Language Universals

[FTF]

I read somewhere that [a] was a sound in every language. But now I have nagging doubts...

I think that's /a/, not necessarily [a], and all /a/ really means in this case is "a phonemic low vowel of some sort".

[Yiuel Raumbesrairc]

I think the problem between /e u A/ and /a i u/ is one of distinction.

a vs i + u is one of pure height : a is low, i and u are high (this explain that a language might have only /a/ and /i/). Then, you have between i and u the same division, only here is it frontness, i is front, u is back. Basically, this means that you make dichotomies, the easiest distinction one can make in languages. the problem of /e u A/ is not one of space, but one of distinction : you must distinguish three levels of high, and with that distinguish between front and back. So you could do 6 distinctions, yet you only have three vowels. (the /a i u/ pattern only show 4 possible distinctions, so it is, phonologically, easier to distinguish)

[schwhatever]

I think the problem between /e u A/ and /a i u/ is one of distinction.

a vs i + u is one of pure height : a is low, i and u are high (this explain that a language might have only /a/ and /i/). Then, you have between i and u the same division, only here is it frontness, i is front, u is back. Basically, this means that you make dichotomies, the easiest distinction one can make in languages. the problem of /e u A/ is not one of space, but one of distinction : you must distinguish three levels of high, and with that distinguish between front and back. So you could do 6 distinctions, yet you only have three vowels. (the /a i u/ pattern only show 4 possible distinctions, so it is, phonologically, easier to distinguish)

This potentially also explains the tendency of trivowel systems to slightly front /a/, since the realization of "/a/" as front would reduce the needed qualities even more.

"Standard" Trivowel: (2 by 3)

Code: Select all

i|.|u
.|a|.

"Fronted" Trivowel: (2 by 2)

Code: Select all

i|u
a|.

Jburke's Suggestion: (3 by 3, because of distinction between back rounded and back unrounded)

Code: Select all

.|.|u
e|.|.
.|A|.

[abeygail]

In parts of Hawaii (like Kauai, or Tauai, as I should say, /k/ And /t/ were not distinguishesd.

[Drydic]

In parts of Hawaii (like Kauai, or Tauai, as I should say, /k/ And /t/ were not distinguishesd.

They were in fact in rather free variation. The realisation of *t as [t] was primarily on the islands of Niʻihau and Kauaʻi (as far as I've been able to determine), while Hawaiʻian's realisation of it as [k] was spread across the other islands. Rather makes sense when you think about how far Kauaʻi is from Oʻahu.

[Vuvuzela]

You're pretty much going to find an exception to every rule. As you all are probably aware, until recently overt recursion was considered a universal. However, Daniel Everett proved that Pirahã lacked it. However, when trying to make a language seem naturalistic, it's good to use "universals" as guidelines.

[finlay]

However, Daniel Everett proved asserted that Pirahã lacked it.

[Vuvuzela]

However, Daniel Everett proved asserted that Pirahã lacked it.

You're right, I should have said "claimed". However, I'm not the one who spent the better part of my life studying Pirahã, so I can't quite be sure. I tend to side with Everett, however, because most claims to the contrary have been made using outdated grammars written by Everett himself.

[finlay]

True, but I don't think that thinking that is particularly helpful: we really need a second opinion on Pirahã. And yet the paradox remains that they're not friendly to outsiders, and I just sort of suspect that Everett's knowledge of their language is still incomplete.

[Vuvuzela]

True, but I don't think that thinking that is particularly helpful: we really need a second opinion on Pirahã. And yet the paradox remains that they're not friendly to outsiders, and I just sort of suspect that Everett's knowledge of their language is still incomplete.

Probably. Everett made one grammar based on his Chomskyist opinions (which Chomskyists consider unbiased) and then made another based on his new opinions (which he and his followers consider unbiased), but no-one has made a grammar that we can agree is unbiased.
I don't think Everett's current work gives us the full picture, but it's the best picture we have right now. When you've got a culture of 350 monolinguals who live in the middle of nowhere and who aren't exactly interested in furthering linguistic research, you probably can't expect it to be well-studied.
Still, I hope more research is done, if only to prevent the entire field of linguistics from becoming a giant flamewar.

[Wattmann]

Still, I hope more research is done, if only to prevent the entire field of linguistics from becoming a giant flamewar.

Linguistics will become a science when linguists begin standing on one another's shoulders instead of one another's toes.

[Vuvuzela]

Still, I hope more research is done, if only to prevent the entire field of linguistics from becoming a giant flamewar.

Linguistics will become a science when linguists begin standing on one another's shoulders instead of one another's toes.

I use "to become" very loosely.