S. Johansson, Origins of language (rozdział 6, 7).pdf

Origins of Language

Constraints on hypotheses

Sverker Johansson

University of Jönköping

John Benjamins Publishing Company

Amsterdam

Philadelphia

The paper used in this publication meets the minimum requirements

of American National Standard for Information Sciences – Permanence

of Paper for Printed Library Materials, ansi z39.48-1984.

Library of Congress Cataloging-in-Publication Data

Sverker Johansson

Origins of Language : Constraints on hypotheses / Sverker Johansson.

p. cm. (Converging Evidence in Language and Communication

Research, issn 1566–7774 ; v. 5)

Includes bibliographical references and index.

1. Language and languages--Origin. 2. Human evolution. 3.

Biolinguistics.

P116.O76

2005

401--dc22

2004066031

isbn 90 272 3891 X (Eur.) / 1 58811 629 8 (US) (Hb; alk. paper)

No part of this book may be reproduced in any form, by print, photoprint, microﬁlm, or

any other means, without written permission from the publisher.

John Benjamins Publishing Co. · P.O. Box 36224 · 1020 me Amsterdam · The Netherlands

John Benjamins North America · P.O. Box 27519 · Philadelphia pa 19118-0519 · usa

C HAPTER 6

ANIMAL COMMUNICATION IN THE WILD

While language in the strict sense may be uniquely human, numerous other species

have their own means of communication, many of which appear to share at least

some, if not all, of the properties of language.

‘The inﬁnite use of ﬁnite means’ is a crucial property of human language (Hum-

boldt, 1836), but it is also a property of the songs of certain birds and whales, at

least in the limited sense of their combining a set of elements in an inﬁnite variety

of permutations. Recursivity may be found in some songs (Li & Hombert, 2002),

and the coining of new elements has been reported for some birds (Clemmons,

1991). The majority of animal communicative acts may be non-symbolic sig-

nals, but there are examples of vocalizations where this is not self-evident (Marler,

1998, further discussed below). Hauser (1997) gives a thorough review of animal

communication, in an evolutionary perspective, and Hakansson (1995) provides a

popular overview.

Birdsong, just like human language, can be culturally transmitted, and geo-

graphical ‘dialects’ are common (Wiener, 1986; Baker, 1996). Some birds appear

to have an innate ‘song acquisition device’ (Whaling et al., 1997), in analogy with

the proposed language acquisition device of humans. Birds of many species need

to hear the songs of other birds of the same species during a sensitive period while

they grow up, or they will not develop normal singing abilities (Wiener, 1986;

White, 2001). And Okanoya (2002) reports that there are patterns in the song of a

Bengalese ﬁnch that can be modeled by grammar-like rules.

But whatever similarities there may be between birdsong and human speech,

they have to be produced by parallel evolution — the underlying hardware is com-

pletely different, both in the vocal apparatus (Goller, 1998) and in the brain. It

is interesting for comparative studies (Wiener, 1986), but if we are looking for

possible precursors of human language, we had better turn to mammals.

Whale songs and calls are culturally transmitted as well (Whitehead, 1998;

Noad et al., 2000; Yurk et al., 2002), and some researchers argue that whale

songs have something resembling a hierarchical grammar (Seife, 1999), though

this remains controversial. The overall style of whale songs more resembles bird-

120 Origins of language

song than speech, so their relevance for language is not totally obvious. Gibbon

songs also resemble birdsong more than they resemble language, even though gib-

bons are our fairly close relatives (Hauser, 2000; Geissmann, 2002). Still, Cowley

(2002) and Ujhelyi & Buk (2001) suggest that there may be links between gibbon

song and human language.

The sounds produced by dolphins have more language-like features, but their

quantitative study is still in its infancy. McCowan et al. (1999) ﬁnd promise in an

information-theoretical approach, akin to Zipf’s classical work on human language

(1935; 1949), but the available ‘dolphinese’ corpus is still insufﬁcient for any ﬁrm

conclusions. The classiﬁcation and discrimination of dolphin ‘words’ is also a

non-trivial task for humans (or human-built computers), since the ‘phonology’ of

dolphins is so different from ours (Janik, 1999).

6.1

Do animal calls mean anything?

“What do animal sounds mean” asks Marler (1998, p. 2). The traditional view,

among biologists as well as linguists, has been that they have only affective mean-

ing, expressing only the emotions of the ‘speaker’, without symbolic referents.

