Pragmatic Roles - Topic and Focus

The best topics from Languages & Linguistics, kept on a permanent basis.
Post Reply
chris_notts
Avisaru
Avisaru
Posts: 275
Joined: Wed Dec 15, 2004 9:05 am
Location: Nottingham, England
Contact:

Pragmatic Roles - Topic and Focus

Post by chris_notts »

This issue has been coming up more and more in the areas which I've taken an interest recently, especially things like anaphora and voice (basically, my current greatest linguistics interest is referent tracking and related areas). But what frustrates me so much is the fact that various terms like "focus" and "topic" are thrown around with vague descriptions that clearly aren't sufficient... or often well defined.
For example, I now have my own copy of Grammatical Voice by Klaiman (another christmas present). The main part of the book that interests me is Ch 5 on "Information Salience Voice Systems"... the author categorises voice according to whether they are basic (the middle voice especially), derived (voices that remap core roles, eg passives and anti-passives as normally concieved) or information salient, ie based on the pragmatic roles of the verbal arguments (in terms of things like topic or focus) and/or their animacy/control/other property (eg inverse systems).
An example of the latter type is for example Ixil, which has a system that allows NPs recieving (contrastive) focus to be fronted and their role marked on the verb rather than with the normal preposition. For example:

A-k'oni in ta'n uula
2nd.erg-shoot 1st.abs with sling

You shot me with a sling

uula a-k'oni-b'e in
sling 2nd.erg-shoot-index me

with a sling you shot me

The language also has separate processes for topic fronting that don't involve verbal morphemes indicating role of the fronted NP.
Klaiman then tries to claim that the trigger systems of the Phillipine languages like Tagalog are also focus orientated systems, but, he says, "Interestingly, there are also important differences between Phillipine focus and the information salience termed salience-F in the previous section's discussion of Mayan languages. For instance, according to section 5.1, salience-F is generally emphatic or contrastive. However, this does not necessarily hold of focus in Phillipine languages". Now, it may be just me, but where some kind of contrastive meaning (even if not heavy) seems like a prerequisite for what's generally termed focus. If there is no contrast (no other option that it's opposed to in the minds of the speaker and/or listener) then there's no need to focus it. From the other things said, it seems extremely dubious to me that trigger selection is based on anything near the same notion as the kind of movement illustrated in Ixil, despite Klaiman's insistence to the contrary.
I've done a lot of searching in books at the university, on the internet, everywhere for a well written and convincing account of topicality, focus, and other pragmatic roles, and I've failed to find a single one. Every single description is vague and fails to conver significant areas of use in any group of languages all of which are said to have some kind of grammaticalization in the form of morphology or syntax of the notions of topic, focus, or other pragmatic statuses. I'm beginning to doubt that such notions are universal at all: I accept that the topic may be well defined in a single language such as Japanese, and that it may be possible to describe the core nature of the status it marks in that language, but it doesn't seem clear to me that something else termed Topic elsewhere will have a clear underlying identical or similar concept. The whole area seems both:

i) under-researched

and

ii) taken for granted too much. People take for granted that the categories and ideas they talk about in pragmatics are in fact real without doing serious study of them.

What do you think? And do you know any good book on such elements of pragmatics that don't just wave away the problem with vague useless descriptions like

"the topic is what the clause is about"

or "the focus is the new information"?

EDIT: I've also over the years done a lot of searching for a convincing description of the exact pragmatic role of the trigger in Tagalog and related languages, and failed to find anything truly believable about that either. Every author seems to disagree, with some (such as Klaiman) characterising it as focus, some trying to identify it with topic, some insisting that the only thing it indicates is referentiality (and in particular definiteness) and some proposing other explanations completely different.

