I don't think it's fair to dismiss analytic philosophy--as a methodology or personality of academic inquiry--because of the opinions of the majority of analytic philosophers. There is a lot of diversity within analytic philosophy, and new ideas take time to percolate throughout.
Second, I can think of two things that may explain the slowness of analytic philosophy to accept the kind of radical thinking your espousing.
First, it doesn't do to get rid of a theory that works ok but has some problems if you don't have a theory that is just as functional and well-defined to replace it with. Can you explain the theory of language that you would like see popularized in a concrete, determinate way? Marc may be on to something. If it works, more power to him, but I gather that his theory is complex and relatively new. Give it time.
Second, a lot of philosophers have commitments that go beyond language. It is, in some sense, their duty to find answers that don't result in skepticism or radical relativism or nihilism, because these are all in some sense self-refuting positions. So yes, "they cannot believe there is not a truth 'out there' in the universe to be discovered"--but what is truth, if this is not the case? That has to be well-defined. If not, then it's questionable whether any of this discussion--even your original post--can actually get anywhere. If post-structuralism--and I'm still not quite sure what you mean by that or what ideas in includes--can't address that more basic philosophical need, then it's important that analytics continue to hold up the fort.
---------
Out of curiosity and a little pugnaciousness, could elaborate on the what you mean by:
- Reference isn't deterministic
- Language doesn't have a strict dependence on the world
I'm not sure what you mean by this.
My impression of vagueness is that it is a problem to the prevailing theories of reference--as in, it's not accounted well by them. But it's not something that needs to be excised from language itself, it needs to be explained.
As far as "how a word is used," that is to many people (not me) orthogonal to what a word actually means, so I'm not sure that really is relevant to those ones. Meanwhile, the Wittgensteinian language-is-use tradition I think would have a problem with having a better description of how use is determined. (Although it has a lot of problems of its own, in my opinion--not the least of which is (you'll like this) that it's behavioristic. However, there have been some good ammendments to it--recently I've gotten really psyched about conceptual role semantics.)
I'll let Cristi answer before I do, but I just wanted to say that the theory I'm coming from is not particularly new; it stems from the work of a French linguist from the first half of the XXth c. (Guillaume). I did, in all fairness, put a more contemporary spin to it by using analogies such as the strange attractor. But since I actually studied physics before linguistics, this analogy quickly came to mind when my profs where trying to explain their view of meaning. As to whether it is a complex theory, I guess it is. Or rather, it has been overly complicated.
first i'd like to thank seb for, as always, keeping me honest.
second, a few caveats: the analogy i made here is obviously a pretty lose one. i don't believe analytic philosophy is a totally homogenous field (as i'm sure einsteinian phyisics wasn't). but the idea of truth conditional semantics is pretty darn universal among all those different percolating ideas. i also tend to use 'post-structuralist' pretty losely. it is also not a homogeneous field by any means. i probably mean, for the most part, 'derridian.'
i think saying that a philosophy of language has to be WELL-DEFINED or replaced by something functional and well-defined is a major problem. claiming that this is a methodological committment and not a philosophical one is VERY dangerous. what is it analytic philosophers want to say about language? when it is 'correct' to use a certain word? when a certain sentence is 'true'? these things seem to indicate that there should be a correspondence between states of the world and the language we use in those situations. clearly, at best this is a tendency, or a matter of probability.
take vagueness as an example: it doesn't actually need to be explained unless we're using a truth-conditional semantics. what needs to be explained is how we manage to communicate anything in a world where things are ALWAYS vague, where we can use language in any situation we want, where we can speak ironically or 'infelicitously' or impolitely, etc... the thing i really like about 'post-structuralist' theories (especially ruthrof's - i'll plan to write a post soon detailing his theory), is that they don't begin by excluding things like metaphor, vague predicates, fictional reference. most analytic philosophy works by taking the most 'basic' or 'central' cases, and then trying to work 'outward' to the more 'confusing' types of speech. ruthrof is exciting because he actually addresses the tradition of analytic philosophy and his theory has the potential for real explanatory power and, once you accept a few traditionally 'post-structuralist' ideas, great simplicity. again, i'll have to get back to it.
haven't really read enough wittgenstein to comment, although you're right that the conflation between meaning and use is troublesome for a lot of people. i guess i'll just close with another teaser by saying that the idea that a word 'has' meaning that people 'convey' is troublesome to me.
"the thing i really like about 'post-structuralist' theories, is that they don't begin by excluding things like metaphor, vague predicates, fictional reference" to me, that makes the most sense. Anyone who observes actual language will see that there's a lot of that going on. To begin by excluding those would be, in my view, like trying to come up with a coherent theory of biology based on humans, with anything else as fringe manifestations.
