Review of The Evolution of Language
|AUTHOR: W. Tecumseh Fitch
TITLE: The Evolution of Language
PUBLISHER: Cambridge University Press
Anne Reboul, Laboratory on Language, the Brain and Cognition, Institute for
Cognitive Sciences, Lyon, France
The present book is intended as an introduction to the evolution of language,
conceived of as the unique preserve of mankind. Fitch intends it to be a
dispassionate exposition of the various theories advanced by various scholars
and the data which supports or contradicts those views. The book consists of an
introduction and four distinct and sections of unequal length: ''The lay of the
Land: an overview of disciplines and data relevant to language evolution'', ''Meet
the ancestors'', ''The Evolution of Speech'', and finally ''Evaluating phylogenetic
models of language evolution''.
The ''Introduction'' insists both on the complexity of language and the resulting
diversity of perspectives about it, ranging from Chomskyan linguistics to the
psycholinguistics of language acquisition via neuroscientific approaches, and
advocating an inclusive rather than divisive approach, combining the valuable
insights in the different approaches and aiming at a comparison-based synthesis.
From there, the first section opens with a chapter devoted to ''Language from a
biological perspective''. Fitch pinpoints the inability of chimpanzees, man's
nearest relatives phylogenetically, to acquire language and advocates a
comparative approach to language based on a multi-component view: in other
words, though no animal species has anything remotely like human language, some
species may have this or that component or, at least, a predecessor to it. Fitch
then introduces two meanings of the phrase ''faculty of language'', a broad sense
which ''encompasses ALL the mechanisms involved in language acquisition and use''
(21, his emphasis), and a narrow sense targeting ''those mechanisms that are both
unique to humans and special to language'' (22). Fitch argues that most
components are indeed shared with other species, which opens the road to a
comparative approach. He then distinguishes ''communication'' and ''language'', the
former common to all species while the latter is uniquely human. Language can be
fairly described as an ''instinct to learn''. Fitch then distinguishes between
I-language (the cognitive system underlying language in each individual's brain)
and E-language (the languages spoken by different human communities).
Chapter 2, ''Evolution: consensus and controversy'', begins with a historical
description of the emergence of Darwin's natural selection, based on three
factors, variation, inheritance and differential survival. Fitch distinguishes
and describes three kinds of selection, sexual (one sex, usually the females,
picks the sexual partners, based on criteria which will then predominate in the
species), kin (based on the notion of inclusive fitness, where the degree to
which genes are shared can boost cooperation) and group selection (extending
inclusive fitness to the group). The comparative method is a tool for the
description of evolutionary history, leading to the construction of phylogenetic
trees, based on clades. It allows the identification of both homologous traits
(shared through inheritance from a common ancestor, showing the species'
phylogenetic history) and analogous traits (shared through independent evolution
in different species, showing the function of a trait). Fitch concludes with the
evolution of behavior, reminding the reader of Tinbergen's famous four-layered
approach to animal behavior, in terms of ''mechanism'', ''function'',
The third chapter, ''Language'', begins with a link between the existence of
sensitive periods for language acquisition and the biological basis of language.
It then naturally turns to biolinguistics, after a brief description of the
Chomskyan approach to linguistics as mental (cognitive), formal and biological.
Fitch points out that though parts of language can be seen as encapsulated in a
Fodorian sense, that is, as informationally closed (e.g., speech), other parts
are executive and unencapsulated (e.g., semantics and pragmatics). In other
words, there is room in language for both individual and cultural levels of
explanation and contemporary evolutionary theory can bridge them, through the
coevolution of cultural and biological systems. Fitch then outlines the
components of language, following the traditional distinction between phonology,
syntax, semantics and pragmatics.
The fourth chapter turns toward ''Animal cognition and communication'',
repudiating behaviorist approaches and defending cognitive achievements such as
episodic memory, categorical perception and mindreading in animals, as well as
more wide-ranging capacities, such as categorization and learning,
planification, inference and reasoning, numerosity, cross-modal matching and
serial ordering. Fitch also emphasizes the existence of ''animal cultures''.
Animals thus evidence a number of cognitive abilities potentially relevant for
language, a conclusion which is reinforced by the fact that different vertebrate
species, from birds to sea mammals to dogs and apes have been shown to be able
to master ''large referential vocabularies''. Fitch then turns to animal
communication systems, highlighting the fact that most animal signals are innate
(in contrast with words) and have an emotional basis though they can still be
under voluntary control. However, some animal signals have a rule-based
structure, and some even have a phonological syntax, though meaningful syntax is
fairly sparse (a two signal-concatenation in some species of monkeys). Though
there are functionally referential signals (alarm or food calls, notably), these
are not intentional in a Gricean sense, but the interpretation of such signals,
especially by non-conspecifics, can be inference-based.
Section 2, ''Meet the ancestors'', opens with an eponymous chapter going from the
origins of life to modern humans. The goal of the chapter is to introduce the
notion of ''latest common ancestor'' or LCA (between humans and chimpanzees, their
Chapter 6, ''The LCA: our last common ancestor with chimpanzees'', describes the
characteristics which may legitimately be attributed to the LCA on the basis of
man-chimpanzee comparisons: communication, sociality, tool use, hunting and
medicine, violence, and limited male parental care.
Chapter 7 is devoted to ''Hominid paleontology and archaeology'', and it gives a
clear and detailed overview of the state of the art, outlining three stages in
human evolution from the LCA: the emergence of bipedal apes (early
Australopithecines) from about 4-2 million years ago (MYA); an increase in body
and brain size and the use of more complex tools comes with the birth of the
genus ''Homo'' (by 1.9 MYA); and the final stage with so-called ''archaic Homo
sapiens'', including Neanderthals, and finally the appearance on the scene of
anatomically modern Homo sapiens or AMHS (by between 100-200 KYA) which spread
all over the globe leaving their birth place, Africa, by 50 KYA. This date
provides the latest possible date of the evolution of modern language, given its
Section 3, ''The evolution of speech'', opens with chapter 8, ''The evolution of
the human vocal tract''. Fitch begins by distinguishing between speech and
language, noting that though language is massively realized in speech, the
existence of signed languages shows that the two are distinct. He describes the
physiological apparatus by which animal vocalizations (including speech) are
produced, noting the commonalities among vertebrates. He then turns to the
descended larynx, pointing out that despite common belief, it is not specific to
humans, and is more likely to have been the object of sexual selection (for size
exaggeration in males) than a specifically linguistic adaptation. Fitch
discusses the hyoid bone in various hominid fossils. However, given the
existence of a mobile or descended larynx in many nonhuman species, it is
probably not relevant to any discussion on the evolution of language. He notes
that phonological categorical perception seems widespread in vertebrates and
thus is, again, unlikely to be a specific adaptation for language.
In chapter 9, ''The evolution of vocal control: the neural basis for spoken
language'', Fitch identifies the ability for complex vocal imitation as crucial
for human speech and, after having pointed out the limited existence of that
ability in nonhuman species (where it is mainly limited to some birds, and to
sea mammals), he turns to the brain mechanisms which underlie human speech.
Though most of them are common among mammals (birds have different kinds of
brains), there are novel cortico-laryngeal connections in humans, which argue
for a discontinuity between primate and human vocal control. Additionally, the
discovery of the FOXP2 gene and its importance for the development of complex
vocal motor control -- this gene, though widely shared with other mammals, seems
to have evolved in its human form about 400 KYA and, when deficient, produces
specific language impairment -- argues for the same conclusion.
Chapter 10, ''Models of the evolution of speech and phonology'', discusses four
models of speech evolution, from Lieberman's model (which attributes a central
role to the basal ganglions), to MacNeilage's (which outlines the link between
vocal motor control and motor control for feeding), to Deacon's (a selective
pruning of cerebral connections leading to the cortico-laryngeal connections
typical of humans), ending with Carstairs-McCarthy's view that vocal motor
control was a product of bipedalism. Fitch then turns to the question of the
evolution of the phonological structure itself, discussing perceptual
constraints, vocal imitation, glossogeny (or cultural transmission), pidgins and
creoles and describing computer simulations of ''invisible hand'' models for
Section 4, devoted to ''Evaluating phylogenetic models of language evolution'',
opens with a ''Historical overview: Western theories of language origin before
Darwin''. After a short mention of the biblical Genesis story, Fitch turns to
Herder's onomatopoetic theory of word origins, to various expressive theories,
to theories of language as a social tool and to Müller's anti-Darwinian theory.
He then outlines Darwin's theory of language evolution, based on the idea of a
sexually selected musical protolanguage, which would lead to a meaningful modern
language through increased intelligence.
Chapter 12, ''Lexical protolanguages'', turns to various accounts of a putative
pre-syntactic proto-language, having as prerequisites both vocal imitation and a
drive for referential communication. Fitch discusses Bickerton's model (under
which infant speech, ape language and pidgins are supposed to be models of
protolanguages), but rejects Bickerton's notion that syntax appeared through a
single ''macro-mutation''. He turns to Jackendoff's model, which insists on
lexical acquisition tools (based on the limited size of ape vocabulary) and a
combinatorial phonology. Both models assume the drive for communicating
information. However, such a drive would clearly involve cooperation and
cooperation is something of an evolutionary mystery. Fitch then outlines several
theories for the evolution of cooperation before turning to Dunbar's social
theory (gossip, leading to social sanctions against free-riders, is a good way
of promoting cooperation), rejecting some aspects of Dunbar's views, but keeping
others. He then turns to Deacon's account in terms of the evolution of the
social structure and feeding patterns of hominid groups, before outlining his
own theory, in three stages: kin selection for information exchange, leading to
reciprocal altruism through gossip, and, finally, syntax.
Chapter 13 deals with ''Signs before speech: gestural protolanguage theories'',
describing several accounts which propose a gestural protolanguage, concluding
that they all meet with a major hurdle: why, if language was gestural, and given
that sign languages show that full communication can be achieved through
gestures, did the shift to vocal communication occur? The recent discovery of
mirror neurons does not solve the problem.
Chapter 14, ''Musical protolanguage'', outlines Fitch's favorite candidate for the
evolution of language. He reminds us of several arguments for strong links
between language and music, including shared cerebral structures, and of
Darwin's account. The problem with the notion of a musical protolanguage, which
would not be meaningful, is how to account for a passage from meaningless
holophrases to meaningful combinatorial and discrete units. Here, Fitch turns to
Wray's account where semantic meaning is supposed to be ultimately derived from
coincidences between meaningless and noncombinatorial sounds in holophrases and
potential meanings. He defends it against criticisms by outlining Kirby's
computational model of the holistic/analytic transition, showing such a
transition to be possible.
Chapter 15 concludes, reminding the reader of Fitch's sympathy for the musical
protolanguage hypothesis and of the rapid growth of data relevant to the field
of the evolution of language.
This is an excellent book, which gives a detailed and precise account of the
current state of the field, as well as historical background to both the central
questions and different solutions which have been proposed. The comparative
perspective is wide ranging and always well-informed, if slightly optimistic on
animal capacities: for instance, though there is no doubt that animals have
categorical abilities -- indeed, it would be hard to imagine how they could be
entirely deprived of them --, this does not mean that their conceptual abilities
are in the human range and the existence of conceptual hierarchies seems fairly
limited. (Apes seem to be the only candidates to date: see Vonk & MacDonald
2002, 2004.) Again, it seems slightly misleading to say that animals can have
important vocabularies (''hundred of items'' are mentioned more than once), when
in fact the upper bound seems to be between 200 and 250 items, which is very far
from, for instance, first-graders' lexicons (estimated at around 6000 words).
Finally, though Fitch claims to maintain an aloof stance and not to push any
theory, he clearly defends musical protolanguage theories, and some arguments
given seem rather flimsy. Though no one disputes the existence of ''formulaic''
utterances in present day languages, this has nothing to say about the
probability of a holistic/analytic transition given that obviously the
transition in that case goes in the reverse direction: idioms or formulas are
made up of discrete units which come in some phrases to lose their individual
meaning, and this, if anything, illustrates an analytic/holistic transition.
Nevertheless, the book is impressive thanks to its erudition, clarity and
precise discussions of data, theories and arguments. It is a must for anyone
interested in the evolution of language.
Vonk, J. & MacDonald, S.E. (2002) ''Natural concepts in a juvenile gorilla
(Gorilla gorilla gorilla) at three levels of abstraction'', in Journal of
experimental analysis of behavior 78, 315-332.
Vonk, J. & MacDonald, S.E. (2004) ''Levels of Abstraction in orangutan'', Journal
of comparative psychology 118, 3-13.
ABOUT THE REVIEWER
| ABOUT THE REVIEWER:
Anne Reboul is a Senior Researcher at the French Center for Scientific
Research (CNRS) in France. She has a Ph.D. in Linguistics (EHESS, Paris)
and a Ph.D. in philosophy (University of Geneva, Switzerland). Her books
include an Encyclopedic Dictionary of Pragmatics and numerous papers in
French and English, on pragmatics and/or philosophical subjects. She has
developed an interest in recent years in both language evolution and animal
cognition and communication and has recently published a book on language
and human cognition.