|
|
|
|
CNL Lunch
Talks
Takuya
Goro, Utako Minai, Tomo Fujii
The anti-reconstruction effect and the
acquisition of scope ambiguities
Thursday September 15th 2005, 12:30 PM, 3416 Marie Mount Hall
|
|
A sentence with two scope-bearing elements (e.g. ""every boy" and "not", etc) is occasionally associated with
two different readings (i.e. scope ambiguity). The phenomenon called the anti-reconstruction effect is one of
the cases where scope ambiguity we might expect to hold does not obtain. The following English sentence, which
involves an infinitival complement, is ambiguous between two readings: |
|
(1) Mary forgot [to close all the windows]
|
|
It means either (i) all the windows are such that Mary forget to close them (all>not; wide scope reading of
"all") or that (ii) what Mary forgot to do is to close all the windows (not>all; narrow scope reading of
"all"). However, some languages including German and Japanese have a very similar infinitival/non-finite
complement construction where "forget" cannot scope over the universal quantifier. In adult Japanese, the
sentence in (2) below lacks the narrow reading of "all the
windows", which contrasts with (1) from English.
|
|
(2) Mary-ga [mado-o subete sime]-wasure-ta
"Mary-NOM [window-ACC all close]-forget-PAST"
|
|
(2) always means that for all the windows, Mary forgot to close them (all>not). Put another way, the universal
quantifier that is thematically associated with the complement predicate must take higher scope than the matrix
predicate "forget". In this talk, we will discuss three things. First, we will examine the nature of the
effect. Some syntactic and semantic observations suggest that the difference between (1) and (2) is structural.
Second, we will discuss what kind of problem the lack of ambiguity in question poses for language acquisition.
Finally, we will report the results of an experiment we conducted with Japanese speaking children, which
investigated how Japanese children behave with respect to the restriction on scope interpretation under
consideration.
|
|