Larry Trask, a world expert on Basque linguistics and the history of the Basque language, passed away on March 28, 2004. Larry contributed extensively to several online communities, including Basque-L and the Indoeuropean list. This collection of his postings is dedicated in his memory. To learn about Larry, see this article.

It occurs to me that I may be bewildering the non-linguists on this list -- almost everybody -- with my asterisks. Maybe I'd better explain how we use asterisks in linguistics.

In fact, we use asterisks in three quite different functions in linguistics: two of them expressly historical, the third not. But, in all cases, the central sense is roughly 'non-existent' or 'unrecorded'.

First, we use an asterisk to mark a linguistic form which is nowhere recorded, but which has been reconstructed or proposed by somebody as having formerly existed. We do this regardless of the degree of confidence we can place in the proposal.

For example, the former existence of the unrecorded Late Latin *<caveola> 'small enclosed place' is regarded as certain: no Romanist doubts that such a word existed. Likewise, in Basque, we routinely reconstruct *<ardano> 'wine' and *<bini> 'tongue' as the ancestors of the modern forms. Again, no one in the field doubts the former existence of these forms. But all are unrecorded, and so all must get asterisks.

But we also use the asterisk to mark more uncertain, more doubtful, or more speculative proposals. For example, Knörr's proposal of an ancient Basque *<egu> 'day' is supported by no evidence at all, and it is not generally accepted, but the proposed form must still get an asterisk. So must my *<(h)iLe> 'moon', where the general reconstruction is not in doubt, but where the precise form is uncertain. So must my very tentative proposal -- still under investigation -- that Basque <lanabes> 'toolkit, tools' might derive from an unrecorded Romance form *<las naves>, or something similar.

Second, we use the asterisk to mark a form which, according to some line of thinking, might have been expected to exist but does not. For example, we note that Latin <cattum> 'cat' should, by the usual rules, have yielded Castilian *<cato>, even though it didn't: the observed form is the unexpected <gato>. In the same way, I pointed out the other day that a Romance origin for the Basque word for 'six' should have yielded a Basque *<seits> or *<seis>, and not the observed <sei>. This is evidence against the suggested Romance origin.

Third, and non-historically, we use the asterisk to mark a linguistic form as ungrammatical or as ill-formed in some way. For example, I might note that the attempted sentence *<Noa etxera> is ungrammatical in Basque, or that *<galb> is not a possible word in Basque, since it violates the rules for building Basque words.

Normally, the context is enough to make it clear in which function the asterisk is being used.

Larry Trask COGS University of Sussex Brighton BN1 9QH UK

larryt@cogs.susx.ac.uk

Tel: 01273-678693 (from UK); +44-1273-678693 (from abroad) Fax: 01273-671320 (from UK); +44-1273-671320 (from abroad)

This page is part of Buber's Basque Page and is maintained by Blas Uberuaga. Please report any problems or suggestions to Blas. Eskerrik asko!