Re: A Greenbergian Tidbit

On Sun, 18 Feb 2001 15:24:13 -0600 Ed Heil <ejh@...>
writes:
> I was reading a little book of old essays by Greenberg, and came
> across
> this interesting tidbit.  I thought it was quite interesting, and
> probably
> true, although invoking Skinner in linguistics went severely out of
> fashion decades ago...
>
> When you are training a person or an animal, and you are trying to
> train
> them to do one of a set of responses B after a stimulus A, it is
> very
> feasible for A to be many and B to be few but very difficult for
> the
> reverse to be true.
>
> For example, it's easy to train a dog to jump after you either
> wink,
> blink, or nod, but it's difficult to train a dog to either jump,
> bark, or
> wag after you wink.
>
> If we generalize this to series of behaviors, we can understand why
> in
> general suffixing is much more common than prefixing in language.
>
> If it's easier to go from (large group of possibilities) -> (small
> group
> of possibilities), it's probably easier to go from (free morpheme)
> ->
> (bound morpheme) and therefore easier to suffix than to prefix.
>
> Not an absolute, but a tendency.
>
> (The observation about suffixing being more common than prefixing
> in
> general comes from Sapir, according to Greenberg.)
>
> Neat.
>
> Ed
>
My intuition says that the foregoing should be true in languages
which have free word order.  But in languages with strict word-order
rules, it should be easiest for a prefix to tie what it prefixes to a
word
that's already appeared in the sentence, because the prefix would
then be alerting the listener to the fact that the word he's about to
hear is tied to a particular word that he's already heard.

--Tommie