"Gricean Maxims": What don't you need to say?

A forum for discussing linguistics or just languages in general.
Post Reply
User avatar
eldin raigmore
korean
korean
Posts: 6352
Joined: 14 Aug 2010 19:38
Location: SouthEast Michigan

"Gricean Maxims": What don't you need to say?

Post by eldin raigmore »

I have a few questions that, in my view, are simple, and "quick-ish" if not "quick" (YMMV); but I suspect the answers are neither simple nor quick.

First Set of Questions:
In some languages, speakers can't say, in a single, simple, monotransitive clause, which of the two participants was the Agent or Actor, and which was the Patient or Undergoer.
To illustrate by a made-up example, let's assume they have a verb (written "verb") which we can gloss as "kiss" (or "kist" or other inflexional variants).
Then the clauses "Jack Jill verb" and "Jill Jack verb" (or "verb Jack Jill" and "verb Jill Jack", or "Jack verb Jill and "Jill verb Jack")
all mean "One of Jack or Jill kissed the other".
At least some of these languages also have detransitivizing morphology. For instance, they may have an antipassive form of verb, such that
"Jack verb-ANTIP" means "Jack was the Actor in a 'kissing' incident".
Or/and, they might have a passive form, such that
"Jill verb-PASS" means "Jill underwent a 'kissing' event".

1A: If a speaker says "Jack Jill verb" and "Jack verb-ANTIP", would the average addressee assume Jack kissed Jill?
1A' and how safe would that assumption be?
1A" and would the speaker assume the addressee would make that assumption?

1B: If a speaker says "Jack Jill verb" and "Jill verb-PASS", would the average addressee assume Jack kissed Jill?
1B' and how safe would that assumption be?
1B" and would the speaker assume the addressee would make that assumption?

1C: If a speaker says "Jack verb-ANTIP" and "Jill verb-PASS", the two clauses now contain no semantic information in common except the verb. Would the average addressee assume Jack kissed Jill?
1C' and how safe would that assumption be?
1C" and would the speaker assume the addressee would make that assumption?

1D: If a speaker says all three clauses -- "Jack Jill verb" and "Jack verb-ANTIP" and "Jill verb-PASS" -- would the average addressee assume Jack kissed Jill?
1D' and how safe would that assumption be?
1D" and would the speaker assume the addressee would make that assumption?
1D+ would they be likelier in case D than in cases A and/or B, and likelier in case A and/or B than in case C, to make the relevant assumptions? (that is; addressee assumes Jack kissed Jill, and speaker assumes addressee will make that assumption.)

Second Set of Questions:
All natural languages have both univalent intransitive and bivalent monotransitive verbs/clauses. Many languages also have trivalent ditransitive clauses; but many don't.
Suppose I want to use a verb meaning "send" or "sent", and want to talk about Don the donor and Thelma the theme and Rhys the recipient.
In such a language, I can neither affirm nor deny, in a single, simple, clause, that Don sent Thelma to Rhys.
But the verb can be inflected in at least two of three ways:
"Don Thelma sent-X" means "Don sent Thelma to some unspecified someone",
"Don Rhys sent-Y" means "Don sent some unspecified someone to Rhys",
"Thelma Rhys sent-Z" means "some unspecified someone sent Thelma to Rhys".

Possibly not all three forms are available in some such language.

2A: If a speaker says "Don Thelma sent-X" and "Don Rhys sent-Y", would the average addressee assume Don sent Thelma to Rhys?
2A' and how safe would that assumption be? (Note Don is the only entity participant in both clauses.)
2A" and would the speaker assume the addressee would make that assumption?

2B: If a speaker says "Don Thelma sent-X" and "Thelma Rhys sent-Z", would the average addressee assume Don sent Thelma to Rhys?
2B' and how safe would that assumption be? (Note Thelma is the only entity participant in both clauses.)
2B" and would the speaker assume the addressee would make that assumption?

2C: If a speaker says "Don Rhys sent-Y" and "Thelma Rhys sent-Z", would the average addressee assume Don sent Thelma to Rhys?
2C' and how safe would that assumption be? (Note Rhys is the only entity participant in both clauses.)
2C" and would the speaker assume the addressee would make that assumption?

2D: If the language allows a speaker to say all three of "Don Thelma sent-X" and "Don Rhys sent-Y" and "Thelma Rhys sent-Z", each two participants would appear together in a clause.
Would the average addressee be even likelier to assume Don sent Thelma to Rhys?
2D' and how would that assumption be even safer than in situations 2A and 2B and 2C?
2D" and would the speaker be even likelier to assume the addressee would make that assumption?

[hr][/hr]
[hr][/hr]

Thanks!
Last edited by eldin raigmore on 27 May 2017 14:46, edited 2 times in total.
Sumelic
greek
greek
Posts: 566
Joined: 18 Jun 2013 23:01

Re: Pragmatics: What don't you need to say?

Post by Sumelic »

This seems odd to me. I don't have experience speaking a language with this kind of commonly ambiguous structure, but I wouldn't think repeating a verb as in your examples would be a common strategy for disambiguation. I feel like it is more likely that there is some kind of word order or animacy hierarchy defaults, and if not that, some method of marking the subject or object of a verb within a single clause (e.g. using the antipassive and using a preposition before what is usually the direct object).
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

Re: Pragmatics: What don't you need to say?

Post by Creyeditor »

I fail to see how this relates to pragmatics. Pragmatics usually depend on the context. But still let's try to do this is German (assuming a zero-affix for the antipassive and periphrastic passive). We can get the whatever-direction effect in complement clauses with special prosody and two feminine gender arguments with articles (slightly southern.)

1A:
dass die Marie die Vera küsst ..., also die Marie küsst.
that the Marie the Vera kisses, that.is the Marie kisses
Yes
A': very safe
A'': Yes
The sentece sounds weird though. As if Vera somehow did not want to be kissed or try to fight it.

1B
dass die Marie die Vera küsst ..., also die Vera wird geküsst.
that the Marie the Vera kisses, that.is the Vera AUX.PASS kiss.PTCP
Yes
B': very safe
B'': Yes
This is a normal sentence to me, if I do it with the right prosody.

1C
dass die Marie küsst ..., also die Vera wird geküsst.
that the Marie the Vera kisses, that.is the Vera AUX.PASS kiss.PTCP
No
1C' mildly safe
1C" probably not
This sentence sounds somehow confusing. As if someone tells a story and does not know where to start. Sounds a bit like someone drunk speaking

1D
dass die Marie die Vera küsst ..., also die Vera wird geküsst und die Marie küsst
that the Marie the Vera kisses, that.is the Vera AUX.PASS kiss.PTCP and the Marie kisses
Yes
1D' very safe
1D'' yes, certainly
1D+Okay, this is ordered in decreasing likelyhood D > B > A > C
This sentence sounds perfectly natural to me.

Note that this is not the only possible structure for German and it only works under a certain prosody. German also does not have overt detransitivizing morphology. So, technically it does not meet your high standards. Still I think this might be a good idea to look at, because it shows that these situations are real. Other native speakers might check if these sentences sound natural.
The second set of questions does not apply to German and I really have no clue about them.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
User avatar
eldin raigmore
korean
korean
Posts: 6352
Joined: 14 Aug 2010 19:38
Location: SouthEast Michigan

Re: Pragmatics: What don't you need to say?

Post by eldin raigmore »

Sumelic wrote:This seems odd to me. I don't have experience speaking a language with this kind of commonly ambiguous structure, but I wouldn't think repeating a verb as in your examples would be a common strategy for disambiguation. I feel like it is more likely that there is some kind of word order or animacy hierarchy defaults, and if not that, some method of marking the subject or object of a verb within a single clause (e.g. using the antipassive and using a preposition before what is usually the direct object).
I agree about the oddness; If it didn't seem odd to me, too, I wouldn't feel the need to ask about it. I think the lack of experience lies at the root of the perceived oddness.

I'm assuming you're mostly talking about the first set of questions? the monotransitive, bivalent clause, in which one can't tell which participant is the Agent and which the Patient without additional help?

Cross-linguistically, I would not be surprised to learn you are right that such strategies are not common.
But I gather at least a few languages exist in which some such strategy isn't really rare!

I wish I had a better memory, or a better-organized library.
I think the language in question may have been in Charles N. Li's "Subject and Topic".
Or maybe it was in some other work with some of the same authors.
I know that doesn't help much.

At any rate, the article said nothing about disambiguating via word-order or hierarchy.
In fact they specifically denied that word-order would help. (If memory serves! I'm almost sure, but I can't be really sure until I can find it again.)
And in the example given, if I remember correctly, there was no useful hierarchical difference between the participants, unless one referred to prior discourse (and in that example one couldn't).

I'm not sure that that particular example was one that disambiguated by one of the strategies I suggested in my post; but I distinctly remember forming the impression that the author(s?) thought at least one of those strategies was used in some language with "monster raving loony" alignment (my words, but it's what I thought they meant).

They did mention that the speaker could always disambiguate, if the addressee asked them to.

I can't even think of a linguistic (that is, non-paralinguistic) way to disambiguate that doesn't involve saying another clause.
But it is accepted as an axiom of linguistics, that any thought that can be expressed in one language, can also be expressed in any other language.
It may be a lot wordier in the second language, but, it can still be expressed.

[hr][/hr]

I don't remember ever seeing any evidence that all three strategies existed in any one natlang.
I just wonder whether such a language would in fact be usable.

[hr][/hr]

Did you have any thoughts about how, in a language without ditransitive or even tervalent clauses, to express something that needs a ditransitive clause in languages that do have such clauses?

Again, I think one needs to say at least two clauses.

Their verbs needn't be morphologically related to each other.
My "example" may have caused some readers to think that I meant to imply they do need to be morphologically related.

But I think one must go rather deep into P.I.E. roots to find the common ancestor of "give" and "have" and "keep" and "receive".

So let's just say the language has at least two of the following three verbs:
"Don verb1 Thelma" means "Don sent Thelma (to someone)";
"Don verb2 Rhys" means "Don sent (someone) to Rhys";
"Rhys verb3 Thelma" means something like "Rhys received Thelma" (because someone sent her to him).

As I understand it, languages without ditransitive verbs, do have some two of those concepts expressed by monotransitive verbs. I don't know that any language has all three; at least, not consistently for every situation that some other language would use a ditransitive verb for.

[hr][/hr]

Am I understanding you right?
Am I making myself understood?

-----

Thanks for your reply!

I see others have replied as well; I haven't read them yet, but I'm about to.

[hr][/hr]
[hr][/hr]
[hr][/hr]
Creyeditor wrote:I fail to see how this relates to pragmatics. Pragmatics usually depend on the context.
I was thinking of the Gricean maxims, two of which are "say everything you have to say" and "don't say any more than you have to say".
Am I wrong about it, or aren't Gricean maxims usually considered part of "pragmatics"?
Creyeditor wrote:.... I think this might be a good idea to look at, because it shows that these situations are real.
I really appreciate your examples! For more than one reason; the "it shows that these situations are real" reason is only the most noteworthy.
Creyeditor wrote:.... German also does not have overt detransitivizing morphology. So, technically it does not meet your high standards. ...
I did not mean to get people tripped up on the detransitivizing or valency-reducing morphology.
As far as I know the language doesn't have to have any; or, at least, the disambiguating clauses don't have to be created by using such morphology (for instance, maybe the detransitivized verbs may be suppletive rather than transparently inflected).
I can tell I misled at least two responders; but I still can't figure out how I might have avoided doing so. I wish I could. (I probably will, eventually, in the middle of a sound sleep!)


BTW:
Creyeditor wrote:This sentence sounds somehow confusing. As if someone tells a story and does not know where to start. Sounds a bit like someone drunk speaking.
I'm pretty sure I can be that confusing with only caffeine for a drug.
Last edited by eldin raigmore on 30 Mar 2017 18:02, edited 2 times in total.
Salmoneus
MVP
MVP
Posts: 3030
Joined: 19 Sep 2011 19:37

Re: Pragmatics: What don't you need to say?

Post by Salmoneus »

I don't understand. If you don't know the name of the language you're discussing, how do you expect anyone to know what speakers of the language would assume any given sentence in that language to mean?
User avatar
Creyeditor
MVP
MVP
Posts: 5091
Joined: 14 Aug 2012 19:32

Re: Pragmatics: What don't you need to say?

Post by Creyeditor »

eldin raigmore wrote:
Creyeditor wrote:I fail to see how this relates to pragmatics. Pragmatics usually depend on the context.
I was thinking of the Gricean maxims, two of which are "say everything you have to say" and "don't say any more than you have to say".
Am I wrong about it, or aren't Gricean maxims usually considered part of "pragmatics"?
You are right, but these maximes are dependent on the linguistic and extralinguistic context; I think this is important here. So, some of the sentences might get less weird if you provide a context for them.
Creyeditor
"Thoughts are free."
Produce, Analyze, Manipulate
1 :deu: 2 :eng: 3 :idn: 4 :fra: 4 :esp:
:con: Ook & Omlűt & Nautli languages & Sperenjas
[<3] Papuan languages, Morphophonology, Lexical Semantics [<3]
Post Reply