Monday, August 15, 2011

Implicature and lying

Philosophers say things like: "Asserting 'There is no conclusive proof that Smith is a plagiarist' implicates that there is a genuine possibility of Smith's being a plagiarist." (And yet taken literally "There is no conclusive proof that Smith is a plagiarist" is true even if no one ever suspected Smith of plagiarism.) However what one implicates not only has propositional content, but also illocutionary force, and both both the content and the force are implicated. So if we want to be more explicit, we should say something like: "Asserting 'There is no conclusive proof that Smith is a plagiarist' implicates the suggestion (or insinuation or even assertion) that there is a genuine possibility of Smith's being a plagiarist." Which of the forces--suggestion, insinuation or assertion--is the right one to choose is going to be a hard question to determine. Maybe there is vagueness (ugh!) here. In any case, we don't just implicate propositions--we implicate whole speech acts. A question can implicate an assertion and an assertion a question ("It would be really nice if you would tell me whether...").

I used to wonder whether the moral rules governing lying (which I think are very simple, namely that it is always wrong to lie, but I won't be assuming that) extend to false implicature. I now realize that the question is somewhat ill-formed. The moral rules governing lying are tied specifically to assertions, not to requests or commands. One can implicate an assertion, but one can also implicate other kinds of speech acts, and it only makes sense to ask whether the moral rules governing lying extend to false implicature when what is implicated is an assertion or assertion-like.

And I now think there is a very simple answer to the question. The moral rules governing lying do directly extend to implicated assertions. But just as these moral rules do not directly extend to other assertion-like explicit speech acts, such as musing, so too they do not directly extend to other assertion-like implicated speech acts, such as suggesting. The rules governing an implicated suggestion are different from the rules governing an explicit assertion not because the implicated suggestion is implicated, but because the implicate suggestion is a suggestion. If it were an explicit suggestion, it would be governed by the same rules.

That said, there are certain speech acts which are more commonly implicated than made explicitly--suggestion is an example--and there may even be speech acts, like insinuation (Jon Kvanvig has impressed on me the problematic nature of "I insinuate that...") that don't get to be performed explicitly (though I don't know that they can't be; even "I insinuate that..." might turn out to be a very subtle kind of insinuation in some weird context).

I think the distinction between the explicit speech act and the implicated speech act does not mark a real joint in nature. The real joint in nature is not between, say, explicit and implicated assertion, but between, say, assertion and suggestion (regardless of which, if any, is explicit or implicated). Fundamental philosophy of communication does not, I think, need the distinction between the explicit speech act and the implicated speech act. That distinction is for the linguists--it concerns the mechanics of communication (just as the distinction between written and spoken English, or between French and German) rather than its fundamental nature.

15 comments:

Heath White said...

I think I disagree with the thrust of this post, for three reasons.

First, there is an important difference between explicit assertions and implicated assertions, namely that the former is not cancellable while the latter is.

Second, I think the distinction between "asserting" and "suggesting" is overdrawn. Roughly, I would say that suggesting that p is equivalent to asserting that "p with low probability" or "p is worth considering as possibly true". We can put the tentativeness into the content or into the act but fundamentally it doesn't make much difference.

The third reason is a piece of evidence from practice, not theory. There is a famous story about I think Athanasius who was fleeing down the Nile from some soldiers. They asked him where the bishop was (not recognizing him) and he said something like "Not far away!" The point of the story is supposed to be that Athanasius cleverly avoided lying but saved his skin nevertheless. However if your theory were correct (and the story is true) then Athanasius implicated that he, the speaker, was not the person they were looking for, since "not far away" implicates "not right in front of your nose". So I don't think your theory can be what has traditionally been behind the rules against lying, and hence it cannot be the traditional view of (norms of) assertion.

Alexander R Pruss said...

1. "I would like a hamburger" could be seen as an autobiographical assertion that implicates a request, but the assertion can be canceled, for instance by saying: "Actually, I hate hamburgers, but I promised a friend that I would try one."

2. Maybe, though I am not sure. Taking it this way forces one to change the propositional content. To suggest that p, then, is to assert a different proposition with a different subject matter. I am not sure that's right.

3. I don't think the Athanasius statement implicates the assertion that he is far away. It implicates the hint that he is far away.

One should probably also distinguish between intentional and non-intentional implicature. We could, for instance, see Athanasius' statement as an attempt to assert a truth that would be unhelpful but that wouldn't lead to further questions, without any attempt to assert the further claim that he's not right there. (Double effect might even apply.)

Alexander R Pruss said...

Heath:

Another thought on Athanasius. Clearly the norms governing what he did with respect to the proposition <Athanasius is not here> are different from the norms governing an assertion of that proposition. Hence, he did not assert that proposition.

So the question is: Is "implicated" an alienans? If it is, then we can still say that he made an implicated assertion of that proposition, even though he did not make an assertion of it. I was, however, assuming that "implicated" is not an alienans, in which case from the fact that the he did not assert the proposition it follows that he did not implicate an assertion of the proposition either.

Here is a little argument that "implicated" is not an alienans. Someone who makes an implicated request is still requesting something. (Actually, here the line between explicit and implicated request blurs. In current English, "It would please me very much if p" typically implicates a request for p, but we can easily imagine the phrase fossilizing into an explicit request.) Someone who implicates a practical commitment is still committed. Someone who implicates a question has still asked. OK, these probably beg the question.

Suppose I am wrong about all these, so that the norms governing implicated assertions/requests/questions are not a special case of the norms governing assertions/requests/questions. We can then imagine people who have special explicit forms of speech governed by the same norms by which our implicated assertions, implicated requests and implicated questions are governed. Very well: our implicated request then is just their shmequest, say. The difference between the implicated and the explicit remains merely a matter of the mechanics of communication--whether you're like us who make shmequests by using implicature or you're like them who make shmequests explicitly--rather than being a matter of the fundamental philosophy of communication.

The mechanics of communication concerns, inter alia, how illocutionary forces are marked. They can be marked by combinations of all sorts of markers: tone of voice, posture, context more broadly understood, choice of synonym, performative verbiage, and, yes, the choice whether to implicate or not.

Alexander R Pruss said...

I should say, however, that there is a distinction in the vicinity of the distinction between the implicated and the explicit (in a different sense from that in Brandom) that I think is a matter of fundamental philosophy of communication, and that's the distinction between what a person asserts and what one can (defeasibly) deduce about the person's beliefs and intentions from the fact that she made that assertion. Thus, from someone's saying "Athanasius is not far from here", one can deduce that the speaker doesn't know that Athanasius is here. This deduction is a contingent empirical deduction--people who say things like that tend not to know that Athanasius is right there. And of course the speaker can manipulate the deductions about her mental states that people draw from what she says. But that manipulation need not be a matter of communication, any more than a robber's leaving false fingerprints (e.g., imprinted on one's gloves) is a matter of communication (though it could be, if we had different conventions).

It's of course important that implicature not be a matter of mere empirical deduction. This is particularly important for implicated consent, I think. Suppose the doctor asks me if she can perform some operation. I say: "I would be very glad if you did it." The doctor can deduce from that that I want the operation. But that I want the operation is insufficient to allow her to perform the operation: consent, and not mere desire, is needed. And the consent must be given, and not merely internal, in the case of a competent patient able to communicate.

Here, think of a case where I really want the operation because it would save my life, but I believe the operation is unethical and hence I do not consent to it. Maybe in the face of temptation, I even wrongfully make an inner consent to it--but what the doctor needs is externally manifested consent. However, externally manifested implicated consent is enough, if the implicature is clear enough, as it is in the case of "I would be very glad if you did it."

Heath White said...

At this point, I think the notion of “implicated assertion” has become excessively rarefied.

Another thought on Athanasius. Clearly the norms governing what he did with respect to the proposition are different from the norms governing an assertion of that proposition. Hence, he did not assert that proposition.

We could make this argument about nearly any implicature, yes? I ask you for a letter of recommendation, and you write one that says, “Heath shaves regularly. Sincerely, Alex Pruss.” Obviously you are implicating that I am no good as a philosopher. But by an argument parallel to the above, we could argue that you are not implicating the assertion that I am no good as a philosopher, because the norms governing what you wrote and what you implicated are different. I think that shows, rather, the traditional view, namely that norms governing implicatures and norms governing explicit statements are different.