EDIT EDIT:

"the topic is what the clause is about" is clearly not an adequate description because about can mean so many things. Take for instance

"John hit Pete"

what is that clause about? On the one hand, it is about John, because it tells me something about his past actions and possibly what kind of man he is, but on the other hand it tells me something about Pete because it tells me what state Pete was (or is) in. I could expect after hearing such a statement that Pete might be injured in some way when I next see him, perhaps having a black eye.
And that's the point: in most cases, any (realis) clause will tell you something about most of its arguments, oblique or core, so aboutness is not relevant to the topic role in and of itself. Topicness seems to have more to do with the point of view adopted by the speaker than it has to do with any abstract notion of aboutness in most languages with topics I've seen, although even that description is too vague to be useful.
Try the online version of the HaSC sound change applier: http://chrisdb.dyndns-at-home.com/HaSC

User avatar
nebula wind phone
Sanci
Sanci
Posts: 67
Joined: Wed Mar 02, 2005 10:58 am
Location: Austin, Texas, USA
Contact:

Post by nebula wind phone »

Well, pragmatics covers a much broader subject than syntax or phonology. Phonology is about the structure of sounds in language, and syntax is about the structure of sentences, but pragmatics is about the structure of human interaction. Human interaction is big and hairy and vague and impossible to pin down. So I'm not surprised that theories of pragmatics turn out big and hairy and vague too.
"When I was about 16 it occurred to me that conlanging might be a sin, but I changed my mind when I realized Adam and Eve were doing it before the Fall." —Mercator

User avatar
gach
Avisaru
Avisaru
Posts: 472
Joined: Mon Feb 17, 2003 11:03 am
Location: displaced from Helsinki

Post by gach »

This is a subject I too have had some iterest in and done a little research about. Unfortunately pretty much all I've found are same as the problems you've ran across. The overall picture I've got is that the exact meanings of Topic and Focus may vary much between different langs and that considering the language specific roles instead of the theoretical cross linguistical roles makes often much more sense. In addition, as I'm interested for my conlangs of using pragmatic marking to mark what's important in a clause, I've been puzzled by the fact that different langs seem to use the T/F marking in wildly different ways though I haven't seen the reasons for this anywhere. Some clarification would indeed be good.

The following quote gave me some deeper understandment about the subject, though it still leaves many things very open. The reference Lambrecht (1994) seems to be quite a central work, though I don't know to tell about it.
Elena Maslova in 'Information focus in relational clause structure', ([url]http://www.stanford.edu/~emaslova/FocusSplits.pdf[/url]) wrote:2.1 Terminological and notational conventions

Following Lambrecht (1994), I will use the term focus structure to refer to different mappings between the propositional contents of a contextualized clause and its pragmatic articulation. There are three basic types of pragmatic articulations, topic-focus, focus presupposition, and thetic. The topic-focus articulation singles out one discourse referent (topic), t, and the information conveyed by the clause (focus) is construed as information about t. Hence, the focus includes all components of the clause except for the expression referring to t (topic expression) (Strawson 1964; Kuno 1972; Reinhart 1982; Gundel 1988; Lambrecht 1994; inter alia). In the focus-presupposition articulation, the information conveyed by the sentence is contained within one nominal constituent (narrow focus), whereas the remainder of the clause is construed as its ragmatic presupposition (Chafe 1976; Prince 1978; Lambrecht 1994; inter alia). In the thetic articulation, the whole clause is presented as the focus (Kuno 1972; Sasse 1987; Shibatani 1990: 262-264; Lambrecht 1994). In what follows, I refer to all components of pragmatic foci as focal elements; similarly, the notion of topical element subsumes topics proper and nominal elements of pragmatic presuppositions.

...

Lambrecht, Knut. 1994. Information Structure and Sentence Form. Cambridge: Cambridge University Press.

gear
Sanci
Sanci
Posts: 18
Joined: Thu May 15, 2003 10:02 am

Post by gear »

This was a very good question. To my thinking, you need to consider language in use to come up with any kind of answer at all. Topic is something you have when you are actually talking with someone, i.e. using the language, and he or she is interested in listening to you.

- Hey, Chris, I was thinking about topics and comments the other day.
- Yes?
- Well.. as for comments, they usually follow topics. I wonder why that is.
- Erm...
- Of course, there I had comments before topics, because my topic was comments, and my comment involved topics.
- Eh..
- Blah blah blah blah

And then you would probably hit me to get me to shut up. That's what happens when people violate the gricean maxims of communication. It really can get ugly.

What I'm after is that every time you say something, you give information about a host of things. Topic and comment is about how you structure that information so the other guy can receive it and know what's important. SO you take into account what he or she already knows, or what you expect he or she to know, and what is the new information.

You probably don't just walk up to people and say "John hit Pete" to them without any context. Though if you want to have some innocent fun, you can do that. Watch out that they don't hit oyu, though.

