How to do things with JL Austin
May 20, 2008 8:04 PM
What's the big deal with JL Austin's "How To Do Things with Words"?
I'm about fifty pages into this torturous book. I don't see what the big deal is. So some utterances are performative. Why were people as far apart as John Searle and Judith Butler so excited by this? What can be done with this discovery?
I'm about fifty pages into this torturous book. I don't see what the big deal is. So some utterances are performative. Why were people as far apart as John Searle and Judith Butler so excited by this? What can be done with this discovery?
I agree with Ms. Saint's characterization, and would add that, from an academic standpoint, Speech Act Theory, as derived from the works of Searle & Austin, is also important as forming part of the basis for the field of linguistic pragmatics. Pragmatics (broadly speaking) studies language in use, as opposed to the ideal sentences constructed in other branches of linguistics. So for somebody like Judith Butler, who is concerned with challenging traditional notions of how we define and understand gender, pragmatics (and Speech Act Theory along with it) helps her argue that gender is performed through discourse. In other words, we don't *have* gender, we *do* gender, and if we change the way we talk, then we can (potentially) change the way we do (and think about) gender.
posted by DiscourseMarker at 9:25 PM on May 20, 2008
posted by DiscourseMarker at 9:25 PM on May 20, 2008
Adding to the heap here that the important dimension in Austin's work is the pragmatic one. To some extent, this can be seen as following from themes Wittgenstein introduced. But if you want a sense of why such a diverse cast of characters looks back to Austin, you need to see that he was the big figure for many people in suggesting that you could both recognize the pragmatic dimensions of language and treat them systematically. Most attempts to reveal some deep and important structure in language in the analytic tradition up to that point had been inspired by Frege, i.e. uncovering some eternal and unchanging proposition expressed by ordinary sentences. While that appeals to the frustrated mathematician in many people, it also came to seem rather musty and ossified to philosophers by mid-century. At the same time, those who had been emphasizing pragmatic dimensions of language in a strongly Wittgensteinian way tended to make those practical purposes sound mysterious and unanalyzable. Austin, for many people, was the first figure to cast the pragmatic in a way that had all the rigor and insight of other forms of analysis, but retain the vivd practical character of actual use.
posted by el_lupino at 11:09 PM on May 20, 2008
posted by el_lupino at 11:09 PM on May 20, 2008
I think the problem you're having is that Austin's basic thesis ("Saying things does shit") sounds blindingly obvious.
It's hard to believe, but the generation of logicians who went before him mostly ignored that fact. Ms. Saint gives a good discussion of why they were wrong to ignore it — as does Austin himself, for that matter — but it's also worth thinking about why they did ignore it.
Modern semantics came about in a sort of ass-backwards way. The first semanticists weren't linguists at all, but logicians and logically-minded philosophers. At the start of the 20th century, people were incredibly optimistic about formal logic. Some big-deal foundational works in mathematics had just come out (The Grundlagen and the Principia Mathematica were the biggies) and Gödel's incompleteness theorems hadn't dropped yet. For centuries, math had been this sort of haphazard trial-and-error affair with no real rules, but for about half a century in there, between the 1880s and the 1930s it looked like that had changed for good: people thought they were on the verge of creating a formal system that would generate all the mathematical truths and none of the falsehoods. Turns out they were wrong, but if they'd been right it would have been incredibly exciting news.
So the next question some people asked was, If we can do it for mathematical truths, can we do it for other kinds of truth? The Vienna Circle, in particular, were interested in using formal logic to generate philosophical and scientific truths. The idea was, Well, if anything expressible in logical form can be proven true or false, then all we need is a way to translate natural-language sentences into logical form, and then we can prove which sentences are true and which ones are false. Again, it turns out their project was doomed to failure — in any logical system that's worth two shits, some sentences are undecideable, but they didn't know that yet.
Anyway, that project of translating sentences into logical notation is where modern semantics comes from. Eventually, Gödel's theorems hit the scene, and people realized that semamtics wouldn't solve all their philosophical problems. But by then, they'd also realized that semantics was interesting in its own right, and they kept at it. For the most part, they kept at it with one of the Vienna Circle's original assumptions intact: that all sentences could be translated into logical propositions, even if some of 'em were undecideable, and that a sentence's meaning was determined by its truth conditions. Using that assumption, they generated some nifty and powerful results (read up on compositionality if you're interested), but it was a false assumption, as it turns out, and it limited what they could do. They couldn't, f'rinstance, say anything coherent about basic social uses of language like promising, naming or apologizing.
So that's the environment in which Austin was writing. The semanticists of his day needed a big fat reality check: someone needed to point out that "What time is it?" and "Fuck you!" and "I apologize" don't really have truth conditions, and that their meanings can't be expressed fully in formal logic.
The other nice thing about Austin's work, though, is that it didn't force you to throw the old stuff about truth conditions out the window. He acknowledged that some sentences do have a layer of truth-conditional meaning, and that the work people had done on truth conditions might still be valid. He just proposed some additional layers of meaning ("illocutionary force" and "perlocutionary force" are the usual terms, and IIRC they're the ones he uses) that cover the practical, social, non-truth-conditional stuff. That was actually a pretty nifty trick, although again it seems obvious in hindsight.
(Jeezis Christ that was long. Sorry. I do this stuff for a living, more or less, if grad school counts as a living. Short version: What you can do with Austin's discovery is do semantics in a way that works. Before Austin, semantics had a huge fucking hole in it, and everyone was doing their damndest to ignore the hole. What excited people was that he suggested a way to fix the hole without starting from scratch.)
posted by nebulawindphone at 11:14 PM on May 20, 2008
It's hard to believe, but the generation of logicians who went before him mostly ignored that fact. Ms. Saint gives a good discussion of why they were wrong to ignore it — as does Austin himself, for that matter — but it's also worth thinking about why they did ignore it.
Modern semantics came about in a sort of ass-backwards way. The first semanticists weren't linguists at all, but logicians and logically-minded philosophers. At the start of the 20th century, people were incredibly optimistic about formal logic. Some big-deal foundational works in mathematics had just come out (The Grundlagen and the Principia Mathematica were the biggies) and Gödel's incompleteness theorems hadn't dropped yet. For centuries, math had been this sort of haphazard trial-and-error affair with no real rules, but for about half a century in there, between the 1880s and the 1930s it looked like that had changed for good: people thought they were on the verge of creating a formal system that would generate all the mathematical truths and none of the falsehoods. Turns out they were wrong, but if they'd been right it would have been incredibly exciting news.
So the next question some people asked was, If we can do it for mathematical truths, can we do it for other kinds of truth? The Vienna Circle, in particular, were interested in using formal logic to generate philosophical and scientific truths. The idea was, Well, if anything expressible in logical form can be proven true or false, then all we need is a way to translate natural-language sentences into logical form, and then we can prove which sentences are true and which ones are false. Again, it turns out their project was doomed to failure — in any logical system that's worth two shits, some sentences are undecideable, but they didn't know that yet.
Anyway, that project of translating sentences into logical notation is where modern semantics comes from. Eventually, Gödel's theorems hit the scene, and people realized that semamtics wouldn't solve all their philosophical problems. But by then, they'd also realized that semantics was interesting in its own right, and they kept at it. For the most part, they kept at it with one of the Vienna Circle's original assumptions intact: that all sentences could be translated into logical propositions, even if some of 'em were undecideable, and that a sentence's meaning was determined by its truth conditions. Using that assumption, they generated some nifty and powerful results (read up on compositionality if you're interested), but it was a false assumption, as it turns out, and it limited what they could do. They couldn't, f'rinstance, say anything coherent about basic social uses of language like promising, naming or apologizing.
So that's the environment in which Austin was writing. The semanticists of his day needed a big fat reality check: someone needed to point out that "What time is it?" and "Fuck you!" and "I apologize" don't really have truth conditions, and that their meanings can't be expressed fully in formal logic.
The other nice thing about Austin's work, though, is that it didn't force you to throw the old stuff about truth conditions out the window. He acknowledged that some sentences do have a layer of truth-conditional meaning, and that the work people had done on truth conditions might still be valid. He just proposed some additional layers of meaning ("illocutionary force" and "perlocutionary force" are the usual terms, and IIRC they're the ones he uses) that cover the practical, social, non-truth-conditional stuff. That was actually a pretty nifty trick, although again it seems obvious in hindsight.
(Jeezis Christ that was long. Sorry. I do this stuff for a living, more or less, if grad school counts as a living. Short version: What you can do with Austin's discovery is do semantics in a way that works. Before Austin, semantics had a huge fucking hole in it, and everyone was doing their damndest to ignore the hole. What excited people was that he suggested a way to fix the hole without starting from scratch.)
posted by nebulawindphone at 11:14 PM on May 20, 2008
I seem to remember Austin also throwing out the suggestion, although not arguing for it, that there may be no such thing as pure description, that all language might be performative to some extent. This would fuck with other subjects in philosophy like truth and rationality and science that often seem to rely on the possibility of pure description of the world. Anyway, yeah, the book seems like small change because he's just talking about different types of linguistic utterances, but that's what a lot of philosophy of language comes down to. Maybe if you read Wittgenstein's Philosophical Investigations you'll get a sense of what sorts of problems motivate philosophy of language in the first place.
posted by creasy boy at 12:27 AM on May 21, 2008
posted by creasy boy at 12:27 AM on May 21, 2008
The great thing about Austin's philosophy is that it works with ordinary language -- the sort of things we say every day, without necessarily reflecting on what it is we're saying. This is a very appealing and liberating way to do philosophy, because it suggests that pretty much everything we need to know is already available to us in the words we use; all we have to do is unpack it. It's also liberating because it focuses on what we have in common (our language), rather than suggesting, as earlier philosophers (Russell, Ayer) had done, that we are all the prisoners of the 'sense-data' inside our own heads.
But your question is a very good one. 'What's the big deal with speech-act theory?' is a question I've often asked myself. I came to all this via my postgraduate work on early modern religious history, which got me reading a lot of sixteenth- and seventeenth-century theology -- and it didn't take me long to notice that Protestant theologians like Calvin seemed very familiar with the idea of a performative utterance (e.g. a king saying 'I pardon you', Jesus saying 'this is my body'), which they used to illustrate the notion of a sacrament (something which, in Calvin's words 'performs what it signifies', like a seal on a contract). I'm not a medievalist, but I'm pretty sure one could trace the origins of this idea back to Thomas Aquinas and the medieval scholastics. So, with all due respect to Ms Saint, I think it's mistaken to say that 'Austin created the idea of performative utterances' or that 'before Austin there just wasn't an explanation of what happens when the bride and groom say 'I do''. I have a suspicion, which may be unfair, that Austin & co knew less than they should have done about the history of philosophy.
But even if Austin didn't create the idea, he certainly created a philosophical framework for it which has proved very useful in all sorts of ways. One area where it has had a lasting impact is cultural anthropology. If you'd asked me, I'd have said that Clifford Geertz's theory of 'thick description' came from Austin -- though, looking again at Geertz's Interpretation of Cultures to refresh my memory, I see it actually comes not from Austin but from his Oxford colleague Gilbert Ryle. Nevertheless I think the point stands. Geertz's theory of culture as (a) expressed in nods, winks, smiles and gestures as well as words, and (b) deriving its meaning from a shared cultural context, like the Balinese cockfight, where all the participants understand what's going on, seems to me to have a lot in common with Austin's theory of performative utterance, where (a) what matters is not just what is being said but what is being done, and (b) the same words may have radically different meanings in different contexts. You can see this in Geertz's example of the Bororo Indian who says 'I am a parakeet', which, as Geertz points out, is not a statement that can simply be refuted by saying 'no you're not, you're a human being'.
The other great thing about Austin is that he is extremely witty -- and here I do agree with Ms Saint, if you don't find him entertaining, then you're missing something fairly basic. It's hard not to like a philosopher who illustrates the notion of 'pretence' with the example of someone at a party pretending to be a hyena, who gets down on all fours and takes a bite out of Austin's leg (to which Austin's response is: 'Not much pretence about that, is there? There are limits, old sport'). Christopher Ricks's essay on Austin (in Essays in Appreciation) is a good introduction to Austin's wit. My favourite Austin story, which is also a perfect illustration of the English use of irony, is about his work for the British intelligence services during the Second World War. A colleague put up a proposal for a commando raid inside enemy lines, which he optimistically claimed would 'leave the enemy surprised'. Austin wrote in the margin: 'Very surprised.'
posted by verstegan at 2:23 AM on May 21, 2008
But your question is a very good one. 'What's the big deal with speech-act theory?' is a question I've often asked myself. I came to all this via my postgraduate work on early modern religious history, which got me reading a lot of sixteenth- and seventeenth-century theology -- and it didn't take me long to notice that Protestant theologians like Calvin seemed very familiar with the idea of a performative utterance (e.g. a king saying 'I pardon you', Jesus saying 'this is my body'), which they used to illustrate the notion of a sacrament (something which, in Calvin's words 'performs what it signifies', like a seal on a contract). I'm not a medievalist, but I'm pretty sure one could trace the origins of this idea back to Thomas Aquinas and the medieval scholastics. So, with all due respect to Ms Saint, I think it's mistaken to say that 'Austin created the idea of performative utterances' or that 'before Austin there just wasn't an explanation of what happens when the bride and groom say 'I do''. I have a suspicion, which may be unfair, that Austin & co knew less than they should have done about the history of philosophy.
