majority of lexical items acquired by children are not of this nature. Even 2- year olds
possess words that refer to specific individuals (Fred), substances (water), parts (nose),
properties (red), actions (give), and so on. This motivates further constraints that
determine how words can relate to one another within the lexicon; these can lead children
to override the whole object and taxonomic constraints.
2.3.2.3.1 Mutual exclusivity constraint
According to the mutual exclusivity constraint each object can have only one label
(Markman & Wachtel, 1988). This assumption does not hold for adults, as pairs of words
such as dog and pet or dog and Fido are not mutually exclusive. But it could be argued that
children are biased to assume that words have mutually exclusive reference and only give
up this assumption when there is clear evidence to the contrary. The fact that children
appear to have some difficulty with class inclusion relations (where categories exist at
different levels of abstraction, such as dog and animal) has been taken as evidence that
children are reluctant to abandon this assumption (Markman, 1987).
Further evidence from Markman and Wachtel (1988) illustrates the role that this
assumption can play in language development. When children are given a novel word
describing a novel object, they will interpret the word as referring to that object (following
the taxonomic and whole object constraints), but when given a novel word describing an
object that they already have a name for, they will move to other, less favoured
hypotheses, such as construing the novel word as a name for a part of the object or a name
for the substance that the object is composed of. In a study by Gollinkoff, Hirsh-Pasek,
Lavalle and Baduini (1985), children were shown two objects, one familiar and the other
unfamiliar (e.g. a cup and a pair of tongs). When told: “ Point to the fendle” they would
tend to point to the unfamiliar object, suggesting that they assume that fendle could not
mean “cup”, a result predicted by mutual exclusivity.
Nelson (1988), in citing the work of others (Merriman, 1986; Mervis, 1984) has suggested
that mutual exclusivity may be developmental in nature and may not be evidenced until
the age of 3 years. Merriman’s subjects were less than three years old. When given novel
names for familiar objects with known labels along with unfamiliar objects they did not
systematically apply the novel words to the novel objects. Mervis (1984) demonstrated
44