CNL Lunch

Andrea Gualmini

Children don't lack any knowledge

Thursday September 27th 2001, 12:30pm, 3416 Marie Mount Hall

ChildrenÕs understanding of indefinites and negation is the subject of numerous experimental studies. One line of research has focused on childrenÕs interpretation of sentences like (1).

(1) Pluto didnÕt eat some apples.

Adult speakers of English interpret (1) as paraphrased in (2a) where some has scope over negation, and therefore accept (1) in a context in which Pluto ate two of the four apples that were available. By contrast, children as old as five are claimed to interpret (1) as (2b) and therefore reject (1) in the same context, on the grounds that Pluto did indeed eat two apples.

(2) a. There are some apples that Pluto didnÕt eat.
b. It is not the case that Pluto ate some apples.

To explain childrenÕs behavior, previous studies have concluded that children have incomplete knowledge of constraints on the interpretation of the indefinite some. In particular, children do not distinguish between the positive polarity item some and its negative counterpart any (see Musolino, 1998; Musolino et al., 2000; Kraemer, 2000).

An independent line of research casts doubts on this conclusion, however. De Villiers et al. (1979) argued that both children and adults only use negative statements to point out discrepancies between the facts and the listenersÕ expectations (e.g., the number of apples that Pluto is expected to eat). Drawing upon this observation, childrenÕs non-adult interpretation of sentences like (1) could result from the use of a linguistic construction that is inconsistent with childrenÕs expectations. If so, childrenÕs non-adult interpretation of sentences like (1) should disappear if the context established that Pluto had to eat all the apples.

To evaluate this prediction, a Truth Value Judgment task was conducted with 22 English speaking children (age: 4;0 to 5;6 Ð mean: 4;10). Children were presented with stories in which one character had a precise task to carry out. On a typical trial, children were told a story about Robocop, who had to deliver four pizzas to GroverÕs house. Grover told Robocop that he would get a big tip if Robocop delivered all the pizzas. Unfortunately, on the way to GroverÕs house two pizzas fell off the delivery truck, so only two pizzas were delivered. At the end of the story half of the children (Group I) were presented with sentence (3), and half (Group II) heard (4).

(3) Robocop didnÕt deliver some pizzas.
(4) Robocop didnÕt drop some pizzas.

Notice that (3) expresses a mismatch between the final outcome of the story and the expectation that Robocop would deliver all the pizzas. By contrast, (4) is infelicitous in the same context, since the common ground does not include the expectation that Robocop would drop the pizzas. The main finding was that children in Group I consistently accepted the target sentences (89%), while children in Group II accepted the target sentences only 52% of the time. This shows that childrenÕs interpretation of indefinites in the scope of negation is fully adult-like when the felicity conditions associated with negative statements are satisfied.