Re: Dictionaries of agglutinating languages

> From: Roger Mills
>
> >Adrian Morgan wrote:
> >>
> >> Robert Hailman wrote:
> >>
> >> > I haven't seen any dictionaries of any languages of a
> similar nature to
> >> > yours, but my idea is to only put in words the meaning of which can't
> be
> >> > derived easily by the root & the affixes; or perhaps you could only
> >> > include the words using uncommon affixes.
> >>
> >> Well, quite. The problem is in defining 'derived easily' and 'common'.
> >> These qualities /can't/ be defined except as one end of a very fuzzy
> >> and subjective continuum -- but to compile a dictionary I have to be
> >> definitive.
> >
> >Very true. Another idea: Provided all these affixes are regular in how
> >they alter the meaning of the root, just put the root words in the
> >dictionary, and have an appendix at the end with a table of all the
> >affixes.>
>
> This last would be my approach.  I'm not familiar with how Turkish
> dictionaries are organized, but Indonesian, while not agglutinative, has
> very consistent, transparent (and limited) derivational
> morphology.  So you
> will have the head word XXXX:  followed by its various derivs., as well as
> compounds, idioms etc.     There has to be occasional cross-referencing,
> since some few forms might derive ambiguously-- e.g. (*)
> /peNatak/ could be
> < peN+katak, peN+atak or even pe(N)+Natak. (Base-initial /N/ is
> quite rare,
> and often the result of "wrong division".)   The dictionary I use has no
> separate list of affixes (though that might be a good idea).
>
> ObConlang:  The Kash dictionary is organized along similar lines.
The approach I took with amman iar was to organize the dictionary/lexicon by
root followed by the derived forms.  See a sample of this at
http://graywizard.net/Conlinguistics/amman_iar/ai_lexicon.htm.

David