524 ROBERT D. VAN VALIN, JR. & DAVID P. WILKINS
"throat" is ahentye, but it is important to realise that the throat is commonly
considered to be the seat of various cognitive acts in a number of areas of
Australia, and among MpA speakers it is perceived as being the seat of
desire and wanting. Indeed, the verbs meaning "want, desire" are ahentye-
ne- (throat-sit) and ahenty-irre- (throat-INCH-). While we would argue for
the universality of the notion of mind, there is no contingent argument con
cerning where the seat or seats of the conscious mind are. What is of
interest here is that, diachronically the verb itelare- is based on a perception
verb {are- "see"), it is etymologically related to the verb meaning "think",
and it suggests the view of the throat as the seat or instrument of a cognitive
act.^6
Comparing the psych-action meaning of English remember with the
definition of itelare-, we find that the meaning of the MpA verb is contained
within this sense of remember. The features it lacks are BECOME, again,
and from.before. All of these features can, however, be inferred in certain
MpA constructions in which itelare- occurs. In other words, these features
are added by pragmatic rules to give it the psych-action sense of remember.
For instance, one commonly hears it in the emphatic imperative itelar-aye!,
which bilingual informants gloss as "remember it!" (as opposed to *"know
it!"). This is to be expected, since by using the imperative one would be tel
ling someone to start thinking about something they know, with the
assumption that they were not thinking about it before that time. This
would add the BECOME component. Similarly, it adds the again and
from.before components, because if one believes that the addressee can
actually start thinking about something particular that they know, then it
must be something from previous experience, i.e. from before; hence they
must become aware of it again. It is to be stressed here that we are talking
about semantic components being added through pragmatic processes of
inferencing to derive a final interpretation in context, and that we believe
that cross-linguistic comparison of roughly equivalent forms must take great
care to distinguish between that which is semantically coded in the forms
and that which can be pragmatically derived through contextualisation of
the forms. This argument harks back to an early discussion of the related
verb remind; in discussing Postal's (1970) paper "On the surface verb
remind" Bolinger (1971:522) cautioned that "[a] line must be drawn
between meanings which are the reference of a word and other meanings
which are inferred from its use in particular contexts, even though the infe
rential meanings may in time become referential."