User avatar
gach
Avisaru
Avisaru
Posts: 472
Joined: Mon Feb 17, 2003 11:03 am
Location: displaced from Helsinki

Post by gach »

Tsumu wrote:You probably don't just walk up to people and say "John hit Pete" to them without any context. Though if you want to have some innocent fun, you can do that. Watch out that they don't hit oyu, though.
So when no element of this clause is supposed to be already known the whole clause would classify as a focus.

[John hit Pete]F

In a more listener friendly way of telling this you would first introduce one of the participants in a different clause and then go with the original clause, that would now only have a topic eg. like this:

[There was this John]F. [John]T [hit Pete]F.
[There was this Pete]F. [Pete]T [was hit by John]F.

Would this be a correct analyse?

Another question I have is that, as different langs mark topics while others foci (ough what word :? ), can we suspect there to be some remarcable differences in the typology of the langs or is this just a random choise?

gear
Sanci
Sanci
Posts: 18
Joined: Thu May 15, 2003 10:02 am

Post by gear »

After thinking about poor Pete and that aggressive ba**d John, I realize there's a pragmatic situation where you might just say exactly "John hit Pete." and have it all in focus as gach suggested. Say that you were reporting what just happened, perhaps as an urgent call for help; this would be something you'd yell at people.

(Help! Come quick!) John just hit Pete!

It might be interesting to think from a conlanging perspective what a language that did not follow the topic comment pattern at all would be like. What we know is that some natlangs grammaticalize it, but wouldn't it be fair to say that all languages in use work that way, in the sense of pragmatics.

I am not sure it would be possible to conceive of such a language. After all, if you look hard enough, you can always find something that looks a lot like topics and comments.

P.S. You can use focuses, I think. Classical plurals are not mandatory after all. Using them carelessly results in many *dilemmae. (should be ?dilemmata or dilemmas)

User avatar
Radagast
Lebom
Lebom
Posts: 97
Joined: Sat May 21, 2005 9:46 am
Location: Denmark
Contact:

Post by Radagast »

I have just read a very interesting (and I believe correct) interpretation of the topicalisation-focus system of nahuatl. It touches upon some questions on the use of certain (p)articles that I have recently discussed shortly with !Pap. I will be using this interpretation in my thesis and I think it would be a help to me to write a bit about it inorder to make sure that I can verbalize my understanding of it. Might I post a few considerations on it here, or should I make another thread?

Anyway the theory is made by Michel Launey who is a funtional linguist dealing with classical nahuatl almost exclusively (and I think some Guyanan languages as well). The basic idea is that since nahuatl is a predicate language it need some kind of system to distinguish topics from comments. I.e. Because every noun is in fact a predicate and there are no copula then how do you know which of two juxtaposed nouns or verbs constitute the topic and which the comment. Nahuatls word order is grammatically non-configurational which makes it even harder to follow.

I'd like to know how other non-configurational and predicative languages work around these seemingly problematic areas.
Last edited by Radagast on Thu Dec 29, 2005 11:07 am, edited 1 time in total.
[i]D'abord on ne parla qu'en poésie ; on ne s'avisa de raisonner que long-temps après.[/i] J. J. Rousseau, Sur l'origine des langues. 1783

User avatar
Rik
Lebom
Lebom
Posts: 134
Joined: Fri Jul 04, 2003 2:57 pm
Location: London, UK
Contact:

Re: Pragmatic Roles - Topic and Focus

Post by Rik »

chris_notts wrote:What do you think? And do you know any good book on such elements of pragmatics that don't just wave away the problem with vague useless descriptions like "the topic is what the clause is about" or "the focus is the new information"?
Chris - if you ever find an answer to this one, please let me know. I use topic/comment and focus stuff in Gevey, but I'm very aware it has no sound basis in theory - pragmatics is a subject I've steered clear from to date, probably to the detriment of my conlang.

Rik

User avatar
Jipí
Smeric
Smeric
Posts: 1128
Joined: Sat Apr 12, 2003 1:48 pm
Location: Litareng, Keynami
Contact:

Re: Pragmatic Roles - Topic and Focus

Post by Jipí »

Just for what it's worth ... As for trigger systems, I have heard that some see the system as working with modes rather than cases, but I guess you know that. I haven't completely understood yet why, though.

