Language Log

« previous post | next post »

At a Semantics Workshop in the Rutgers Philosophy Department last weekend, my job was to comment on an excellent paper by Alex Lascarides & Julian Schlöder, "Understanding Focus: Tune, Placement, and Coherence". Here's the opening section of my presentation:

We modulate our linguistic performances in many ways, expressing our state of mind and our attitudes towards the interaction and towards the content of the message.

These modulations include officially phonologized aspects of prosody, like phonological phrasing, pitch accents, and junctural tones.

But they also include local and global modulations of other aspects of pitch, of speaking rate, of voice quality, of gesture, posture, gaze, etc. …and in text, there's layout and typography.

Scholars have recognized for at least a century that English intonation includes elements that are tropic to stress (“pitch accents”), and elements that are tropic to boundaries (“”boundary tones”). Current theory quantizes and phonologizes these elements in a particular way, turning them into tonal symbols that are integrated into phonological representations. There are also obviously many other communicatively-relevant aspects of prosody, which are treated as paralinguistic modulation, or ignored.

But posture, gesture, gaze, etc. include elements that are tropic to stress and elements that are tropic to boundaries. These are not generally quantized or phonologized in current linguistic theory.

Are these the right choices?

Here's a nice illustration of stress-aligned and boundary-aligned gestures, from Donald Trump's inaugural address:

And from Trump's 100 Days rally last weekend in Harrisburg, here's a nice illustration of some communicatively-relevant paralinguistic (?) aspects of prosody:

as you may know
there's another big gathering taking place tonight
in washington D C did you hear about it?
a large group
of hollywood actors
and washington media
are consoling each other in a hotel ballroom
in our nation's capitol right now
they are gathered together
for the white house correspondents' dinner
without the president

The most striking example in this passage is the low level pitch of the last phrase:

(Note also the expressive use of gesture, facial expressions, and posture…)

Trump's prosodic decoration of that last phrase reminded me of the poster for an upcoming movie, "Everything, everything", where the last word of the title is decorated in an expressive way that's obviously not just encoding some combination of phrasing, pitch accents, and boundary tones:

To sum up this aspect of my presentation, I quoted from from Dwight Bolinger, “Intonation and Analysis”, Word 1949:

Is English intonation susceptible of phonemic analysis? If not now, will it be so later? What kinship is there between tonal and phonemic events, and is there such a thing as a tonal segment?

It seems obvious that (at least some aspects of) expressive prosody, like expressive posture and gesture and typography, are a sort of decoration of linguistic performances, whose interpretation must be integrated with the interpretation of words and phrase structures, without being rigidly codified as part of "phonological representations".

I believe that it's time to return to Bolinger's question, reconsidering this null hypothesis for all aspects of intonation in languages like English. In the end,  the answer to his question is probably either "Yes" or "The question presupposes a false dichotomy".  But we need to recognize that the answer implicit in current prosodic orthodoxy is a failure, on both practical and theoretical grounds, and has become an impediment to progress.

That's enough for this morning. There were some other fun things in my presentation, to be posted another day.

 

May 2, 2017 @ 4:51 am · Filed by Mark Liberman under Philosophy of Language

Permalink