Re: On Relating Languages

Nicolas Walker wrote:
>
> I have three languages on the go, seperately concieved, yet structured
> (unintentionally) around similar 'design principles'. Upon realising that
> these languages were indeed related in some obscure ways, I thought
> perhaps I might be able to actually relate one to the other.....
> So,
>
> (1) In what ways may languages show relationship? I am aware of the main
> ones, but cannot find a comprehensive list or extended explanation.
(snip 2 questions that have already been discussed, and are trivial)...What
are the
> really 'interesting' ways a langauge can be related to another?
Aside from genetic relationship, which I assume you understand, there are
ways in which languages can become _similar_ but not related--
-- Contact, which can lead to massive vocabulary borrowing from one to
another, leading to superficial resemblance. Grammatical elements are less
frequently borrowed, but it's probably fair to say that English is a
somewhat divergent Germanic language due to influence from French post-1066.

Contact can also lead to pidgins, then to creoles. Not a field I know much
about.

Contact can also produce a _linguistic area_ ("Sprachbund"), a rather
mysterious phenomenon where the languages (unrelated) of a given area come
to resemble each other, especially in phonology. Examples are India, the
Caucasus, the langs. of the NW Pacific/Canada

-- Typology, which IMO really isn't worth the effort when it comes to trying
to _relate_ languages. There are just so many ways to handle, e.g. S, V, O--
German is SOV, but so is Basque. It proves nothing.

>
> (2)The phonology of each language is different, although they all
> fairly 'normal' (I suppose that is to say Indo-European). The approximate
> real word equivalent off each sound system would be (1)German(ic) (2)
> Arabic (3)Gaelic. Therefore,
> (a)In what main ways do sound systems diverge, in particular those
> mentioned above - i.e. Gaelic, German, Arabic.
> (b)How might I show relationship between, for example, one language with 5
> vowels ([a],[e],[i],[o],[u]), another with only 3 ([a],[e],[u]), and a
> third with a long list of glides, diphthongs and 'soft' (as in Russian)
> prononunciations. I suppose what I am getting at is, at what point is a
> language UNRELATABLE to the next?
If the vocabularies show few if any cognate items and few if any discernible
phonological developments in common.  A good first impression can be gotten
from the 100- and 200-word "Swadesh List" but it's impressionistic;
presumably any two languages that show less then 10% or so in common can
safely be called unrelated. But there are exceptions, which can only be
uncovered by using vastly more vocabulary, and rigorous comparison.

If you're dealing with 3 already created conlangs, you may have to make a
lot of adjustments, fancy phonological footwork etc. to make them appear
clearly "related".
>
> (3)What is the 'correct' academic fashion to explain and express
> relationship and langauge families. What are the accepted norms when it
> comes to comparative method,
You probably know that the comp.method deals primarily with phonology- to
establish _regular sound correspondences_ between languages. Once these are
established, we can look for morphological and grammatical resemblances,
which are often much harder to establish.

>family trees, dialects, orthography etc.?
Family trees are not always depictions of the actual truth, but they can be
handy. Dialects can be tested as to mutual intelligibility. Orthography is
unimportant.
>