User avatar
Radagast
Lebom
Lebom
Posts: 97
Joined: Sat May 21, 2005 9:46 am
Location: Denmark
Contact:

Post by Radagast »

(my rephrasing of my Proffessor Una Cangers interpretation of Michel Launey)

Michel Launey posits the interpretration that nahuatl syntax is largely understandable from a pragmatic viewpoint rather than a grammatical one. In nahuatl both nouns and verbs can function as a predicate of a sentence.

In classical nahuatl the noun tla:catl can be translated both as "man" and "he/it (who) is a man" depending on context.

When studying classical nahuatl wordorder there are hardly any sequences of predicates and their arguments that aren't found. This suggests that nahuatl is grammatically non-configurational and that the tendencies that are found are governed by pragmatics. Studying pragmatics from texts in a language that is no longer spoken is necessarily difficult to the extreme because most of the mechanisms that are normally studied in the field of pragmatics pertain to language spoken
in context. However studying nahuatl Launey finds some clues to an understanding of how classical nahuatl dealt with pragmatic issues, namely two particles "in" and "ca" which working together with the order of phrasal constituents provide the basis for the two pragmatic principles that Launey have dubbed thematisation and emphasis/focus.

The two particles have basic functions that are expanded two function syntactically as well as pragmatically. "In" is basically a definite article preceding known arguments that are known and definite. "Ca" is a particle that is optionally used preceding a noun to show that it is functioning as a predicate.

For something to be thematized to Launey means that it is made the specific subject of the discourse, it is information presupposed to be known to the speaking parties. The fact that it is known is shown by the particle "in" and the thematization is shown by fronting of the known subject or known object.

This makes the sentence structure (in N Pred)
in cihua:-tl ?-cochi
DET woman-ABS 3sS-sleep
"the woman, she sleeps" (or maybe better transliterated as "oh the woman, she sleeps" or "as for the woman she sleeps")
It also works if the predicate is a noun:
in Pedro ca ?-mex?ca-tl
DET Pedro ca 3sS-mexican-ABS
"Pedro, he is a mexican"
or if the object is thematized:
in cal-li ni-qu-itta "the house, I see it"
DET house-ABS 1sS-3sO-see
If thematising first or second persons the free pronouns are fronted:
(in) n?huatl (ca) ni-mex?ca-tl
DET I ca 1sS-mexican-ABS
"Me, I am a mexican" (answering what are you?)
but also pronouns for the third person can be used to thematize something that has already been mentioned:
(in) y?huatl (ca) ni-c-nequi "as for that, i want it"
DET He/she/it ca 1sS-3sO-want

Emphasis/focus is used when the verbal proces is known whereas the new information is what is expressed as subject or object: The arguments that receive emphasis are never definite (preceded by the (p)article "in") subject or object functions as predicate whereas the verb becomes subject. More precisely there are two predicates: a main predicate and a subordinate predicate with a coreferential subject.

(ca) ?-cihua:-tl in ?-tz?tzi
PRED 3sS-woman-ABS SUBORD 3sS-scream
"it is a woman who screams/the one screaming is a woman" (answer to "what is it that is screaming?")

Also what is semantically an object can receive emphasis and work like a main predicate in a construction with in+verb as a subordinate predicate.

(ca) ?-naca-tl in ni-c-cua "it is meat that which I eat"
PRED 3sS-meat SUBORD 1sS-3sO-eat

Thematization is proces in a which a known argument of a new predicate is fronted and preceded by the Definite article "in" in order to make it the specific theme of the discourse.

Emphasis is a proces in which a new argument of a known predicate is fronted, turning the known predicate into a subordinate predicate with the Subordinating particle "in", in order to emphasise it as the new information and possible new theme of the discourse.

