tendency for languages to avoid synonymy. Bréal (1897), for example, proposes a
law of differentiation in which several forms that are associated with a single mean-
ing come to distinguish themselves so that their meanings are no longer synonymous.
The law predicts, in other words, that however two words come to mean the same
thing, the language will evolve so that eventually the two words will no longer be
synonyms—thus preserving bi-uniqueness.
Once we have noted that languages seem to change to avoid synonymy, the next
step is to attempt to determine if there is a psychological or cognitive explanation for
this tendency. In the case of synonymy, language researchers have indeed discovered
evidence for a psychological tendency that may account for the rarity of synonymy in
language. The “disambiguation effect” (Merriman and Bowman 1989) occurs when
children are asked to find the referent of an unfamiliar object in a set of objects that
contains one familiar and one unfamiliar object. In such a situation, children nor-
mally choose an object whose name they do not know. To account for this effect (at
least in part), researchers have proposed several principles—including the Mutual
Exclusivity Principle (Markman 1992), Novel Name for Nameless Categories
(Golinkoff et al. 1992), the principle of contrast (Clark 1983), and Grice’s (1975) Co-
operation Principle (Gathercole 1989). Although there is disagreement about the ex-
act nature of the cognitive principle that is responsible for the disambiguation effect,
what is important for the present discussion is that children assume by default that a
many-to-one mapping between form and meaning is less likely than a one-to-one
mapping. Given that children have such an assumption, to possibly learn a synonym
they must first overcome their bias against many-to-one mappings.
Where a single form is mapped to many meanings, the case is not so clear. For
one thing, homophony is not uncommon in languages—especially if one considers
grammatical morphemes as well as words. Along these lines, Clark (1983, 70) points
out that the English
–s
morpheme is used for possessives, marking plural forms on
nouns and marking the third person singular form on present-tense verbs. Moreover,
polysemy, which involves different but related senses, seems to be the norm for both
lexical items and grammatical morphemes (Lakoff 1987; Taylor 1995). Nevertheless,
there is some research that suggests that at least homonyms, though not impossible,
may be dispreferred in lexical acquisition. Research on near-mergers, for example,
indicates that words that seem to be converging phonetically over time often will
change (or be changed) to make their differences more acoustically salient, or one of
the pair will be lost entirely (Labov 1994, esp. Part C). Research into processes of
phonological change has yielded similar conclusions. For example, Gurevich (2004)
found that among nearly 300 processes of lenition in 170 different languages, only
20
Do'stlaringiz bilan baham: