<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=iso-8859-1">
<META NAME="Generator" CONTENT="MS Exchange Server version 5.5.2654.45">
<TITLE>RE: summary of productivity (quite long)</TITLE>
</HEAD>
<BODY>
<P><FONT SIZE=2> I've had quite a few interesting replies on this and have been asked to summarise them.</FONT>
</P>
<P><FONT SIZE=2>Firstly I should say a couple of things which I didn't say before: I'm not working on English, but on two languages that have previously not been studied wrt language acquisition - Kiswahili, and Kigirama, a related language spoken also in coastal Kenya. Also, I am currently in coastal Kenya with no access to a library (beyond about 3 books which I have with me) and the postal service is very slow and unreliable. So thanks to everyone who suggested books and/or offered to mail me stuff... forgive me if I appear rude! But if anyone wants to send me stuff electronically I'm really happy to get it.</FONT></P>
<P><FONT SIZE=2>I'm currently just collecting pilot data with a view to getting a larger grant to fully characterise acquisition in these languages. So I have 100-200 utterances each from about 5 children all in the 2-3 year age range.</FONT></P>
<P><FONT SIZE=2>Some people have used my indicators a) [walk + walked] and b)[walked + kissed] as measures of productivity. </FONT>
</P>
<P><FONT SIZE=2>However various people pointed out quite correctly that my indicator c) [drives + finished] would not be a good measure, especially considering work by Tomasello and others on frozen forms.</FONT></P>
<P><FONT SIZE=2>Various people pointed me to Roger Brown's book A First language (sadly not in the collection I have here) and pointed out that the absence of a contstruction in an obligatory context was a good indicator of non-productivity. This might be quite useful given the data I have.</FONT></P>
<P><FONT SIZE=2>Some people suggested a nonword elicitation task (wug - wugs). Nice for the future I think but not a first step with an undescribed language possibly - also difficult given the characteristics of the children - none have been to nursery school and none would be used to any kind of formal testing or teaching situation.</FONT></P>
<P><FONT SIZE=2>A couple of people also said that the only way to be sure of productivity is overregularisation. Nice if you have a language with many irregularities but I can count the irregularities in Kiswahili on the fingers of one hand for the verbs [the present tense of the verb "to be" and the imperative of the verb "to come"] and one for the nouns... [salt, sickness, tea]. However one of these children overregularised "come here" so maybe he knows how to do imperative...</FONT></P>
</BODY>
</HTML>