CROTCHETS & QUIDDITIES

Good Vibrations: The Silent Symphony of LifeKENNETH WEISS

Look at the life around you. Every-where, you are confronted with mod-ularity and repetition, from the leavesand branches on plants, to your hairand the skeletal and other structureswithin you. Modularity, segmenta-tion, and repetition are, like measuresand tones in music, the way in whichthe living opus has been assembled bythe composing processes of evolution.In his elegant book Evolution Emerg-ing, W.K. Gregory (1951) likenedmodular (“polyisomeric”) organiza-tion to “the notes in an octave . . . evo-lution emerging has involved an infi-nite number and variety of naturalpolyisomeres in both space and time.”This has long been known, but littleanalyzed, in modern evolutionaryterms until very recently. To continuethe musical analogy, we can view pat-terned structures as a harmony oforganization, with many parts inte-grated to form an organism, the waythe violins, horns, flutes, and so on,form an orchestra. In fact, old ideasand new facts suggest that musicalsimiles are relevant to the symphonyof life.

Soon after Darwin published his Or-igin of Species, opponents of his viewsbegan marshaling their evidence. One

of the first prominent biologists to as-semble objections was St GeorgeJackson Mivart (1870), previously re-spected by and on good terms withDarwin, Huxley, and the rest of thebiological “in” crowd. The main issuewas whether adaptive natural selec-tion could explain species variationand evolution. Mivart so irritated Dar-win that he responded at great lengthin the 6th edition of Origin. Some ofMivart’s objections involved the na-ture of inheritance and homology inregard to discontinuous variation,such as repetitive, serially homolo-gous structures. Mivart became a betenoir who was ostracized by the Dar-winian community then becoming themainline of biology. Mivart also triedto reconcile evolution and Catholi-cism, and was excommunicated fromthat church, too, poor fellow. But histheme was picked up by W.K. Brooksin the US who credited Mivart andthen by William Bateson (1892, 1894,1913, see Webster, in Bateson, 1894;Webster and Goodwin, 1996).

Bateson coined the term “genetics”and was a strong promoter of Mende-lian genetics, but he did not think thatthe darwinian evolution by gradualnatural selection could generate spe-cies or their diversity (e.g., Bateson,1913). He pointed out problems in in-terpreting modular (repetitive, seri-ally homologous, or meristic) struc-tures: “Segmentation . . . is almostuniversally present . . . greater or lessrepetition of various structures is oneof the chief factors in the compositionof animal forms.” (Quoted in Web-ster’s preface to Bateson, 1894). Butdiscrete segment numbers cannot

evolve incrementally. What is actuallyinherited? A different gene for eachhair or vertebra? That made no senseto him, and it may have been a reasonfor the non-specific suggestion a cen-tury later that evolution was a “punc-tuated” process.

One of his ideas about this1 is strik-ingly modern in concept. The sugges-tion drew biological parallels withconcepts of fields and vibrations oroscillations borrowed from physics.As we’ll see, Bateson used a musicalanalogy that makes some evolution-ary genetic points easy to understand.There were then no actual genesknown and his views were ignored oreven ostracized during the decades ofascendance of the neodarwinian syn-thesis. But the basic ideas are enjoy-ing a justified revival in a profoundadvance in our understanding of therole of genes in the evolution and gen-eration of complex morphologies. Wecan trace these ideas, at least fanci-fully, back to the famous Geoffroy-Cuvier debates in 1830 on the natureof animal form itself, and perhapseven push the musical analogy back tothat time as well.

A FEW BASIC QUESTIONS

How Fundamental Is ModularPatterning?

It’s worth stressing how very modu-lar structure is in the construction oflife. If the canonical essence of evo-lution is descent with modificationamong individuals across genera-tions, to me an equally importantprinciple is duplication with variationwithin the individual. From DNA toproteins to morphology, among plantsand animals (even bacterial colony

Ken Weiss is Evan Pugh Professor of An-thropology and Genetics at Penn Univer-sity.

Evolutionary Anthropology 11:176–182 (2002)DOI 10.1002/evan.10034Published online in Wiley InterScience(www.interscience.wiley.com).

1Bateson’s reference to Chladni figures was mademost clearly in a letter to his sister (see B. Bateson,1928).

A century ago a leading biologist suggested that repetitively structured biolog-ical traits resembled interference patterns used in tuning violin plates. Are suchideas in tune with modern biology?

176 Evolutionary Anthropology

formation), from cells to ants in a col-ony, there is modular organization ev-erywhere. This is built by gene dupli-cation to form the genome, proteinpolymerization or multimerization toform functional biochemical units,and at the morphological level, begin-ning with cells themselves.

Across the plant and animal world,if you look at structures, organs, ororgan systems you will see how com-monly they are built of repeatedsubunits (themselves sometimeshierarchically modified). To makethis happen, sets of differently pro-grammed cells are produced in peri-odic or episodic fashion, and becomethe precursors of each unit within thestructure. In this way, duplicationwith variation has allowed life to be-come complex, and organisms toachieve larger size and functional spe-cialization.

Recognition of modularity is thor-oughly built into modern biology. Weroutinely search for gene family mem-

bers to explain comparable structureswithin an organism and for repeatedregulatory sequences around genesexpressed in similar cellular contexts.Wavelike patterning mechanismshave been a widely observed phenom-enon in development.

How Does Modular StructureRelate to Genes?

Bateson’s “vibratory” theory of pat-terning was that repeated structureswere manifestations of sympatheticvibration or similar interferencephenomena. His imagery was that ofChladni figures. These are the wave-like interference patterns that formwhen a source of oscillating energydiffuses through a material of somesort (Waller, 1961). The pattern is afunction of the location, frequency,and energy level of the source of vi-bration. Ernst Chladni was a Leipziglawyer, musician, and amateur scien-tist. As early as 1787, he had reported

a way to make the vibrations causedby sound waves visible. He coveredglass, metal, and wooden plates withsand and ran a violin bow againstthem. The vibrations moved the sandinto patterns that are known today as“Chladni’s figures.” The vibration jos-tles the powder to areas or nodes ofthe plate in which vibration wavescancel each other out and there’s nonet motion.

The idea of using a tuning fork asa source of oscillating energy sug-gested an interesting way to illus-trate the phenomenon. A physiciannamed Felix Savart was interested inapplying Chladni’s notions to thisproblem, by dusting the plate withblack powder, applying a tone to theplate and observing the interferencewaves or nodes, as shown in Figure3. Plates of great violins have consis-tent patterns within, and consistentdifferences between them. Makers ofgreat violins “tune” the top and backplates by shaving small amounts ofwood here and there until appropri-ate frequencies generate the stan-dard “modes.” An important point tonote is that the same plate has dif-ferent patterns if the energy, loca-tion, or frequency of the source ischanged.

In 1952, Alan Turing (known for de-ciphering the Nazi cryptographic sys-tem during World War II and ideas ondesigning programmable computers)suggested that two interacting sub-stances diffusing through a uniformfluid could generate wavelike interfer-ence patterns. One substance isknown as an activator, and diffusesfrom some source, inducing its ownactivity as well as that of the secondsubstance, known as an inhibitor. Insuch a “reaction-diffusion” process,the inhibitor reduces the level of theactivator. Depending on the produc-tion and diffusion levels of the twosubstances, and their interaction dy-namics, an initially uniform area or“field” will generate wavelike patternsof high and low levels of the activator.

Figure 1. William Bateson in youth A; and hisinterests, B. Supernumerary premolar inupper but not lower jaw of Ateles mon-key. (sources: (A) http://post.queensu.ca/�forsdyke/bateson1.htm# with permissionof Donald R. Forsdyke; (B) from Bateson,1894).

(Virtual) Figure 2. Modular structure in life:left-right symmetry, hair, pores, teeth, cusps,papillae on the tongue, . . .

CROTCHETS & QUIDDITIES Evolutionary Anthropology 177

Turing did not acknowledge thatBateson had thought of this in regardto biology, but several biologists no-ticed the relevance of Turing’s ideas todiverse repetitive or wavelike biologi-cal traits. Bateson (1913) referred to

the nodes and internodes as “the seatof appropriate and distinct chemicalprocesses leading to the differentia-tion of the parts” of organisms. Struc-tures can develop at the peaks, for ex-ample, if expression of developmental

cascades occurs when activator levelscross some threshold, with the valleysthe structure-free inhibition zones.

