dissociations at the neurological level seems to lead to the hypothesis of
specific modules for the temporal and melodic components of music,
themselves composed of distinct submodules for,on the one hand,meter
and rhythm,and on the other,contour,pitch,and intervals (Peretz 1993).
To these diverse components,it seems necessary to add a semantic
component.
We know of the great difficulty in specifying the nature of musical
signification,vigorously challenged by the formalist tradition from
Hanslick to Stravinsky and the latter’s famous axiom,“I consider music,
by its nature,incapable of expressing anything.”I believe that to have a
less artificial and less inexact idea of musical signification,one must
abandon “great”music and instead turn to contemporary and primitive
forms of dance music,from ritual to disco.The issue is not about repre-
sentational semantics but about what I call rhythmo-affective semantics,
which involves the body,its movements,and the fundamental emotions
that are associated with them.This point seems to be essential:our
conception of music,based on the production,perception,and theory of
“great”European classical music,distances ourselves irremediably from
the anthropological foundations of human music in general.
Let us now attempt the same exercise in the area of language,a
field in which resistance is much greater,because it is difficult for us to
think that language does not constitute,according to the formulas of
Saussure,an organism in which everything is internally connected.Yet,
language is not,any more than living organisms,a perfectly organized
totality or a formal system:both are made from the pieces and fragments
that evolution,bit by bit,adapted to the world,and coadapted among
themselves.It is thus not certain that all of the components of language
appeared at the same time (but see Bickerton,this volume,for a differ-
ent viewpoint).The constitutive dimensions of language are well known,
but I would like to emphasize especially those aspects that are most often
underestimated.
We classically distinguish a phonetic-phonological level,a grammati-
cal or morphosyntactic level,a lexicosemantic level,and a pragmatic
level.Concerning the first level,an essential point is that this itself is com-
posed of two sublevels:a segmental level of phonemes,and a supraseg-
mental level corresponding to the phenomena of accent,intonation,and
duration.One sees here the appearance of a first point of meeting
between language and music:the suprasegmental level of language
depends on mechanisms close to those that are operative in the melodic
component of music (see Brown,this volume).Moreover,language pos-
sesses,like music,a temporal and rhythmic component,essential for
speech,and that appears,for example,in the fundamental unit of the
syllable.A universal definition of the phoneme (if there is one and if we170 Jean Molino