Still, there are some implicated assertions that do have the norms of explicit assertions, e.g. around medical consent, and we could call these “implicated assertions” and have a special theory of them. But there may be other ways to understand them. (a) As particularly strong idioms in English, e.g. “I would like…” for “I hereby request/consent to…”, which leave no practical doubt about one’s intentions. (b) As ambiguous (I think this how to understand the hamburger example). “would like” or “want” in English is ambiguous between (phenomenological) “would get pleasure from” and (intentional) “satisfies a goal I have”, and the latter is sufficient for a request or consent.

I think the distinction between what one can deduce from the content of an explicit assertion , and what one can deduce from the act of asserting it, just is the distinction between the semantics and the pragmatics of language, and implicatures generally fall on the pragmatic side. That is why they can be cancelled and why they are much more context-dependent than the content of explicit assertions.

It is also true that any pragmatic communication could be codified into an explicit form, even the explicit form of just the words used to pragmatically communicate it. But it doesn’t follow that they are so coded now, or that if they were so coded that the semantics of what they say and of what we say would be the same. (I hope that makes sense and is relevant.)

Alexander R Pruss said...

1. I don't know that I'd say this about every implicature. There may be cases where we are governed by the full norms of assertion. For instance, it might be that a part of the point of the oath to speak the truth, the whole truth and nothing but the truth is to make one's assertion-like implicatures be governed by the full norms of assertion. Even if that's not so in fact, we can certainly imagine a context like that.

2. In the end, I am not sure there is any philosophically significant difference between "particularly strong idioms" and implicatures. There is only the difference that the phrasing in the idioms is less flexible.

3. Minor point: It's not enough for consent to a procedure that the procedure satisfy a goal that the patient has. (A particularly interesting family of cases to think about is ones where the procedure satisfies the someone's goals iff only if the procedure is not consented to.)

4. "I think the distinction between what one can deduce from the content of an explicit assertion , and what one can deduce from the act of asserting it, just is the distinction between the semantics and the pragmatics of language, and implicatures generally fall on the pragmatic side."

I don't think so. The stuff that's in the pragmatics of language is norm governed in ways that what can merely be deduced is not.

Consider this: there are reasons of trustworthiness not to engage in false implicature; but these reasons do not apply to, say, laying down misleading footprints in the sand.

Suppose research shows that people who talk about cockroaches are more likely to have an undiagnosed cancer. The inference from "Smith spoke about cockroaches" to "Smith probably has an undiagnosed cancer" doesn't seem to be a matter of the pragmatics of language.

Heath White said...

Suppose research shows that people who talk about cockroaches are more likely to have an undiagnosed cancer. The inference from "Smith spoke about cockroaches" to "Smith probably has an undiagnosed cancer" doesn't seem to be a matter of the pragmatics of language.

You’re right. It has something to do with inference based on interpreting someone as an intentional agent (“what was he up to in making that assertion”) rather than as a causal mechanism.

Consider this: there are reasons of trustworthiness not to engage in false implicature; but these reasons do not apply to, say, laying down misleading footprints in the sand.

I think, at most, this is the point at issue. In any case I can think of where I would be willing to lay down misleading footprints, I would also be willing to engage in misleading implicatures, and for the same reasons.

Alexander R Pruss said...

I think more than just as an intentional agent, but as a linguistic agent. Suppose there is a machine that gives a drink to anybody who makes the noise "green". Sam, knowing this, intentionally utters the sentence "Grass is green" near the machine. We can infer that Sam wants a drink. But it's not a language pragmatics involving inference. Something to do with meaning is needed.

I am inclined to think that the false implicature violates some kind of a conversational norm, but the false footprints violate no conversational norm.

Heath White said...

Well, Grice has those maxims of cooperation and anyone engaging in false implicatures is violating those.

I would think it is analytic that leavers of misleading footprints are not violating any *conversational* norms. Other sorts of norms, maybe.

The question remains about the moral status of each of these.

Heath White said...

Here is another way to think about this topic. I am wondering if we are not confusing two things when we speak of implicatures and speech acts.