In recent years, the ability to detectcell-specific levels of expression ofspecific gene products (proteins ormRNA) has put these ideas to directtests. Embryonic tissue is tested forspatiotemproal distribution of signal-ing factor molecules diffusing fromcells of origin across the tissue. Ini-tially broad patterns narrow to stripesand then periodic spots surroundedby expression of diffusible inhibitors.For example, hair and feathers de-velop in zones expressing the Fgf andWnt signaling factors, but are inhib-ited in areas where Bmp factors arefound. These gene products havepatchy distribution that presages thelocation and spacing of future feath-ers, sometimes appearing first as aninitial line of expressing cells (e.g., thedental lamina), that resolves intospots, and then spreads laterally toform other spots. Short distance spac-ing among adjacent cells may be pat-terned by additional activator-inhibi-tor interactions (e.g., the Notch-Deltasystem). These are highly conservedprocesses; for example, the sameNotch-Delta signaling helps patternthe ommatidia in insect eyes, teeth,and feathers. These patterning sys-tems then activate “selector” genesthat initiate cascades of gene expres-sion that lead to organ formation(see Gilbert and Gerhart references).

In addition to direct molecularexamples, activation-inhibition pro-cesses have been shown by computersimulation to apply in varying ways tocoloration in mammals, fish, sea-shells, butterfly wings, the location ofmammary glands, feathers, scales,and digits in animals and flower parts.This is probably just the first peek atthe wide distribution of such pattern-ing mechanisms in nature.

The concepts are highly relevant toanthropology. My own interest is indental patterning, in which it seemslikely that the number, location, anddifferential morphology of teeth alongthe jaws are due to such mechanisms.Indeed, the same genes demonstrablyexpressed in this way in featherand other vertebrate patterning pro-cesses are expressed in teeth in waysconsistent with activation-inhibition

Figure 3. A: Chladni’s figures shown here are 12 engravings of these acoustically producedpatterns (Source: Chladni, 1787, copied from http://www.sil.si.edu/Exhibitions/Science-and-the-Artists-Book/phys.htm). B: Cross section of vibrating plate to show nodes (N) of nomotion. C: Chladni figures of the 7 classical “modes” of a handmade violin (tonal frequen-cies 91, 138, 196, 231, 306, 312, 392 Hz, respectively). (Source: J. Wolfe, http://www.phys.unsw.edu.au/�jw/chladni.html)

178 Evolutionary Anthropology CROTCHETS & QUIDDITIES

patterning. (Jervall and Jung, 2000;Weiss et al., 1998). In elegant experi-ments that I wish I had done ratherthan they, Jukka Jernvall, Soile Ker-anen, and others in Irma Thesleff’s labin Helsinki have shown that gene ex-pression and expression-manipula-tion are consistent with this generaltype of patterning process. Zonesknown as enamel knots expressing Fgfand other growth factors appear alongthe dental lamina; cusps form whenthis induces down-growth in sur-rounding tissue, while the knots them-selves are self-inhibiting (Fig. 5B). In-deed, an earlier round of patternedexpression of similar genes appears tobe involved in the serially patternedlocation of tooth germs.

Salazar-Ciudad and Jernvall havesubsequently shown by computersimulation that activation-inhibitionprocesses modeling the behavior ofenamel knots can generate strikinglyrealistic molariform cusplike crownpatterns (Salazar-Ciudad and Jern-vall, 2002; Fig. 5). In an importanttest, the authors tried the darwiniangradualism experiment of simplycomputer “morphing” a tooth from itsinitial to its final states (both of whichmatched between program and actualtooth embryos), and compared the re-

sulting intermediate states with the in-termediate states of real tooth germsthat are closely imitated by the dy-namic process simulation (J. Jernvall,personal communication). As far asthe morphed intermediates go, artdoes not imitate life. Only the dy-namic patterning process resembledthe real thing.

Bateson could only speculate aboutsuch matters, but a century later wehave at least some examples vindicat-ing the idea at the molecular level.Dynamic patterning mechanisms likethis seem to a ubiquitous means bywhich complex animals and plants areproduced from single fertilized eggcells. Of course, to credit Bateson forhis prescience does not mean that wewould expect to see his—or Turing’s,or anyone’s—ideas fulfilled exactly,and repetitive patterning is broughtabout in many ways.

How Does Modular StructureEvolve?

A long-standing question of the “tem-po and mode” of evolution has been theway complex traits evolve. Argumentsabout this pitted darwinian gradualismversus saltational evolution. As notedearlier, this problem bothered Batesonand others and kept them from adopt-

ing Darwin’s solution to the speciesquestion. In the 20th century, RichardGoldschmidt (1940) suggested that mu-tational hopeful monsters may fromtime to time be produced that have newstructures or traits. On rare occasions,these might pass the selective screen asadaptations to a new form of life. Sal-tational evolution was and is generallyheld to be a serious kind of heresy (ormadness). Nobody thinks wings canevolve suddenly from legs, but pat-terned organ systems do normally varyin their number and morphology of el-ements (teeth, cusps, feathers, limb-bones, regions of the gut, etc., Figure1B). This could address Bateson’s ob-jection that neither darwinian gradual-ism nor simple mendelian segregationwere consistent with the evolution ofmeristic traits. As Bateson (1928) said,an eight-petal form is to a four-petalform as one octave is to another. A mu-tationally derived shift in the dynamicsof an interaction process can bringabout such differences, and there aremany examples.

Recognition of dynamic patterningprocesses has a second important im-plication for evolutionary thinking. AsBateson (1894) said, “Of course, he-redity becomes quite a simple phe-nomenon in light of this.” There is nosingle gene “for” a specific feather ortooth or intestinal villus or nephron.All the elements in each system ex-press essentially the same genes. Inprinciple, all that need differ along thejaw would be the position of the peaksof expression of say odontogenicgenes. Not only need there not be sep-arate genes for each iteration of astructure, but what may differ be-tween the dentitions of carnivores andherbivores may mainly, or only, in-volve the interaction dynamics—things like diffusion rates, activationor inhibition efficiencies—of the samepatterning genes.

What About Homology?

All of this affects notions of homol-ogy (see Abouheif, 1997; Hall, 1999).It may be evolutionary non-questionsto ask whether the diastema in amouse jaw represents “missing” teeth,or what the homologous digits are in abird and mammal. Homology may re-ally lie in the patterning process it-self—a common generative mecha-

Figure 4. Peaks erupting from a simple reaction-diffusion process. Wire-screen is concentra-tion of activator (this could be proportional to growth, for example), grey-scale of inhibitor(courtesy Brian Lambert).

CROTCHETS & QUIDDITIES Evolutionary Anthropology 179

nism, in Bateson’s terms. The pattern-ing process and genes may be shared,and the patterning genes may be thesame, but there may be no specificgene or gene combination that is aparticular member and can form thebasis of taxonomic homologies. An ex-ample may be the transitory dentalrudiments that appear in the (upperbut not lower) diastema region of anembryonic mouse jaw (Keranen et al.,1999), but are not obviously identicalto the teeth absent in the mouse.

It seems likely that one problemthat taxonomy has is its tacit relianceon a gene-for notion of the relation-ship between genes and traits. Suchassignments may not explain dynam-ically patterned serially homologoustraits, of which there are so many. We

have the same problem assigning ho-mology to members of gene familiesthat have undergone repeated dupli-cation, such as globin or photore-ceptor genes between birds andmammals. Should we call teeth andfeathers or insect eyes homologousbecause at some point in their devel-opment the same patterning circuitryis used?

The Hairs on Your Head MayBe Numbered, But WhatAbout the Wave?

Dynamic patterning processes aretypically nested. For example, hairsare individually periodic structures,but pelage is also striped in mam-

mals—including humans (Figure 6).2

Hair form is regionally differentiatedand individual hairs can themselvesbe striped. Thus, hair is patternedby several simultaneously occurringwave-generating processes. Murray(1993) provides a Chladni-like simula-tion of haircolor patterning in justthat way.

Similarly nested patterning is seenalong the mammalian dentition, infeathers, intestinal epithelial struc-ture, tongue, limb, and vertebral pat-terning, bristles in various parts of afly, and many others. As in tooth de-velopment, the same activation-inhi-bition process may be involved inmultiple traits at different times orplaces in the same organism.

THE MUSIC OF EVOLUTION

Chemical “vibration” is harmoniousto the organism and has propertiessimilar to those of music. Ideas aboutthis relate to age-old debates in biol-ogy. These ideas were applied to pale-ontological, comparative, and embry-ological data, and they may even havebeen relevant to early 19th century dis-cussions of animal form. In 1830,Georges Cuvier and Etienne GeoffroySt. Hilaire held famous debates inParis (Appel, 1987). Cuvier believedthat complex animal form was due toindependent adaptation (at that time,in the functional, not darwinian,sense of the term) of different bodystructures, such as limbs, claws, teeth,and so on. Geoffroy held that varia-tions among animals reflected under-lying body plans that could be dis-torted and modified but representedfunctional wholes.

These were bitter and celebrated de-bates. A new edition of Chladni ap-peared in 1830, and I have speculated(Weiss et al., 1998) that Felix Savart,as a member of the Paris intellectualset, may have attended the debates.Might he have seen the similaritiesbetween biological patterning and theway Chladni figures demonstrate vari-ation on an underlying plan (the basicstructure of a violin plate), and sidedwith Geoffroy?

Ideas about serial homology werearound for most of the 20th century,

2I can personally attest that in their younger years,these stripes were clearer.

Figure 5. Simulation of tooth crown shape by a class of reaction-diffusion-like mechanism. A:Basic model of simulated spatial units showing activator (A) and inhibitor (I) relative toenamel knot (dark) and differential cell growth. B: Predicted (simulated) and observedmouse (left) and vole (right) first molar crown surfaces (with permission from Salazar-Ciudadand Jernvall, 2002).

180 Evolutionary Anthropology CROTCHETS & QUIDDITIES

though suggestions as to cause did notget very far, and were often deni-grated as untestable, or unscientific.But that was then and this is now.Dynamic patterning processes are afundamental part of the developmen-tal toolkit, that are used and re-used,and deeply conserved phylogeneti-cally. But it is tempting to over-extend

the implication of these generalities.Similar genes are used in diversetraits, but this does not mean that thesame genes will be used in any partic-ular trait, or even in the same traitamong different species. Two logicallysimilar patterning mechanisms caninvolve different, perhaps entirely un-related sets of genes. For example,similar periodic patterning processesare involved in plants, but not withanimal signaling factors.

Interestingly, a recent paper has ex-amined various patterning differencesbetween teosinte and maize (Lauterand Doebley, 2002). A long-standingquestion is how these traits, that areinvariant within each species, couldhave evolved. Did it involve a Gold-schmidt leap? We can’t say, but Lau-ter and Doebley have shown by clevercross-breeding experiments that thereis variation within teosinte that doesnot lead to trait variation, but thatcould be the latent source of rapidevolution of the trait. A small amountof mutational change might have suf-ficed to reconfigure this silent back-ground variation to jump teosinte tomaize form, creating this most impor-tant cultivar—just as a Chladni figurecan jump when the sound frequencychanges. The developmental timing ofa single gene can change the shape ofa leaf from simple to complex(Bharatham et al., 2002).

The musical analogy is probablynot a bad one. Throughout an em-bryo as it develops, a highly orches-trated program of “vibrational” pat-terning mechanisms enables a single

cell to become a harmonious com-plex organism, with tones and over-tones. This is a developmental or-chestra that all organisms can play,but none can hear.3 We are probablyjust beginning to discover the degreeto which we owe our nature to thegood vibrations of this silent sym-phony of life.

NOTES

I welcome any comments on thiscolumn: kmw4@psu.edu. I maintain aCrotchetyComments page at www.anthro.psu.edu/rsrch/weiss_lab.

I thank Anne Buchanan and JukkaJernvall for helpful comments.

TO READ

Most things discussed here can beprofitably explored by web searching.

Abouheif E. 1997. Developmental genetics andhomology: a hierarchical approach. Trends EcolEvol 12:405–408.Appel T. 1987. The Cuvier-Geoffroy debate. NewYork: Oxford Press.Bateson B. 1928. William Bateson, F.R.S. natu-ralist. Cambridge: Cambridge University Press.Bharathan G, Goliber T, Moore C, Kessler S,Pham T, Sinha N. 2002. Homologies in leaf forminferred from KNOXI gene expression during de-velopment. Science 296:1858–1860.Bateson W. 1894. Materials for the study of vari-

3DNA sequences can be made into much morethan a 4-tone exercise in minimalism, but transla-tions to date are more aesthetic than biological.CDs and software are available; for example http://algoart.com/dnamusic/ or http://education.llnl.gov/msds/music/midi-dna.html or http://www.dnamusiccentral.com/.

Figure 7. Pattern variants in teosinte andmaize that are phenotypically invariantwithin strain but with underlying variation re-vealed by gene mapping in crosses. Arethese differences due to genes for dy-namic-patterning mechanisms rather thanfor specific traits? (from Lauter and Doeb-ley, 2002, courtesy John Doebley).

Figure 6. Variably striped humans.2 A: George Milner, archeologist. B: James Wood, demographer. C: Andrew Clark, geneticist.

CROTCHETS & QUIDDITIES Evolutionary Anthropology 181

ation. Reprinted Baltimore, Johns HopkinsPress, 1992.Bateson W. 1913. Problems of genetics. (Reprint-ed, 1979). New Haven: Yale University Press.Chladni Ernst FF. 1787. Entdeckungen uber dieTheorie des Klanges [Discoveries concerning thetheory of sound] Leipzig.Gerhart J, Kirschner M. 1997. Cells, embryos,and evolution: Toward a cellular and develop-mental understanding of phenotypic variationand evolutionary adaptability. Malden, MA:Blackwell Scientific.Gilbert S. 2000. Developmental biology 6th ed.Sunderland, MA: Sinauer.Goldschmidt R. 1940. The material basis of evo-lution. New Haven: Yale Press.Goodwin B. 1994. How the leopard got its spots.New York: Charles Scribner.Gottlieb T, Wade M, Rutheford S. 2002. Potentialgenetic variance and the domestication of maize.BioEssays 24:685–689.Gregory WK. 1951. Evolution emerging. 2 vols.New York: Macmillan.Hutchins CM. The acoustics of violin plates. Sci-entific American, October 1981, 170–186.

Jernvall J, Jung H. 2000. Genotype, phenotype,and developmental biology of molar tooth char-acters. Ybk Phys Anthropol 31:171–190.

Jung H-S, Francis-West P, Widelitz R, Jiang T-X,Ting-Berreth S, Tickle C, Wolpert L, ChuongC-M. 1998. Local inhibitory action of BMPs andtheir relationships with activators in feather for-mation: implications for periodic patterning. DevBiol 196:11–23.

Keranen S, Kettunen P, Aberg T, Thesleff I, Jern-vall J. 1999. Gene expression patterns associatedwith suppression of odontogenesis in mouse andvole diastema regions. Devel Genes Evol 209:495–506.

Lauter N, Doebley J. 2002. Genetic variation forphenotypically invariant traits detected in teos-inte: implications for the evolution of novelforms. Genetics 160:333–342.

Meinhardt H. 1996. Models of biological pat-tern formation: common mechanism in plantand animal development. Int J Dev Biol 40:123–1345.

Meinhardt H, Grier A. 2000. Pattern formationby local self-activation and lateral inhibition.BioEssays 22:753–760.

Mivart St. George Jackson. 1870. The genesis ofspecies. London: Macmillan.

Murray J. 1993. Mathematical biology, 2nd ed.Berlin: Springer Verlag.

Salazar-Ciudad I, Jernvall J. A gene networkmodel accounting for development and evolutionof mammalian teeth. Proc Natl Acad Sci USA,99:8116–8120.

Turing A. 1952. The chemical basis of morpho-genesis. Phil Trans Roy Soc London, Ser B 237:37–72.

Waller MD. 1961. Chladni figures. A study insymmetry. London: G. Bell & Sons.

Webster G, Goodwin B. 1996. Form and trans-formation. Cambridge: Cambridge UniversityPress.

Weiss K. 1990. Duplication with variation:metameric patterning in evolution from genes tomorphology. Ybk Phys Anthropol 33:1–23.

Weiss K, Stock D, Zhao Z. 1998. Dynamic inter-action and dental patterning. Cr Rvw Oral BiolMed 9:369–398.

© 2002 Wiley-Liss, Inc.

Forthcoming Articles

• Crotchets & Quiddities: How the Eye Got It’s BrainKenneth M. Weiss

• The Behavioral Ecology of the Spectral Tarsier, Tarsius spectrumSharon Gursky

• Selection, Species, and Spandrels: Divergent Views of EvolutionaryTheoryIan Tattersall

• Whither the Evolution of Human Growth and Development?Daniel E. Lieberman

• Practice Random Acts of Aggression and Senseless Acts ofIntimidation: The Logic of Status Contests in Social GroupsJoan B. Silk

182 Evolutionary Anthropology CROTCHETS & QUIDDITIES