Soon after he stops, having run out of breath, Monica drops her pitch to the top of his
usual range, and gives a short laugh (‘ha!’) at that pitch.
Even without understanding of Italian, the sound recording of this episode makes sense
as a brief period during which two people good naturedly mock a third one, and do so
together. The prosodic details just identified help make sense of why this interpretation
is so easy. Aldo and Monica are identifiably ‘together’ because their utterances
harmonise, showing a brief allegiance in the same way as bodily orientation shows
acceptance or rejection. Their vocalisations are identifiably ‘about’ Rosa’s partly because
the pitch on which they converge is indexical of the end of her last utterance, and
because Aldo’s unusual starting pitch is also indexical of her typical range, rather than
his own. Monica’s laugh in turn indexes Aldo, again by being pitched into his normal
range. These latter two co-ordinating properties are probably less noticeable to people
who don’t know the utterers, but are evidence of the ways in which prosodic patterns
between people with histories of shared intimacy are modulated by that history, as they
can also be by shared cultural experience. In this case, crucially for our purposes, the
gentle mocking which is so clearly accomplished doesn’t involve even a single standard
‘word’.
Similar forms of indexing can be found by looking beyond pitch, and attending to the
ways in which, inter alia, accent, timing, and loudness and various kinds of visible
movement play out in utterance-activity. Although the ‘oeu’ example just discussed is
very striking, prosodic detail of the same type is all but ubiquitous in utterance-activity
at all ages, and occurs in word-based speech as well as in response cries.
Conclusion
We opened this paper with the assertion that utterance-activity should be regarded as
continuous with language, and went on to suggest that approaching our ‘how’ question
from the perspective of distributed cognition would suggest ways of re-evaluating the
argument from the poverty of the stimulus. Most of the preceding section is descriptive,
rather than argumentative, consisting of an account of how we are inclined to see a
number of examples, and in the first two cases, the cognitive and behavioural transitions
of which they might be paradigmatic. A question naturally arises, regarding how one
sympathetic to our way of describing the episodes might begin to make sense of them.
Here is a somewhat speculative suggestion. In a provocative paper on emotions Ross and
Dumouchel (MS) argue that emotions should be understood as strategic signals, having
the particular effect of encoding preference intensities (which are more difficult to infer
than preference orderings) in ways that, unlike standard commitment devices, do not
have explicitly to be constructed in advance of strategic interaction. By having
preference intensities thus (even if roughly) publicly represented, otherwise intractable
strategic problems can be negotiated, and mutually uncongenial prisoners’ dilemma type
situations, sometimes, avoided. Focussing on the first of these possibilities, the idea is
that negotiations between agents who are mutually affectively legible involve lower
computational demands for each agent’s individual strategic decision making. As they
say:
On our interpretation of the role of the emotions in bargaining, their status as
social conventions enables their expression to be used as early moves in games,
ruling out certain outcomes which might otherwise be thought by other parties to
be possible equilibria. This can be expected to influence the other party's choice
of strategy so long as the structure of the game is such that the other party has a
choice at all.
15 / 23