Sociological Linguistics

[Larry Trask]

On Tue, 18 May 1999, Patrick C. Ryan wrote:

[PR]

>>> All they needed was to perceive and intend to communicate in
>>> another way.

[JS]

> > -- they didn't do that, either.  All they intended to do was talk,
> >and they did -- and you can talk just as effectively in any
> >language, in any era of the human race.

> > Middle English is not one iota more "effective" at communication
> > than Old English.  It's just different.  Languages change because
> > they do.

[PR]

> Everything in life of which we have knowledge shows a development
> from the simple to the complex.

Not so.  Blatantly not so.

A good counterexample is dedicated parasitism, in which the parasite
loses all structures required for locomotion, perception, self-defense,
pursuit of prey, and whatnot, and is reduced to a mere sac of tissue
able to do nothing but to absorb nutrients from its host and to
reproduce.

Languages are also good counterexamples.  The earliest recorded or
reconstructible languages are in no way simpler than contemporary
languages.  And perhaps no recorded IE language possesses an
inflectional morphology as complex as that of PIE.  Does this make the
modern languages in any way inferior to PIE?

> My own studies and common sense decree that, at some point after the
> onset of linguistic communication, languages were simpler than they
> are now; and hence, less explicitly expressive.

It is not possible to make "studies" of the languages of our earliest
ancestors, since no data exist.  As for "common sense", well, I take
Einstein's view: `common sense' is merely a label we apply to something
we believe only because we want to believe it.

Our linguistic methods allow us to penetrate no more than a few thousand
years into the past, even in the most favorable cases, and they reveal
earlier languages in no way "simpler" than modern ones.  We have no way
of reaching back to the remote antecedents of language, tens or hundreds
of thousands of years ago, and we can't guess what these were like.

> As just the simplest example, a language which is unable to
> designate the plural form of a noun, is bound to introduce an
> *ambiguity* into a statement that a language which can does not
> exhibit.

Not remotely true, I'm afraid.  You are confusing grammaticalization
with expressive power.  A language can express distinctions of number
perfectly well without grammaticalizing some of these distinctions.

English is not more ambiguous than Classical Arabic or Fijian because,
unlike them, it does not grammaticalize dual number.  We use words like
`both' or `two'; they use inflected forms of nouns or pronouns for the
same purpose.  There is no difference in expressive power.

The North American language Kwakiutl grammaticalizes visibility:
different pronouns must be selected depending on whether the referent
is, or is not, visible to the speaker at the moment of speaking.
English does not do this.  Does this fact make Kwakiutl "more
expressive" or "more complex" or "less ambiguous" than English?

The tense language English requires `I saw Susie yesterday'; a tenseless
language like Mandarin Chinese has, literally, `I see Susie yesterday'.
There is no ambiguity: there are merely different choices as to which
information should be built into the grammar, as opposed to being
expressed otherwise.

> As another, certain languages have morphemes that have a much
> greater range of semantic inclusion than other languages. This also
> is a source of potential ambiguity that is not shared by languages
> that have differentiated semantic ranges more finely.

Not so.  English has only a single past tense, and `Washington crossed
the Delaware' can denote any temporal period between a moment ago and
the beginning of time.  Some other languages grammaticalize much finer
distinctions of past time: a moment ago, within the last hour, earlier
today, yesterday, recently, within the last few months, within the last
few years, many years ago, before I was born, and so on.  Is the African
language Bamileke-Dschang superior to English because it distinguishes
five different past tenses in contrast to our single one?  Is it less
ambiguous?

> Please do not introduce "feel good" sociology into an ostensible
> linguistic discussion.

I know nothing about sociology of any kind, but I do know that there
exists no case for claiming that any living, attested or reconstructed
language is more or less complex than any other, or more or less
expressive, or more or less ambiguous.  That's just a plain fact.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk