Pierre Weiss, with his mysterious molecularfield and Weiss magnetons, who explainedferromagnetism? Or was it, as I believe,Werner Heisenberg, with quantum theory?Laughlin misleadingly accuses two unnamedphysicists of predicting that superconductorsbe limited to below 30 K (the actual figurewas 40 K), when what they said applied spe-cifically to a particular mechanism for whichit is true.

In my experience, which incidentally isgreater than Laughlin’s, underlying causesoften enlighten our conceptual thinking asmuch as precise numbers do, something thatLaughlin seems to deny. After condemningastroparticle types for overemphasizing deepthoughts and broad vistas, he seems to reveala certain measure of ‘particle envy’ and dis-taste for the messy, quarrelsome but absorb-ing ways of doing the sciences in the realworld. What made Bardeen great, as indeedhe was,was his stubbornness and experimen-tal taste,and Laughlin dismisses these values.

Those who devour the work of Greene,or decorate their coffee table with Hawking,will find this book a useful antidote.It shouldspike the interest of those who read thephysics popularizers, although in its person-alized coverage and opinionated style it is suigeneris.My message is this: buy the book. ■

Philip Anderson is in the Condensed MatterPhysics Group, Department of Physics, PrincetonUniversity, Princeton, New Jersey 08544, USA.

Don’t talk to the animalsDoctor Dolittle’s Delusion: Animalsand the Uniqueness of HumanLanguageby Stephen R. Anderson Yale University Press: 2004. 368 pp.$35, £22.50

Neil Smith

Doctor Dolittle was the hero of a series ofchildren’s books written by Hugh Lofting(1886–1947). The doctor’s ability to talk toevery animal in its own language had aseductive appeal that finds current expres-sion in the widespread belief that the com-munication systems of animals, from beesto bonobos, are essentially similar to humanlanguage. So we should be able to learntheir languages or, just as good, they couldlearn ours. Stephen Anderson pours coldwater on this belief, arguing convincinglythat it is a delusion.

He provides a masterly overview of whatis currently known about the communica-tive abilities of a wide range of creatures:the dance of honeybees, the communicativecroaking of frogs, the warning cries of mon-keys, and the remarkable cognitive abilitiesof bonobos and parrots. Much of this is

superficially familiar from other popularaccounts, but Anderson’s synthesis providesilluminating comparisons with the infinitelymore sophisticated resources of human lan-guage, whether spoken or signed. There areundeniable parallels between humans andother animals, but the differences are equallystriking and confirm the view that our lan-guage is qualitatively different from theirs.

Bees famously indicate the direction, dis-tance and quality of sources of pollen to theirfellow workers by means of a ‘waggle dance’,which is often taken to show that they have a ‘cognitive map’ of the local terrain. In ameticulous dissection of the properties ofthis dance,Anderson undermines this claim,showing that the bees’perception of distanceis largely a function of differences in theirvisual experience. When bees are made to flythrough tunnels with visual patterns on thewalls, the distance they indicate correspondsto the complexity of the pattern to whichthey have been exposed. The dance, then,reflects the bees’subjective experience, ratherthan a map of the external world.

We are evolutionarily rather remote frombees; closer parallels to human language andhow we learn it can be found in birds. Cur-rent views of first-language acquisition inchildren treat it as a process of selection,rather than instruction. Humans are bornwith a set of principles known as ‘universalgrammar’ that define the notion of possible

books and arts

702 NATURE | VOL 434 | 7 APRIL 2005 | www.nature.com/nature

The maxim that the best science possesses anartistic grace has probably never been applied to the down-and-dirty world of palaeontology.First there’s all the digging, and then there’s the arguing over how the fragmentary findingsshould be interpreted and slotted into the bigpicture. But a new painting by Damien Hirst, A New and Diminutive Species of Human BeingHas Been Discovered, brings the face of Homo floresiensis (the ‘hobbit’) — one ofpalaeontology’s most iconic recent images — to the gallery.

Photorealism — the painstakingly faithfulreproduction of photographic images on canvas— is a surprising new direction for Hirst,arguably the doyen of the laddish Britart sceneof the 1990s. Better known for pickling animalsinside glass boxes, he has often engaged withscience before. One of his trademark ‘dotpaintings’ travelled to Mars aboard the ill-fatedBeagle 2 lander and, given a smoother landing,would have been used to calibrate the craft’sonboard cameras.

Hirst’s other photorealist creations includeimages of the Iraq conflict, vivisection and a haunting reproduction of a British police anti-drugs campaign poster featuring the gaunt features of a now-dead crack addict. But what inspired him to take on H. floresiensis,

unveiled in Nature last October? “It’s just anexcuse to paint skulls,” he says. Perhaps, butthe hobbit’s discoverers may nonetheless beamused to see the fruit of their labours raised to an art form.

A New and Diminutive Species of HumanBeing Has Been Discovered is part of theexhibition ‘The Elusive Truth’, which can be seen at the Gagosian Gallery in New York until 23 April. Michael Hopkin

Exhibition

Hirst’s hobbit

DA

MIE

N H

IRST

/GA

GO

SIA

N G

ALL

ER

Y

7.4 books 701 MH 31/3/05 5:11 pm Page 702

Nature Publishing Group© 2005

human language and a set ofparameters that characterizepossible variation amonglanguages. The acquisition oflanguage consists largely offixing the values of theseparameters on the basis ofclues in the input. Import-antly,this theory predicts thatthere are logically possible,but linguistically impossible,mistakes that children can-not make.