But even if Austin didn't create the idea, he certainly created a philosophical framework for it which has proved very useful in all sorts of ways. One area where it has had a lasting impact is cultural anthropology. If you'd asked me, I'd have said that Clifford Geertz's theory of 'thick description' came from Austin -- though, looking again at Geertz's Interpretation of Cultures to refresh my memory, I see it actually comes not from Austin but from his Oxford colleague Gilbert Ryle. Nevertheless I think the point stands. Geertz's theory of culture as (a) expressed in nods, winks, smiles and gestures as well as words, and (b) deriving its meaning from a shared cultural context, like the Balinese cockfight, where all the participants understand what's going on, seems to me to have a lot in common with Austin's theory of performative utterance, where (a) what matters is not just what is being said but what is being done, and (b) the same words may have radically different meanings in different contexts. You can see this in Geertz's example of the Bororo Indian who says 'I am a parakeet', which, as Geertz points out, is not a statement that can simply be refuted by saying 'no you're not, you're a human being'.
The other great thing about Austin is that he is extremely witty -- and here I do agree with Ms Saint, if you don't find him entertaining, then you're missing something fairly basic. It's hard not to like a philosopher who illustrates the notion of 'pretence' with the example of someone at a party pretending to be a hyena, who gets down on all fours and takes a bite out of Austin's leg (to which Austin's response is: 'Not much pretence about that, is there? There are limits, old sport'). Christopher Ricks's essay on Austin (in Essays in Appreciation) is a good introduction to Austin's wit. My favourite Austin story, which is also a perfect illustration of the English use of irony, is about his work for the British intelligence services during the Second World War. A colleague put up a proposal for a commando raid inside enemy lines, which he optimistically claimed would 'leave the enemy surprised'. Austin wrote in the margin: 'Very surprised.'
posted by verstegan at 2:23 AM on May 21, 2008
Thanks, everyone. I hadn't dreamt of receiving such thoughtful and elaborate replies. I suppose that part of the problem for me is that I've heard of performative utterances before I could attach the term to a specific philosophical achievement, so that it's now difficult for me to imagine a time when people were not aware of them. As for Austin's humor and readability, I'm not entirely convinced -- it's a little too Oxbridge for me. But I'll give it another go. Thanks again.
posted by limon at 5:24 AM on May 21, 2008
posted by limon at 5:24 AM on May 21, 2008
People have given very good answers here so I will just try to add one thing. People learning semantics/pragmatics for the first time often have the feeling that some of what they are learning is "blindingly obvious" as nebulawindphone puts it, and I think your question is an instantiation of this.
However, what you have to understand is what often seems obvious to us about meaning really isn't from any kind of a priori or neutral viewpoint. This is because we are native speakers of a natural language and have all sorts of mechanisms that go along with this to deal with meaning, and we deal with natural language meanings constantly and automatically. However, all or most of this is (apparently) highly human specific and highly specialized. So if your goal is to develop a theory that models human behavior, or develop a computer that makes use of semantics to understand natural language, many things that seem obvious to native speakers really aren't at all. Austin may not have been explicitly trying to model human behavior, but that is the way in which his aims fit into the modern study of semantics within linguistics, viewed as a cognitive science (along with the other ordinary language philosophers.) So, the idea that one can use a communication system in all sorts of ways to "do things" is highly non-trivial, and only seems so to you because you have that kind of communication system built into you. In fact, it is pretty non-trivial to understand how humans conceptualize what it means to "do things" in the first place.