"the conflation between meaning and use is troublesome for a lot of people" That's why I like the idea of potential meaning: (vague) boundaries that encapsulate the actual uses of a word.
i think saying that a philosophy of language has to be WELL-DEFINED or replaced by something functional and well-defined is a major problem. claiming that this is a methodological committment and not a philosophical one is VERY dangerous.
Dangerous for what?
I can see how a prior commitment to well-definedness might be problematic to philosophy of language, depending on how you...defined..."well-defined."
But I would put the emphasis on functionality. A good theory of language needs to be useful in some way. It, at the very least, seems to need to be useful in a particular way (see below). And if that philosophy contains a lot that is ill-defined, it becomes less useful, I think--in the same way that any other theory becomes less productive when it loses formality and precision.
I guess what I'm wary of here is any theory that undermines it's own basis for application. (see below)
what is it analytic philosophers want to say about language? when it is 'correct' to use a certain word? when a certain sentence is 'true'?
What a philosophy of language absolutely needs is some account of how it can be true. For example, consider the following theory of language (or parody of one):
P: There is no such thing as universal truth. All truth is relative to the society one is in.
P is, at least in the first-pass analysis, self defeating. Why? Because if P is true, then P is true only in some societies. That means there is at least one society in which P is false, which means there is some universal truth--something true for all people--which contradicts P.
So something has gone terribly wrong with this theory. And while I'm not suggesting that what you're proposing is anything close to as crude as this, there is the danger that even a well-intentioned theory can slip into this sort of trap.
Also, there needs to be some account for why, if there's a dog in the room, and I say "There's a dog in the room," that's different from if I said "There's no dog in this room." Because if we don't have some notion of what right and wrong mean, then...that's bad. It leads to a kind of radical skepticism that is again self defeating and, for pragmatic reasons--for the reasons motivating the theory itself--untenable.
Now, I'm not saying that a post-structuralist solution will fail along these lines--hell, I don't know anything about post-structuralism, and look forward to your explanation of it. Ruthrof does sound exciting, and bringing metaphor etc. into the fold sounds like a step in the right direction. Personnaly, I've got big problems with the way truth is handled by the majority of analytic philosophers as well. (Although my hunches as to solutions, as you know, are with less emphasis on language, not more)
But there is a lot at stake here, and I think it's important to acknowledge the motivations of the proponents of truth-conditional or model-theoretic semantics as valid. Quantum theory won out against newtonian mechanics because it was better according to some well-defined criteria. (Begin science studies, Kuhnian debate here) What the test of success for theories of language ought to be is less clear, but what I haven't seen yet is an explanation of what post-structuralist theory does better than the alternatives. Sure, it may include "metaphor, vague predicates, fictional reference"--all victories, to be sure!--but what does it do with them? Can it do this without becoming inapplicable, or--worse--self-defeating? Can it honor any of the other motivations that seem relevant to philosophy of language? I haven't seen this work yet, which is why I remain skeptical.
Marc, a question about the theory you espouse--psychomechanics?
Is the idea that meaning is a strange attractor, or that the the meaning of a particular word is an attractor?
Maybe another way of asking this is: in the analogy to an attractor working on an N-dimensional space, what are the dimensions? What is a point in the space?
Wow, wonderful commentary from all posters. A very stimulating exchange. Glad to have found this blog. Kudos to all of you!
thanks, christian, for the encouragement! thanks to seb & marc too, for keeping this thing going.
i know i'm overdue for another actual post, but i'm in phoenix packing up family photos...
i guess the biggest problem i see here is that the 'functionality' of most scientific theories is based on their predictive capacity. i don't really see this as a possibility for a theory of language, since we can't ever completely predict how people will use language. if there's a dog in the room, i can say "there is a dog in this room" or "there is not a dog in this room" depending on my whim. after the fact, an explanation of how the two differ would have to take in (i would argue) an infinite amount of facts: to whom am i speaking? am i crazy? am i an adult? am i joking? am i generally known as a joker? am i lying? do people think i am lying? what if i only think its a dog? am i being fooled? am i blind? and so on.
i feel like i'm shooting myself in the foot here, because now i'm not sure WHAT a philosophy of language is supposed to do.
the biggest problem, in my mind, with truth-conditional semantics is that i don't believe in truth outside of language. that's not to say i'm any kind of relativist. i believe that to compare a representation of something (a word, sentence, drawing, etc., thought if you're a post-structuralist), to the 'real' world (note the quotes) first requires a representation. that is, the limits of our ability to represent things are also the (theoretical) limits of our ability to determine the truth of something. only representations have truth values. which is why i think a philosophy of language (representation) HAS to precede any truth-based philosophy (which is to say all of western philosophy). i guess that would be my goal for such a philosophy: to map the limits of representation (or signification, to use the parlance of our times) and tendencies within the field of representation.
again, i owe you all a post. thanks to everyone who's scrolled down this far into the comments!
Yes, it is psychomechanics, although not the hard-core variety.
"in the analogy to an attractor working on an N-dimensional space, what are the dimensions? What is a point in the space?"
