The Origins of Music: Preface - Preface

primates in our own case,so that it is clear that vocal learning evolved

quite separately in the two cases.Indeed it also appears to have evolved

at least twice elsewhere among birds (parrots,hummingbirds) and three

times among mammals (seals,whales and dolphins,bats;Janik and Slater

1997).Although vocal learning is a key feature in the evolution of sound

communication,and of complex repertoires in particular,I will not

discuss it further here.It is reviewed extensively elsewhere (Catchpole

and Slater 1995;Slater 1989;Whaling,this volume).

Repertoires and Their Use

Repertoires have a wide range of sizes even among close relatives.

Thrushes,for example,range from the European redwing (Turdus iliacus),

which only has a single type,to the song thrush (T.philomelos) with its

repertoire of over 200 (Ince and Slater 1985).Here we are talking about

distinct songs,with little if any sharing of elements between them.

However,some birds build up what sounds like a formidable repertoire

from just a few elements assorted in all sorts of different ways.Catchpole

(1976) suggested that a male sedge warbler,in the course of his life,may

never repeat exactly the same sequence of elements twice.This is because

the song is long,and the few dozen elements follow each other in highly

varied orders.At one level (that of the song) his repertoire size is enor-

mous,but at another (that of the element) it is not very large at all.

Of course,if a bird does not often sing the same song type twice it may

be either because it has a huge repertoire of types or because it is impro-

vising so that its repertoire is continuously changing.The latter may

occur in some species,as suggested for the Sardinian warbler (Sylvia

melanocephala;Luschi 1993),but it is certainly rare.In some other

species,songs change more slowly from one part of the season to another

or from one year to the next (e.g.,thrush nightingale;Sorjonen 1987).

When they appear to generate new songs much more rapidly than this,

so that the same one is seldom repeated,it may be because a limited

repertoire of elements is reassorted between songs.Jumbled and ram-

bling though many birdsongs may seem to be,examination of sonograms

reveals that they are in reality usually far from that;exactly the same

elements or song phrases occur again and again,albeit perhaps in very

different orders.

When a repertoire is large,it may be quite difficult to measure by

looking through the bird’s output and searching for repetitions

(Kroodsma 1982).The brown thrasher (Toxostoma rufum) from North

America is the best example here,and it is the current record holder for

song repertoire size.As each bird has over 1,000 song types,arriving at

52 Peter J.B.Slater

MUS4 9/14/99 11:59 AM Page 52