criteria of comparison

[Klaus Laalo]

On Sat, 16 Dec 2000, Yonata Levy wrote:

> Similar issues need to be addressed when one tries to count MLU in Hebrew.
> Dromi and Berman have proposed a system many years ago. In our studies we
> have adopted much of what they propose but also changed it to some extent.
> Here are some examples of what we do for Hebrew:
> In view of the problems that Dan raised wrt determining productivity and
> since morphemes are conflated and since the aim is to arrive at a measure
> which will be comparable to other languages so that one can do
> cross-linguistic studies the decision was to never give any single word more
> than - 2. Furthermore, certain forms will never receive more than 1 because
> they can be too easily picked out from the input. For example, the form that
> the child is being addressed with. So a girl using an adjective in the
> feminine will only get 1 for it. A boy will get 2. Nobody gets more than 1
> for the canonical verb form (which in Hebrew is 3rd singular) etc. I believe
> such decisions are needed, even if they seem ad hoc at times, since the aim
> is to arrive at a reliable comparative measure.
> Yonata.
>
I would like to add one point about canonical verb forms from the point of
view of Finnish language: the 3rd singular present indicative is in
Finnish the basic verb form, and it consists of the stem + vowel
lengthening, e.g. sano+o 'says/is saying', otta+a 'takes/is taking'. I
would count 2 morphemes for these forms, because the vowel lengthening is
analogically spread in child language to verbs which don't have it in
Standard Finnish, e.g. to one-syllabics (e.g. juo+o 'drinks/is drinking';
in my view, these analogies show that the child can find the morpheme of
3rd person, the vowel lengthening.
Klaus Laalo
professor of Finnish language
University of Tampere