Azkue's dictionary

Larry Trask larryt at
Wed Dec 22 10:49:37 UTC 1999

[ moderator re-formatted ]

Jon Patrick writes:

> a response to Lloyd Anderson about the use of Azkue's dictionary. I comment
> that his objections were centred firmly around its ease of use for
> etymological studies not on the merit of the content itself and by default
> Larry's answer justifies Lloyd's assertion, namely:

>>  On the whole, it seems from Larry Trask's review of it that Jon Patrick is
>>  on very solid ground in using Azkue's dictionary as a basis for
>>  analytical studies.

> Larry reponded:

>> But only, I think, if a detailed knowledge of Basque linguistics is first
>> superimposed on the raw data.

> Larry's response to Lloyd's comment struck me as an oblique but transparent
> attack on the merit of my contribution to the analysis basque, being the only
> person on the list without a "detailed knowledge of Basque linguistics", who
> is debating the topics.

Well, it appears I have offended Jon.  If so, I apologize, but I wasn't in fact
trying to attack his work in any way.

If your only interest is in analyzing the content of Azkue's dictionary, then
Azkue's dictionary is the best source I can think of.  But what I was trying to
point out is that Azkue's dictionary cannot be used at face value for
historical work, for a number of reasons.

First, Azkue's policy is to lump together unrelated homophones under a single
entry, while at the same time providing separate entries (usually without
cross-references) to variant forms of a single word -- sometimes as many as 18
or 25 separate entries for variants of one word.

Azkue enters everyday words along with rare, obscure and doubtful words, all on
equal footing.  He enters a large number of loan words.  He enters quite a few
hapaxes.  He enters words for which the only source is somebody else's earlier
-- and sometimes dubious -- dictionary.  He enters words for which the sole
authority is the unsupported report of an outsider, such as the Dutchman van
Eys or the Spaniard Hervas y Panduro.  He enters neologisms coined by Basque
writers.  He enters compounds and derivatives without reference to their

And he makes mistakes.  Some of his entries are errors copied from other
written sources, and others are errors of hearing or of interpretation
committed by him.  And some of them are non-existent words, fantasies invented
by Azkue in pursuit of his etymological speculations.

All this means that the entries in Azkue's dictionary cannot be taken at face
value, or treated on equal footing, for historical work.

To be fair, Azkue is usually punctilious about providing provenances and
sources for his entries, and this information is potentially valuable.  But it
is useless as long as it merely sits there on the page, or on a database.  This
information has to be interpreted by a human user, which means that decisions
have to be made about what to do with it.

Now, I have no quarrel at all with Jon's enterprise, and I wish him well with
it.  But I wouldn't want to see anybody drawing any historical conclusions on
the basis of Azkue's list of headwords.

Larry Trask
University of Sussex
Brighton BN1 9QH

larryt at

More information about the Indo-european mailing list