(of course the particle is the same but it has two functions, which is really the same function: showing part of the information is known like a normal definite article, but since it can also be used in fornt of a known verbal predicate its function also becomes subordinating, something which is also seen when in is used as introducing relative clauses)
[i]D'abord on ne parla qu'en poésie ; on ne s'avisa de raisonner que long-temps après.[/i] J. J. Rousseau, Sur l'origine des langues. 1783

chris_notts
Avisaru
Avisaru
Posts: 275
Joined: Wed Dec 15, 2004 9:05 am
Location: Nottingham, England
Contact:

Post by chris_notts »

Radagast wrote: Michel Launey posits the interpretration that nahuatl syntax is largely understandable from a pragmatic viewpoint rather than a grammatical one. In nahuatl both nouns and verbs can function as a predicate of a sentence.
The Salishan languages also seem to work in the same way... I've read arguments that they do distinguish in some ways between noun and verbs (although it's less than obvious, and I'm not entirely convinced) but what they actually seem to distinguish is the follow: subordinate (ie argument, whether clause or noun, of another verb) vs not subordinated, and on the other hand predicate (which may be something translated as a noun in English, not necessarily something verbal) vs the rest.
In classical nahuatl the noun tla:catl can be translated both as "man" and "he/it (who) is a man" depending on context.
What about the modern descendents of nahuatl? Has spanish contact influenced them into making a bigger (more clear-cut) distinction between nouns and verbs?
If thematising first or second persons the free pronouns are fronted:
(in) n?huatl (ca) ni-mex?ca-tl
DET I ca 1sS-mexican-ABS
The definite article occurs with first and second person pronouns?
(ca) ?-cihua:-tl in ?-tz?tzi
PRED 3sS-woman-ABS SUBORD 3sS-scream
"it is a woman who screams/the one screaming is a woman" (answer to "what is it that is screaming?")
Would a good rendering be "the woman she is the one who screams"? Incidentally, how does Nahuatl form headed relative clauses? Can (this is a guess) ?-cihua:-tl in ?-tz?tzi without the ca also mean "the woman who screams"?
The funny thing is, a lot of these processes parallel processes from languages as diverse as Tagalog (Phillipine languages) to Lisu (Sino-Tibetan). For example, the following is a Lisu example from a pdf I have (sorry about the screwed up formatting and missing tones):

asa la d@-a ma nya ale
asa to hit-dec one topic ale

"The one who hit asa is ale"

This is almost the required answer to the question:

asa la d@-a ma nya ama-a
asa to hit-dec one topic who-question

"Who hit ale?"

And the question itself may occur in two basic forms, both of which force the question word out of the scope of the topic marker because the form with the question word possibly topical (rather than being the non-topic marked part of the clause as above) is ungrammatical. The similar Tagalog clause is:

Sino ba ang gumawa noon
who question trigger do that

Who did that?

Where again (if we interpret ang as marking some kind of topicality or definiteness, kindof like "in") the entire clause except the question word is in the scope of the topical marker, and the answer must occur in the same form with the questioned NP remaining focused but not topical. If we interpret ang as marking focus (as some people have tried to claim) then we have to answer the question of why exactly in these kinds of structures tagalog patterns in exactly the opposite way to other languages which clearly mark topic and focus in things like questions and answers.
This is all assuming that topic and focus are well defined of course. :D
Try the online version of the HaSC sound change applier: http://chrisdb.dyndns-at-home.com/HaSC

TomHChappell
Avisaru
Avisaru
Posts: 807
Joined: Wed Dec 28, 2005 2:58 pm

Re: Pragmatic Roles - Topic and Focus

Post by TomHChappell »

chris_notts wrote: This issue has been coming up more and more in the areas which I've taken an interest recently, especially things like anaphora and voice (basically, my current greatest linguistics interest is referent tracking and related areas). But what frustrates me so much is the fact that various terms like "focus" and "topic" are thrown around with vague descriptions that clearly aren't sufficient... or often well defined.
-- SNIP --
And that's the point: in most cases, any (realis) clause will tell you something about most of its arguments, oblique or core, so aboutness is not relevant to the topic role in and of itself. Topicness seems to have more to do with the point of view adopted by the speaker than it has to do with any abstract notion of aboutness in most languages with topics I've seen, although even that description is too vague to be useful.
Hi, Chris.

As for Focus;
There seems to be a difference between "emphatic" focus and "empathic" focus.
"Emphatic" focus seems to refer to "what the speaker thinks is the most important part of what he/she is saying".
"Empathic" focus seems to refer to "the participant from whose point-of-view the event is being presented". (A.k.a. "perspective").

Siewierska, in "Person", listed several kinds of "emphatic" focus, one of which was "contrast". Another was question-word focus; another was "answer" focus; and there were more. One kind of focus suggested was "truth-value" focus, in which the speaker intends the focus to be on the entire clause.

I don't think she intended her list to be exhaustive, though I gather she thought she was listing everything major she already knew about at that time that she considered to be well-established by then.

"Empathic" focus, or point-of-view or perspective, is, classically, what "grammatical voice" is about ("Classic" in this case may refer to Greek and Latin and Sanskrit more than to English and German and French and so on.). I think "trigger languages"'s "triggers" probably have to do with these "empathic" foci. Some writers would rather call them "topics", and some would rather call them "subjects".

It has been suggested that in a relative clause, the head of the RC is always the empathic focus of the RC. It has been suggested that in any clause there is a strict, linear, total ordering of participants from most "empathic" to least "empathic". In simple sentences it might not be very important to know much more about this than which participant is "the most empathic"; but in every complex sentence, any shared participants between a matrix clause and a subordinate clause, must be in the same order-of-empathy in both clauses.

-----

As for Topics;
there are again two things to consider; "aboutness" and "givenness".
I'm not going to say much about "aboutness" here. I will mention, however, that a clause may have an "about"-topic; the sentence of which it is part may have the same or a different "about"-topic; the paragraph of which that sentence is a part may have the same or a different "about"-topic; and a larger segment of discourse of which that paragraph is a part may have yet another "about"-topic.

In the literature there seem to be about seven different kinds of "givenness". Two of them make good sense to me; "knowledge-given" and "attention-given".

I think it was Chafe who said "knowing about something and thinking about something are two different mental states."

Something is treated as "knowledge-given" if the speaker assumes the addressee already "knows about" it; something is treated as "attention-given" if the speaker assumes the addressee is already thinking about it.

A couple of points:
Usually anything that is attention-given is also knowledge-given; but often not vice-versa.
Usually knowledge-givenness does not fade; once something is knowledge-given, it stays knowledge-given.
Attention-givenness, however, does fade; if neither of us mentions, for long enough, something to which we have been paying attention, then one of us wants to re-introduce it as a topic (sorry), we may need to be pretty explicit about it so as not to be misunderstood.

------
Does any of that help?

-----
Tom H.C. in MI

chris_notts
Avisaru
Avisaru
Posts: 275
Joined: Wed Dec 15, 2004 9:05 am
Location: Nottingham, England
Contact:

Re: Pragmatic Roles - Topic and Focus

Post by chris_notts »

TomHChappell wrote: As for Focus;
There seems to be a difference between "emphatic" focus and "empathic" focus.
"Emphatic" focus seems to refer to "what the speaker thinks is the most important part of what he/she is saying".
"Empathic" focus seems to refer to "the participant from whose point-of-view the event is being presented". (A.k.a. "perspective").
HHm. I've been doing a lot of reading since I posted this (Lambrecht in particular), and I'm not sure that Empathic focus actually is focus, if you see what I mean. Pragmatics is a very interesting area, and one that is still not well investigated... which is unfortunate, since it seems to me to offer more scope for true linguistic universals that morphological or syntactic typology.
Siewierska, in "Person", listed several kinds of "emphatic" focus, one of which was "contrast". Another was question-word focus; another was "answer" focus; and there were more. One kind of focus suggested was "truth-value" focus, in which the speaker intends the focus to be on the entire clause.
I own this book. I will have to re-read the relevant section of it... I think that Siewierska's truth value focus would be Lambrecht's "Sentence Focus Structure", and from Erteschik-Shir's point of view a clause where the stage is the topic (since she does not accept that declarative clauses with no topical elements at any level can exist, and it is well known that topics are often not overt, she posits that a relevant spatial-temporal location is the topic in clauses which otherwise do not seem to have one).
I think "trigger languages"'s "triggers" probably have to do with these "empathic" foci. Some writers would rather call them "topics", and some would rather call them "subjects".
I don't believe so, at least not in Tagalog. Tagalog chooses somewhat odd triggers if the selection criteria is purely topicality or empathy, at least from the examples I've seen (do you remember the discussion in Klaiman about this issue?). The trigger in Tagalog seems to me to be a somewhat odd mix between focus and topic... on the one hand, it has to be referential and generally identifiable (which is odd if it is focus determined) but on the other hand it seems to choose really odd topics if it is topic determined. It often seems to me to often mark locus of attention, which is not generally something associated with topics (that is, topics are generally active, but in the clauses they occur in they do not generally become *activated*, ie dragged from longer term memory or brought into a new context).
It has occured to me and others that perhaps it marks neither of these notions ("Topic", "Focus").
It has been suggested that in a relative clause, the head of the RC is always the empathic focus of the RC. It has been suggested that in any clause there is a strict, linear, total ordering of participants from most "empathic" to least "empathic". In simple sentences it might not be very important to know much more about this than which participant is "the most empathic"; but in every complex sentence, any shared participants between a matrix clause and a subordinate clause, must be in the same order-of-empathy in both clauses.
Yes, I agree that generally the head of a relative clause must be treated as a topic within the relative clause. Incidentally, this might provide further evidence for your hypothesis that Tagalog triggers truly are topics: the head must occur in the trigger role in the relative clause, which is what you would expect if it were topical but not if it were focal.
Something is treated as "knowledge-given" if the speaker assumes the addressee already "knows about" it; something is treated as "attention-given" if the speaker assumes the addressee is already thinking about it.
"Topic" in the work of most linguists typically seems to mean this kind of givenness. That is, a topic is usually a referent already in "short term memory" and easily accessable to the addressee, rather than something the addressee knows about but is not currently active.
Try the online version of the HaSC sound change applier: http://chrisdb.dyndns-at-home.com/HaSC

