Analytic languages and their function. (3)

A. Katz amnfn at
Sat May 27 06:05:42 UTC 2006

Steve Long wrote:

>Early language is a different question.  If you imagine a language with
>the minimal possible use of grammatical differentiation -- one case, one
>tense, one voice, etc. -- you are essentially imaging a language with
>little need for inflection or fusion or any other synthetic feature.  It
>may be that such a language would need some modicum of syntax and some
>minimal number of prepositions to be comprehensible.  But its fundamental
>character would be recognizable as analytic.  And if such a language were
>to move in the direction of differentiating out the grammatical features
>we recognize in modern languages, it would be moving towards the

Why prepositions and syntax? It still seems as if you are imagining early
language as a pidgin working its way up to a creole. But animal cries,
even those of our closest relatives, tend to be holistic, in the
sense that a whole sentence is communicated with a single word, as is the
case in the most synthetic of human languages.

Imagine the minimal set of messages to be divided into two categories:
(1) warnings (about predators) and  (2) invitations (to feast on specific

The coding possibilities could be something like this:




The warning or invitation effect would be created by pitch and frequency
of repetition effects superimposed on formant combinations -- that is
combinations mostly of different vowels the animal is capable of making.

How did such a system arise? Partially, it came about through genetically
transmitted compulsions. The animal making the cry does not gain from
making it -- its surviving relatives are the ones who transmit the system
to the next generation. The system is autonomous and requires no theory of
mind for its interpretation. On seeing a predator, the speaker feels
compelled to make the corresponding cry. On seeing food, it cannot help
but declare it. (More evolved individuals of the group may realize the
danger to themselves and try to stop their mouths with their hands, to
keep the involuntary cry from coming out.)

But as unplanned as this system may be in terms of the individuals who use
it to encode and decode information, it is surprisingly autonomous as a
code, in the sense that sentences which have never been spoken would
immediatley be understood.

"Warning: fruit" would be a possible sentence whose meaning might be
something like the attack of the killer tomatoes.

(The creatures are omnivores, so it is possible that some animals in their
list of arguments can serve as both predator and prey.)

As theory of mind develops gradually in more evolved descendants of this
group, they gain voluntary control of the inbred communicative system, may
add vocabulary, and even learn to lie. As the system becomes more complex,
sentences will become less tightly bound, less holistic, more

In human languages, the larger the lexicon, the more isolating the
language is. Yet most linguists imagine an isolating starting point, where
people talk in something resembling a pidgin. Having separate words for
predicates and arguments is something that doesn't happen until a language
has acquired a very extensive lexicon.

>But, once again, we would expect the process of going synthetic to be
>forgotten, so that a language that goes to analytic is NOT going back to
>its earlier state.  We would not expect English speakers to recognize
>that the possessive "s" was once "his" and to go back to the old usage. So
>it's also not really a "cycle" in that sense.  You can't go back to an old
>analytic state.  You can only go on to a new one.

We don't need to confuse the re-cycling of particular morphemes with the
recycling of typologies. It is possible to fall into an older typology
without having the original morphemes performing their original function.
This is consistent with a unidirectional cycle of typologies.

That said, I would like to point out that sometimes the same morphemes do
return to roughly the same function through a series of changes. This does
not require speakers to know the history of the language. Sometimes they
just accidentally stumble into an older usage.

My 1996 dissertation showed that copulas turn into pronouns and pronouns
turn into copulas through a normal process of grammaticalization in many
unrelated languages. In Hebrew, there is even a full cycle from a third
person singular version of an inflected copula to a third person singular
pronoun to a third person singular copula once more. (The same word
was interpreted as a copula when the third person pronouns were null, and
as a  pronoun when null pronouns came into disuse, and then back to a
copula when null copulas were in disfavor.)

Not knowing the history of your language can sometimes doom you to repeat



Dr. Aya Katz, Inverted-A, Inc, P.O. Box 267, Licking, MO
65542 USA
(417) 457-6652 (573) 247-0055

More information about the Funknet mailing list