> At 7:15 pm +0100 20/4/02, And Rosta wrote:
> >Ray:
> [snip]
> >> Yes, but the two goals of (i) brevity (using the Roman script) and (ii)
> >> IAL-hood were those of Reginald Dutton when he designed and developed
> >> Speedwords.  It was as a result of dissatisfaction with his solution that
> >> work on BrSc began.  There clearly has to be compromise and trade off
> >> between the two goals - that's the challenge.
> >
> >Key things to maximize brevity:
> >
> >(1) Maximize the combinatorial symbols.
> >    -- but here for IAL reasons you're choosing c. 26 letters
> >(2) Minimize combinatorial restrictions on basic symbols.
> >(3) Maximize homonymy.
> >(4) Make relative length reflect relative frequency.
> (4) is something that Dutton consciously attempted to do.  He used
> frequency list drawn up by Professor Ernest Horn of Iowa University.

Thanks to modern technology, better data is nowadays available.
Michael Rouse once compiled a pan-web frequency list (based
purely on orthography, so further work would be needed to
work out frequency of different senses associated with the
same orthographic form).

Did Dutton apply frequency data only to roots/primitives, or
to other aspects of Speedwords?

Another frequency-related issue that I think is worth bearing
in mind is that a sequence of separate words may have higher
frequency than some single words. Accordingly, analogues of
netspeak's "iirc", "otoh", "iow" etc. would enhance brevity;
that is, make allowance for words that abbreviate frequent

> >The crucial issue is whether the element that distinguishes the polysemes/
> >homonyms from one another must be marked. The Lin A/N/V distinction is
> >not marked; it is recovered from grammatical context.
> It's partly marked by cements also. If two words are bound by an internal
> cement, the first must be a qualifier (adjective/adverb) and the second
> must be a noun or verb.  If they are bound by an external cement we have
> either N+V or V+N according to position.

Okay. I'm not so much trying to make a point about Lin as to
make the point that a very good way to achieve brevity is
make use of grammatical context as well as phonological form
in determining lexical contrast.

> But the difference between a word with same spelling but different meaning
> according to whether it's used as a noun or verb seems to me analogous to
> the English {sow}
> a) noun /s{w/  = 'female pig'
> b) verb /s@w/  = 'to scatter seed', 'to plant by scattering seeds'
> We'd normally consider these as two separate words, surely?

Yes, these are two separate words. The sort of thing I'm thinking
of is the to/too/two contrast. If English were syntactically
unambiguous then these would not give rise to ambiguity and could
happily share the same phonological form while remaining distinct