Introduction

It was shown in Chapter 4 that the order of emergence of CAN forms in spontaneous speech was:

can't (D) → can (D) → can (Q) → can't (Q)

This sequence contravenes the general trend found in Section 6.1 for affirmative forms to precede corresponding negative forms. However, given reports of negatives such as ‘can't’ and ‘don't’ as early rote-learned items (Bloom, 1970; Ervin, 1964; Fletcher, 1979; McNeill, 1966), and considering that we are talking of emergence rather than acquisition, it is hardly surprising.

What was less predictable was that in the imitation test ‘can’ (D) emerged later than both ‘can't’ (D) and ‘can’ (Y/N) for all children (see Section 4.4.3). The explanation suggested was a tendency to attribute an exclusively interpersonal, in particular a regulatory, meaning to utterances (in this case the model sentence) whose semantics are not fully understood. No such meaning could be attributed to the ‘can’ (D) item, hence its late appearance. This interpretation is consistent with the view that early modals are ‘without exception interpersonal and action-oriented’ (Fletcher, 1979, p. 282; see also Stephany, 1986).

On the other hand, it will be shown that there is a deep-seated inconsistency in the literature on this subject. If comprehension and production are based on interpersonal function, why does Wells (1979a) find meanings of permission and ability to emerge before the performative meanings? One possibility is that this is an area of individual differences in functional orientation, similar to Nelson's (1973) Referential/Expressive distinction but at a later stage. Another explanation lies in a possible mismatch between the development of the child's interpersonal aims and the researcher's reading of the underlying modality of utterances.