How to do things without words: Infants, utterance-activity and distributed cognition.

Even so, as a linguist Halliday may have brought additional (and charitable) interpretive
resources to bear on the question whether Nigel, on any two separate occasions, was
making the ‘same’ sound again. By doing so he, perhaps somewhat more than parents
without linguistic training, was lowering the demands on Nigel’ s behaviour insofar it
could be taken as producing labels which Halliday himself could then go on to take as
significant. Although the much younger child taken as ‘asking to be picked up’ in the
episode described above undoubtedly had less vocal control than Nigel, Halliday’ s
criteria for sameness of utterance is similar to that parent’s regarding the successive
vocalisations of her child as attempts to say ‘up’. Both cases have in common a
movement in the direction of less multi-modal behaviour (one largely gestural, the other
largely vocal), and towards producing more effective labels.

In the ‘thula’ case the behaviours we described are likely to be seen as too far from
language to count as relevantly related to it. In the present case we need to guard against
the opposite tendency, that is to regard Nigel’s various [na]s and [b0]s as too much like
mature language. Halliday himself regards the vocalisations as uses of ‘protowords’,¹⁷
and treats them as expressions of relatively well-formed intentions, perhaps even
propositional attitudes, to the effect that Nigel wants the bird, or wants some other
present object. Thibault (2000) for his part, regards the data as evidence that Nigel has
crossed the threshold to indexical reference. We have just seen, though, how infant
responses to attempts to quieten them down can be taken by care-givers as indicators of
how ‘good’ the child is, and how such ascriptions need not find counterparts in the
cognitive world of the infant. Is a similarly deflationary approach possible here?

Clearly it is. Nigel need not initially ‘want’ the bird, any more than the child just
described need ‘want’ to be lifted. What is required is that the child be capable of
learning the correlation between some aspect of its own behaviour and the regularities
produced by attentive adult responses. Nigel could have just gone [b0] at some time
when he was shortly after pleased to be presented with the bird toy, and thereafter gone
on to learn that [b0]s were reliably followed by bird-givings, and adult utterances of
‘bird’ which partly echoed his own vocalisations. (At the same time Nigel was, of course,
acquiring a kind of expertise appropriate to his being in a situation in which 10 month
old children get to order parents about at all!) Indexical reference on Nigel’ s part can be
one product of ongoing interaction, scaffolded by Halliday’ s production of regularities in
the environment, but it need not be the case that Nigel’ s initial behaviour be so
motivated.

There are, though, important differences between the ‘thula’ case, and that of [na]∕[b0].
Nigel, unlike three month old infants, is capable of behaving in ways which produce
highly salient label candidates, not naturally related to affective states in the ways that
smiling or crying are, and hence amenable to being conventionally associated with goals,
desires and so forth. At his age Nigel also initiates interactions, and, encouraged by
caregivers, to engage in active exploration of the world. The regularities in his vocal
behaviour, coupled with his greater tendencies to manifest agency, mean that Halliday’ s
(likely) overinterpretations will produce specific opportunities for Nigel, relevant to his
level of maturation, and through his exploitation of these opportunities, genuine
indexical relationships can come to be established.

Oeu!

The discussions of the preceding two examples leave open an interpretation of what we
are saying which we wish to dispel. That interpretation would have it that what we are
describing is a developmental phase, or perhaps series of phases, during which motor-
centric aspects of utterance-activity play an important role because abstraction-amenable

13 / 23

<<   10   11   12   13   14   15   16   >>

More intriguing information