Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site sdcc6.UUCP
Path: utzoo!watmath!clyde!bonnie!akgua!sdcsvax!sdcc6!ir44
From: ir44@sdcc6.UUCP (Theodore Schwartz)
Newsgroups: net.ai
Subject: Re: Langauge Evolution
Message-ID: <1791@sdcc6.UUCP>
Date: Tue, 30-Oct-84 20:55:23 EDT
Article-I.D.: sdcc6.1791
Posted: Tue Oct 30 20:55:23 1984
Date-Received: Fri, 26-Oct-84 09:07:37 EDT
References: <13190@sri-arpa.UUCP>
Organization: U.C. San Diego, Computer Center
Lines: 72

There are some ideas around on why languages become simplified over time
as well as why they arrived at some initial point of complexity. It has
been suggested that simplification takes place where there is contact
between speakers of different languages who must communicate.
Pidginization results. In Pidgin languages, and there have been many,
such as current Melanesian pidgin English or current Indonesian, a 
language develops, sometimes to facilitate trade, sometimes where there
is domination by one group over the other, or where one population is
conquored or enslaved (early American slave English was a Pidgin
language). Often part of the vocabulary of the language of the dominant
group is combined with a simplified syntax (in the case of Melanesian
pidgin, derived from Melanesian). Many grammatical distinctions are 
dropped, sometimes even number and gender, irregular verbs are
regularized, as in "I is, you is, he is, we is, etc." Sydney Ray, early
this century, argued that Melanesian languages, were already pidginized
and locally developing some new complexities, when they encountered 
European traders and colonizers. Pidgin English grew up between them,
with contributions from both sides and without much conscious planning.
Pidgin languages spread wherever Europeans went but with both some
carry-over and much local influence in each area, so that Melanesian
pidgin includes some features from the China-coast trade pidgin and
some words from Portuguese African pidgins (like "pikinini" for child,
from "pequen\~a, "little.") Similarly, English is a pidginized language
having lost considerable complexity in comparison to other Germanic
languages.

The next step is to argue that simplification will occur, less rapidly
and dramatically, even in response to internal borrowing and exposure
to different dialects within a language community. Especially where
speech and verbal memory were the pricipal media for communication and
storage, differences would develop as a language community became 
larger and more spread out as well as more internally differentiated
by political and other cleavages. Communication across such gradients
would also lead to the sort of simplification that would facilitate
learning and intercommunication. Conversely, it has been suggested
that the greatest phonological and grammatical complexity develops
in relatively small, relatively isolated language communities, in 
part because internal gradients do not develop. Ancient Greek, referred
to by my predecessors in this discussion, may have reached maximum
complexity during a period when the population was small and relatively
isolated and progressively lost this as that population spread,
internally differentiated (within the language community) and entered
into communication with many groups who learned and simplified Greek.
I don't know this-- I'm merely fitting it to the above argument. It
could be that this is all we need and perhaps a historian could test
it, that complexity either develops or is conserved in relatively
small, isolated language groups. Why is this degree of complexity
needed in the first place? (A Melanesian language that I know, for
example, not only has singular, dual, and plural, but also trial
for small sets of persons or objects, not necessarily three, and
distinguishes inclusion or exclusion of the person addressed in the
first person pronouns). Why the complexity? Who needs it? Obviously
such complexity increases the redundancy in the communication of
ideas. Each distinction also implies constraints on the selection
or identification of referential objects or of other terms. One 
suggestion for which I can't argue very far, is that for people 
depending entirely on oral-verbal communication and memory, the 
additional redundancy may be useful. Much of the grammatical
complexity and referential specificity (e.g., no general set of
numbers except as inflected by numeral classifiers depending on
the many classes of objects counted, such as long, thin objects,
animate objects, etc.) Such specificity has proformal functions,
that is, the affix tells you something, narrows the range, of the
object referred to. It is likely also that small, isolated language
communities provide the container within which innovations can 
accumulate. There must be on-going complexifying processes as 
well as entropy or we would have nothing but pidgins around. Pidgins,
by the way, as you would expect from their minimal redundancy, require
extensive circumlocution to achieve the specificity of reference
(not only to objects but to more complex ideas) that the more complex
languages condense into words.