The parallel with birdsongis striking. Many birds havesongs that develop appropri-ately only after interactionwith conspecifics: the song is partly innate and partlydependent on experience.Nightingales, like children,only make mistakes that correspond to patterns that could occur aspossible song elements for their species, justas children only make mistakes that arelicensed by universal grammar.

We are evolutionarily closer to birds thanto bees, but we are closer still to other pri-mates. Their systems of communication,however, are less similar to human languagethan is popularly supposed. Vervet monkeyshave distinct alarm calls for leopards, eaglesand snakes. These calls can be extended tonew types of threat — humans, for instance— and they are under some degree ofvoluntary control, yet they do not seem to ‘refer’ to the respective animals in the way werefer with our language. The calls can affectbehaviour but not knowledge. Similarly,attempts to teach American Sign Languageto chimpanzees have made it clear that,although human infants read intentions into the actions of others,chimps never do.

To complement his critique of ‘animal lan-guage’,Anderson also outlines what is specialabout human language: in a word, syntax.Our vocabularies are dramatically largerthan those of other animals, and our soundsystems are more complex, but the essentialdesign property of human language is syntax— the way we use combinations of words toconvey meaning. This concept is alien to thecommunication systems of other species.

What animals learn is impressive andtheir cognitive abilities may be remarkable.But they never master anything like a humanlanguage and seem incapable of doing so:the complexity of their grammar is notremotely comparable to ours. This com-plexity is exemplified at length in the book,but two examples should suffice. First, theessence of syntax is recursion: the possibilityof including one sentence inside another adinfinitum.For example: [Anderson discusses

the claim that [many people think that [ani-mals can talk]]]. Second, we all have subtleand consistent intuitions not only of what ispossible in our language, but also of what isimpossible. In [John expects to visit him],“John” and “him” must refer to different people, but consider this: [I wonder who[John expects to visit him]]. Here, “John”and “him”can,but need not,refer to the sameindividual. Judgements such as these have no parallel in the communication systems ofother animals.

Anderson’s elegant book contains a hostof other insights and observations. He con-cludes that, just as the dance of bees, the song of birds and the calls of monkeys areunique to their respective species, so humanlanguage is unique to us. ■

Neil Smith is in the Department of Phonetics and Linguistics, University College London,Gower Street, London WC1E 6BT, UK.

NATURE | VOL 434 | 7 APRIL 2005 | www.nature.com/nature 703

The grizzly truth? Only humans use syntax, so sharing a language with animals is left to films such as Doctor Dolittle.

20T

H C

EN

TU

RY

FO

X/R

ON

ALD

GR

AN

T A

RC

HIV

E

The Royal Society has announced the shortlistfor this year’s Aventis Prizes for Science Books’General Prize, which celebrates the very best inpopular science writing for adults.

Critical Mass by Philip Ball (William Heineman)takes a look at the application of physics to thecollective behaviour of society. “This book isimpressively clear and breathtaking in scope…For anyone who would like to learn about theintellectual ferment at the surprising junction ofphysics and social science, Critical Mass is theplace to start.” Steven Strogatz (Nature 428,367–368; 2004).

Robert Winston’s The Human Mind (BantamPress/Transworld) is an examination of theworkings of our brains for an adult audience,inspired by his recent television series. His book What Makes Me, Me? (Dorling Kindersley),

which takes a wider view of how the human bodyfunctions for a younger audience, has beenshortlisted for the Junior prize.

In The Ancestor’s Tale (Weidenfeld & Nicolson),Richard Dawkins “views species as pilgrimsmarching into the past, joining each othergenetically on a 3-billion-year journey toevolution’s Canterbury: the first ‘replicator’”.Jerry Coyne (Nature 431, 903–904; 2004).

In Matters of Substance (Penguin, Allen Lane),Griffith Edwards presents a lucid account of druguse and control, taking the radical view that theeffect of any drug is just as dependent on thesocial, historical and psychological context as on its chemical structure.

In The Earth (HarperCollins), Richard Fortey“offers a clear, graphic and entertaining

exposition of the manner in which, over an eon,the observed geological phenomena haveachieved their present state. And he forcefullyreminds us that events remotely embedded indeep time may yet be highly relevant asdeterminants for the lifestyles of modern humancommunities.” Gordon L. Herries Davies (Nature428, 697–698; 2004).

A review of Why Life Speeds Up As You GetOlder by Douwe Draaisma (Cambridge UniversityPress), an examination of the nature of memory,will appear in next week’s issue of Nature.

The General Prize judging panel consists of authorBill Bryson, who won in 2004, weather forecasterLisa Burke, Sian Ede,who is a renowned authorityon art and science interactions, neurophysiologistMark Lythgoe, and poet Ruth Padel. The winnerswill be announced on 12 May 2005.

Aventis Science Book Prize shortlist announced

7.4 books 701 MH 31/3/05 5:11 pm Page 703

Nature Publishing Group© 2005