The analogy to a strange attractor doesn't really have to do with N-d space, but rather the idea that the meaning of a word usually centres around one or two core(s) and the fact that, like an attractor, the meaning has no defined boundaries, but will really go outside of an observable zone.
But Marc---"the idea that the meaning of a word usually centres around one or two core(s) and the fact that, like an attractor, the meaning has no defined boundaries, but will really go outside of an observable zone."
What is a core, and what sort of space is the zone around them?
In order for their to be even ill-defined boundaries, or movement, or a region, there has to be a space of some kind. But is that a space of usages, of referents, of senses, of cognitive roles, or what?
The core (don't really like the term but I'm stuck with it now) is an idea, a concept. In the case of words designating "real-life" objects, it would be the idea of this object.
The space around this core (or chewy centre -- sorry, a bit hungry at the moment) is a conceptual space, where some aspects of the main idea are done away with, where we can just use some of the caracteristics of the core. The furher away you get, the more abtracted it becomes.
It could be construed as a space for uses, references, senses, in that these are actualizations of meaning.
Ok, great! That makes sense to me!
Although now I'm not sure that psychomechanics provides the kind of answers that Cristi wants.
Why? Because that theory still requires that there be "an idea, a concept. In the case of words designating "real-life" objects, it would be the idea of this object." This core has to be some kind of representation with a determinate meaning. The claim, as I understand it, is that words have these fuzzy zones around them which account for the flexibility of language. But here, concepts play a different role. They are stable in a sense that language is not.
I think to be consistent, Cristi needs to reject this theory, since she seems to be commited to the conclusion that all representation is linguistic, and all language has the fuzziness and flexibility that you attribute to the strangely attracting words, but not to the conceptual cores.
the biggest problem, in my mind, with truth-conditional semantics is that i don't believe in truth outside of language.
When I read this, I want to disagree until I remember that you're using "language" to refer to all kinds of representation.
only representations have truth values. which is why i think a philosophy of language (representation) HAS to precede any truth-based philosophy (which is to say all of western philosophy). i guess that would be my goal for such a philosophy: to map the limits of representation (or signification, to use the parlance of our times) and tendencies within the field of representation.
But an argument goes the other way as well: when we make claims about the limits of representation, we want those to be true claims. We need to know what it means for something to be true before we can make any progress on what might be true about our representations. So a philosophy of truth must precede a philosophy of language.
I don't agree with this line of reasoning totally, but for now I'm just representing the downtrodden analytics.
"concepts play a different role. They are stable in a sense that language is not" Actually, I don't believe they are. They're not exactly the same from one person to another, from one language to another (chair is not the same concept in French as in English, for example). But you are right, psychomechanics probably doesn't offer the kind of answer Cristi wants, unformtunatly. It is more concerned with semantics than philosophy. It stops just short of defining concepts (as far as I can remember). But, although it might not have seemed that way from what I said earlier, it does not profess static, unwavering concepts/cores. More like chewy centres. The core meaning of a word, would be the sense people would most readily agree upon; not that that would be the one true sense.
I recall my semantics prof. comparing words (and morphemes) to software: you can't actually predict or define all that someone can o with, say, Photoshop, but there are boundaries, one cannot do everything with it. In that sense, the meaning of a word would have more to do with conditions (or constraints). These, if we were to plot the range of uses/senses, would yield a strange attractor of sorts.
my dear seb! you seem to forget that you have an account and can make your own posts here. you know how much that would please us. :)
Of course, Scott, of course. I'm just intimidated by the task, that's all. But I'll shape up and make a complete post soonish if I get to it.
again, i PROMISE to post again soon... but i'm on vacation :)
marc, i did disagree with the post where you described the 'core', but if you're willing to concede (as i think you have) that those conceptual meanings also have this electron-cloud type of arrangement, then I like this a whole lot.
seb, you've hit the nail on the head: no philosophy can operate outside of language (of course i mean 'representation'). i think this is the root of your 'this theory compromises its own application' objection, but i also think it's something we have to deal with. i believe that we can't operate outside language.
The problem with linguistic theory as it has evolved from the days of "Syntactic Structures" and associated tomes is that it presupposes, falsely in my view, that we have grammars in our heads, and that in speaking and hearing we use algorithms that somehow use this grammar.
If one looks at actual linguistic data -- real world conversations, which ought to be the focus of study since that is what our having language is all about, namely talking to each other -- and focuses as much on speech production as speech perception, one is led, I believe to the view that the most useful approach to the study of language is computational. Linguists approach language analytically, just as listeners do, but it takes very little time focusing on speech production to show that what drives the language machine is pragmatics (usually the tail end of analytic work) -- what is communicated, and that ain't reducible to any truth-conditional approach to meaning even given the new improved versions of that approach. I hereby promote my book on speech acts and conversational interaction for a long account.
Post a Comment