ANIMAL BEHAVIOUR, 2004, 67, 369e378

doi:10.1016/j.anbehav.2003.10.001

Book and Software Reviews

From Hand to Mouth: The Origins of Language. By MICHAEL the movements of the lips, tongue, the soft palate and
C. CORBALLIS. Princeton, New Jersey: Princeton University larynx are all controlled muscularly as are the facial and
Press (2002). Pp. xii+384. Price $27.95. manual expressions that often accompany speech. De-
scribing a study by Michael Tomasello et al., Corballis
writes that captive chimpanzees’ gestures, produced
In From Hand to Mouth: The Origins of Language, Michael without training, are almost universally directed towards
Corballis argues that language expressed in the medium of attentive individuals, but their vocalizations are not
gesture emerged in human evolution prior to spoken directed towards any particular individuals. In this way,
language. As someone who studied dance from an early the dyadic nature of chimpanzee gestures appears more
age and for whom dance is the body-in-language, I find akin to language, inviting interaction and response in
this a very intriguing hypothesis. Dance’s gestures com- contrast to the less directed chimpanzee vocalizations,
municate as much to me as a well-written poem or story and this may suggest why gestural forms of communica-
about ideas, intimacy, relationships, and the full range of tion are more successfully taught to great apes. However,
human emotions. Likewise, for Corballis, gestures embody another interpretation of Tomasello’s results is that the
the message. attentiveness of purposely observing gestures is an artefact
In this book, Corballis presents an intricately woven of the requirement to look at the gesturer to be able to see
and well-researched argument for the a priori emergence the gesture, while vocalizations can often be perceived
of language-as-gesture. The detailed evidence presented without looking at the sender.
and the author’s interpretations will fascinate both Could human ancestors, like the other great apes, have
researchers and students. Corballis’ witty humour and been more capable of communicating gesturally rather
lucid writing style make the book eminently readable and than vocally? Corballis speculates that our hominid
enjoyable. Moreover, the author illustrates that science is ancestors may have used gestures in a systematic, rule-
a human endeavour that extends beyond itself: Corballis’ based manner comprising language of some sort, or at
inclusion of insightful literary quotes highlights how least a protolanguage, but uttered vocalizations in the
artistic and philosophical commentaries on the world same manner as present-day monkeys, referentially
parallel scientific questions and how they are inextricably expressing emotional states to specific stimuli. Protolan-
entwined through the capacity for language and thought. guage, a concept credited to the linguist Derek Bickerton,
Language, as a tool of precise communication of is described as distinct from a true language, relying on
complex ideas unrestricted temporally or spatially, may a primitive syntactical foundation that uses words
have been the greatest invention or development (de- representational of objects or events engineered into short
pending on one’s perspective) in all of human evolution sequences. Corballis describes the typical protolinguistic
and may in fact be causal or instrumental in facilitating animal as being able to organize words into two- or three-
other developments that characterize human evolution, word sentences comprising a command, an object and
such as complex tool production, pedagogy and art. a location. However appealing the concept may be of
According to Corballis, the ability for language was a protolanguage existing before true language, it may have
probably responsible for early humans’ successful migra- little explanatory power. Gould (1980) hypothesized that
tion out of Africa about 50 000 years ago and their incremental steps in the adaptive process were not always
subsequent survival in new and challenging habitats. helpful to the organism. Protolanguage similarly is a pro-
Could language have originated in gesture rather than posed intermediary step in the evolution of true language.
vocalized sounds? Furthermore, is language as argued by One needs to question the usefulness of being able to
Corballis to some degree, unique to humans? Languages express oneself with less precision. The resulting mis-
currently in use across human populations are not all communication could be highly disadvantageous to the
confined to being spoken. Like spoken languages, signed individual doing the communicating and to group
languages are processed in Broca’s and Wernicke’s areas, members or kin on the receiving end. Moreover, the fact
relying on the same neural substrates. They are made up of that humans develop the use of gestures in language when
the same bases as spoken languages utilizing syntax, speech and hearing are impaired is not necessarily
generativity and symbolism, although some signs are evidence that vocalized speech evolved from gesture.
more iconic than symbolic. Thus, gesture can be the entire Instead, it points to the flexibility of human behaviour to
medium for expression of complex ideas and events generate an alternative form of communication. Similarly,
extending beyond the temporal and spatial scope of the if language needs to develop in a language-rich environ-
individuals engaged in the act of communication. As ment and in a socially shared context, then how did it
Corballis notes, spoken language is itself a gestural system: evolve before such an environment existed, and how did
369
0003e3472/03/$30.00/0 Ó 2003 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved.
370 ANIMAL BEHAVIOUR, 67, 2

protolanguage incrementally make its way towards lan- seem to advance beyond simple associative learning’ and
guage without such an environment to support its use if it that ‘The human ability to project into the minds of
was at all imprecise? And if protolanguage was precise others may be another example of a principle that
enough, then why did it evolve to true language? chimpanzees do not grasp, a principle that enhances the
It is unlikely that sequenced gestures conveying impor- ability to interpret other people’s activities and modify
tant information, as in a sentence, could have evolved in one’s own behavior accordingly. But perhaps this ability
groups of humans unless they remained in spatially was also crucial for that other capacity that seems to be
tightly knit groups most of the time. Cooperative uniquely human, language!’ (page 59). There are a number
behaviours organized using gesture between a few indi- of problems with these conclusions. First, modifying one’s
viduals would be extremely limited for tasks such as own behaviour, whether human, chimpanzee or rat, by
gathering or hunting or even travelling. Moreover, as responding to social cues can be explained purely through
Corballis points out, human hands are specialized units instrumental means and need not refer to a higher-order
with flexible utility. Given our human ancestors’ probable processing such as ‘theory of mind’. To assume that when
nomadic way of life, their hands were likely to be busy or we are referring to humans we must assume theory of
full. Would it be adaptive to have to choose between mind rather than an operant principle at work is
either losing the ability to communicate because both anthropocentric. Second, it would be a mistake to take
hands were full or dropping what was being carried? a few studies by the same researcher using the same
Recursion, an important concept in this book, is the sample of captive chimpanzees and draw universal
basis for the overarching rules to produce and compre- conclusions about either the lack of abilities of an entire
hend embedded structures found in language and for species, both captive and wild, or the evolution of human
applying and deducing syntax. Recursion as a cognitive language. Finally, it is entirely contradictory to state that it
capacity is applied more ubiquitously by Corballis to is possible for Kanzi, a captive bonobo used in ape
explain why humans possess ‘theory of mind’, i.e. the language research, to possess a rudimentary form of
ability to know that we know what another knows and to recursive thinking (‘.Kanzi.demonstrates memory
manipulate this information, usually for our own benefit. of past events, future intentions.and even ‘‘theory of
Having an understanding of one’s own mind as well as the mind’’, an awareness of the feelings of others.[he] seems
minds of others includes the embedded quality of re- the equivalent, in both linguistic and social skills, of
cursive thinking and the rules that govern it, whether for a two-and-a-half- to three-year-old human child, except,
language or for social behaviour. Recursive thinking as of course, that he cannot talk.’ (page 35)), but then some
applied to theory of mind, Corballis concludes, may 20 pages later state as a possibility that chimpanzees
separate humans from other animals, including the great cannot go beyond simple associative learning.
apes. However, as Corballis suggests, if recursive thinking Along the same theme, Corballis writes that wild
has a globalized intelligence factor and is generalized to bonobo and chimpanzee vocalizations ‘are no more
problem solving in social manipulation, it is not logical to complex than those of other primates’ (page 29),
assume language as unique from other forms of cognition, equating them to the referential calls produced involun-
in parallel with Chomsky’s and Pinker’s theoretical tarily to emotional events and specific stimuli such as
positions on this subject. approaching predators. On the other hand, he states that
Referring to Daniel Povinelli’s research on pointing and the ‘natural communications of apes have not been
eye gaze in captive chimpanzees, Corballis writes that decoded’ (page 37). At times it is difficult to discern
these chimpanzees were sensitive not to the accuracy of where the author stands on issues concerning nonhuman
pointing at the desired object but to the pointing hand’s animal linguistic or cognitive abilities and therefore it is
vicinity to the desired object. Corballis describes a similar difficult to concede to his interpretation of the evolu-
study by Povinelli in which chimpanzees were given tionary origins of language. At one point, Kanzi’s
a choice of having a blindfolded person, a person whose impressive ability to understand spoken English is
head was covered by a bucket, or one without hindrances highlighted, including the bonobo’s ability to understand
to retrieve an object that the chimpanzee pointed to. The nested or embedded instructions which require decipher-
chimpanzees did not appear to distinguish easily between ing using rule-based knowledge (i.e. recursive thinking).
the blindfolded/bucketed person’s ability to retrieve the Contrastingly, Kanzi’s lesser abilities in producing fully
desired object and the ability of the person without the grammatical sentences cause the author to make an
headgear. (The experimental paradigm forces one to analogy between Kanzi and the Clever Hans phenome-
question the evolutionary and ecological validity of this non. And then, confusingly, the author refers to
situation to any species other than perhaps humans, who Chomsky’s and Pinker’s position that language is a bi-
are the only species to wear clothes and head ornaments, ological adaptation unique to humans, but insists that in
although not commonly buckets.) In a further study cited this new 21st century, we should ‘yield a little, and give
by Corballis, however, the chimpanzees appeared to have animals their due’ (page 36).
no problem distinguishing between reliable individuals The strongest evidence that Corballis presents for the
when one of the individuals was actually facing away emergence of gestural language in human evolution is
from the chimpanzee, i.e. with averted gaze (a much more related to Giacomo Rizzolati’s work on mirror neurons.
evolutionarily valid problem). I mention the inclusion of Mirror neurons were found in Broca’s area in macaques
this series of experiments because Corballis writes conclu- and process both the production and perception of
sively, that, ‘If Povinelli is correct, chimpanzees do not specific movements; i.e. they respond to a monkey’s
 BOOK AND SOFTWARE REVIEWS 371

own arm movements (towards an object) as well as to the doi:10.1016/j.anbehav.2003.10.002

available online at http://www.ScienceDirect.com
same movements made by another individual (towards
the same object). If gestures were the next logical
evolutionary step in complex communication leading to Mating Systems and Strategies. By S. M. SHUSTER & M. J.
language, it is possible that mirror neurons were the WADE. Princeton, New Jersey: Princeton University Press
gestural neural substrate precursors to spoken language in (2003). Pp. x+533. Price $35.00 paperback.
Broca’s area. Corballis speculates that at some point during
hominid evolution the control over both gesture and
vocalization was paired, and in humans this pairing Every offspring of sexual species has a father and a mother.
became lateralized to the left hemisphere. The existence Consequently, unless the sex ratio is biased, the average
of human left asymmetries in Wernicke’s area, specifically fitness of males and females is equal. Recently, Kokko &
in the planum temporale, necessary for language compre- Jennions (2003) highlighted the fact that researchers often
hension, has also been found in the other great apes. overlook this simple rule when attempting to explain the
Similarly, more recent studies by Cantalupo & Hopkins female bias in parental care that is evident among many
(2001) not included in this book have shown that both animals. Indeed, much theory regarding mating systems
species of chimpanzee as well as gorillas also exhibit a left- and strategies appeared to have developed with little regard
hemisphere asymmetry in Broca’s area (Brodmann’s area to quantitative genetics. This observation was the catalyst
44) that has been correlated with language production in for Mating Systems and Strategies by Stephen Shuster &
humans, and all three species show a right-hand bias in Michael Wade. The authors set themselves an ambitious
producing gestures. Cantalupo and Hopkins conclude objective: to provide a comparative, conceptual and stat-
that, ‘From an evolutionary standpoint, therefore, asym- istical framework for studying alternative mating strategies
metry in area 44 may be associated with the production of in natural populations. This volume establishes a solid
gestures accompanied by vocalizations in great apes, an theoretical foundation for the topic and provides testable
ability that eventually selected for the development of predictions that challenge current views. It follows a logical
speech systems in modern humans.[and] indicates that progression from the proposed theory to important pre-
the origin of asymmetry in language-related areas of the dictions and a new classification scheme for mating sys-
human brain should be interpreted in evolutionary tems and strategies. The book comprises 12 chapters, each
terms rather than being confined to the human species’ of which begins clearly by outlining the main aims and
(page 505). contains a useful summary of the key themes.
In From Hand to Mouth, Corballis presents a complicat- The first chapter uses mathematics to resolve the
ed line of reasoning and interpretation of evidence, apparent ‘Paradox of Sexual Selection’. Sexual selection is
theory and experimental results, from the importance of paradoxical because it is a potent evolutionary force despite
bipedalism and hominid fossil endocast evidence to the being restricted to a single sex, is opposed in the other, only
behaviours of living great ape species, discussing the affects one fitness component, and is exposed to selection
kinds of selective pressures and changes that may have for a relatively short time. Using the statistical approach of
taken place leading to humans’ abilities for gestural and Crow (1958) the authors demonstrate that the paradox can
then spoken language. Although there is a place for be resolved if you consider that the strength of sexual
gestural language in human societies, whether a formal- selection is driven by the variance in male mating success
ized signed language or a choreographed dance perfor- and the lack of this variance in females. They argue that the
mance, overall this reader is not entirely convinced by sex difference in the strength of sexual selection determines
the author’s argument that gesture was the originating the mating system. The appropriateness of using this
medium of expression for language. On the other hand, parameter to identify the strength of selection is convinc-
comparative psychology as a field can only gain from ingly argued in the face of several criticisms, and this
Corballis’ piecing together such a complex argument framework is established as the underlying model for the
and providing readers with a great deal to ponder and predictions and arguments that follow.
build upon. The next three chapters investigate the influence of
females on mating system evolution and extend on the
FRANCINE L. DOLINS
mating system classification scheme of Emlen & Oring
Department of Psychology,
(1977). In chapters 2 and 3, the authors argue that the
School of Humanities and Social Sciences,
strength of sexual selection depends on the combined
Polytechnic University,
temporal and spatial distribution of females, and therefore,
Brooklyn, NY 11201, U.S.A.
experimentally manipulating these parameters provides
a test of the nature of this relation. In general, an increase in
the synchrony of receptive females reduces the strength of
sexual selection, because variance in male mating success is
reduced, but an increase in female spatial aggregation
References increases variance in reproductive success among males. In
chapter 4, the authors discuss the implications of multiple
Cantalupo, C. & Hopkins, W. 2001. Asymmetric Broca’s area in mating by females within the theoretical framework
great apes. Nature, 414, 505. described earlier. They use a simple hypothetical example
Gould, S. J. 1980. The Panda’s Thumb. Middlesex: Penguin. to demonstrate that sexual selection should favour males