Last Word or First Word

What degree of authority did Austin think pertained to ordinary usage? What "ultimate" wisdom did he think implicit in the idioms that it accepts and those it rejects? Austin’s representations on these scores were repeatedly very modest, but the critics of ordinary language philosophy seem determine to saddle Austin with the view that ordinary usage delimits the pale and the boundary of sense. Here is a passage from “A Plea for Excuses.” Judge for yourself whether Austin regards ordinary usage as the final arbiter, or “last word”, on what we can and cannot say.

Certainly ordinary language has no claim to be the last word, if there is such a thing. It embodies, indeed, something better than the metaphysics of the Stone Age, namely,…the inherited experience and acumen of many generations of men. But then, that acumen has been concentrated on the practical business of life. If a distinction works well for practical purposes in ordinary life ( no mean feat),…then there is sure to be something in it. Yet this is likely enough not to be best way of arranging things if our interest are more extensive or intellectual than the ordinary….
And it must be added that superstition and error and fantasy of all kinds do become incorporated in ordinary language, and even sometimes stand up to the test of survival… Certainly, then, ordinary language is not the last word: in principle it can everywhere be supplemented and improved upon and superseded. Only remember, it is the first word. [ Phil. Papers 3rd Ed., p. 185 ]

Please note that the door stands wide open to other ways of talking " if our interests are more extensive and intellectual than ordinary." And talking not just in other ways, but in better ways : ordinary language "can everywhere be supplemented and improved upon and superseded." Austin is not merely making a concession in principle. He goes on to cite some chapter and verse. Modern psychology, through its studies displacement and compulsive behaviours, is in the process of supplementing ordinary language's virtual silence on these ways of acting. ( p. 204 )

Philosophers who feel their enterprise menaced by Austin's programme do not understand it. Austin is not out to put philosophers out of business because they speak in "other ways." He asks only that they first pause to understand how ordinary language addresses the issues they wish to discuss, and then , if they wish to propose new usages, be prepared to show that they have in fact coherent improvements to offer. The idioms established in ordinary language have already proven themselves clear enough and useful enough to compete successfully in the struggle for survival against other idioms ( linguistic Darwinism ). Passing this test of survival bestows some authority on the idioms of ordinary speech. Austin does not claim more.

Gentle readers:

I crave the benefit of your refined sensibilities in a matter of usage.
In his famous 1956 prolegomenon“A Plea for Excuses”, Austin offers us these four sentences to illustrate the ways in which the position of an adverb can affect its meaning.

a1. He clumsily trod on the snail.
a2. Clumsily he trod on the snail.
b1. He trod clumsily on the snail.
b2. He trod on the snail clumsily.

Austin claims that first two sentences ( a1,a2) imply that the treading was incidental to the performing of some other act, while the second two examples (b1,b2) imply that it “very likely, his aim or policy” and "what we criticize is his execution of the feat." [ Phil. Papers 3rd Ed., p.199 ]

The position of an adverb can indeed affect its meaning.
In a1 the adverb is in what linguists call the medial position. In a2 in the initial position. In b1 in the initial-end position ( or postposition). In b2 in the end position. Different kinds of adverbs prefer different positions, and different positions have different connotations, especially when the positioning is uncommon.

B2, the grammar books tell us, is the common position for an adjunct adverb of manner. Shifting "clumsily" to initial position in a2, called "fronting", is probably the least common positioning for this kind of adverb.

But setting these general considerations aside, and looking just at the four sentences above—uttered, should we assume, in the same context—do you agree with Austin’s call that that the a’a are incidental and the b’s deliberate?

Once you have stabilized your intuitions on that question, there are two further pitches from Austin to swing at. Consider the result of setting off the adverb in b1 with commas:

c1. He trod, clumsily, on the snail.

C1 “might be used for a1”, Austin says.

And a2 might be used as “a poetic inversion for b2.”

I shall contain myself and invite your comments.

( One issue: is acting "clumsily" or "awkwardly" something one can do as a matter of policy? I suppose in some unusual circumstances someone who is normally skilful and careful could be accused of being "intentionally clumsy", but that pairing sounds very odd, almost oxymoronic. The implication is perhaps that one is pretending or feigning clumsiness. Normally, however, clumsiness denotes a lack of dexterity and flexibility that afflicts a person who most definitely does not wish to be so afflicted. No author or artist cultivates a deliberately clumsy style. )

Truth and Infelicity

"Infelicity” is a portmanteau term that Austin favored for the variety of the misfortunes that can beset our attempts to do things with words. An infelicity may mar or compromise or completely undermine the warning or promise or command we are trying to utter. A useful taxonomy of infelicities appears at the beginning of HOW TO DO THINGS WITH WORDS (Lecture II), to which I direct the interested reader.

My interest in what follows is with the ways infelicities can affect the truth value ( and illocutionary force ) of our utterances. This is a big topic that would not be exhausted by a score of dissertations, so perforce my efforts will have a narrow focus, and the “big picture” will often be out of focus. But the big picture includes understanding how the descriptive/evaluative notion of truth actually functions in a natural language like English. A project worth spilling a little ink over, I think.

As part of this project, in another place, I have begun to investigate how grammatical flaws affect the truth value and illocutionary force of utterances. No general conclusions can be drawn at this point, but I have been surprised to find how tolerant users of natural languages are of many types of grammatical error. Someone who reports, for example, “each of these books have a dust jacket” is allowed to have said something true if each of the books referred to in fact has a dust jacket. The conspicuous solecism of pairing a singular subject with a plural verb form is not regarded as compromising the fact that he has truly reported that all the books in question have dusk jackets.

In this place, I have set myself the task of beginning to look at how a speaker’s use of an “inappropriate” referring expressions affects the truth value and illocutionary success of his utterance. I shall confine my attention to referring expressions that are being used to identify someone present at the time of speaking. These expressions that are presumed to have what linguists call “situational reference”. I shall explain what I believe is comprised in the use on an “inappropriate” referring expressions.

Suppose Bill and I wish to talk about a woman who is sitting in plain sight about 30 feet away from us. How can we refer to her? If both of us know her name, then we can use her name ( “Susan” ) to refer to her. Or, if there is no other woman present to confuse the reference, we can use the pronoun “she.” Or we could the noun phrase “that woman” ( perhaps accompanied by a nod or a gesture ). Or the description “the woman sitting over there on the terrace.”

The use of the name “Susan” to refer to this woman would be inappropriate if her name is not Susan. The use of “she” or “that woman” would be inappropriate if the person were actually a man. The description “the woman sitting on the terrace” would be inappropriate if the person were actually sitting next to, but not on, the terrace. In general, a speaker’s use of expression in an attempt to refer to someone immediately present is an inappropriate use if the term does not apply to the person in any of the ways just enumerated.

It is obvious that the inappropriate use of a referring expression can completely sabotage the meaning and illocutionary success of an utterance. “That woman is Russian,” said in plain view of several burly men standing at the bar, identifies no one and conveys no clear message.

In some cases, though, the inappropriate use of an referring expression does not necessarily lead to illocutionary failure. I tried to exhibit such a case in a previous post ( “Trouble at the Pussycat Lounge” ). In that example, a speaker’s attempt to identify the woman he intends to speak about did not fail, despite his use of the mistaken desription “the woman drinking champagne,” because both listener and speaker shared the very reasonable belief that the woman was drinking champagne ( and no other woman was ). The woman was drinking a clear liquid from a champagne flute. And even if the listener had had some special reason to doubt that the woman was actually drinking champagne from the flute, he quite naturally took the speaker’s use of “the woman drinking champagne” to be situationally equivalent to the wordier, more guarded expression “the woman who appears to be drinking champagne from a flute.” There was in the circumstances no confusion in anyone’s mind as to the intended referent.

Nevertheless, although we grant that the utterance was fully successful as a warning, we are disinclined to say that the speaker’s utterance “the woman drinking champagne over there is carrying a gun” is true---or, for that matter, false. It’s true, we agree, that the woman I believed to be and spoke of as “drinking champagne” was carrying a gun. But that true proposition is not said the one I uttered. My way of referring to her clearly assumes a misdescription of her as a champagne drinker. Arguably, I didn’t make a false statement, either, by alleging anything false about the woman in question. She was in fact carrying a gun, as I asserted, but not drinking champagne, as I assumed.

[to be continued]

Since I seemed to have been "outed" as someone with closet procivilities toward ordinary language philosophy--just don't tell my mother-- somewhat before I had planned to announce this site publicly, may I ask the reader to realize that these essays are all works in progress that want serious revision before I would turn them loose in the world. I am trying to explore the possibilities of a very new medium as a vehicle for serious philosophical discussion. Some people--Maverick, Dissoi, Laudator-- have found their feet and voice quickly here, but I am still experimenting.