User avatar
Leo
Niš
Niš
Posts: 10
Joined: Sun Jul 09, 2006 9:29 pm
Location: France/Spain

Post by Leo »

There should be much to learn about this topic by comparing good and bad writers.

Good writers don't get you lost or stuck, they know what happens in your mind while you read, and they lead you smoothly.

Bad writers change track and ideas without warning, making you read back with the false impression you've missed a step.

Maybe some researchers did make studies about pragmatics in literature?

TomHChappell
Avisaru
Avisaru
Posts: 807
Joined: Wed Dec 28, 2005 2:58 pm

Re: Pragmatic Roles - Topic and Focus

Post by TomHChappell »

chris_notts wrote:
TomHChappell wrote:
As for Focus;
There seems to be a difference between "emphatic" focus and "empathic" focus.
"Emphatic" focus seems to refer to "what the speaker thinks is the most important part of what he/she is saying".
"Empathic" focus seems to refer to "the participant from whose point-of-view the event is being presented". (A.k.a. "perspective").

HHm. I've been doing a lot of reading since I posted this (Lambrecht in particular), and I'm not sure that Empathic focus actually is focus, if you see what I mean. Pragmatics is a very interesting area, and one that is still not well investigated... which is unfortunate, since it seems to me to offer more scope for true linguistic universals that morphological or syntactic typology.

"Empathic Focus" may not be really a "Pragmatic" concept. That is, "given vs new" and "topic vs comment" and "focus vs (back)ground" all have to do with "Information Structure", and some with "Discourse Structure"; whereas "empathic focus" or "perspective" or "camera-angle" or "point-of-view" has more to do with "Information Packaging" than with structure. I think it's more of a "semantico-syntactic" concept; though perhaps the "semantico-" part is "semantico-pragmatic". In any case it has more to do with "information packaging" than with other areas of pragmatics.
BTW "Empathy" in this use doesn't have quite the same meaning as "empathy" normally has (outside of linguistics).

I mentioned it more as an example of the confusion with which the term "focus" is used, than as a different pragmatic concept also called "focus".

I had brought up some of these ideas some time ago on the CONLANG-L list, and someone had disagreed with my use of the term "focus" to mean "emphasis"; to him/her, "focus" meant "perspective". It turns out it means one or the other (or, maybe, something else?), depending on what you are reading at the moment.

See the thread beginning with
http://listserv.brown.edu/archives/cgi- ... 3&m=110427

and in particular
http://listserv.brown.edu/archives/cgi- ... =0&P=40579

and
http://listserv.brown.edu/archives/cgi- ... 3&m=128978

(BTW I noticed you posted some very interesting contributions to that thread.)
chris_notts wrote:
TomHChappell wrote: I think "trigger languages"'s "triggers" probably have to do with these "empathic" foci. Some writers would rather call them "topics", and some would rather call them "subjects".

