Iyionaku wrote: What do you consider to be a "sentence"? Because I just read all the replies on that post and I'm a little curious why so many conlangers seem to have >1000 words but almost no sentences. Is there some "definition" when something said in that conlang is a "sentence", matching in this context?
Seems to me this might be a good question.
I don't have an exact answer, and nobody else seems to have tried to answer -- or, at least apparently no-one posted such an attempt.
My first guess or first approximation would be what's in
http://www-01.sil.org/linguistics/Gloss ... ntence.htm.
(I'm assuming readers can find out what "clause" means.)
I would probably go on to the internal links SIL puts on that page.
them wrote:What is a complex sentence?
What is a compound sentence?
What is a matrix sentence?
What is a simple sentence?
What is a construction?
And I would look at
https://en.wikipedia.org/wiki/Sentence_(linguistics), and particularly at
https://en.wikipedia.org/wiki/Sentence_ ... sification, though
https://en.wikipedia.org/wiki/Sentence_ ... associates might also be a particularly useful part of that article.
I'd also mention that, whatever "sentence" means, any translation into your conlang of any of the 218 English sentences at
https://cofl.github.io/mirror/cstc.html, is a sentence in your conlang.
So you could get to 100 sentences by translating some 100-member subset of those.
[hr][/hr]
Lambuzhao wrote: I didn't notice if this question was asked already , but in your word count metric, are you including all attested conjugated/declined forms of words (that do that sort of thing), or just the bare dictionary forms?
I'd say, count it each way, and any time your conlang reaches 500 in any version of "word-counting-metric", let us know.
The strictest would be; Only count root morphemes of inflectable lexical-content words in large open word-classes -- "substantives" and verbs, for instance. Or, noun-roots if you have nouns (who doesn't), verb-roots if you have
rootsverbs (who doesn't) (but there might not be much difference between noun-
roots and verb-
roots), adjective-roots if you have adjectives, and adverb-roots if you have adverbs. Maybe you do, and maybe you don't, want to also count adposition-roots (if you even have any adpositions, and your adpositions even have any morphology).
I think one might consider pronouns, conjunctions, articles or determiners, and interjections, (and, depending on the language, maybe also adpositions), to be function-words rather than lexical-content words. Some might not consider pronouns to be function-words rather than lexical-content words, in at least some languages. Interjections are an open word-class, and possibly a large one, but they are still not lexical-content words.
You might include all of the function-words too -- or not. And/or, you might restrict yourself to the open word-classes -- or not. For instance, in most languages with adpositons, adpositions are a closed class, and in most languages with pronouns, pronouns are a closed class. And in some languages with adverbs, adverbs are a small closed class.
Maybe you could count them separately.
First the root morphemes of inflectable lexical-content words in large open word-classes;
then relax each of those restrictions on the word-class {i.e. inflectable vs particle, lexical-content vs grammatical-function, large vs small, open vs closed}, one at a time, then two at a time, then ... etc.;
then go ahead and count major derivations of the roots as different words -- for instance, counting verbal nouns and verbal adjectives separately from the verbs from which they are derived.
I guess that's all a non-answer.
You might look up the definitions of "word" in the SIL Glossary and/or Wikipedia and/or Trask, and maybe those would help.
[hr][/hr]
Actually, I think I personally would be more interested in how many and which morphemes have been completely defined, and whether a complete set of rules for combining morphemes into words has been thought out and written out. If the language has non-concatenative morphology, or has suprasegmental morphemes such as tonemes, chronemes, or stressemes, I would want to know about them, too.