Excluding data

[Jon Patrick]

Lloyd Anderson's message of 19th Aug expresses, much better that I could, the
method I intend to apply in the study of early BAsque. For me the key factor
is to present a record of ALL words available for analysis and record HOW I
can classify them. In the case of sound imitative words it is important to
retain them in the database and show how their phonological profile as a class
is similar or not to non-imitative words. In keeping with my last message I
find the idea of excluding data because they don't conform to someone's
particular expectation about the data inappropriate in trying to produce a
generalised stochastic profile of the language.

On a specific item mentioned by Lloyd, namely

    [Expressive and sound-symbolic words are also very much under-recorded
    for many languages and language families.  Many of them are not known
    to learned scholars who are not native users of the languages, because
    they are used in language registers which are never the domain of activity
    of those scholars.  So this criterion is PARTLY defective or circular.
    There is a partly definitional relation between sound-symbolic and
    narrowness of attestation in recordings.]

In basque we have the word for the sound of the heartbeat as <taup> and the
word for heartbeat as <taupa> as reported to me by native speakers. My
suggestion here is that potential sound symbolic words are so close to the
actually words for natural elements that their usage in an analysis of early
basque is potentially vitally important to understanding the phonological
structure of the language.   Has such a potential been found and realised in
any other languages?
cheers
Jon Patrick