I don't believe so, at least not in Tagalog. Tagalog chooses somewhat odd triggers if the selection criteria is purely topicality or empathy, at least from the examples I've seen (do you remember the discussion in Klaiman about this issue?).

I think I remember some of it.
I also remember some of what was in Li and Thompson's "Subject and Topic".
chris_notts wrote:
The trigger in Tagalog seems to me to be a somewhat odd mix between focus and topic... on the one hand, it has to be referential and generally identifiable (which is odd if it is focus determined) but on the other hand it seems to choose really odd topics if it is topic determined. It often seems to me to often mark locus of attention, which is not generally something associated with topics (that is, topics are generally active, but in the clauses they occur in they do not generally become *activated*, ie dragged from longer term memory or brought into a new context).
It has occured to me and others that perhaps it marks neither of these notions ("Topic", "Focus").

Actually I believe that the trigger in trigger languages generally, and in Tagalog in particular, are really thought to be almost the perfect examples of "empathic focus" a.k.a. "empathic subject"; at least according to some of the grammarians of such languages. Look at the articles in Li and Thompson and at the works those articles refer to. They say something like "the [trigger] is the participant from whose [point of view] the action is [viewed]."
chris_notts wrote:
Yes, I agree that generally the head of a relative clause must be treated as a topic within the relative clause. Incidentally, this might provide further evidence for your hypothesis that Tagalog triggers truly are topics: the head must occur in the trigger role in the relative clause, which is what you would expect if it were topical but not if it were focal.

Thanks.
I want to demur a little from the phrase "your hypothesis"; while I have not made up my mind either to accept or to reject that hypothesis, I know I didn't invent it.
chris_notts wrote:
TomHChappell wrote: Something is treated as "knowledge-given" if the speaker assumes the addressee already "knows about" it; something is treated as "attention-given" if the speaker assumes the addressee is already thinking about it.

"Topic" in the work of most linguists typically seems to mean this kind of givenness. That is, a topic is usually a referent already in "short term memory" and easily accessable to the addressee, rather than something the addressee knows about but is not currently active.
If one already has the "given" vs "new" distinction, it seems a shame to waste the "topic" vs "comment" distinction on givenness instead of "aboutness".

These distinctions are not going to be statistically independent anyway. (There's good reason to believe -- and it is widely believed -- that they will be logically independent.) Topics are ordinarily going to be the same as the topic of the previous remark; thus they will ordinarily be attention-given. But often a new topic will be brought up; it will then be attention-new. Sometimes it will even be knowledge-new. Ordinarily the topic will be in the background rather than in the focus. But if I bring up a new topic, or re-introduce one that isn't the same as that of the previous remark, I will usually focus on it. Ordinarily anything new I say will be in focus; and ordinarily anything given I say will be in background. But sometimes I will want to focus on something which is knowledge-given and may even have recently been attention-given, because I think it is insufficiently attention-given at-the-moment.

Also in some languages there are three degrees, rather than only two, of some of these distinctions; for instance there can be given matter, new matter, and, between them, transitional matter. Also, in some languages an utterance can have more than one focus-of-contrast or focus-of-attention or emphatic focus (focus-of-emphasis). And there can be both overlaps and gaps; part of a sentences topic can be part of its focus, or there may be part of a sentence which is neither part of its topic nor part of its focus.

If you track down the CONLANG-L thread I was talking about, you'll see I said many (even most modern) writers seem to believe there are at least two dimensions; but there seem to be three main candidates for what these two dimensions are.
"Given" vs. "New";
"Topic" vs. "Comment";
"Focus" vs. "(back)ground".

(See the thread beginning with
http://listserv.brown.edu/archives/cgi- ... 3&m=110427

and see in particular
http://listserv.brown.edu/archives/cgi- ... 3&m=128978
)

Each of these has several interpretations.

If we allow the "given" vs "new" distinction to have its own label, what does the "topic" vs "comment" distinction mean that the "given" vs "new" distinction doesn't already cover? Answer; when "topic" isn't just a synonym for "given", it mostly means "what the comment is about".

----

I also said, in that thread, that several writers think the two-dimensional figure resembles a triangle more than a square; but that I think conlangers could profitably go ahead and pretend it is a cube.

See in particular
http://listserv.brown.edu/archives/cgi- ... 3&m=128978

_________________

Post Reply