Anglin (1977) examined two- to six- year-old children’s comprehension and production
of labels at subordinate, basic and superordinate levels. He reported that although his
subjects readily produced basic level terms, fewer than half were able to produce
superordinate or subordinate level terms. Other researchers, focusing primarily on
comprehension tasks have given further substance to this claim. Macnamara (1982)
introduced preschool children to individual objects (e.g. a dog) and asked whether these
objects could be labelled at both the basic level (e.g.“Is this a dog?”) and at the
superordinate level (e.g.“Is this an animal?”). Most children accepted the basic level
terms, but most denied that the objects could also be described in superordinate level
terms.
The data reviewed thus far, indicate that young children tend to accept one label (typically
the basic level label) for objects and to reject other, non basic level labels. Yet, this single
level assumption is certainly not inviolable (Gathercole, 1987). There are clearly some
cases, even in the studies outlined above, in which children acknowledged that objects can
be labelled at multiple levels. For example, Blewitt (1994) reported that children at a very
early age do accept both basic and superordinate labels for objects in comprehension tasks.
Au and Glusman (1990) present similar data based on two-year-old children’s
comprehension of novel labels. Moreover, some investigators have suggested that
children may be more willing to supplement their basic level labels with subordinate level
labels than with superordinate level labels (Clark, Gelman & Lane, 1985). Also, Waxman
and Hatch (1992) found that three- and four-year-old children were able to produce
multiple, hierarchically related labels for a given object.
Mandler (1993) has questioned whether there is any rule which can reliably identify basic
categories, and she argues instead, that basic categories are a result of social convention
rather than conceptual analysis. Thus, an item is at a basic level if adults and children use
that level to talk about referents.
2.3.2.2 Criticism of the Constraint theory
The constraints proposals discussed in the previous sections provide some possibilities
about the acquisition of words. However, disagreement remains in relation to the
48