fill in the missing word

[Brian MacWhinney]

Jean,
  I agree that allowing children to fill in a word at the end of the
sentence rather than in the middle helps a bit, but I wonder if that is
enough by itself.
  By using the "what are these?" and "what did he do?" format I was able to
get Hungarian children to respond with inflected plurals and verbs for nonce
words down the age of 1;8 (pigák as the plural of the nonce form piga),
whereas I believe that your original study mostly showed competence in
English speaking children at age 4 and perhaps a few just before age 4.  Of
course, comparing Hungarian with all its morphology to isolating English is
not quite fair, and I did notice that German children took a few more months
to tune into this, so that I first saw plural productivity about 2;3.  But
my point is that, by using the more direct question-answering format, you
can tap into knowledge perhaps earlier and perhaps more directly.  If we are
talking about children with SLI or WS, this is important, since we want to
be sure that we are not just giving them a task which allows them to "get
away with zero-marking."  (see Leonard and others on the issue of
optionality and omissibility)
  Another important difference to note is that I used toys to display the
action right in front of the child.  You may be right that the child could
still say "miff" as a past in my task, but somehow the immediacy of the
action makes this a bit less likely.
  If we compare the answers on the plural test to the answers on the past, I
think I see your point.  For the plural, if I say "What are these?" it is
fairly difficult for the child to just say "wug" as if two objects were one.
However, for the verb, if I say "what did he do?" the answer "miff" instead
of "miffed" seems a bit more acceptable.  I agree that, by age 4 children
begin to understand that they have to adapt their verb to the "Yesterday he
...? frame, but again that is age 4, not age 2.
  I guess what we need here is a study that actually runs the two methods in
a head on comparison.  Of course each method has a combination of
potentially important features.  In particular, the question-answering
method makes use of toys that the child can hold, whereas the fill in the
missing word method seems to rely on totally verbal presentation, without
even pictures as a support.
  What surprises me is that this crucial methodological issue has never been
discussed in the literature.  Or have I missed something?

--Brian MacWhinney