As to practical implications, aside from the cognitive science questions about understanding human behavior and cognition (which you may or may not find practical), this stuff does get used occasionally in computer science, to model communication in multi-agent systems (here's a paper I pulled up via google). I don't know offhand if it gets used in natural language processing, and I suspect not on any wide scale (for cultural reasons), but it could be.
posted by advil at 9:46 AM on May 21, 2008
However, what you have to understand is what often seems obvious to us about meaning really isn't from any kind of a priori or neutral viewpoint. This is because we are native speakers of a natural language and have all sorts of mechanisms that go along with this to deal with meaning, and we deal with natural language meanings constantly and automatically. However, all or most of this is (apparently) highly human specific and highly specialized. So if your goal is to develop a theory that models human behavior, or develop a computer that makes use of semantics to understand natural language, many things that seem obvious to native speakers really aren't at all. Austin may not have been explicitly trying to model human behavior, but that is the way in which his aims fit into the modern study of semantics within linguistics, viewed as a cognitive science (along with the other ordinary language philosophers.) So, the idea that one can use a communication system in all sorts of ways to "do things" is highly non-trivial, and only seems so to you because you have that kind of communication system built into you. In fact, it is pretty non-trivial to understand how humans conceptualize what it means to "do things" in the first place.
As to practical implications, aside from the cognitive science questions about understanding human behavior and cognition (which you may or may not find practical), this stuff does get used occasionally in computer science, to model communication in multi-agent systems (here's a paper I pulled up via google). I don't know offhand if it gets used in natural language processing, and I suspect not on any wide scale (for cultural reasons), but it could be.
posted by advil at 9:46 AM on May 21, 2008
Austin's funniest comments are usually in his footnotes. From Sense and Sensibilia:
We are about to watch, from seats high up at the back of the stadium, a football match in which one of the teams is Japanese. One of the teams comes running into the arena. I might say (1) "Hullo. They look like ants", or (2) "They look like Europeans". Now it is plain enough that in saying (1), I do not mean either that I am inclined to think that some ants have come on to the field, or that the players, on inspection, would be found to look exactly, or even rather, like ants. (I may know quite well, and even be able to see, that for instance they haven't got that very striking sort of nipped-in waist).
posted by KRS at 10:54 AM on May 21, 2008
We are about to watch, from seats high up at the back of the stadium, a football match in which one of the teams is Japanese. One of the teams comes running into the arena. I might say (1) "Hullo. They look like ants", or (2) "They look like Europeans". Now it is plain enough that in saying (1), I do not mean either that I am inclined to think that some ants have come on to the field, or that the players, on inspection, would be found to look exactly, or even rather, like ants. (I may know quite well, and even be able to see, that for instance they haven't got that very striking sort of nipped-in waist).
posted by KRS at 10:54 AM on May 21, 2008
« Older How do they rank how popular a website is? | Where to find intelligent American female voices? Newer »
This thread is closed to new comments.
The big deal is that Austin created the idea of performative utterances. Before Austin, it was assumed that any thing you could say would be either true or false -- that everything about language could be understood in that way (as in, for philosophers. I know nothing bout linguists or anyone else). So, everything about how humans used words was interpreted as saying something true or saying something false, no other options, no other interpretation. It was a very binary view about how human beings are capable of using language.
Think about it: everyone had thought language was for the purpose of making statements that were true, but sometimes we messed up and said something false. End of the story, for these previous philosophers. What Austin is doing is saying that how we use words (this basic part of human life) is completely different from that. He's saying something about how people interact with each other, how we interact with truth, and how we interact with the world around us. This changes our whole understanding of how people work and what words are.
Think about a marriage ceremony. Before Austin, there just wasn't an explanation of what happens when the bride and groom say, "I do." It was assumed what they said was "true," but how exactly could that be? What would it mean to say it is true that "I do" in that context? Does it make sense to turn to the groom, scoff, and say, "That is false!" after he says "I do"? Not really, no. This means there was a failing in how marriage ceremonies were understood. Something so very common as a marriage, and we didn't know how to account for it.
Or something even more common! "I apologize for insulting you." Again, no account of what could make this true or false. It just isn't right to respond to that with "That is false!" Instead, you might say, "No, you don't." This difference just doesn't make sense without an Austinian account. Or, say, "I bet you ten dollars" -- what are the truth conditions for this beyond you saying the words felicitously? And on and on and on... All of these very common, very normal, very important ways that we use words and, before Austin, people just didn't have any way to make sense of them. It was a giant, gaping hole in our understanding of human communication.
If you're not convinced that this matters.. If you think it's not enough just to gain a greater understanding and want, instead, some hard facts about how this theory can impact human lives... Well, then why are you reading philosophy?
(...But I'm not saying there aren't practical implications of Austin's work. I just don't know them, personally. And feel free to MeMail me if I haven't done good enough of a job explaining the Austin-love, or if I missed your point, or whatever.)
posted by Ms. Saint at 8:34 PM on May 20, 2008