Christophe Grandsire-Koevoets, On 13/07/2015 15:01:
>> g. In no language can an affirmative be turned into a negative by changing
>> the intonation contour (Horn 1989).
> Ja ja... In Dutch, the sequence "ja ja": "yes yes" can have an affirmative
> or negative meaning purely depending on the intonation contour used. Plenty
> of other languages feature similar phrases. And if you want a more
> "traditional" type of example, I believe that some language of Africa
> (Igbo?) can mark negation purely through the change of the tone of the verb
> (it's a tonal language in any case).

I think what is meant is analogous to yes-no question-formation in many lgs. The Dutch case is one idiom, not a regular process. The Igbo case as you describe it sounds like inflection with tone; tone is not intonation.

>> l. In every language in which there is a person and number inflection,
>> there is also a tense, aspect, and mood inflection (Bybee 1985; 267).
> Here, I can answer more authoritatively: the Shizunai dialect of Ainu
> definitely breaks that one: it has person and number inflection, and can
> (weakly) be argued to have aspect and mood inflection (although most, even
> arguably all, of it is through auxiliary constructions), but it definitely
> lacks any form of tense as a grammatical category.

I think "a tense, aspect and mood inflection" means "an inflection for tense, aspect or mood (TAM)".

> There's a very good reason why many of us are wary of any UG claim.

Those were claims about universals, not necessarily claims of UG. UG is what is a genetic characteristic of the species. To me, Statements A--F look to arise necessarily from the fundamental architecture of language, while Statements G--Y look to be statistical tendencies, where any absence of exceptions is an accident of insufficiency of number of languages (absolutely or in samples). I haven't read Newmeyer's book, so don't know what he says about them.