Henrik Theiling wrote:
> Hi!
> 
> Herman Miller writes:
>> ... a lot of interesting stuff ...
> 
> So your main criterion would be predictability of semantics?  If
> predictable => no new word, if not predictable => new word.  This
> seems, well, very reasonable for composing a lexicon.  Of course there
> will be difficult cases, but let's ignore them for now.

At least it's something to fall back on in the absence of specific 
information about a language. Stress patterns, inflections, and so on 
may be useful in some languages. But I don't know if I'd consider it as 
the main criterion.

> This means that for counting a conlang's words, we probably should:
> 
>   - also count phrases ('bubble sort algorithm') and idioms
> 
>   - not count lexicon entries that are due to irregular forms
>     ('saw' cf. 'see')
> 
>   - count polysynthetically constructed words several times,
>     excluding structures that are semantically clear operations,
>     but counting all irregularly derived concepts
> 
> This seems quite reasonable.  Do you also think it's a good way of
> counting?  It also looks undoable since the lexicons are generally not
> structured like this.

Right. Minza for instance has various predictable forms listed 
separately since their English equivalents aren't always formed in the 
same way: chengiła "shame" is derived from an adjective (chengi) meaning 
"shameful", for instance.

I don't know if that's the best way of counting. Take the Minza phrase 
"tlahzi kiřvi" -- literally it means "yellow tea" (referring to the 
color of the liquid) but it's translated as "green tea" (referring to 
the color of the leaves) in English. It has a place in the lexicon, but 
should it count as a word? I think the native Minza speakers (if any, 
since Minza doesn't have a conculture) might see this as a transparent 
phrase consisting of two words.