Re: Ural-Altaic

Helmut Richter wrote:

1. As far as I know, the existence of a common Ural-Altaic language
branch was accepted well into the second half of the 20c., that is,
into a time when rigour standards for scientific methodology were
not lower than nowadays. How could it be regarded as proven when it
is now regarded as disproven?


Standards were possibly as rigorous, though maybe it's more exact to
say that the ways in which they lacked rigour were not more numerous
than today, only different and hopefully few. However, science is like
that. Newton's gravity was superseeded by Relativity, which in itself
isn't a complete account of anything. In this case, from what I
gather, the relationship between those language groups was asserted
mainly on typological similarity. Nowadays, knowing more about the
process of language change, it's become accepted that typology is not
correlated to descent. Taking typology out, the similarities between
those language groups don't seem to be bigger than between any of them
and some other outer group, say IE. So, until further notice, there is
no reason to posit a relationship.

Unlike what may appear, language groups are not obtained by trying to
divide the set of known languages. Instead, languages are studied and
compared to see whether they have more in common than either with any
random language. *If* there is something, then that may be the
beginning of establishing a relationship between them.

IOW, it's not a matter of looking at the 'ural-altaic' languages and
trying to find a justification for the phylum. Rather, every known
group within the ensemble is compared to the others. If there is no
special similarity found, there is no basis for clobbing them together.

2. Is not the same process underway today for the Altaic languages
which, after haven been regarded as provably related, are now
regarded as two or three distinc branches (Turkic, Mongolic,
Tungusic)?


Look at it the other way: whereas no special similarity was found
between uralic and 'altaic' languages, within the latter there is a
reasonable amount of hints that they may form a group (altaic*).
However, the evidence is not conclusive. So, it seems there may be an
altaic whole, though it's not certain. While, unless something new**
appears, it seems there is no 'ural-altaic' group.

(*) Within Uralic, of course, the similarities are so good as to
warrant certainty.

(**) New languages, living or inscriptions, new ways to look at known
data, etc.


- [f] which evolved into [p] in B - [p] which evolved into [f] in A
 - something else which evolved into [f] in A and into [p] in B

What criteria could one have to discern between these three cases? 

None foolproof, again it's all probabilities:

.. Some sound changes are more likely than others
.. There may be other languages which help decide
.. There may be other phonemes or other features within the languages
which help decide

When all is said and done, one can't know, and historical linguistics
doesn't claim it, what exactly was the original value of the sound.
For instance, Proto-IE's laryngeals are posited, but there is no
certainty to what sounds they amounted. One may speculate with
likelyness, maximal differentiation of phonemes (for instance, if you
know a language has 2 and 2 only vowels falling in the range [a A & E
e], it's more probable that they are well apart than close*), but
that's speculation.

(*) Arabic, for instance, has /a i u/, but there's no need it be so.
However, I'd be surprised at a language which had only 3 vowels if
they were /a & E/.

Historical Linguistics cares more about phoneme contrasts than their
particular realisations.
--
am

laurus : rhodophyta : brezoneg : smalltalk : stargate
.

Relevant Pages

  • Re: Success of VB
    ... we never had to do garbage collection in the first place ... To be honest I tend to work in whatever language I'm ... This kind of debate goes on in all language groups with ... so much and it's a royal pain to develop a rich, good looking, modern, ...
    (microsoft.public.dotnet.languages.vb)
  • Re: Language
    ... mother and father, women and children, are old forms and are similar ... children and women are a plural form that is no longer used elsewhere ... Different language groups may be older ...
    (talk.origins)
  • Re: Language
    ... mother and father, women and children, are old forms and are similar ... children and women are a plural form that is no longer used elsewhere ... Different language groups may be older ...
    (talk.origins)
  • Re: Origin of Ogam
    ... I'm thinking that until recently that language and the people who ... language and genetics. ... There are areas where language groups are very small, ... at the physical diversity of speakers belonging to the larger language ...
    (sci.archaeology)
  • Re: The theory of Mr. Carrasquer
    ... >> an individual linguist's interpretation of the language, ... > to translate the corpus, and that is just one person's interpretation, ... "grammar and dictionary". ... Nowadays there are no excuses. ...
    (sci.lang)