Talking is speech at its most expressive,most interactive,and most
rhythmically repetitive.The best models for it are the kinds of fast,lively,
collaborative,overlapping talk we find in Lorna Marshall’s (1976) de-
scription of !Kung conversation,Erickson’s (1981) description of Italian-
American dinner table talk,Scollon’s (1981) analysis of conversational
ensemble,Tannen’s (1989) accounts of high-involvement,rhythmically
repetitive talking voices,Feld’s (1994) account of Kaluli “lift-up-over-
sounding” putting talk together,Coates’ (1996) collaborative floor
talking jam sessions of women friends,and Chafe’s (1997) polyphonic
topic development.
I am following the principle that richness and diversity of behavior at
the beginning are the best things to look for if one wonders what really
complicated behaviors such as human language developed from.We
should not look for impoverishment at the beginning,as a general trend
can be seen in the evolution of all sorts of things,whether in biological
(Gould 1991) or cultural (Nietzsche 1887),that complexity always comes
from previous,but different,complexity.Behavioral complexity is never
created out of nothing.
Specifically,language always comes from previous language;all human
utterances are repetitions,versions of previously spoken utterances.
And the succession of repetitions with variations of previously spoken
utterances goes back indefinitely into the distant past.Talkers who are
maximally expressive,maximally interactive,with great repetition and
rhythmic grooves,with the greatest richness and diversity of these forms,
such as we see today in talkers referred to above,are the best candidates
for a model of evolving and fixing and playing around with the thousands
and thousands of formulas that eventually worked their way into socially
constructed language.In contrast are talkers who are dumbly laconic,
who do not say much,with little range of expressive forms,with little
back-and-forth interaction,with little or no repetition of self or others,
with no rhythmic repetition and thus no rhythmic grooves in which one
expects certain forms to fill upcoming temporal slots,with no help from
a huge variety of emotionally expressive intonation melodies and rhyth-
mic riffs to individuate sequence formulas.Because these laconic talkers
are monotonously nonexpressive of emotions,they are precisely the
worst candidates we could imagine for developing the richness of spoken
formulas we see in human language.
More generally,as far as human language discourses are concerned,
all sorts of specialized forms of human discourse,such as complex speech
acts,telling stories,speaking monologically in diatribes,or telling people
what to do,and eventually written discourses—the whole diverse range
of specialized forms of discourse people have developed over thousands
of years of history—can be easily seen as narrow,specialized forms that308 Bruce Richman