© Copyright JASSS


Thought Contagion: How Belief Spreads through Society

Aaron Lynch
New York, NY: Basic Books
Cloth: ISBN 0-465-08466-4; Paper: ISBN 0-465-08467-2

Order this book

Reviewed by
Paul Marsden
Graduate Research Centre in the Social Sciences, University of Sussex, Falmer, Brighton BN1 9RH.

Cover of book

Social contagion theory is not new. In fact, it has been an established explanation for the clustering and spread of human behaviour for nearly two centuries. In his classic study of suicide at the turn of the century, Emile Durkheim (1897) critically reviewed a century of 19th century contagion theory, from Lucas' Imitation Contagieuse to Despine's Contagion Morale to the then influential and contemporary theory of Tarde (1903). In another landmark study, Le Bon (1895) used contagion as the central mechanism to explain crowd behaviour, and Baldwin (1894) used the concept, qua imitation, as "... the law and the only law of the progressive interaction of the organism and its environment." (Baldwin 1894, p. 54) Both Bagehot (1869) and James (1880) provided an evolutionary dimension this ecological framework by arguing that differential propagation could be accounted for in terms of selection processes. Similar evolutionary-ecological approaches to culture were subsequently (re)-proposed and developed by various authors including Zipf (1935), Lovejoy (1936), Medawar (1959), Siegfried (1960), Sperry (1965) and Monod (1971). More recently, Dawkins (1976) coined the term meme - in preference to Zipf's actemes or Lovejoy's unit-ideas - to denote the unit of focus for such an approach.

The differential replication and selection of non-genetic information (culture) over time has been the specific focus of evolutionary epistemology (Popper 1974, Campbell 1974, Hull 1988, Plotkin 1982), whilst the differential replication of culture through both space and time has been the focus for contemporary innovation diffusion (Rogers 1995) and social contagion models (Levy and Nail 1993, Marsden 1998) of culture. The evolutionary-ecological approach today competes with rational choice theory as the most useful approach for understanding and predicting the spread of human behaviour.

Recently, a number of models exploring different aspects of the evolutionary dynamics and biases of culture have been developed under the broad umbrella of evolutionary culture theory (ECT). These include the coevolution model (Durham 1991), the dual inheritance model (Boyd and Richerson 1985), the cultural transmission model (Cavalli-Sforza and Feldman 1981), the gene-culture transmission model (Lumsden and Wilson 1981), the i-culture and m-culture model (Cloak 1975) and an emerging memetic model (Dawkins 1993, Dennett 1990, 1991, 1995, Plotkin 1994, Marsden 1998, Blackmore 1999).

Although ECT has tended to focus on the ideational aspect of culture, including the epidemiology of representations (Rindos 1985, 1986, Sperber 1994, 1996), a number of alternative conceptualisations of culture have been proposed, including those emphasising behaviour (Langton 1979, Gatherer 1998), norms (Allison 1992), practices (Runciman 1998), rules (Burns and Dietz 1992) and strategies (Marsden 1999).

Enter Aaron Lynch's Thought Contagion: How Belief Spreads Through Society: The New Science of Memes, a book which I hoped would provide an accessible introduction, overview and perhaps an assessment of a part of this exciting emergent field. Lynch's express aim in the book was one of "putting memetics on more serious footing in the sciences". (p. vii)

Thought Contagion is accessible and entertaining, and it has two particular strengths. Firstly, it demonstrates clearly the memetic shift in focus from the internal cognitive processing of social agents, to the identification of patterns of cultural information transmission. Secondly, Lynch shows how such a framework or stance can be used to generate new sets of hypotheses which do not intuitively follow from traditional perspectives. In this way, Lynch does a good job of demonstrating how memetics could be used to generate a novel research programme that is both informed and guided theoretically.

However, Lynch's aim is severely handicapped by his decision to ignore over a century of ECT in particular and social science in general. Instead, Lynch declares that he independently reinvented memetics in 1978 (p. vii) and embarks on a confused and largely uninformed account of how, according to him, ideas about masturbation, breast fetishism, circumcision, psychoanalysis, abortion, firearms, dieting and crucifixion are supposed to spread through society.

In this exposition, Lynch draws on little besides Dawkins' "meme" label. He redefines memes somewhat mysteriously, and without elaboration, as "actively contagious ideas". (p. 2) A modicum of clarification is gained 148 pages later when he states: "Though not exactly a contagion of thought, the addiction to a specific drug can become a replicating brain condition, and hence a meme." (p. 150) Confused? Well, things get even more confusing when Lynch talks of 'ownership' as "a kind of "transcendental meme"" (p. 19) and focuses on what he calls thought contagions, a sort of sub genus of memes, which (we are told) are made up of memes that do not spread through "passive adaptation". (p. 25) For the reader who requires further clarification, Lynch points to an "extremely technical" journal article "filled with propagation diagrams and mathematical equations" on his home page at <http://www.mcs.net/~aaron>. I have read this paper and, perhaps because of my own shortcomings, I am no more enlightened. A review of this paper can be found in Gatherer (1998).