Trouble at the Pussycat Lounge

I’m not going to lead you on. The trouble I’m talking about is what Austin would call a (possible) rhetic misfire. Firearms are involved, but no one is actually harmed in the telling of this story. Hang on for a few paragraphs and you will understand.

It is Saturday night, and Bill and I are sitting at our usual table at the Pussycat. Our table faces the bar, not the stage, because we are here to discuss metaphysics and semantics, not anatomy. I notice that Bill is paying attention to one real looker at the bar. I noticed something earlier that I think it would be useful for Bill to know something about her. The woman is carrying a automatic pistol in her handbag, a Glock .45 cal, I think, from my brief glance at it when she opened her purse.

The woman has been drinking rather conspicuously from a champagne flute. The Pussycat is not noted for its fine cellar, but presumably it is decent vintage. Probably a 2004 Korbels, I’d guess. My problem is how to identify this woman for Bill so he will know which of the woman at the bar I’m talking about. No one else is drinking from a flute so I decide to go with
“The woman over there drinking champagne is carrying an automatic pistol in her bag.”

Unbeknownst to me or Bill, the woman is in fact drinking Perrier in the flute. But Bill has no difficulty identifying whom I’m referring to. He says, “I’ve been watching, her. You noticed, ah? What a looker! “
“Yes,” I say, “she’s very attractive. But it’s real gun, Bill. The slim .45 cal Glock. Not a little toy.”
“You saw it?” he asks.
“I saw it, “ I reply.

I’ve warned Bill that the woman he was showing an interest in is packing, and packing something dangerous. As a rule, the ladies at the Pussycat don’t go in that sort of hardware. My warning illocution has succeeded and has had the intended perlocutionary effect: Bill is alerted and cautious.

But what about the fact that I used as my referring expression a definite discussion that is not literally true of the lady in question, a so-called vacuous definite description? Did that, or could, that have caused any problems? The first thing to notice is that it in fact caused no problem for Bill in identifying the woman whom I was speaking about. He immediately understood whom I meant and that I was warning him that that woman with the flute was packing a gun. There was no miscommunication, no uncertainty about reference or meaning on his part. My meaning and reference were crystal clear.

“But the woman wasn’t actually drinking champagne and that could have caused a problem for Bill.” How? How did Bill understand my use of “the woman drinking champagne”? She was obviously drinking something from a champagne flute. I could have said instead “the woman whom I assume to be drinking champagne from a champagne flute”, but why bother with his qualified, wordy description when it obviously didn’t matter exactly what she was drinking from the flute. Didn’t Bill understand that that was what I meant? Didn’t he understand that I, like he, was making a reasonable assumption from her use of a flute? The wordy, qualified description is otiose and distracting in a circumstance where it doesn’t matter for the purpose of identifying the proper referent.

“But suppose Bill had known that she was drinking Perrier rather than champagne? Couldn’t that have confused him” How could he have known, sitting here with me? And even if somehow he had, would he have automatically assumed in the circumstances that I just meant the only woman drinking a clear bubbly liquid from a champagne flute? Again, how was it plausible for either of us to know it wasn’t champagne, and what did it matter whether the liquid in the flute was champagne or Perrier? The piece of information was irrelevant to the function of identifying the intended referent. My communication was completely successful.

In this situation, “we have a substantial intuition that the speaker said something true of the [wo]man to whom he referred in spite of his misimpression [ about what she was drinking]. So says Saul Kripke in “Speaker’s Reference and Semantic Reference ( 1977).” But that is not quite right. Several of the things we believe here are clearly true. I believe that I identified the right woman for Bill with the description “ the woman drinking champagne.” I believe that I warned him that she had a handgun. I believe he understood me completely. All of these are true. But from these intuitions it does not follow that what I said was literally true. What Bill assumed I meant to say, “The woman drinking what I assume to be champagne is carrying a gun.” was clearly true. And so was “The woman drinking from a champagne flute…” But, I submit, that we have no clear intuition that the statement I actually uttered, “the woman drinking champagne is carrying a gun” , was true here. She was not, after all, drinking champagne. The surprising lesson seems to be that that fact did matter to the success of my communicating what I wanted to communicate.

Suppose the woman drinking from the flute is a Russian national. I know this and assumed that Bill does also. The rest of the women at are hometown girls. So I say to Bill, “The Russian babe over there is carrying a gun in her purse.” It’s true that she is the only Russian at the bar and she is carrying a gun. But in fact Bill hasn’t a clue whom I referring to. He thinks the woman I’m talking about is a hometown girl. So “the Russian babe” means nothing to him. He assumes I’m trying to refer to some woman present, and possibly warn him about her gun-toting ways, but whom I’m referring him escapes him. My communication has seriously misfired because I’ve used a “true” description that Bill does not connect to the right woman. Much better to have a description such as “the woman drinking champagne” which, though literally false, is strongly credible and has a clear unique referent in this situation.

When a speaker wishes to tell his companion something about another person who is present and in view of both of them, he has a range of referring expressions to chose from. He can use a proper name ( Natasha Smith ), a pronoun ( she ), a definite noun phrase ( the woman ) or a definite description (The woman drinking champagne ). If the speaker is confident his listener knows the name of the woman, “Natasha” may be his easiest choice. Or if there is only one woman in view, “she” or “the woman” may be simplest. If there are several women is view, only one of whom appears to be drinking champagne, he may prefer “the woman drinking champagne”. All of these expressions are meant to serve the purpose of identifying clearly and simply the person about whom the speaker wishes to say something. If what he means to say about the woman in question is actually true, then it is true that what he said of her is true. But if his referring expression does not actually apply to the woman in question, there is a serious problem with saying that what he said was true.

A final point. Ironic descriptions are often literally false. Suppose that the woman I wished to refer to was stunningly beautiful, and I believe that my listener, Bill, shared this judgment. Then, since Bill also knows I am given to ironic comments, I might risk "That really unattactive woman over there is carrying a gun", using an ironic and literally false description. The fact there is not, in anybody's judgment, a really unattractive woman in sight does not matter to the success of communication.

Can you do it?

“Can you do x?” often asks no more than do you know how or have the ability to do x. But in some contexts, especially where doing ( or being asked to demonstrate ) x is in immediate prospect, “can you do it?” asks me for a pledge that I will do it.

“Can you sink that putt? I have ten dollars that says no.” The situayion is that I’m facing a twelve foot putt and I’m pretty good at that distance. On the practice green, I normally sink 2 out of three at that range. So there is no question whether I am able to sink a 12 foot. The question and challenge is whether I will sink this one. Do I have $10 worth of confidence in my ability on this one putt? That question has less to do with my ability than with other things, as we all understand.

“Yes, I can do it” is not a relevant reply here. I’ve been challenged to bet that I will. Either I accept ( “Your on” ) or decline ( “no bets” ).
But compare the same question asked over dinner by a novice looking for advice on his poor putting game. That is about my ability and wants an answer such as "Yes, I can. Two things are vital. You must..."

“He could have done it.” ( crime )

A detective is reviewing his list of suspects in an art heist. The theft from a wealthy home seems to be an inside job. John is a nephew of the owner and was staying at the house. The detective concludes: “John could have done it. He had the ability and the opportunity. But, although he might be our thief, I would put three other people in front of him as more likely suspects.”

In effect, the detective concedes “could have, but probably didn’t.” And since we don’t know whether he did or didn’t, there is no connection to any counterfactual suppositions about what John would have done if…. “Could have” in this usage is particularly detached from the counterfactual “would have if…”.

Notice how “could have” focuses on ability and opportunity while leaving other “springs” of action open. The detective could say “He could have done it, but I don’t see a compelling motive”, or “He could have done it, but it would be completely inconsistent with his character” . But the detective could not say “He could have done it but he had no opportunity to do it” or “He could have done it but he was not physically able to do it.”

“He could have done it “ascribes ability to an agent at a past time. Notice that it is the context and background information that determine whether there is an implication that he did not do what we suppose he was able to do. Grammar does not require that “could have” be uttered in reference to something that didn’t happen. “Could have done x ” is not necessarily, or even usually, the apodosis verb in a counterfactual conditional with a suppressed protasis.
“Could have” works fine ascribing an ability in a context like this where there is no counterfactual supposition in play about the use of that ability.

“I’ll do it, if you like.”

