Re: Cowan on morphology

[Mr Andrew Rosta <ucleaar@UCL.AC.UK>]

John Cowan writes:
> Mr Andrew Rosta writes:
> I believe that the best word to characterize the relationship between
> morpheme meaning and compound-word meaning is "motivated", in more or
> less Lakoff's sense: the meanings of morphemes do not determine the
> meanings of compounds, but do give excellent general guidelines to that
> meaning.  Not everything in the world can be fitted into a dichotomy
> of "arbitrary" vs. "predictable".

(a) The grammar at present does not require compounds to be motivated.
(b) Compounds can be motivated without self-segregating morphology.

> > Since the meaning of Lojban compounds is not predictable, the
> > unambiguous morphology is not really very useful, and it adds
> > extraordinary complexity to the language and to the task of
> > acquiring vocabulary.
>
> I don't see this.  If you treat the compound words as beyond useful
> analysis, they are no harder to acquire than any other system of
> construction, including the most arbitrary.  We acquire English words
> easily, but have to learn morphemic analysis painfully -- and in many
> cases, even the learned may be in doubt.

Lojban compounds have lots of 'allologues' (I don't know the technical
term, if there is one): the same word may be spelt & pronounced in
numerous different ways. This makes them especially tough to learn.

----
And.