Theiling Online    Sitemap    Conlang Mailing List HQ   

Re: OFFLIST: Re: Topic/Comment, Focus/Background, Given/New.

From:Aidan Grey <taalenmaple@...>
Date:Thursday, December 15, 2005, 5:24
Hi Tom!

  This got sent to me on accident, I think. Just a heads up!


Tom Chappell <tomhchappell@...> wrote:
  Hi, and thanks for your reply.

I hope you get this one faster.

Your reply has prompted me to write out some thoughts
I have been having.

If you look into those references by Mark Steedman and
Ivana Kruijff-Korbayova, you'll see they say that most
modern researchers propose that a sentence's
information structure varies along two dimensions; a
"givenness" dimension (given vs new) and an
"aboutness" dimension (topic vs comment).

Some apparently think one of these dimensions just
"splits one end of the other dimension", leading to a
kind of "triangle"; others apparently think the two
dimensions are quite independent, leading to a kind of
"square". Steedman and Kruijff-Korbayova haven't made
up their own minds, apparently; they led an annual
conference basing it on one assumption in 2001(?), and
led it again in 2004(?) basing it on the other
assumption. (If I understood correctly.)

For purposes of conlanging, it seems proper to allow
ourselves the maximum flexibility that hasn't actually
been ruled out; so, I propose some languages may be
allowed to have a "square" information structure.

Another difference between what I was proposing for
conlangs, and what actually takes place in the
professional literature about natlangs, was my
proposing three levels of I.S. instead of only two.

Square example:
There is an example among the references I sent you of
a sentence whose Topic and Comment each have a Ground
and a Focus, according to the linguist who analyzes
it.* The analysis supposes that the sentence answers
an implicit wh-question. The Topic selects the
wh-question, and the Comment selects the answer.
There is, implicitly, a set of possible alternative
answers; their common part is the Ground of the
Comment, and the part unique to the actual alternative
selected is the Focus of the Comment. There is also,
implicitly, a set of alternative questions; their
common part is the Ground of the Topic, and the part
unique to the actual alternative selected is the Focus
of the Topic.

*For three such, see pages 11, 12, and 13 of
plus the preceding and following pages to make sure
you understand what the heck she's talking about.

Triangular example:
A different guy among those references thinks
sentences have an optional Ground and a mandatory
Focus; the Ground has an optional Tail and an optional
Link. He likens a sentence to a file-update command.
The Focus tells what new information to put on the
file card; whether this merely adds information, or
replaces old information. If there is no Ground, you
just "add a line" to the "current card". If the
Ground contains a Tail, that tells you which "file
card" to update; if the Ground contains no Tail, the
"current card" is to be updated. If the Ground
contains a Link, that tells you which "line" of the
"card" is to be "written over" with the new
information; if the Ground contains no Link, you
simply add the new information to a "new line" on
whichever "card" the Tail, or lack of a Tail,

Different Types of Givenness and Newness
One of those references did go into quite an analysis
different types of "givenness" and "newness".

They got the idea, I think, from
Prince, E.F. (1981) `Toward a taxonomy of Given-New
information', in: P. Cole (ed.) Radical Pragmatics,
Academic Press, New York, p. 223-255.

This URL
contains a scan of a xerographic copy of Prince's

Here's my own independent distillation of what a few
of the papers had to say about that;

A speaker makes assumptions about:
1) what the addressee knows about, and
2) what the addressee is thinking about.

Now, "knowing about something" and "thinking about
something" are different mental states. (That's
almost a quote from someone; so, I'm plagiarizing --
but I have no choice, because I can't remember who I'm
plagiarizing from.)

An item may be assumed by the speaker to be
Knowledge-Given for the addressee, or to be
Knowledge-New for the addressee; that is, the speaker
may assume the addressee already knows about it, or,
the speaker may assume the addressee does not already
know about it.

An item may also be assumed by the speaker to be
Attention-Given for the addressee, or to be
Attention-New for the addressee; that is, the speaker
may asssume the addressee is currrently thinking about
it, or, the speaker may assume the addressee is not
currently thinking about it.

(Note that, while Knowledge-Given is a permanent
status, once attained, Attention-Given is a status
that can fade during discourse.)


Topics are always Knowledge-Given; but they may be
either Attention-Given or Attention-New.

Comments (or should I say Foci?) are always
Attention-New; but they may be either Knowledge-New or

If the speaker states the Topic, it is to ensure that
the addressee's attention is correctly directed prior
to the speaker stating the Comment.

If the speaker states a Comment, it is either to
inform the addressee of something the addressee does
not already know, or, to draw the addressee's
attention to something to which, as far as the speaker
knows, the addressee may not currently be thinking

So, you see, if you have a sufficiently complicated
sentence, it could have a Topic (all Knowledge-Given)
which had an Attention-Given part as well as an
Attention-New part; and also have a Comment (all
Attention-New) which had a Knowledge-Given part as
well as a Knowledge-New part.

_That_ one was _my_ idea; if it's hogwash, I have
nobody else to blame.

--- Jonathan Knibb wrote:

> Tom, > > firstly, thanks very much for your email. Secondly, > apologies for not having > picked it up until today - I rarely access this > account. > > Your reply was indeed very helpful (and not I > suspect by chance). The > difficulty I have is being able to imagine that > Given and Topic on the one > hand, and New and Comment on the other, can fail to > align in those pairs. In > your example with the three (or four) pistoleros, I > agree that I believe > that there are three men, and therefore this > information is in a sense > 'given' for me. However, when I tell you there are > three men, I do so > because I believe that you think there are four. > That is, the men's > three-ness is news to you, and therefore New > information in the context of > my communicative intent to you. > > In effect, I'm saying that Given information is > "what the speaker expects > the addressee already knows", but not, as you > suggest, "or already should > know". If I think you *should* know something, but I > believe that at present > you do not know it, then surely I will treat that as > New information when > speaking to you. > > Your other example was 'the salt sea', where I agree > that the fact that the > sea is salty is presumably Given. But is it Comment > rather than Topic? If > you are referring to sentences of the form "The salt > sea rose over the side > of the ship.", we can accept 'sea' as Topic, but its > saltiness is relevant > neither to Topic nor Comment information - in fact, > I would almost say that > it is not information at all. The motivation for the > use of the word 'salt' > here is literary convention; it serves no > communicative function. > > What about "The salt sea burned my wounds.", where > the saltiness is > relevant? This I find very difficult. Part of the > reason is perhaps that > sentences of this form are very unusual in > conversational speech - topical > referents are almost always referred to using > pronouns. "Its salt (or just > 'it') burned my wounds." would be much more likely, > for example, if the sea > is Topical. I would almost say that, in > conversational speech, a referent > expressed by a full noun phrase is ipso facto not > Topical. If it were > topical, what would the speaker's motivation be for > expressing it thus? And > in a literary context, it's quite possible I think > to have a sentence > without an expressed topic. For example, what is > "The salt sea burned my > wounds." really about? I would say that most likely > it's 'about' my feelings > or state of mind, less likely the sea or the wounds > themselves. > > Hrm - I need to think about this some more! > > Thanks again for your help, > best wishes, > Jonathan. > > >
__________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around --------------------------------- Yahoo! Shopping Find Great Deals on Holiday Gifts at Yahoo! Shopping