The mistle thrush of Europe engages in similar behavior (Marler 1959),
and learned birdsongs provide many other examples.
Some of the most complex songs of all are found in birds that,as they
acquire and develop their repertoire,take this process to extreme.Mock-
ingbirds and their relatives create hundreds of distinctive sequences
using phrases that are both invented and acquired not only from their
own species,but from other species as well,all recast into mockingbird
form and tempo (Boughey and Thompson 1976;Baylis 1982).The record
is held by a male brown thrasher,a relative of the mockingbird,with an
individual repertoire of over 1,000 distinct songs (Kroodsma and Parker
1977).
At some primitive level,the accomplishments of these songsters are
reminiscent of our own speech behavior.The more accomplished song-
birds create huge vocal repertoires,making extensive use of the same
basic process of syntactical recombination or phonocoding that we
use to create words.But of course there is a crucial contrast with
language.Song sequences are not meaningfully distinct,in the referen-
tial sense;they are rich in affective content,but lacking in symbolic
content.
Each of the thousands of winter wren songs that exist means basically
the same thing.Each serves as a kind of badge or emblem,a sign that
denotes identity,population membership,and social status.The diversity
may impress the listener with the performer’s virtuosity,and in some
species certainly enhances his reproductive prospects (Catchpole and
Slater 1995),as is argued for human music (Miller,this volume).Such
functions are important enough from a communicative point of view,and
there may be others.Many wood warblers have two distinct classes of
songs,one associated more with sex and the other more with male-to-
male interactions and aggression,as though there is a contrast in the
quality or nature of the underlying emotional state (Kroodsma 1988).
But as far as I know,no one has suggested that they are in any way sym-
bolically distinct.
Songs have none of the semantic content that some alarm and food
calls possess.The variety introduced by the generation of repertoires
serves not to enrich meaning but to create sensory diversity.We could
think of repertoires as providing aesthetic enjoyment or as alleviating
boredom in singer and listener. But in these learned birdsongs,
phonocoding does not augment the knowledge conveyed,in the symbolic
sense,as is so obviously the case in our own speech behavior.On the
other hand,symbolic functions are less at issue in music,and something
like phonological syntax is also involved in musical composition.Could
it be that more parallels with music than with language are to be found
in the communicative behavior of animals?40 Peter MarlerMUS3 9/14/99 11:58 AM Page 40