Understanding Singular Terms
Imogen Dickie (Toronto/St Andrews)
Abstract
This paper unfolds as follows.
§1 argues for a biconditional connecting the ‘aboutness’ of ordinary beliefs with what I call ‘cognitive focus’.
§2 sketches a kindred account of our understanding of one another’s uses of the singular terms we use to express these beliefs: understanding is joint cognitive focus.
§§3 – 4 use the phenomenon of ‘felicitous underspecification’ involving demonstratives to argue for the view sketched in §2, and gesture towards some additional applications to puzzles about our understanding and use of singular terms.
Though these gestures will be preliminary, I shall try to do enough to provide a backdrop for the following immodest claim: the view introduced in this paper – understanding of singular terms is joint cognitive focus – promises to generate a new and rewarding route across this boggy and uncertain terrain.
Biography
Imogen Dickie is Shelby Cullom Davis Professor of Philosophy at St. Andrews. Before that she was Professor of Philosophy at the University of Toronto. Her research is centred on issues about reference and representation in language and thought. Her book Fixing Reference came out with OUP at the end of 2015. She is currently working on a sequel which extends the inquiry to questions about other minds, linguistic communication, and our conception of the world as objective.
The Goal of Conversation
Zoltan Szabo (YALE)
Abstract
Dickie (2020) presents an argument against the traditional, broadly Gricean view of conversation. She argues that speakers must sometimes be more specific than required for sharing knowledge on a topic of common concern. Her proposed solution is to claim that the goal of conversation is not just sharing knowledge, but also sharing cognitive focus. In response, I argue that her proposal faces both conceptual and empirical difficulties, and that the traditional view can handle the problem of non-specificity by acknowledging that in order to sustain mutual trust, conversational participants should be less then optimally efficient.
Biography
Zoltán Gendler Szabó is the John J. Saden Professor of Philosophy at Yale University. As an undergraduate, he studied mathematics and philosophy at the Eötvös Loránd University in Budapest and he received his PhD in philosophy from the Massachusetts Institute of Technology. Prior to his current appointment, he taught at Cornell University. Szabó’s work focuses on the intersection of philosophy of language and metaphysics. He has published on a wide variety of topics, including the semantics of descriptions, modals, tense, aspect and the expression of mental attitudes, on the role that conversational context plays in interpretation, and on the limits of systematicity in assigning meanings to complex linguistic expressions. His main current project explores the way in which natural languages incline us towards certain metaphysical beliefs over others.
The supplementary volume for the conference will be available from 3-13 July.
Graeme A Forbes (Kent)
10th July 2020 at 5:14 pm
Hello, I’m going to be ‘chairing’ these comments.
Mainly what that means is that I will trying to keep the conversation going, and to gently remind you that, despite being on the internet, this is a conference talk and the usual norms of politeness, relevance, and charity still apply. Also try to be patient. The discussion will develop over the course of the weekend; it is not a requirement that people sit clicking refresh for 72 hours straight.
I’d also like to thank Imogen Dickie and Zoltan Gendler Szabo for two excellent papers. I’m sure that there’ll be lots for us to discuss.
Zoltan Gendler Szabo
10th July 2020 at 5:58 pm
Hi Imogen and Graeme,
happy to be here; I will be checking in regularly.
Graeme A Forbes
10th July 2020 at 11:35 pm
Imogen (if I may). I’m a little confused about the relation between same-saying and reference on your view. Shared cognitive focus gives us same-saying, and will give it to us felicitously in some cases where reference is indeterminate. But reference still seems to be part of the story; you are offering an alternative to the causal and descriptivist accounts of reference. I fear I’m getting myslef confused, so I’m unable to see the wood for the trees.
imogen dickie
11th July 2020 at 11:51 am
Hello Graeme, hello Zoltan, hello anybody else…
I’m writing up a few points in response to Zoltan, and will post them later today.
Here’s a first response to Graeme. The idea is supposed to be that the answer to the ‘Felicitous or not?’ question with respect to an utterance of ‘That is F’ is coming from whether the speaker has done enough to enable the hearer to join in a shared cognitive-focus-sustaining activity. So a case where there’s felicitous underspecification is one where a single activity generates focus relations on more than one object. In that case, there’s not a unique answer to the question ‘What’s the referent of the speaker’s use of “that”?’.
Graeme A Forbes
11th July 2020 at 1:13 pm
Thanks. I’m with you so far. I suppose I’m confused about what more general lessons we take from this (perhaps my mistake is to try to go off-piste from the specific cases we’re dealing with here). In the case of underspecification we have sufficient shared cognitive focus to achieve felicity, but the problems seems to be that there are *too many objects* that we might be referring to; no *unique* answer.
Can we have shared cognitive focus without there being any objects? (e.g. The Greeks worshipped the same gods as the Romans under different names, but none of those gods are real). It looks like we can have felicity with shared cognitive focus, when there are no referents at all.
Perhaps the metaphor of a focal point is useful; we can have two telescopes focussed from different angles on the same point, even if there is no object there. But what is crucial here is that some counterfactual holds (e.g. ‘were Zeus to be real, ‘Jupiter’ would succeed in referring to him’). So shared-cognitive focus, even when focussed on empty space, still allows for same-saying. Does that sound right?
[Sorry for the flat-footed questions. I’m tempted to introduce this material to my Philosophy of Language course in the Spring, and want to make sure I’ve got my head round it].
imogen dickie
11th July 2020 at 5:16 pm
That’s not flat-footed at all.
I’ve actually been thinking a lot about aboutness-failure and empty names lately. The core of what I want to say is that, because the information-marshalling activity is focus *seeking*, you can engage in it – and your engagement in it can be in good normative order – even though there’s not actually any object that you’re focussed *on*. (It’s as if you reach out your hand intending to pick something up but the world conspires against you: your act of reaching is in good normative order but there’s nothing there for you to grab.) In *Fixing Reference* I was thinking that this gives us a way to make sense of Frege’s notion of a ‘mock thought’: a case of empty singular thought where you marshall information in a way that *would* lock to an object if you weren’t unlucky is a case where there’s normative good order, but no truth conditions.
The natural extension for linguistic understanding will follow a similar pattern. If you and I are using what we take to be co-referring names – suppose you’re using ‘Instanbul’ and I’m an historian so I keep saying ‘Constantinople’, but we’re crunching information in ways that are justified and will tend to get the city’s properties right, that’s a case of shared cognitive focus. If the names are in fact empty, but we’re crunching information in a way that is focus-generating unless the situation is unlucky, then we’re engaging in focus-generating activity which fails to deliver the goods because the world conspires against us. So depending on what you mean by ‘same-saying’ I think we’re…same-saying with respect to this extension of the framework.
Naomi Osorio-Kupferblum
11th July 2020 at 1:55 pm
Thank you for a great talk! I was wondering whether the force of the examples hinges on their being cases of counter-examples to negative existence statements, where the fact that there is at least one counter example is all that is needed? / If this is right, I wonder whether the shared cognitive focus is on F / the (set of the) Fs / anything F rather than the o(s) which happen(s) to be F. – Would there be other examples for felicitous referential underdetermination?
(I find it striking that in a related setting of “name a few”-questions, referential underdetermination would not be felicitous: “Where can I get x?” or even: “What here is inexpensive?” and “Who makes this class worthwhile?”)
imogen dickie
12th July 2020 at 3:11 pm
Hi Naomi. I’m not sure I’m seeing your point with respect to the first part of the question. The QUD in the ‘that’s cheap’ case is an existence question (‘Is there anything you’d say is cheap?’. But in the ‘She’s really clever’ case it’s whether the person whose seminar it was thinks that the class is worthwhile.
With respect to other examples – I think there are certainly examples involving proper names: a famous person’s name might be used to refer to that person, or to that person + their entourage, with the context such that it doesn’t matter which.
Daniel Morgan
11th July 2020 at 11:43 am
Hi, Thanks to both for great talks! I find the ‘intelligibility’ requirement Zoltan mention plausible, especially as applied to the ‘She’s a great student’ example. I’m not sure I got the example offered as the original motivation for the requirement though, the help-with-moving example. What exactly is infelicitous about being asked for help with moving, and then saying ‘Sure’, and then immediately moving on to a different topic of conversation? Even the ‘No I can’t’ response I would find at worst a little brusque.
Zoltan Gendler Szabo
11th July 2020 at 2:57 pm
Thanks. It looks like you agree that saying ‘No, I can’t’ in response to the request is a little brusque, and hence that it violates some sort of norm. The question is only whether it is a conversational norm or some broader social norm. I think it’s the former for two reasons. First, the only non-conversational norm in the neighborhood is the one that says that when one inconveniences another one should normally apologize. But in this case he refusal remains brusque even if it is ‘Sorry, I can’t’. Second, what does make the response smooth is that one proves a reason for saying ‘No, I can’t’, as in ‘No, I can’t. I am out of town that day.’ And providing reasons for one’s utterances seems like a conversational requirement.
I agree that the case of saying ‘Sure’ in response to the request is murkier. To my ear this too feels incomplete without some way of indicting how the promise will be met, i.e. how it fits into the speaker’s plans.
imogen dickie
11th July 2020 at 12:50 pm
This is about what I take to be Zoltan’s central challenge to my proposal. In many cases, what gets communicated in an ordinary, successful, conversational transaction is something general (for example, I might say ‘Sheep are not very bright’). So how can it possibly be that part of the speaker’s goal in making an utterance is to get the hearer to engage in an activity that is directed at cognitive focus on a particular thing or things?
Let me say first that this challenge brings out an important respect in which my view needs to be clarified. Suppose we look at things from the point of view of the hearer. The speaker makes an utterance. The hearer’s language-understanding information processing goes to work. According to what I call the ‘conservative’ view, what this information processing is doing is (something like) looking for a proposition that is specific enough to answer the question under discussion. What I propose instead is that this information processing is trying (a) to achieve cognitive focus on the speaker, and (b) to join with the speaker in achieving joint cognitive focus on things in the world.
I’ve been thinking of cases of general utterances as all to be dealt with in an (a)-ish kind of way: I say ‘Sheep are not very bright’; you process my remark in a way that brings me into cognitive focus for you. If I say ‘That’s not very bright’, demonstrating a sheep, you process the remark in a way that brings me into cognitive focus, and puts us in a relation of joint cognitive focus with respect to the sheep.
This wasn’t made clear in the paper (‘(a)’ is just tucked away hiding….)
OK. That’s a first move.
Zoltan Gendler Szabo
12th July 2020 at 3:00 pm
Hi Imogen. We agree that when someone is making a general conversational move (e.g. says ‘Sheep are not very bright’) she is not trying to share cognitive focus with her addressee on any particular sheep. But I don’t think the goal in this case could be simply to bring the speaker into cognitive focus for the addressee. If I utter ‘Sheep are not very bright’ I draw attention to myself, but for that goal any other utterance (or even a non-utterance, like coughing loudly) would do.
In my response, I suggested that the practical view you propose could be amended: the goal of conversation might be to build shared mental files. When I utter ‘Sheep are not very bright’ I propose that we introduce new file card into the shared mental file we are building with a (plural) discourse referent labeled N and the information that N are not very bright. (That we have a discourse referent for the sheep is attested by the fact that ‘sheep’ is a proper antecedent for subsequent anaphora. For example, after ‘Sheep are not very bright’ I can say ‘They cannot even do long divisions’.)
imogen dickie
12th July 2020 at 3:33 pm
Thanks Zoltan. This is bringing out the fact that I’ve got a lot of detail to fill in with respect to what the *need* that’s driving everything is actually a need for. At the moment I’m calling it, cornily, a ‘need for fellow-travellerhood’: the speaker is trying to get the hearer *both* to focus on the speaker, and to share aspects of the speaker’s information-marshalling activities. Then I’m (I think) fine about saying that part of what we’re trying to achieve in a cases involving general terms is a shared perspective on, for example, sheep in general….
I do want to say that achieving focus on you (by which I mean, processing your remark in a way that will deliver a right verdict as to your mental states unless the situation is unlucky) will very often involve forming a view as to what you are s
But I do think that the ‘hearer is trying to focus on speaker’ thing can do quite a bit of work. I make an utterance. Your language-understanding info-processing gets to work. What’s it trying to do? It’s trying to deliver a ‘She’s proposing such-and-such update’ verdict that registers as focussed. So the claim isn’t just that I’m trying to draw your attention (so that you ‘focus’ on me in the sense of attending to me). It’s that I am trying to show you (and maybe get you to share) an aspect of my perspective on the world.
Bill Brewer
11th July 2020 at 1:13 pm
Thank you both for excellent papers and talks. I have a question for Imogen.
I can see how the conservative account explains the felicitousness of the package case and fails to explain the infelicitousness of the student case. I can also see how your cognitive focus account explains the infelicitousness of the student case.
But can you say a bit more about how the cognitive focus account explains the felicitousness of the package case. For cognitive-focus-sustaining information-marshalling in relation to the package is quite different from cognitive-focus-sustaining information-marshalling in relation to its contents, at least if the ‘sustaining’ is supposed to take place over some time. The two activities of cognitive-focus-sustaining information-marshalling diverge once the packing is removed and binned, and the expresso machine placed on the hob for use – either because the package just is the packing (now in the bin) or because it’s the packing together with the machine it contains (now either destroyed or scattered).
imogen dickie
11th July 2020 at 4:48 pm
Thanks very much Bill.
Your question brings out something very central. Part of what I’m starting to explore in this paper is a shift from thinking about language understanding in terms of a PRODUCT to thinking of it in terms of an ACTIVITY. In what I call in the paper a ‘conservative’ framework, the hearer’s language-understanding information-processing is looking for a product – a representation/proposition that is specific enough to answer the question under discussion. I’m suggesting that the hearer’s language-understanding information processing is (instead) directed towards engagement in an activity. If I say ‘That arrived yesterday’, pointing at the package, I’m proposing that you join with me in an activity of marshalling information to achieve cognitive focus. Now, a focus relation is focus at a degree of resolution: objects indiscriminable at one degree of resolution might be discriminable at another. In my cognitive focus framework, changing the properties with respect to which you’re making decisions in your information-marshalling will often change the resolution of the focus relation. When I say ‘That arrived yesterday’, I’m proposing that you join me in a focus-sustaining information-marshalling activity where the degree of resolution doesn’t distinguish package from contents. The situation might evolve in ways that change the resolution, so that our joint activity zooms in on the contents (you say ‘Oh great I’ve been looking forward to using it for ages!’) or on the package. At the point of the initial remark, both of these evolutionary paths for the activity are open: we are communicating at a level of resolution which does not force an answer to the question of which object I am talking about.
Bill Brewer
11th July 2020 at 5:11 pm
Thanks very much, Imogen,
So now I want to go back to the infelicitousness of the student case. What blocks an analogue of what you just said about the package in that case. At the point when you say ‘she is very bright’ your colleague joins an info-marshalling activity that is for her not yet at a resolution discriminating amongst the women students in the seminar (but does rule out all the men as the relevant subject matter), but is nevertheless on its way to doing so over the course of the subsequent discussion. How does your view then get the desired result that there is an infelicity?
I suppose the general worry is that distinguishing the two cases on your view, as you admit you must, risks feeling a little ad hoc.
imogen dickie
11th July 2020 at 6:05 pm
Ha.
I *think* the answer is that the resolution of the focus relation that the speaker is proposing the hearer join is riding with the predicates – though it must also be subject to modulation by features of context.
Relativisation to the properties you’re in the business of deciding is there in my notion of cog focus: a strand of info marshalling activity is focussed on o iff, for all F such that the strand of activity puts the subject in the business of deciding F-or-not (adding ‘F’ or ‘not F’ or ‘F to this degree’ to the body of beliefs), the subject will be unlucky if the verdict doesn’t match o and not merely lucky if it does. Felcitous underspecification cases are cases where the properties in the domain of ‘for all F’ are such that there’s more than one o for which this condition is met. In the parcel case, the ‘for all F’ ranges over properties which the parcel has iff the espresso maker does as well: when the conversation advances, this shifts. But the initial domain is coming from the predicate the speaker uses (‘arrived today’). In the student/seminar case, the speaker says ‘She’s a really good student’ – introducing straight away a predicate which may be satisfied by one potential referent while it is not satisfied by another. The information-marshalling activity the speaker is proposing that the hearer join in this case is a joint activity of marshalling info at this level of resolution – the level carried in by the predicate. There’s infelicity because the speaker hasn’t provided enough of a demonstration to get such an activity off the ground.
Bill Brewer
11th July 2020 at 6:10 pm
That sounds right to me, thanks, Imogen. Great stuff.
Karl Egerton
11th July 2020 at 1:36 pm
Thank you to both of the speakers for the very interesting talks! I have a question regarding the two proposed problem cases. I wonder whether another explanation for their infelicity might be available, though I’d be interested to hear if either speaker thinks that what I say is already covered by the views they have proposed.
It looks to me like there is a mismatch between the conversation participants on communicative purposes and thus on context (I’ll just explain my thoughts on one): in the cost-of-living example, it matters to the questioner whether the respondent is pointing at, say, a can of petrol, or a single toothpick, since the former would suggest that the respondent concedes that some normal things are cheap after all, while the latter would suggest that the respondent is either being facetious or responding to a much more specific question (is *literally anything* cheap?). By failing to be specific, the respondent acts as though the context is the very specific question, while the context the questioner is operating within is quite clearly the more general question. Should context-mismatch be part of the story for these cases?
Zoltan Gendler Szabo
11th July 2020 at 3:17 pm
Thanks for this question. I agree that there is room for a mismatch in presuppositions in this example. Specifically, it could be that there is no alignment in what the participants take to be the domain of the question under discussion: one thinking they are debating whether every item in certain domain is expensive, the other thinking they are debating whether every item in a different domain is expensive. This could explain the infelicity of the utterance. However, I think the infelicity remains even if we stipulate that there is no such misalignment and we do need an explanation that works even in those cases. Imogen provides one (by proposing a new goal of conversation); I provide another (by appealing to a new norm of conversation).
Rory Harder
11th July 2020 at 5:25 pm
Thanks all. This is for Zoltan.
First, as background, for a part of my dissertation I have been developing a dynamic semantics treatment of demonstratives (along with their associated demonstrations) as well as other definite expressions. So I found your remarks around p. 78 on how to treat demonstrations within dynamic semantics interesting and helpful. I would just like to present to you some difficult data (going back to MacLaran 1982) that is an issue for the simple picture you give.
In short, if demonstrations open new discourse referents, and if both demonstratives and definite descriptions are simple anaphoric expressions, then why are there the following contrasts?
1. Context: The interlocutors are in an art gallery surrounded by paintings.
a. That painting [pointing to one painting] is beautiful.
b. #The painting [pointing to one painting] is beautiful.
2. There are a couple house plants in front of the interlocutors.
a. John loves that plant [pointing to one plant] and Bill loves that plant [pointing to the other].
b. #John loves the plant [pointing to one plant] and Bill loves the plant [pointing to the other].
There is much more to say here: I think that accommodating this and similar data involves deeply modifying the simple picture that considers demonstrations as indefinites, and also that the perhaps natural moves for modifying the way in which demonstratives and definite descriptions are anaphoric (for instance, as in Roberts 2002) will not work. In any case, I would be interested in hearing what you think about this data!
Zoltan Gendler Szabo
11th July 2020 at 6:34 pm
Hi Rory. I don’t think ‘that painting’ and ‘the painting’ are used the same way, although I believe they are both used to indicate that an old discourse referent is to be updated. The difference is perhaps that we use ‘the painting’ the old discourse reference is easy to identify and ‘that painting’ if we think it will require extra effort. The extra effort is often figuring out what the speaker is demonstrating. But sometimes it’s something else. If I ask ‘Do you remember that crazy teacher from high school?’ instead of ‘Do you remember the crazy teacher from high school?’ I suggest that you may have to do some extra thinking in trying to identify which teacher I am talking about.
This (admittedly vague) distinction may go some way explaining why using ‘the painting’ accompanied by a demonstration (as in 1b and 2b) is odd: the speaker choice of words indicates that a discourse referent can be easily identified, but then she makes a special effort to help the addressee to identify it.
Rory Harder
11th July 2020 at 10:00 pm
Thanks for the reply. I would like to make a point in response (but don’t feel obliged to respond again!).
I agree that the anaphoricity of demonstratives and definite descriptions is different in an important way. I am worried though that your description of the contrast in 1 no longer treats demonstrations as simply introducing DRs (in the way normal indefinites do). If the demonstration, which is constant between 1a and 1b, simply introduces the painting DR, then that DR should be just as easily recoverable in both cases. So assuming that a demonstrative requires that its DR is less easily accessible than for a definite description does not explain the contrast between 1a and 1b.
To fill out what I think you are getting at: in the context of 1, there are a bunch of painting DRs “hanging around”, and then the demonstration selects one of them. And since the demonstration is required to select one of them, that already existing DR is not easily accessible. I think this is approximating the correct account of what is happening. However, note that on this treatment demonstrations are not introducing DRs in the simple way indefinites do; rather, they are selecting among ones already in the context in some sense.
Zoltan Gendler Szabo
11th July 2020 at 11:14 pm
I think our disagreement comes from the fact that you consider demonstrations part of the context and I do not. I think they are part of the utterance.
Let’s consider the utterance 1b. To file this utterance, the addressee must identify a discourse referent associated with the condition ‘painting’. I suggest that it is a conventional implicature associated with the definite article that this identification can be done on the basis of information available in the context. But the fact that the speaker enhanced her utterance with a demonstration conversationally implicates that she did not think so (otherwise, being a rational cooperative subject, she would not have bothered with the enhancement). These two implicatures are in conflict, which explains why the utterance is infelicitous.
You are absolutely correct that this explanation cannot get off the ground if the demonstration is part of the context.
imogen dickie
12th July 2020 at 9:41 am
This is Part (B) of me responding to Zoltan’s paper.
In Part (A) (posted earlier), I stress something that should have been brought out more clearly in my paper.The claim is not supposed to be that ALL conversational moves are directed towards engaging in joint cognitive focus on particular things in the world. I’ll quote from the paper: ‘We have something like a need to see others and be seen by them as fellow travellers, and to stand with others and look out at a shared world: a need to sustain relations of mutual cognitive focus with other people, and to join other people in relations of shared cognitive focus on things in the world.’ The claim is that this need is always at work in driving our language-production and language-understanding information processing along. The ‘shared cognitive focus on things in the world’ part is in play when we’re using and understanding singular terms.
This raises the question of WHAT THE ARGUMENT IS for the claim that language production and language understanding are directed by this more complex goal. And here I need to point to the wider project of which my paper for this symposium is an instalment. I’m working on three lines of argument towards this claim (and trying to clarify the claim by thinking about exactly how the arguments for it go…):
1) The ‘from first principles’ argument: if you take my cognitive focus picture of the aboutness of thought, and add the claim that language use enables us to think about other people and have them think about us and join with them in thinking about the world, you get some kind of ‘language use and is directed towards enabling and achieving and sharing cognitive focus’ picture out the other side.
2) The empirical argument: empirical work on cooperation and joint attention provides some kind of support for my kind of view.
3) The argument from explanatory power. Much of this paper was about how the view explains the relation between felicity and specificity for uses of demonstratives. But I’m thinking of that as just one motivating puzzle. I think the framework also generates improvements on extant treatments of a range of phenomena, including conversational implicature, loose use, metaphor, the distinction between lying and misleading, and various ways we misunderstand one another.
[Part (C) coming up.]
Zoltan Gendler Szabo
12th July 2020 at 3:49 pm
I think it is plausible that part of the goal of conversation is to share cognitive focus on each other. This is compatible with the amendment of the practical view I mention in the paper – all we need to add is that the mental files conversations aim to build always include discourse referents for the participants of the conversation.
You say that in conversations we always aim at sharing cognitive focus on each other but the shared cognitive focus on things in the world part is only in play when we employ singular terms. But I think conversations can employ singular terms without being oriented at things in the world – this happens, for example, when a liar tells a story about her non-existent rich uncle Rick, or when an author tells a story about a fictional character Humpty-Dumpty. I think these exchanges are conversations and whatever the individual goals of their participants might be, they have the same shared goal as other conversations.
imogen dickie
12th July 2020 at 6:19 pm
Thanks Zoltan. I’m currently thinking of cases involving empty singular terms as involving the guidance-of-information-processing-by-need-for-focus thing within the scope of pretence. The case I’ve thought about most is standard/realistic fiction. Here the idea is that when you’re reading a standard fiction, you have flipped a ‘pretend!’ switch, and, within the scope of that, are letting your language-understanding information processing do its focus-seeking thing. Other empty names cases will relate to core cases of motivation by the need for shared cognitive focus in other ways – details not worked out yet. There are hard issues here about how the ‘look for cognitive focus’ motivational structure can count as still at work (still guiding information processing) when the participants’ total motivational states contain elements that push us away from the core case.
imogen dickie
12th July 2020 at 2:38 pm
This is Part (C) of me responding to Zoltan.
According to the traditional/conservative/post-Gricean picture, the felicity of a (declarative) conversational move depends on whether the speaker has done enough to make public a proffered answer to the question under discussion. Zoltan and I are both rejecting this view. We’re disagreeing about what to say instead.
Zoltan proposes that we graft what he calls the ‘Intelligibility’ norm onto a traditional, Gricean, ‘Cooperativity + Publicity’ background. On his view, felicity demands not just that the speaker do enough to make public a proffered answer to the QUD, but also that speakers do enough to make it intelligible why they’ve said what they have in the way that they have.
I’m proposing a more radical departure from the traditional picture. On my view, our language-understanding and language-production information processing proceeds as it does because we are trying to make ourselves available as objects of cognitive focus, and achieve cognitive focus on other people, and (sometimes) achieve joint cognitive focus on the world. In this framework, an utterance is in bad normative order relative to the basic thing we are trying to do in conversation iff, given the context, the hearer has not done enough to put a competent speaker in a position to achieve a ‘here’s what S is trying to get across’ verdict which registers as focussed.
There’s more detail to be filled in on both sides. But let me try to bring out a point of overlap between our proposals and a point of difference.
Overlap: On my view, conversational participants are trying to FOCUS on one another. On Zoltan’s, they expect one another to be INTELLIGIBLE. My notion of focus has a kind of intelligibility built into it. Think about focussing a set of binoculars on something: you take it that you’re focussed when the image achieves a characteristic kind of coherence. Similarly, whether your language understanding information-processing is registered by you as bringing me into focus is going to depend on whether it delivers a ‘message expressed’ verdict that registers my making the utterance as a coherent thing for me to have done.
Difference: I’m discarding – in quite an emphatic way in fact – the Gricean claim that linguistic communication requires cooperation. Zoltan’s keeping it. Obviously, in most cases of linguistic communication we ARE cooperating. But I’m after an account of linguistic normative good order which will allow that an utterance can be both fully felicitous and radically uncooperative at the same time. In Zoltan’s story, cooperation is there in its traditional Gricean role, and it’s also there in the discussion of intelligibility: the suggestion is that speakers must make themselves intelligible enough to sustain hearers’ trust (trust that they are, and will continue to be, cooperating interlocuters). In my story, if I say things that undermine your trust in me, this probably reflects badly on me in various ways. But felicity doesn’t require that my making the utterance in the context put you (the hearer) in a position to see my doing so as ‘intelligible’ in a sense relevant to sustaining your tendency to trust me. It requires that my making the utterance in the context put you (the hearer) in a position to see my doing so as fitting into a picture that registers, from your point of view, as bringing me into focus.
Zoltan Gendler Szabo
12th July 2020 at 4:43 pm
I completely agree with this summary. Imogen and I agree that the standard Grice-inspired theory of conversation has a problem – I try to reform the theory, she tries to replace it.
The heart of the Gricean view is the claim that conversation is a cooperative enterprise. In light of all the non-cooperative behavior within conversations (lies, propaganda, obfuscation, innuendo, euphemism, evasion, etc.) this seems hard to believe. The central piece of evidence that all conversation is cooperative was noted by David Lewis in his paper ‘Scorekeeping in Language Games’. It is this: even in highly confrontational situations, conversational participants tend to accommodate each other and accommodation is paradigmatic cooperative behavior. (As Lewis notes, in a competitive game, such a baseball, the score is never adjusted just because one team recognizes that the other assumes it is different from what they assume it is.)
Those who wish to reject the idea that conversation is essentially cooperative will have to explain accommodation away. Take the classic example of someone saying ‘I have to pick up my sister at the airport’ in a context where it is not common ground that the speaker has a sister. The addressee is entitled to complain: ‘Hey, wait a minute! I did not know you had a sister.’ However, such complaints a rare – when the speaker’s assumption is uncontroversial and does not concern the question under discussion the addressee will typically accept that the speaker has a sister and continue the conversation. Proponents of the standard view explain this by appeal to cooperativity. Proponents of the practical view (with or without the amendment I mentioned in my paper) do not. Instead, they might say that there is some conversational norm that prohibits these sorts of complaints, or that the addressee normally fails to notice what the speaker is assuming.
Graeme A Forbes
12th July 2020 at 9:28 pm
Thanks everyone who commented, and thanks to Imogen and Zoltan for engaging with the questions. A shame we can’t all go to drink coffee together, but we can cognitively-focus on the coffee we would have drunk.