Docsity
Docsity

Prepare for your exams
Prepare for your exams

Study with the several resources on Docsity


Earn points to download
Earn points to download

Earn points by helping other students or get them with a premium plan


Guidelines and tips
Guidelines and tips

Understanding Language and Communication: Gricean Meaning and Dialogue Annotation, Lecture notes of Reasoning

The concept of Gricean meaning in language communication, emphasizing the importance of context, intention, and inference. It also introduces the Dialogue Act Annotation ISO 24617-2 Standard, which categorizes communicative functions and dimensions in dialogue. the history of research on deixis, pragmatic implicature, and conceptual metaphor, highlighting the role of context and inference in understanding language.

Typology: Lecture notes

2021/2022

Uploaded on 09/27/2022

benjamin56
benjamin56 🇬🇧

5

(4)

225 documents

1 / 21

Toggle sidebar

Related documents


Partial preview of the text

Download Understanding Language and Communication: Gricean Meaning and Dialogue Annotation and more Lecture notes Reasoning in PDF only on Docsity! Felicity Conditions  Performative speech is neither true nor false, as we’ve  argued, but it can certainly “fail” in some sense of the  word -- there’s some sense in which a performative must  be uttered under “appropriate circumstances.” By way of  example, to bet is not merely to utter the words “I bet …,  etc.”: someone might do that all right, and yet we might  still not agree that he had in fact succeeded in “making a  bet,” or at least not entirely, succeeded in betting. To  satisfy ourselves of this, we have only to to go to a horse  race and, for example, announce our bet after the race is  over. None of your interlocutors, you may be sure, will  believe you to have succeeded in betting. Nonetheless,  under the right circumstances, saying “I bet …, etc.” is  precisely what betting is.    Sometimes performative speech goes wrong and the  intended act -- marrying, betting, bequeathing,  christening, baptizing, etc. -- is therefore to some extent a  failure. The utterance is then, we may say, not indeed false  but in general ​unhappy​. In trying to classify the ways in  which things can go wrong we arrive at the ​Doctrine of the  Infelicities​. Or, as they are now know, felicity conditions.    (A. I) There must exist an accepted conventional procedure  having a certain conventional effect, that procedure to  include the uttering of certain words by certain persons in  certain circumstances, and further,  (A. 2) the particular persons and circumstances in a given  case must be appropriate for the invocation of the  particular procedure invoked.  (B. I) The procedure must be executed by all participants  both correctly and  (B. 2) completely.  (​Γ. I) Where, as often, the procedure is designed for use by  persons having certain thoughts or feelings, or for the  inauguration of certain consequential conduct on the part  of any participant, then a person participating in and so  invoking the procedure must in fact have those thoughts  or feelings, and the participants must intend so to conduct  themselves, and further  (Γ. 2) must actually so conduct themselves subsequently.    Indirect Speech Acts  [...]  Sometimes explicit evidence of our construal of indirect  speech acts rises to the linguistic surface, and we get  adjacency pairs like:  A: “I’d love to help.”  B: “Thanks for the ​offer​.”  Or:  A: “I could eat the whole cake.”  FeedbackElicitation, Stalling, Pausing, TurnTake, TurnGrab,  TurnAccept, TurnKeep, TurnGive, TurnRelease,  SelfCorrection, SelfError, Retraction, Completion (of partner),  CorrectMispeaking (of partner), InitGreeting, ReturnGreeting,  Apology, Thanking, ​etc.  Across 8 dimensions: ​AutoFeedback, AlloFeedback, Time  Management, Turn Management, Own Communication  Management, Partner Communciation Management, Social  Obligation Management, Discourse Structuring​.    If enough and large enough corpora are so annotated, this  scheme (and others like it) obviously have the potential to  give rise to rich theoretical and practical (engineering)  work in the field of dialogue, broadly construed.    Coming up: reasoning from context  The modern work on deixis really beginning with (John)  Lyons (1977 monumental work on Semantics) and (Chuck)  Fillmore.    The work on pragmatic implicature, conversational and  conventional, really beginning with (Herbert Paul) Grice     The work on conceptual metaphor primarily from (George)  Lakoff, (Mark) Johnson, (Gilles) Fauconnier, (Zoltán)  Kövecses, etc.    The work on politeness theory primarily from (Penelope)  Brown, (Stephen) Levinson, (Dan) Sperber, (Deirdre)  Wilson (the last two a relevance-theoretic interpretation)    Deixis (background)  The concentric circles shift — take, e.g., “The book is over  there [on that desk, in the classroom, in this building, at  UMD]” ​vs ​“All the students are here by now [at the  University of Maryland].” Clearly the speaker-proximal  area is far smaller in the first sentence. In other words, the  meaning of these deictic words and expressions is  non-conventional, and approximately inferred.    Or I might shrink my zone of intimacy/familiarity in a  language that grammatical uses these notions (German  Sie/du, Spanish tú/usted, etc.) — for instance my child in  one moment might be addressed by ​du​, but in the next  (supposing he or she’d done something to upset me, or I  needed to discipline him or her) I might address him/her  as ​Sie​, shrinking my circle of intimacy for the purpose of  driving home the disciplinary nature of my  action/utterance.    Deixis of place  Here/there — “the pragmatically given space,  proximal/distal to speaker’s location at CT(​, that includes  the point or location gesturally indicated​).”  E.g.: “Place it here,” “Place it there.”    This/that — [glossed by Lyons 1977a: 647 as] “the object in  a pragmatically given area close to/beyond the speaker’s  location at CT”  E.g.: “Bring me that book,” “That’s very delicious.”    Empathetic deixis: the speaker can shift the deictic centre  to that of the addressee/hearer to show emotional  closeness or empathy, or perhaps to garner empathy or  help one’s argument to go through by conflating one’s  own thoughts with those of the hearer. Such a shift has to  be recognized contextually; and appropriate inferences  may be drawn therefrom.    Various languages discretize space along this deictic  dimension differently; English has only two pragmatically  given discrete locations, the NW Amerindian language  Tlingit has four (​right here, nearby, over there, way over there​),  and Malagasy has a six-way contrast along the same  dimension.    But it still serves to indicate to the hearer — to assure  them, in a sense — that the intended referent is ​available​,  even in some sense ​salient​ in the contextual space.  Consider the utterance: “The cathedral was built by the  Medicis.” What cathedral? It depends on context,  obviously; but note that now ​how we search for the intended  referent also depends on context​. If we’re in Italy, we’ll likely  assume that the proper cathedral is deictically available  along the spatial axis — perhaps it’s ‘​the cathedral in this  city we’re currently in​.’ If we’re in a classroom in America,  during a history class, and we fell asleep, we might assume  that the intended referent is available along the discourse  axis — perhaps it’s ‘​the cathedral that was grounded a few  utterances ago​.’ Or if we just watched a movie that involved  a cathedral, perhaps the cathedral is temporally proximal  — ‘​the cathedral that we just saw a few minutes ago in the  film​.’    In any case, we can agree that ​the ​indicates that the correct  referent is ​salient​, in the sense (now that we can formally  articulate it) of ​deictically relatively proximal​.    Salience examples: salience decays over time (if two  women were grounded, probably more recent one in  absence of more explicit description); more likely to refer  to say, someone in the room or someone mentioned  recently than say, to suddenly refer to ​Hitler ​or ​Mahatma  Ghandi​, although he’s certainly somewhere in the common  ground (as a part of our collective cultural knowledge).    Frame Semantics  Cf. also the notion of ​profiling ​found in Ron Langacker’s  theory of cognitive grammar (CG), for which consult the  eponymous ​Cognitive Grammar​.    Now at first blush you’re likely to think, “Well this is some  kind of philosopher’s utopia — no one talks like this.” And  certainly people don’t always tell the truth, and sometimes  we ramble on about irrelevancies, and sometimes we get  confused and use more words than were strictly necessary  to convey a thought.    But this isn’t what Grice is talking about. That’s all at a  very superficial level. The fact of it is that even when we’re  at each other’s throats verbally, we’re cooperating in the  most wonderful way.    Suppose I say something to you in the course of a  conversation that’s blatantly false — I violate the maxim of  quality — “Queen Victoria was made of ​solid​ ​steel​.” All of  a sudden a metaphor springs into existence — that is, you  assume I’m not saying what you and I know to be false and  instantly find an interpretation that renders what I say  adherent to the maxim of quality.    Or suppose I appear to say something completely  uncooperative and irrelevant — you ask “What time is it?”  And I say “Well the milkman just came.” Strictly speaking,  this is a non sequitur if we’re to take the second sentence  at literal face value. But you never even ​consider ​that I  might not be cooperating. You find a set of inferences that  connects the two sentences.    Or suppose we were British, and you say in the course of a  political conversation: “What if Putin were to blockade the  gulf and keep all the oil from us?” — “Oh come now,  Britain rules the seas!” Of course that hasn’t been true  since perhaps before WWII, but you don’t suppose I’m  uninformed or lying — you try to find a nearby proposition  that I might really be trying to get across, and you hit upon  the exact negation of what I said — “Britain has no naval  clout anymore; we’re screwed if that happens.”    Other quality flouting:  “Tehran’s in Turkey, isn’t it, teacher?”  “And London’s in Armenia I suppose.”    Some quantity flouting:  proposition (in a sense that we haven’t time to define or  formalize) that he/she could safely infer we meant.    In general, this led to theoreticians beginning to  increasingly notice the (1) pervasiveness and (2)  strangeness of NL metaphor. Combined with notions from  cognitive psychology, in particular a series of studies and  papers in favor of the ​perceptual symbol system​ view of  cognition, and the concomittent assumption of the  fundamental importance of ​embodiment ​to human  cognitive processes and reasoning, these observations  began to coalesce into an important field of linguistics  known as conceptual metaphor theory.    We mention here briefly simply because it rounds out our  picture of the ​hardness ​of a general regime of natural  language processing techniques, and (for our purposes)  completes our short-list of things that a dialogue system  will have to be able to cope with.    Can you think of any other metaphor schemata?  An argument is a ​building​.  We’ve got the ​framework ​for a ​solid ​argument.  If you don’t ​support ​your argument with ​solid ​facts, the  whole thing will ​collapse​.  He’s trying to ​buttress ​his argument with a lot of  irrelevant facts, but it’s still so ​shaky ​that it will easily ​fall  apart ​under criticism.  With the ​groundwork ​you’ve got [​foundation​ you’ve  laid], you can ​construct ​a pretty ​strong ​argument.  Arguments can be ​undermined​ ...    An argument is a ​journey​.  So far​, we haven’t ​covered much ground ​[​terrain​].  This is a ​roundabout ​[​circuitous​] argument.  We need to ​go into this further ​in order to ​see clearly  what’s involved.  As we ​go further into​ the topic, we find …  We have ​come to a point ​where we must ​explore ​the  issues much more deeply.    Understanding is ​seeing ​[combines with the above to  produce …]  [​combine with journey metaphor​]  Having come this far​, we can now ​see ​how Hegel went  wrong.  We will now ​show ​[guides do this] that Green  misinterpreted Kant’s account of will.  We ought to ​point out ​[a guide does this] that no such  proof has been ​found​.  Dig further into ​his argument and you’ll ​discover ​a great  deal.  We can ​see ​this only if we ​delve into ​the issues.  Shallow ​arguments are practically worthless, since  they don’t ​show ​us very much.  [​combine with building metaphor​]  We can now ​see ​the ​outline ​of the argument.  If we ​look ​carefully at the ​structure ​of the argument …  [​combine with container metaphor​]  That is a remarkably ​transparent ​argument.  I didn’t ​see ​that point ​in ​your argument.  Since your argument isn’t very ​clear​, I can’t ​see ​what  you’re getting at.  Your argument has no ​content ​at all -- I can ​see right  through ​it.    Embodied Construction Grammar (ECG)  This is of course work undertaken by (Nancy) Chang and  (Benjamin) Bergen -- originally under the direction of Jerry  Feldman at UC Berkeley. It builds on ​construction ​based  approaches to syntax/semantics, which are due to  Fillmore, Kay, Goldberg, Croft and a number of others --  this is a huge and (still) productive area of research that’s  crossed over into NLP via the algorithmic notion of a  unification grammar.      An insertion sequence might look like:  ...  C: “Can I get a scone?”  E: “Vanilla or strawberry?”  C: “Uhhhh — strawberry”  E: “Sure thing, sir.”    If E were at this point to begin asking about methods of  payment, or talking about the weather, we’d consider this  the start of a new conversation unit, a new sequence. But  there are ways to extend it without opening a new  sequence — using a​ post-sequence ​expansion, like a  farewell/farewell:  …  E: “Sure thing, sir.”  C: “Have a nice day!”  E: “You do the same.”    At this point, it’s probably impossible to start a new  sequence (on E’s part) without some kind of pre-sequence  like another summons/answer (“Hey — umm”, “Oh yes?”).    Or another very common pre-sequence opening that  you’re probably very familiar with is: “​Are you doing  anything tonight?​” The dreaded question. The whole  point of the exchange is what’s coming next turn — some  kind of dreaded invitation to go do something. In fact, we  often answer “Yes” even when it’s not true — and the only  reason we do this is because we recognize this as a  common, polite pre-expansion to the invitation/accept or  invitation/reject sequence. We know what’s coming.    Or consider the following sequence collapse, which only  occurs because the speakers are subconsciously familiar  with the structure of common sequences:  “Do you smoke?”  “Nah I left them [my cigarettes] at home.”  He’s not answering “nah” to whether or not he smokes —  he’s answering “nah” to the part of the sequence that was  collapsed, the fpp (​first pair-part​)​ ​of the request sequence  (“can I bum a cigarette?”). What’s been collapsed is the  answer to the pre-expansion context question, which  serves no other purpose than to determine whether the  request is likely to be felicitous, or is applicable, as well as  the fpp (​first pair-part​) of the request sequence itself.