The Grice Club

Welcome

The Grice Club

The club for all those whose members have no (other) club.

Is Grice the greatest philosopher that ever lived?

Search This Blog

Friday, February 19, 2010

The Langlish Lift

By Lawrence J. Kramer, for the Grice Club.

This is as good a place as anywhere to tell you about an invention from a place where energy is very expensive. The place is called Langland, and the invention is the Langlish Lift.

Like elevators most everywhere, the lifts of Langland are built with a counterweight so that the motor only has to move the difference between the weight of the load (including the carriage) and the weight of the counterweight. But whereas elevators in most places use a counterweight equal in weight to the carriage plus 40% of its capacity, the elevators in Langland use a variable counterweight system. A scale in the floor of the elevator weighs each passenger, and a mechanism attaches just the correct amount of counterweight to balance out the load. That way, very little energy is needed to lift or lower the carriage.

To save time, the elevators are equipped with very sophisticated personal recognition software. If the software recognizes a passenger, it immediately transfers weights equal to the stored weight of the passenger. And whenever that passenger rides, the information in the database is updated. For the most frequent users, there is even a custom counterweight available so that only that one piece need be transferred to get very close to the correct weight. The scale is then used to determine what needs to be added or removed to get the weight exactly right. (In the case of removal, the Langlish use the term “cancellation.”) If the passenger is not recognized, the software may determine that the passenger resembles someone in its database and transfer an estimated amount of weight, pending the reading of the scale.

The Langland Lift’s counterweight system thus has three components:

(i) a counterweight for the carriage, which is the same for all passengers;

(ii) a counterweight determined by the recognition software, which is the same for all of a particular passenger’s rides; and

(iii) a counterweight determined by the scale, which is different for each ride.

If we assume that all utterances are answers to asked or unasked questions, we might consider A, carrying a particular question that U wishes to answer, as a passenger on U’s Langland Lift. Everybody is assumed to be cooperative, and so the counterweights representing Grice’s principle and maxims are always onboard. The context in which the question arises (what U believes A already knows, including their common language) is analyzed by the recognition software, and the appropriate counterweight added. U’s utterance, the final counterweight, consists only of what is needed to balance the load.

Before we get to where La Bardot lives, let’s look at where Thornton Wilder tells us Jane Crofut lives:

Jane Crofut
The Crofut Farm
Grover’s Corners,
Sutton County,
New Hampshire
United States of America
Continent of North America
Western Hemisphere
the Earth
the Solar System
the Universe
the Mind of God.

A gets on the elevator asking “Where does Jane Crofut live?”, and the counterweight system engages:

i. The carriage counterweight never changes.

ii. If U knows only that A is an English speaking human, the context counterweight consists only of the English Language and the parts of the address that fall below “United States of America” (with the interesting exception of the Mind of God). But if U’s recognition software detects that A is an American, the context counterweight may include everything below “New Hampshire.” One can imagine circumstances in which the recognition software moves even higher up the address chain.

U must now figure out how much information A needs above the part of the address that A already knows or would be able to figure out from something above it (e.g., that Jane lives in New Hampshire from the fact that she lives in Grover’s Corners). The recognition software will provide a counterweight for everything that U believes A believes.

iii. U will say the rest.

So, is it correct to say that Bardot lives at 34 Rue de Gaulle? It is if U knows that A knows know in which town she lives and so is obviously looking for a street address and only a street address. Otherwise, it's certainly not enough, and it may be too much. Whether a street address is over-informative depends on the level of detail demanded, how "heavy" the passenger is. We need to refine the context (“Why do you ask?”) or put the question directly on the scale (“How detailed an answer do you need?”) and see how much it weighs to know how much more heft to add to the answer.

I imagine that many different physical systems can serve as metaphors for conversation. All you need is a part that is always the same, a part that is applicable to a class, and a part that is applicable only to the moment. The Langlish Lift works for me because I think the counterweight is an elegant solution to the actual elevator problem. In any event, the metaphor is intended to place the natural and rational origin of language in U’s effort to say the least he can by leveraging (his estimate of) A’s beliefs. That origin not only necessitates rules of syntax and grammar, A’s access to which can then be, and is, part of the context that U takes into account, but also, perhaps, to order the maxims we observe. So, for example, one might argue that the central principle of conversation is “be brief” rather than “be cooperative,” and that the rest of Grice’s maxims can be marshaled in service to the former as well as to the latter.

I assume that all of this is either ridiculous or old hat. I will leave it to JLS and others to tell me which.

19 comments:

  1. No. It is very good. And I'll elaborate on it, and report.

    And thanks for the post. It was genial.

    ----

    ReplyDelete
  2. Adapted slightly:

    "the elevators in Langland use a variable counter-weight system: a scale in the floor of the elevator weighs each passenger, and a mechanism attaches just the correct amount of counter-weight to balance out the load."

    -- This is extremely clever. Thank you! I'm revising the device stage by stage because it is so clever.

    ReplyDelete
  3. Adapting then:

    Where does Jane Crofut live?

    i. At the Crofut Farm.
    ii. At the Crofut Farm, in Grover’s Corners,
    iii. At the Crofut Farm, in Grover's Corners, in Sutton County.
    iv. At the Crofut Farm, in Grover's Corners, in Sutton County, in New Hampshire.


    APPLICATION:

    "A gets on the Langlish Lift and askks, 'Where does Jane Crofut live?' The counterweight system engages. One can imagine circumstances in which the recognition software moves even higher upthe address chain."

    I once gave a lecture in Campinas -- published, which provided 34 answers to

    "Is this your first time in Brazil?"

    I joked on

    "My first time in Brazil, not in Campinas"

    as odd.

    --- Similarly,

    (a) A and B are at Grover's Corners:
    A: Where does Jane Crofut live?
    B: i. In a farm (incorrect)
    ii. In the Crofut farm. (correct)

    A and B are driving in Sutto County.
    A: Where does Jane Crofut live?
    B: In Grover's Corners.

    -- The addition, "In a farm", as per "In a farm in Grover's Corners" may be "interesting and informative" but otiose in the circumstances -- cfr. Grice on "Did you see the tv programme yesterda?" "I was in a blackout city" vs. "No, I was in New York, which was black-out".

    --
    Level 3. A and B are driving in New Hampshire:

    A. Where does Jane Crofut live?
    B. In Sutton County

    -- the addition of the farm is even more otiose at this point. Grover's corner is not, but it could be for Grice (I would never think that mentioning "New York" can be otiose as he says it is in the black-out example above).

    --- Etc.

    ReplyDelete
  4. From wiki:

    "the coordinates of Grover's Corners are given as 42°40′ N. Lat. & 70°37′ W. Long., which would actually put it very close to Rockport, Massachusetts."

    Where does Jane Crofut live?

    (a) 42°40′ N. Lat. & 70°37′ W. Long.

    (b) Not far from the border with Rockport.


    ----- Cfr. Noel Coward, The house where I was born:

    I cannot remember
    I cannot remember
    The house where I was born

    But I know it was in Waldegrave Road
    Teddington, Middlesex

    Not far from the border of Surrey;

    An unpretentious abode
    Which, I believe,
    Economy forced us to leave
    In rather a hurry.

    ---

    ReplyDelete
  5. The Bardot example is indeed a 'clash' as Grice presents it, between

    first-quantitative maxim
    first-qualitative maxim

    -- comment in "Somewhere in the South of France". The thing is defeasible, as Grice knew, in various respects.

    In Kramer's new scenario of the Langlish lift, the constraints are, as they should be, much more general. It's convesational cost/effort, etc. for each utterance. So I will have to elaborate and report back.

    A little caveat: I love 'be brief' trumping over 'be cooperative'. But in a way, the 'brief' of 'brief' is not egotistic brief. It's 'brief' as U will judge you. Since that's the case, it would import a level of 'cooperativeness. Etc.

    ---- This relates to Kramer's and my considerations re: "in no particular order". The other day I was reading an essay on two people on Strawson or something. "By Smith and Willams". "The ordering is arbitrary: this is truly a cooperative venture." But I'm not sure 'aribtrary' will do for me. Etc.

    ReplyDelete
  6. Do you mean brief as A will judge you?

    I would not say that brevity trumps cooperativeness. I would say that cooperativeness is otiose in an Occamesque sort of way: what purpose does it serve if we can adduce all of Grice's maxims in service of brevity? Cooperativeness is perhaps tautological: can one be said to be "conversing" or "informing" if one is not cooperating (on the linguistic level)?

    ReplyDelete
  7. Yes, I'm pretty sure "brief" HAS to be understood as "to A" (A: addressee). This is chapter 7:2.4 of my PhD thesis!

    --- (and of course, PhD theses are always _right_-JLS).

    In that PhD thesis I propose, in ch. 7.1 to consider the cooperative principle in terms of strategic or means-end rationality, and then go one by one over the four categories: quantity, quality, relation, and mode -- hence this is 7.4. and within the modal maxims I consider 'be brief'. I am especially concerned there with FLOUTINGS to the maxims, since I have said, in ch. 6, that they are NOT rules. (What's the good of a rule if the pleasure of it is in the flouting it? Plus bees, and other animals, don't FLOUT, so this is a very human activity).

    The other day I was websurfing for material for the club, and among all the pretentious stuff written on Grice out there found this LOVELY dissertation in China on Grice: the floating of the maxims, it read. I have to save pieces of it, for the club. For they indeed float all over the place. And make YOU float. It's a liberating, exhilarating feeling, as anyone who has learned to float in the ocean, or any swimming-pool (library) can testify.

    --- So, Grice is not just saying, 'be brief'. He is having that as sub-merged (oops) onto 'be perspicuous' (modal maxim) -- this one annoys me sightly in I cannot formulate it if I want my thing to be a decalogue: I have to take it for GRANTED). And the "be perspicuous" under "cooperate!"

    I cannot see how Emily Dickinson would care for "be brief!" unless she's thinking of that hateful editor. Don't you know that ALL editors are always, "cut, cut, cut, cut". It's never brief for the utterer.

    Etc.

    I think it IS conceivable to utter conversational moves which do not abide by the Cooperative Principle, so otiosely called. The Cooperative Principle, as I argued in my publication, "Conversational Feast", only justifies the APPROPRIATENESS of a conversational move, NOT the EXISTENCE or possibility of existence of a conversational move. This is a transcendental justification. But it is a WEAK transcendental justification.

    A transcendental justification, for Kant, Strawson, and Grice, justifies the existence of something. It is pretty strong. Now, in the case of the Cooperative principle, it justifies the existence of an APPROPRIATE conversational move, not the _existence_ of the conversational move (if I may repeat myself).

    Thus, it is conceivable that all the moves will be, between A and B, outright lies, inconsistencies, irrelevancies, and inperspicuities. Let's see category by category:

    In Malagasy, they won't say where their daughter is: they are not required to be 'as informative as they are required' because they are not required to be informative in the first.

    The malignant demon of Descartes was Lying to him, all the time.

    James Joyce sexualised Norah Barnacles by uttering irrelevancies ('booring' ones) to her.

    Grice was seen to go over the top in some of his explanations. Hardly 'be brief', etc. But anyone reading him knew that that was okay, because he has every right, etc.

    --- Later,

    ReplyDelete
  8. ---- When and if you comment, will you use the Langlih Lift thing? I am having SOME considerations using it... and I want more appliances, to other scenarios, too.

    I would suggest you publish it as

    "The Langlish Lift" as a separate new post, too.

    Recall to have the first line of the body of the post, as


    ---------- By L. J. Kramer,
    for the Grice Club, etc.

    --- or something

    So that people _KNOW_, i.e. so that YOU know, when you read it.

    I think it has of course broader applications than the Bardot. So that we can have illustrations as it applies (the Langlish lift does) to the four categories, with the four types of floutings, etc.

    --- I'm in particular interested what the load thing is. I have a few things to say about that. Grice too. He talked (this must be in google). PERE, principle of effort of rational effort, which may relate. But he was thinking of impicit versus explicit in that.

    He does speak of "conversational effort" in the New York blackout example.

    So, we would need to consider the strict applications of the elevator thing in detail. Also, why is it called 'lift' or elevator if it only takes you down? (But that's a minor thing: what comes down, must come up? :)

    Later,

    ReplyDelete
  9. I asked about A judging because your original comment said "U will judge you." Do you know Abbott and Costello's Who's on First?

    U has a mission to inform A. U ought not to care whether A thinks U's been brief unless A's opinion of U has an interpersonal consequence. Otherwise, it's U's time and effort that U values, and so U must be the judge of what is brief enough.

    But even then, I don't believe brevity is subjective. Evolution carries us toward objective brevity as what Dawkins called a evolutionarily stable strategy.

    A will, however, be the benchmark (not the judge) of whether U has been too brief. In other words, U will be as brief as he can be by his own lights while satisfying A's demands for intelligibility. Diff'rent strokes for diff'rent folks.

    I have no idea how extensible the lift metaphor may be, and I make no claims for its being extensible. I'll fiddle with it as we go, but there is no reason to believe that in Langland, what goes up must indeed come down. Langland, at least not via the elevator. It's good exercise to ride up and walk down.

    ReplyDelete
  10. Sorry. Yes, indeed "As A will judge you". Sorry I missed the implicature of your question. As if you needed a tirade on 'brief-A', and 'brief-U', etc. Later.

    ReplyDelete
  11. In some of my writings, e.g. by PhD dissertation, I started to use

    "co-conversationalist"

    to avoid "U" and "A" (utterer and addressee) or "A" and "B".

    So one formulates things as:

    "Be as informative as is required, you think, by the expectations of your fellow co-conversationalist". It can be pretty clumsy, but to consider. Etc.

    ReplyDelete
  12. No, I lean toward U and A, as each move strikes me a assymetrical. And I would eschew "expectations" in favor of "needs." I need only be as prolix as your ability to understand me demands. After that, we're dealing in maxims of grace, not Grice.

    ReplyDelete
  13. Kramer quotes from Dawkins:

    "Evolution carries us toward objective brevity as what Dawkins called a evolutionarily stable strategy."

    Does he use "strategy". Grice did use "Dawkins". "Read "Selfish Gene"", he wrote. We doubt he did.

    In any case, what's the good of those memos. Surely I don't need to remind myself "Read Kant's Cr. Pract. Reas. in German" unless I will forget.

    --- But 'strategy' interested me. I was once giving a talk to teachers of lingos, on the Gricean strategical model. She said, "Surely your model is wrong. While teachers have to FIGHT with students, we are NOT all of us, 'generals', you know".

    'Strategy' has a very politically incorrect name to it: it means to 'kill'. The general is there to kill others, etc. The gene is selfish enough, so we never know. Etc.

    ReplyDelete
  14. I loved your idea of 'riding up, walking down' 'good exercise'. Indeed

    I think it was Newton who said,

    what comes up must come down

    (I don't think lifts were invented then) etc.

    Oddly

    what comes down must come up

    also makes sense, _timeless-ly_, as Grice would have it:

    what comes down at t1 must HAVE
    come UP at time t2 < t1.

    Etc.

    I was wondering beause in your Langlish lift metaphor. (Don't let's keep calling it metaphor. We know it is).

    --- we are supposed to be 'ground floor'. I think you call them "first floor". So 'first floor'.

    A lift you can descend to 'below first floor'. and use stairs to walk up. etc.

    --- Plus, some lifts are horizontal, too. Etc. So we may need to consider:

    where you are
    where you want to go

    sort of thing. You did extend on the navigation metaphor elsewhere (notably THEORIA), so we can drop some of that metaphorisation here too.

    Unlike a navigator (which is freerer) the Langlish lift moves you up. The destination is your comprehending of what the passenger said, load and all. Etc. Will elaborate.

    Glad you didn't like the 'co-conversationalist'. There is a sense in which we may need variables: x and y, for any participant. But I'm glad you stick to U and A.

    Most readers of Grice's "Logic and Converation" have NOT and will not (Ever) read his work on Meaning, so they are in the dark.

    Grice was pretty particular as to these things. I tend to believe (on a bad day) that his "Logic and Conversation" is just exegetical towards his closer 'analysis' of "meaning" as per "Meaning" when he talks of U and A.

    ---

    U and A feature in belief ascriptions. As you are considering A as being a 'benchmark' and not a 'judge', it may pay to consider formalisms for

    'be brief'

    BELIEF (A, BRIEF (U)).

    i.e. A(ddresseee) believes that U is being brief.

    Etc. I will elaborate on this. One should also consider cases of "in the absence of an audience" (or addressee) that Grice considers.

    I knocked my own finger with the hammer and say, "Jesus", "Sh+t", whatever. These phrases are meaningful, yet hardly A-directed. And there are other cases to consider.

    Sometimes

    U = A.

    but this has to be qualified temporally or chronologically, as

    "read Dawkins, Selfish Gene"

    --- i.e. U (Grice) is reminding A (Grice) that A (Grice in the future) read the note that advises him to read the book. Etc.

    He also considers entries in journal:

    "Today, spent day meaning this and that."

    It's NOT the "Dear-Diary" thing, for Diary is Addressee then. This is just a note you write, or U writes.

    In a way, I think Grice was slightly confused about this.

    An 'utterance' seems to be a vehicle for meaning-NN, non-natural, only. The metaphor is that something gets OUTED. But surely something can get outed (a 'flatus' as it were, from various bits of the body) and nobody (i.e. no Addressee) there to find out.

    In some cases, the U should not care less (or should care less?) about the A.

    I will need to revise from WoW:Meaning, where exactly he is starting to have addresse's expectations into the bargain (as we move from meaning-n to meaning-nn).

    ReplyDelete
  15. In a way, his

    "Those spots didn't mean anything to me;
    to the doctor, however, they meant that
    Tommy had the measles"

    --- It seems like we have an interpretant



    measles

    the doctor

    spots

    I think this is what Sharpless calls the semiotic triangle (vide Wharton on Sharpless, etc.).

    Here the doctor is a bit like the addressee even in cases of "natural" "meaning".

    ReplyDelete
  16. While Grice does have "addressee" -- but I'll have to check -- he also has "recipient", which is NOT just a bottle.

    He also has 'utterance' as aiming at a "response" from the recipient, usually a belief.

    Of course linguists will look down on all this and just speak of speaker-meaning and hearer-needs, etc.

    But as a semiotician, I disagree.

    I'm a word-of-finger person. This idiotic expression learned today via Quinion: in the world of the internet, TGIF, tweet-google-internet-facebook, it's word-of-finger.

    Word-of-finger is wordy enough. What I'm onto is pre-verbal gestures, and stuff. Etc. But not wedded to them, either. Etc.

    ReplyDelete
  17. Yes, Dawkins uses "strategy." The ESS (Evolutionarily Stable Strategy) is a central notion of The Selfish Gene

    Further on the subject of extensibility. I'm not sure it pays to extend a metaphor that does not draw from a real thing. There is no such elevator, so we cannot have a common understanding of any aspect of it other than those I have specifically delineated. That seems to me to weaken it as a tool of further analysis.

    I prefer to think of the Langlish Lift as a rhetorical machine, an easier way to make two drier points:

    1. Efficiency is the driving force in the evolution of language, and

    2. Efficient systems have three layers: a part that is always the same, a part that is the same for all members of a class, and a part unique to the current move.

    After that, and being a reductionist, I would actually prefer to speak in terms of efficiency and its layers. I want to relate the emergence of a conversation to the emergence of language, and the emergence of language to the emergence of homo sapiens, ontogeny recapitulating phylogeny if you will, moving from general to specific, and stopping at the least specific level conferring a new advantage. (The last sentence inadvertently uses an elevator metaphor of moving and stopping at a level. It should, of course, go without saying that the Langlish only ride the elevator to as high a level as they need to go. To go higher would be wasteful. Efficiency is all.)

    Although ESS is Dawkins's trope, I was refering to that term, not quoting anything Dawkins had written.

    ReplyDelete
  18. Excellent quotes, thanks, and excellent non-quotes, i.e. original material from you.

    I see that he SHOULD use 'strategy' because a strategos is a general, and his gene is such a clever thing, it should feel it is a general alright. And as for 'kill', that's what genes do. (recall Woody Allen carrying genes as a spermatozoon as he kills his friends in "Everyhing you always wanted to know about sex* (but were afraid to ask)".

    I liked your drier (I wouldn't call it that you see) model of the efficiency thing. I'll elaborate on it.

    I too like your evolutionary view on things, and reference to homo sapiens, etc. I'll elaborate on them, too.

    ReplyDelete
  19. For later considerations, I hope:

    "Efficient systems have three layers: a part that is always the same, a part that is the same for all members of a class, and a part unique to the current move. ... After that, and being a reductionist, I would actually prefer to speak in terms of efficiency and its layers. I want to relate the emergence of a conversation to the emergence of language, and the emergence of language to the emergence of homo sapiens, ontogeny recapitulating phylogeny if you will, moving from general to specific, and stopping at the least specific level conferring a new advantage."

    Oddly my first ecounter with 'layer' is in the "Heavyside layer" as used by Lloyd Webber in "Cats". But I see your point.

    layer vs. level.

    When we (i.e. you) speak of the three layers of the lift, is 'component' alright? Or is 'layer' meant as meant? I tend to think of 'layer' as a stratum. And it seems it's three components (no particular order: the load that is ceteris paribus required: the load of the pragmatic module as it were; but also the load as it comprises A and U; and the special load of U's utterance as a task for A). The layer-thing seems otiose there?

    --- The other bit when you speak of 'level' is a different thing, right?

    While you cleverly have "efficiency" as the Grice of the Grice (versus the grace of the grace), it will do to define it in terms of goals, etc.

    Grice sometimes played with extrinsically weighing (and it's all about weighing as we lever up and lever down) versus intrinsically weighing.

    Your requirement about the Langlish lift being _in_ Langland, where energy is, did you say, a 'valued' item, surely will affect how we _design_ the lift, etc.

    Your philogeny and ontogeny are very apt. I do thing one recapitulates the other but I forget which which. Later.

    ReplyDelete