Conference: Syntax & Semantics in Dubrovnik

[Greg Thomson]

>Some other examples of verbs (or words, more generally) that have converse
>meanings:
>
>   -  English "teach" and "learn" (in some varieties, "learn" is used for
>      "teach"; compare the proverbial answer of a parent to the teacher
>      who complained that the child was not kept clean enough: "You are
>      there to learn 'em, not to smell 'em!")
>   -  English "borrow" and "lend" (some people use "borrow" for "lend")
>   -  Latin "altus" 'deep', 'high'
>   -  Hungarian "ural" 'to dominate' (etymological and original meaning
>      now dying out I believe: 'to consider somebody one's lord')

According to common anecdotal reports and my own observations, a fairly
common error among second language users involves the substitution of
antonyms. Why would an L2 effort to retrieve "happy" end up activating
"sad"?

In relation to Tuggy's question and Edith's response, I think that an
answer could lie in the tendency of the language processor to use partial
information in interpreting utterances. For example, "teach" and "learn"
involve conceptual structures with all the same components. The grammatical
relations signalled, and the argument structure of the verbs, can work
together to constrain the interpretation. However, often these cues are not
necessary, as the discourse context and world knowledge may adequately
constrain the interpretation. If a listener frequently interprets the verb
"learn" without relying on grammatical cues, s/he might also come to
produce utterances in which the verb is used with indifference with to such
cues. That wouldn't interfere with further listeners' comprehension for the
same reason. If it happened enough it could affect the way the verb is
treated when grammatical cues _are_ processed, so that "learn" becomes
ambiguous with respect to it assignment of roles to its arguments.

Greg Thomson