This view is likely to be accurate in the case of most animal communication.

But numerous studies in recent decades, starting with Seyfarth et al. (1980), have

demonstrated that many animals use alarm calls and/or food calls, that for all prac-

tical purposes function as if they carried symbolic referential meaning.

The original study of Seyfarth et al. (1980) concerned the alarm calls of vervet

monkeys. These monkeys have a set of three distinct alarm calls, used for three

different predators (snakes, leopards, and eagles). When a vervet monkey hears

one of these calls, he or she takes appropriate action, different for each alarm

call. They run for cover in bushes when hearing the eagle call, climb up into

the treetops when hearing the leopard call, and stand up to scan the grass when

hearing the snake call. A purely affective call, basically conveying only that the

caller was scared by a predator, could not reasonably have led to such appropriate

actions. Marler (1998) also reviews some interesting data on the call acquisition

of young monkeys — the calls as such appear to be innate, 1 but the association of

a particular call with a particular predator is learned, and the young monkeys start

out by over-generalizing, using the eagle call for anything from falling leaves to

actual eagles, and then gradually learn when it is appropriate to use the call. Adults

use the call almost 2

exclusively when a monkey-eating bird is around.

1 Though there exist other monkey and ape calls, where the calls themselves appear to be learned, since

‘dialect’ differences between groups have been observed (Mitani et al., 1999; Fischer et al., 1998;

Hauser, 1992; Marshall et al., 1999).

2 The rare ‘mistakes’ concern birds that resemble dangerous birds, but are actually safe.

Animal communication in the wild 121

It should also be emphasized that these monkey calls are not, as far as we can

tell, iconic (cf. Section 11.3). They do not resemble any sounds of the predator

they’re referring to (Carstairs-McCarthy, 1996). In Diana monkeys, the alarm calls

of males and females are acoustically different, both from each other and from any

predator sounds, but nevertheless engender the same response in both genders (Zu-

berb uhler, 2000d). This non-iconicity is a central property of human languages as

well, and has been invoked as a deﬁning and uniquely distinguishing property of

human languages. 3 Finding similar non-iconicity in animal communication under-

cuts the human claim to uniqueness.

But in order to regard calls as truly symbolic, referentiality and intentionality

are crucial diagnostic features, that are difﬁcult to operationalize in wild animals.

The phrase ‘ functionally referential’ is often used in animal communication con-

texts, basically in order to sidestep the contentious issue of whether animals intend

to refer to an eagle or whatever, but retaining the implication that these calls for all

practical purposes function as if the caller intended to warn his fellows. We shall

return to the issue of communicative intent below.

Since the original work by Seyfarth et al. (1980), similarly ‘functionally refer-

ential’ calls have been observed in numerous species:

1. Birds:

Chicken, both domestic and their wild relatives (Evans & Evans, 1999; Mar-

ler, 1998; Hauser, 1997).

Several species of passerines (Marler, 1998).

2. Rodents:

Alpine marmots (Marler, 1998), but oddly enough not the closely related

yellow-bellied marmot (Blumstein & Armitage, 1997).

Several species of squirrels 4 (Greene & Meagher, 1998; Marler, 1998).

3. Suricates (Manser et al., 2002).

4. Primates: too many to list here. Reviewed in both Marler (1998) and Hauser

(1997). The list includes a few lemurs (Fichtel, 2004), numerous monkey

species, and chimpanzees. 5

It should be noted, however, that the functional referentiality is not always along

the dimension of different predator species. According to Zuberb uhler (2000d),

squirrels use different calls depending on the urgency of the alarm, and chicken

use different calls depending on predator elevation — a hawk on the ground elic-

its the ground-attack call normally used for foxes, rather than the air-attack call

3 The non-iconicity of human language is not, however, totally self-evident and universally accepted.

Nuckolls (1999), Langacker (2001), and Wilcox (2004) argue for non-negligible iconic components in

human speech, and in sign languages the iconicity of many signs is obvious.

4 Shriner (1998) found that squirrels and marmots also respond to each other’s alarm calls.

5 Hauser (1997) and Marler (1998) disagree on how solid the chimpanzee evidence is, which is rather

remarkable since Marler’s sole reference on this issue is to a study by Hauser. Byrne (2000) and

Crockford & Boesch (2003) add some more evidence of chimpanzee call referentiality, which may

help to resolve the issue.

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: