Sentences

[Dan Everett]

Wally says: "To show that I'm wrong about this, one would have to
examine spoken language more carefully than most people have as yet
done."

There are two things to say about this. First, there are a lot of us,
Wally, who work exclusively with non-written languages and find the
sentence quite a coherent unit and that the mismatch between
intonational sentences (what Pike called 'breath groups' 'etically' or
'contours' 'emically) and morphosyntactic sentences is not, in fact, all
that great. Your remark seems to ignore the existence of fieldwork.
Moreover, many of us don't have the slightest problem talking about
sentences in unwritten languages and have shown in many writings that
sentences are causally implicated in the understanding of the grammar as
a whole, certainly not merely artifacts of writing systems (Piraha, to
just take a random example, seems to offer intonational evidence in
favor of both paragraphs and sentences. On the other hand, every single
thing we study is in a sense merely an artifact of how we look at the
world, languages, grammars, people, etc.)

But second, I agree strongly with Wally that grammar-intonation
'matching' in the study of sentences (or other constructions) itself has
not been well-studied, by and large. One reason for this is the failure
of nearly every regional or intellectual tradition of linguistics to
accord intonation research and documentation equal status in grammatical
description with, say, the study of words, affixes, phrases, etc.

I have come to the conclusion over the years (unfortunately *after*
writing two grammars) that a grammar without a careful and detailed
study of intonation (i.e. involving study of the phonetics and phonology
of intonation, its relation to information structure, etc) is seriously
incomplete, likely seriously flawed. But to incorporate intonation into
grammar-writing and general analytical & theoretical linguistics
requires a model of intonation that is more useful for the replication
of experiments and checking of analyses than most attempts in the past
have been. Interestingly, here, as in many areas of language study,
computational-linguists seem to have done some of the most useful work,
though many non-hyphenated linguists have pioneered (Pike, Halliday,
Bolinger, Liberman, to name a few very random examples of people whose
influence on this has been around for decades). I strongly recommend
that people take a look at the work growing from the research of Janet
Pierrehumbert, Robert Ladd, Carlos Gussenhoven, and others of this line
(e.g. Mary Beckman, Julia Hirschberg, Candy Sidner, etc.).

I am hoping to begin this summer a multi-year project for the
documentation of intonation in Amazonian languages (something I should
have done years ago). I hope that similar studies can be launched in
other parts of the world.

There is a need in fact to add courses on intonation to the general
linguistics graduate and undergraduate curricula (along with more/some
training in instrumental phonetics). Regardless of one's
perspective/theoretical intonation, the failure of the average field
linguist to carefully study intonation is one of the most serious
omissions of our discipline's history. But it is hard work, and for many
of us, certainly for me, it requires serious intellectual re-tooling.

-- Dan