Suppose my wife says to my son, “Your bed is not made.” In this context, her utterance has the force of a command to the effect of, “Make your bed.” We might try to call this an implicated command, on the grounds that the speech act is not explicit, i.e. what she says is not in the imperative mood. But the usual way to distinguish explicit from implicit performatives is whether there is a prefix mentioning the speech act, e.g. “I hereby command you to make your bed” is explicit. Grammatical moods are just a shorthand, somewhere between very explicit speech acts (“I hereby command…”) and very implicit ones (“Your bed is not made.”)

Back to “Your bed is not made.” Suppose we distinguish (relatively) explicit and implicit speech acts. This one is pretty implicit. Nevertheless it is a genuine command and has all the force thereof. We could make the same distinction in assertions or requests, etc.

Implicatures, however, are best construed as propositions. They are the propositions one can infer from the fact that a certain (explicit or implicit) speech act has been performed (with caveats about the importance of meaning in making the inference). The classic examples involve inferences from the making of assertions, e.g. “he is not a good philosopher” from an off-topic letter of recommendation, or “Athanasius is not here” from the assertion that A. is not far away. However there could be such inferences made from other speech acts. One response to an awkward question, for example, is to deliberately change the subject by asking the waiter for another beverage, or to respond with a question. (“How’s your sex life?” “How’s yours?”) One might infer from this that the requester didn’t want to answer the question, and the implication could be quite deliberate. Or if a commanding officer asks for volunteers for an undisclosed mission, one might infer that the mission was very dangerous, as otherwise he would be giving orders. These teeter on the border of being mere empirical generalizations but I think they could become conventional enough that they deserve the title of implicature.

Distinguishing the implicitness of a speech act from an implicated proposition has several advantages. (1) It’s not revisionary. (2) It agrees that implicit/explicit speech acts is not a real joint in nature. (3) It can agree with whatever moral/normative restrictions one wants to put on speech acts, including assertion. (4—at least I regard this as an advantage) The morality of misleading implicatures is just the morality of laying false evidential trails, made specific to certain kinds of linguistic evidence.

Alexander R Pruss said...

Yes, I think this is a helpful way to think about it.

It would explain why it is that the literature has focused on assertion-like implicatures, rather than request-like or question-like implicatures. For the things we can reasonably infer from what people say are more like information they give us (and hence assertion-like) than the effect of some other kind of performative. It would also explain why the content of an implicature is generally described as a proposition, rather than a message (I think of a message as an ordered pair consisting of an illocutionary force type and a content (often, but perhaps not always, a proposition)).

Notice, though, that while your view isn't all that revisionary, it is deflationary. Implicatures just become a special case of the general non-linguistic category propositions we can learn from people's intentional behavior, distinguished only by the fact that the intentional behavior is linguistic and that the meaning is relevant, which doesn't change the nature of the inference.

I do have three reservations, however.

1. We (i.e., philosophers) talk of someone implicating something, and I think that many of the things we so describe do include an element of intentional communication, rather than mere evidence-giving. For instance, the letter of recommendation example may be like that. We would say something like "Reading between the lines (or, the line!), he's telling me that the candidate has no good philosophical qualities." There is communication going on here. This doesn't really challenge your division, which I like, but it does make your division more revisionary--some things that a lot of philosophers would take to be implicatures probably end up being on the side of implicit speech.

More later...

Heath White said...

On 1: It is possible to intentionally lay non-misleading trails of evidence....

Alexander R Pruss said...

2. We can correctly say "She implicated that Leibniz is a second-rate philosopher," even when for us the speaker's assertion is neither evidence for the claim that Leibniz is a second-rate philosopher nor for the claim that the speaker thinks so, for instance because the speaker is such a notorious liar that what she asserts is evidence of its negation. We could try to bring in the normative: the implicated claim is what the speech should be evidence of. But then we no longer have the nice deflationary view.

3. Cases like those in 2 are rare. But rather less rare are cases where the person is so incompetent in a subject area that her implicating that p is no evidence of p. This is a fairly minor point, because you can replace the claim that one implicates p in a speech provided that one (in the right way) gives evidence of p with the claim that one implicates p in a speech provided that one (in the right way) gives evidence that one believes p.


By the way, I think that perhaps the letter of recommendation case should be assimilated to the implicit assertion rather than implication. When in answer to a question Q one states a proposition P, sometimes one is not asserting P, but some related proposition. For instance, if you ask me: "Could you tell me something false about whales?", I can answer with "Whales are fish." But in the context of an answer to the question, what I am asserting is: <<Whales are fish> is a falsehood about whales>.