We saw in a previous post that if clauses sometimes do not state conditions that apply to the main clause. “There are cookies in the kitchen if you want them”, but also, whether or not you want them. Whereas, “there are cookies in the kitchen if I remembered to buy some”, but not, whether or not I remembered to buy some. We can bring out the ellipse and compression inherent in the former sentence if we paraphrase as a conjunction, “There are cookies in the kitchen, and help yourself if you want some.” The first conjunct states a fact, the second makes a polite offer that doesn’t assume the person we are addressing is a cookieholic. Notice that speakers can make two very different responses to this offer depending upon which of the conjuncts they focus on. “No thanks” declines the offer. “No there aren’t. I just looked” rejects the existential statement.

So what about “I’ll do it, if you like”? We are at a departmental meeting and the chairperson asks for someone to help him review a new set of promotion policies send to all Humanities departments by the dean. As a tenured faculty member with light admin duties, I feel it proper to OFFER my services in response to his request for a volunteer. So I say “I’ll do it, if you like.” “I’ll”, we might note, has to be “I will”, not “I would.” “I would do it…” declines to help. What is the function of the if clause in this offer? Why not just say “I’ll do it’?

Suppose I’d said instead “I will do it, if you’d like to assign it to me.” That expansion explicitly recognizes that it is something that I’m offering to do, not something I’m claiming. The chairperson must decide whether he will accept my offer and assign the task to me. He may have reasons for not wanting me to do job. My offer to help, if he sees fit to appoint me, acknowledges the chairperson’, right and authority to decide whom he wishes to assign. I do not presume that he must accept whomever volunteers (first). The if clause in fact states a genuine condition on my doing the review. I will not do whether or nor he wishes to assign to me!

“I might have holed that 35 foot putt”

Not even Tiger Woods, I think, should say of a missed 35 foot putt “I could have holed that putt.” That’s would be boasting. “Could have” in the context of a failure locates that failure within the scope of what one has the ability to do with some reliability. We discussed an example of that in the previous post. “I could have holed that 14 foot putt” implies, first, that I have the ability to make 14 foot putts regularly and, second, that in this particular case there was a definite, fixable cause of my error. I missed the putt because I did Y, and Y is something I can ( and usually do ) correct.

Thirty-five foot putts are a different story. I do make them on occasion. I practice them regularly, but my long putt game is happy if I hole one in twenty at that kind of distance. It won’t be sheer luck if I sink the next one, but it is not part of my game to do so. I might do so, it’s possible that will do so, but to say “I can sink 35 footers” would be boasting of an ability I do not have. Eighteen or nineteen failures in twenty tries does not attest an ability. So, of my failure to sink a 35 footer I can say “I might have holed that putt”, implying that this was one of the easier 35 footers. But I should be not boasting that I could have that holed putt. People are likely to ask me to put my money where my mouth is, and then I shall I have the opportunity to play it again, and fail again, this time loosing money as well.

“I could have holed that putt”

I just missed an important putt. I had been putting well all day and this was only a 14 footer. I had holed a 17 footer on the previous hole. There is no question of my ability to sink puts out to 14 feet with some reasonable probability of success. But of course not every time. In fact, if I can regularly sink three out of four at that distance, I’m doing well. So I missed this one. What needs to be explained? Human abilities and skills are inherently fallible.

Let’s assume I’m not just whining about my failure but trying to critique it. I replay the put in head looking for something I did wrong. I recall that my grip didn’t feel quite right. My right hand was too high, and when my right hand is high, I tend to push the ball to the right, which is what I did. For some reason I didn’t stop and correct my grip, but tried to compensate with my shoulder. “There’s my mistake! That shoulder trick almost never works.” OK, then, diagnose complete. Conclusion: “I could have holed that put.” That is to say, there was an identifiable and fixable cause of my error. To be specific, “I could have holed it if I had stopped and fixed my grip.”

Moore’s claim is that utterances of “I could have done X” are inherently iffy. Does this case support it? It might seem so, but we must pay careful attention to what the speaker is saying. With an utterance of “I could have holed it” he alleges only that there was some definite and fixable cause of his failure. He does not specifically say what he believes that cause was. In this case, as we saw, he’s decided that it was his failure to hold the putter properly. He could go on to say “I could have succeeded by holding the putter corrrectly”, or equivalently, “I could have succeeded if I’d held the putter correctly.” But this further claim, please note, is a more specific than the simple “I could have done it.” That last is not a disguised or truncated form of the conditional. Moore’s account does not hold up.

“I Could” and “I Could Have”

Austin starts to investigate these idioms on page 214 of “Ifs and Cans” (Phil Papers 3rd ed). The view he is opposing, first floated by G.E. Moore, is that “I could/could have” declarations are necessarily iffy, in the sense such declarations always and only occur as the consequents of a counterfactual conditional, though the antecedents are sometimes omitted in ordinary speech as understood.

Austin carries the first half of his (counter)claim very easily. Consider the utterance “I could do that 20 years ago”, referring perhaps to my ability at age 30 to run a 16 minute 5K. (No longer, alas!) There is no suppressed condition here. I claim that I was able to do something at age 30 that I no longer can do. What supports my claim of fleetiness of foot years ago? I ran several timed 5K races and finished under 16 minutes in those days. I could because I did. I could do that 20 years ago. This could is not iffy.

Austin also seems to carry the second half of his (counter)claim easily. Consider the utterance “I could have ruined you this morning (although I didn’t).” That is Austin’s example on p. 215. The parenthetical, I think, is redundant and would be omitted in ordinary speech. The speaker is again describing a real ability he had at a previous point in time. There is no (unfulfilled) contingency associated with this ability , as there would be if I had said “I could have ruined you if I had had one more vote against you.” In this latter case, I allow that I did not in fact have the ability to ruin you. I fell short in the motion to unseat you by one vote. In actuality, I could not ruin you.

“I could have ruined you this morning” , by contrast, is not a counterfactual claim. It is not iffy in the sense of depending on any implied and unrealized condition. I assert that I actually had, not lacked, the ability to unseat and ruin you. I could have ruined you BY calling for a confidence of the trustees who were overwhelmingly against you. The means to do so were in my power.

“I could have ruined you this morning if I had chosen to” is a redundant expansion, and the if-clause does not state a condition upon my ability. I had the ability whether or not I wished to use it. The proper counterfactual expression would be "I would have ruined you this morning if I had chosen to." What is counterfactual is my acting to ruin you, not my ability to do so. There is nothing counterfactual about my ability to have ruined you.

Austin carries the day on both points. But a lot more obviously remains to be said about "I could/would/should/ might have."

“I shall marry her if I choose to” ( a case of illocutionary suicide?)

This puzzling sentence, in a slightly different formulation, appears at the top of p. 214 in “Ifs and Cans.” ( Phil Papers 3rd ed) It is followed by the equally puzzling “I intend to marry him if I choose.” I really haven’t a clue what these sentences are supposed to say or mean.

Austin has just said that “I shall” in such utterances is not an assertion of fact but an expression of intention. I agree, and that’s precisely why I don’t understand these utterances. If John says he intends to commit marriage, I make the assumption that he has already decided or chosen to commit marriage (God help him). Otherwise I don’t understand his saying he intends to marry. An intention to do X assumes a prior decision to do X. How then can I possibly declare my intention to marry someone and at the same time imply that I have not yet decided to do so? As an attempt to express an intention to marry , “I shall marry her if I choose to” seems to commit illocutionary suicide with its if clause.

“I shall marry her if/provided she consents” is fine. I express my intention to marry her, but I note, quite sensibly, that her consent is needed to make this marriage happen. Her consent is not assumed by my intention or a condition of it. Ordinary speakers understand that I’m saying “I intend to marry her, and the marriage will happen provided she consents.”

"There are cookies in the kitchen if you want them."

[ Deleted by the incompetent software at Blogger.com ]

"I paid you back yesterday, if you remember"

I take this example from a footnote in Austin's "Ifs and Cans" ( p. 213 Phil Papers 3rd Ed ). Austin credits it to Peter Geach and describes it as another "excellent' example to illustrate the difference between "the conditional if and the if of doubt or hesitation". I don't disagree that the if clause does not state a sufficient condition for the apodosis, but let's see if we understand exactly what going on with this usage.

Austin suggests as a paraphrase the question "I paid you back yesterday, don't you recall?" That is possible, I suppose, but I can't hear someone asking a question in that idiom. If a person genuinely wants to ask a question, he will say "Do you/do you not recall that I paid you back yesterday?"

