Re: Phonemes

"Peter T. Daniels" <grammatim@xxxxxxxxxxxxxxxx> wrote in
news:42B34FA3.1E47@xxxxxxxxxxxxxxxx:

> David Wright Sr. wrote:
>>
>> "Peter T. Daniels" <grammatim@xxxxxxxxxxxxxxxx> wrote in
>> news:42B319AF.7AB2 @worldnet.att.net:
>>
>> (snip)
>>
>> > David Crystal gives charts of the largest and smallest inventories in
>> > the Cambridge Encyclopedia of Language. Largest is a Khoisan language
>> > with four or five articulations for each of the clicks, totaling
>> > 80-something IIRC; smallest is Rotokas, with 11.
>> >
>> > Why would there be a theoretical maximum?
>> >
>>
>> Thanks for the leads.
>>
>> I would suspect that there would some upper limit, (not necessarily
>> absolute, but variable within some limits), to the number of phonetic
>> variations that could be used without running into serious confusion
>> between them.
>
> "Not necessarily absolute but variable within some limits" is the very
> antithesis of "theoretical maximum"!

Not necessarily, I have a theory that there is a point at which too many
phonemes would be impractical. Therefore, in my theory, there is a
practical upper limit. Also, there is a definite limit in the human ability
to distinguish sounds which are too close to each other. This, in itself,
would put a theoretical upper limit on phonemes. No, don't ask me where I
learned that. I don't remember, and I will concede that I could be wrong,
but I don't think so.

>
> Malayalam has seven points of articulation for the nasals.
>
>> My original basis for asking this was due to a story by my favorite
>> author[1]
>
> He was my introduction to sf; I read all the kids' books because they
> were in the Fort Washington branch of NYPL. Then I found Asimov and
> hardly ever looked back. Everything from Stranger in a Strange Land and
> Moon Is a Harsh Mistress is sheer madness, and for some reason I looked
> back at Starship Troopers and found that it went even back to there.

As far as I can tell from corresponding with a lot of RAH fans over the
years, roughly half prefer the so-called juvies. Another rough half prefer
the ones, post _Stranger_, (or _SST), and another half like everything
that he wrote. I am in that third half. Yes, I know, you can only have two
halves, but I am speaking of an extra-dimensional half. It is also rumored,
that there is a fourth half off in another dimension, that doesn't like
anything he wrote.

> (BTW those bottom-notes are very annoying.)

Are you referring to my signature files? If so, annoying in what way?
Because of the Heinlein references? I admit that I overdid it on the
previous post and I will not do that again. That was in the nature of a
joke. However, unless there is some rule against such, I will continue to
do individual ones. I have used them on a number of newsgroups and no one
has ever complained. I am very proud of my association with The Heinlein
Society. If such references are not acceptable here, then I will say do
svidanijie and leb wohl.

Oh, you could be referring to my habit of footnoting for which, sometimes,
I forget to put the note. If that is the case, then I can only say that I
prefer to footnote things which are not, IMO, specifically germane to the
immediate discussion.

>> written in 1948 called 'Gulf'. In it, the author using the work of
>> Ogden and Richards in Basic English, and Alfred Korzybski in General
>> Semantics, and implicitly, some linguist[2], postulated a language
>> called 'speedtalk'. In
>
> It doesn't really sound like Bloomfield's sort of thing, but Jespersen
> was into conlangs. Sweet wrote dismissively of them in the 11th
> Britannica.
>

I wasn't connecting Bloomfield to any notion of artificial languages.

I was referring simply to the fact that Heinlein was aware of *some*
linguistic theory and I would suspect that it would be from Bloomfield as
_Language_ would have come out at approximately the time that he was doing
much of his early studying in all sorts of fields, especially during his
tours of duty at sea.

>> simple terms, speedtalk was based on the notion of one phoneme per word
>> for
>
> Then it is by definition not a possible human language -- there's no
> duality of patterning!
>

Can you explain what you mean by 'duality of patterning'. I don't recall
ever hearing that one. Why would the lack of it make it a non-possible
human language?

>> the roughly 800 words of a language similar, but not strictly based on
>> Basic Language. In concept, the structure of the language was more
>> along the lines of LogLan.
>
> I should think 800 phonemes would be thoroughly impractical, but not
> that it's "theoretically" impossible.
>

See above

>> The notion of one phoneme per word has very obvious problems, not alone
>> considering what I suspect would be caused by running into a much
>> reduced upper limit, but also having serious problems with a lack of
>> redundancy. IOW, I simply don't believe that it would possibly work. Of
>> course, he did postulating that the speakers of this language were
>> 'homo novis', the beginning of a new sub-species of man whose 'superman
>> ability' was the ability to 'think better'. see:
>> http://tenser.typepad.com/tenser_said_the_tensor/2005/05/gulf_by_robert_
>> .html for a critical review on the story.
>>
>> Anyway, as much as I respect this author and like his works, I think
>> that he was trying to either overlook limitations of real linguistics,
>> or that he simply had an incomplete understanding of linguistics, to
>> make a plot point.
>
> They always did.
>
>> I was just curious as to what people might think on the notion of a
>> practical upper limit.
>

> It's entirely different from the notion of a theoretical maximum.

See above.

--
David Wright Sr.
There is most certainly such a thing as a flea circus. However,
there ain't no such thing as a flea lunch!
To e-mail me, remove 't' from dwrightsr

.

Relevant Pages

  • Phonemes
    ... What language has the largest known number of phonemes and how many does it ... Has any work ever been done to determine a theoretical maximum? ... Prev by Date: ...
    (sci.lang)
  • Re: Phonemes
    ... David Wright Sr. ... >> the Cambridge Encyclopedia of Language. ... >> Why would there be a theoretical maximum? ... > upper limit, but also having serious problems with a lack of redundancy. ...
    (sci.lang)
  • Re: Phonemes
    ... > the Cambridge Encyclopedia of Language. ... I would suspect that there would some upper limit, (not necessarily absolute, ... The next meetings of the Heinlein Readers Group ... The Heinlein Society is full of people who really believe, ...
    (sci.lang)
  • Re: P vs NP: my proof of P != NP
    ... >> next phrase indicates, since any A in P will be accepted by some TM ... We can build such reduction only supposing that ... said that your statement that "there is no upper limit" is incorrect. ... that every language in P is accepted by an Otime-bounded TM. ...
    (comp.theory)
  • Re: Phonemes
    ... David Wright Sr. ... > What language has the largest known number of phonemes and how many does it ... the Cambridge Encyclopedia of Language. ... Why would there be a theoretical maximum? ...
    (sci.lang)
Loading