A letter of recommendation is in part an answer to an implicit question (set by the social practices of writing letters of recommendation) of what all the good job-relevant characteristics of the candidate are. Therefore, by saying that Heath shaves regularly, I am asserting that Heath's shaving regularly is the only job-relevant positive characteristic of Heath. And hence I am simply lying.

Heath White said...

A quick recursion to first principles: Grice introduced the idea of implicature to save Russellian truth-conditional accounts of meaning from a certain criticism of them on behalf of Wittgensteinian use-accounts of meaning. This was the “what we would say” criticism. When someone asserts, “The present king of France is bald” we would not say “that’s false” but “you’ve misunderstood”, and the W’n takes this to refute Russell. Grice argued, on the contrary, that this does not prove that the assertion isn’t false, it just proves that the assertion implicates something which is false, and we might respond to that implicature rather than to the strict meaning of the assertion.

The “what we would say” criticism assumes a competent and trustworthy interlocutor; if “we” are grossly ignorant or pathological liars, there is no telling what we would say. So I think the concept of implicature is a relation between a particular assertion-type (or speech act-type) in a particular context-type, and an implicated proposition, assuming a quasi-idealized competent and trustworthy speaker. Implicatures are not usefully construed as relations between particular token assertions, by possibly incompetent or lying speakers, and implicated propositions. We can of course use expressions like “she implicated p” but this should be taken to mean “the sort of thing she said, in the context she said it, implicates p.”

(FWIW, I have no objection to my views being deflationary.)

I think we ought to distinguish pretty sharply between assertion and communication, with “telling” being somewhat ambiguous. If we think Grice was at all on the right track, then what one asserts is defined by the assertion’s truth-conditions. E.g. “Heath shaves regularly” is a lie only if it is not true that Heath shaves regularly. This no doubt communicates something much different in the context of a recommendation letter. But “communication” is pretty broad—we can communicate by laying all sorts of trails of verbal or non-verbal evidence, with the intention that the other person come to understand what we are doing. (This is more or less how Grice wanted to understand meaning, which is puzzling now that I think about it.)

Example: In LOTR, one of the hobbits has been kidnapped by orcs and deliberately drops a tell-tale brooch. The intention is not merely that the pursuers will find it but that they will know that he dropped it in order for them to find it, and the hobbit knows that if it is found the pursuers will know his motives, and the pursuers know that the hobbit knows the pursuers know his motives, and so on. This is communication of a sort: the hobbit is “telling” the pursuers where he has been (conditional on them finding the brooch). But it is not an assertion. And if the orcs (or the hobbit, acting deceitfully) had taken the brooch and laid it somewhere misleading, it would not be a lie.

Alexander R Pruss said...

I more or less agree about the brooch case. The force here is assertion-like (you could imagine a case where you drop an engagement ring, and the force is request-like), but it's not quite assertion. However, I distinguish between the hobbit and the orc dropping it deceitfully. If the hobbit drops it deceitfully, and the deceit is meant to work like typical lies do (i.e., those who find it infer that the hobbit wanted to communicate something with it), it's lie-like.

If the orc drops it deceitfully, it may not be lie-like, because those who find it aren't supposed to infer that the orc meant to communicate something with it. The orc doesn't solicit trust in himself with an intention of breaking it, in the way the hobbit may be soliciting trust in himself. (Here I am following Jorge Garcia's excellent account of the wrongness of lying as consisting in soliciting-and-breaking trust.)

Here's another interesting case. Suppose an orc, imitating (successfully or, more likely, not) the hobbit's voice calls out: "It's me, Pippin, I'm right here", in a place where instead of Pippin there is am ambush. I am not sure that's a lie. That's because I am not sure the orc is soliciting trust in himself, but relying on the listener's trust in Pippin. But it's a tricky thing, because the "I" probably refers to the orc ("He said he was Pippin, but he wasn't"), and so we might say that the orc is soliciting trust in himself, under a false pretense.

An intermediate case: A hacker installs a speech content distorter in a cell tower that waits for you to utter some particular sentence, and distorts it into a falsehood. Did the hacker lie, or was she merely deceiving?