"I paid you back yesterday, if you remember", first of all, asserts that I paid you back. Your recalling at this time that I did so is obviously no condition of that having happened in the past. The utterance also suggest to the listener, politely, that he take some time and recall what happened. Less, or perhaps much less polite, would be the suggestion "You will remember that I paid you back yesterday." That directive, I suppose we'd call it, comes close to be a command or a demand, in a situation where giving commands or making demands is not appropriate. We prefer an idiom which does not demand the listener remember something, but suggests that if he think about it, you will perhaps remember that I paid you.

That is the paraphrase I would offer. "I paid you back yesterday. If you will take a minute to reflect, perhaps you will remember that I did so." But I don't ordinary speakers need a paraphrase. They understand quite well what the speaker is doing and saying.

I don't like Austin's description of this usage as an if of doubt or hesitation. The speaker is signaling no doubt or hesitation. He is making a polite suggestion or request rather than a demand. But should look at some similar usages to confirm this.

"I can if I choose to" ( Part II)

Austin notes in a footnote ( p. 213, Phil. Papers 3rd Ed ) that "I can if I choose to" used to mean predominantly something like " I can and I have the choice", where "can" is referring to my (legal) right to do the thing in question. I cannot confirm the historical part of Austin's observation but the usage he alludes to persists today.
My neighbor owns a riverfront property on Florida's Inland Waterway. Building on these lots is heavily restricted. We visit his property and I ask him, "Can you build a home on this lot?" He replies, "I can if I choose to." He means that his title allows him to put a residence on the property, and he may do so. He hasn't yet decided. He does not mean that he has the right IF he chooses to.do so. Perhaps at law some right arise by fiat, but that is a bizarre supposition in this case and we all discount it. I understand my neighbor to have said, in a standardly elliptical way, " I have the right to build here, and if I decide to, I shall. But that decision is not made."
This agrees exactly, I believe, with our analysis in the previous post. There is not odd or paradoxical about a conditional used in this way.

"I can if I choose to"

Austin ponders the peculiarities of this conditional in "Ifs and Cans." It fails to allow contraposition. It does allow an inference to the apodosis, and to "I can whether or not I choose to." In short, it declines to behave like an ordinary sufficient condition/causal conditional in which the antecedent guaranteees the consequence. So what's going on? Is this some deviant usage that needs to be diagnosed and treated with a healthy dose of grammar? What does this case tell us about the English conditional?

Nothing profound, I believe, or even puzzling. Like many other utterances in ordinary language it is compressed and elliptical usage. Logicians, focused only the words uttered, miss the rest. Logic is like studying icebergs: if we look only at the words spoken, we will miss most of the meaning.

I ask a neighbor, who has a very long drive to clear of snow, "can you get that cleared this morning?" He replies, "I can if I choose to." What has he said to me? Has he said or implied something pretty strange: that his ability to clear the snow quickly depends on choice to do so? No one, I think, takes it in this sense. He asserted, first of all, that he can clear it . He has the ability, he is able. His tone of voice supports his claim of ability. He could have belabored his point with " I can do, don't doubt that for a minute, but that is IF I choose to." But that expansion is unnecessary. He is saying, plainly enough, that it is his actual performance, not his ability to perform the task, that remains in question.

He is telling me, secondly, that he wasn't yet decide to tackle the task. If does, I can be sure, he will get it done this morning. But the decision is not yet made. He could said less elliptically, " I can, and I shall if I chose to." But this conditionality is well enough understood by his listener in the shorter form of the utterance. No one thinks he saying that his choice will suddenly create an ability that is otherwise in doubt. This would be a bizarre claim in the circumstances, would it not? So we discount it. But his actually doing what he is able to do is not unconnected to his choice to act. I have the ability, he is saying, and I choose to it, it will get done.

Ordinary speakers are naturally economical in the form of expression they use. Needless words and phrases get dropped because we do not need them to understand the speaker's meaning. But logicians come along and forget this, and look at "I can if I choose to" and think it a great puzzle. It isn't. It is compressed idiom, like much of natural speech, in which an unnecessary and understood clause ( "and I shall" ) is left unspoken. Put it back in, and all of the logical puzzles go away.

Ifs and Cans

I take my title from J.L. Austin's justly famous paper of the same name, celebrating its 50th anniversary this year and none the worse for a little graying. A young philosopher interested in the logic of conditionals and modals, as well as puzzles about human abilities and actions, could do much worse than to begin here and absorb the lessons Austin offers.

I shall use this place to investigate some of those lessons, and inevitably I shall arrive at conclusions that conflict with Austin's. I'm sure he would approve.