Lynch goes on to suggest seven modes of transmission (see Appendix) which are largely at odds with (and uninformed by) anything in ECT, social psychology or social science in general. Of these seven modes, Lynch focuses overwhelmingly on what he calls "quantity parental" and "efficiency parental" modes for his analysis. The reasoning goes something like this: Belief B causes the holder of B to have more children than average in a particular population. Given effective parental socialisation, this means that the children of the holder of B are more likely to hold B too, and thus the prevalence of B will increase with each generation. For example, Lynch suggests that a masturbation taboo may result in an increase in sexual intercourse, presumably to reduce frustration, producing an increase in fertility, which (given effective belief transmission down genetic lines of descent) results in more offspring inheriting the masturbation taboo. (If you find this sort of reasoning too preposterous to swallow, see p. 4 and p. 90 to check it, and note that this is one of Lynch's more sensible, if unsubstantiated claims.)

The problems with this sort of reasoning are manifold and are indicative of the standard of argumeny throughout the book. Firstly, Lynch provides not one jot of evidence to substantiate the proposed correlation between masturbation taboo holders and fertility rates. In fact, the entire book is devoid of evidence which substantiates any of his wild speculations. Even if there were a correlation between the taboo holders and fertility, an elementary knowledge of social psychology should have suggested to Lynch that publicly expressed beliefs and private behaviour are often completely at odds with each other (LaPiere 1934). Beliefs and attitudes are generally rather poor predictors of behaviour and have been largely abandoned by predictive models in favour of intentions, within an overall logic of utility maximisation (Petty and Cacioppo 1981). So, for Lynch's central argument to hold, he would first have to show that holding a masturbation taboo was predictive of masturbation behaviour and then show that masturbation and fertility were inversely correlated. He does neither. Furthermore, Lynch's argument would only hold if there was evidence of this correlation and if effective parent to child belief transmission could be demonstrated. In the case of sexual mores and norms (which change radically between generations), it is naïve in the extreme simply to assume that this is the case. In fact, Lynch ignores the hugely important effect of belief transmission between genetic lines of descent, through schooling, peer groups, the mass media and other institutions. The net result of this approach is a tower of unsubstantiated supposition often producing quite incredible conclusions.

To be fair, Lynch does recognise other cases in which beliefs are transmitted between genetic lineages, and he introduces what perhaps is the most interesting concept in the book, that notion that thoughts have a specific "propagative profile" made up of the modes by which they are transmitted. Unfortunately, Lynch does not develop or exploit this theme, choosing instead to continue with a series of examples, including "girlish helplessness" and "family-man finder" memes that challenge the credibility of the book. For example:

Now there is nothing wrong in using a memetic or any other framework to generate hypotheses, however wild, but they are of no value until evidence is provided to support them. But even in the absence of evidence, Lynch's speculations seem to be devoid of any rationale. For example:

The problem I have with all this uninformed speculation is not that there is the slightest risk that anybody could possibly take it seriously, but rather that it might be construed as representative of work that is being done in ECT and memetics, because it is not. Fortunately, suspicion is likely to be aroused in even the most uninitiated reader when Lynch includes a discussion of the science fictional "psychohistory" in his brief misrepresentation of the various social sciences, in the process of claiming that his brand of memetics is "the missing link".

I have argued elsewhere (Marsden 1999), as have others (Gatherer 1998, Flinn 1997) that, when taken literally as Lynch does, the thought contagion metaphor is quite disingenuous. I will not repeat the arguments here, suffice it to say that even if thoughts are anything other than a heuristic description in folk psychology, we know nothing about how they combine or recombine and it is certainly not ignoring cognitive processing and endowing thoughts with contagious properties that makes the memetic approach any more viable. When Lynch conflates the partial metaphor of thought contagion with population genetics by arguing that "Population memetics is the study of how proliferating memes combine and separate in a population", which "roughly parallels" population genetics (p. 12), his whole approach becomes even more confused.

What, for example, is it about the "mandatory breast fetish meme" that qualifies it as a meme? Just in what sense is it "actively contagious"? Or how is it a replicating brain condition? We know that folk psychological concepts do not match specific brain states; there is no breast fetish brain neuron or state that replicates and is identical across different brains. How does this meme recombine and perhaps then separate from a 'non mandatory nipple revulsion meme'? If they actually exist at all, thoughts and their 'meanings' are certainly contextually and subjectively dependent (Heyes et al. 1993, passim), and this fact alone strongly suggests that there can be no laws of combination and separation.

Does Thought Contagion demonstrate that an epidemiological or evolutionary-ecological approach to understanding non-genetically transmitted information (culture) is ultimately redundant, vacuous or false? The answer, I suggest, is no simply because exposure to culturally transmitted information clearly does influence the likelihood of an agent (re)-transmitting the information in that or another form. Call it social, vicarious or observational learning, replication or imitation, the fact is that our behavioural strategies are in part determined by the behavioural strategies of those around us. (See Marsden 1998 for a review.) From this simple observation, it is possible to adopt an alternative stance, a memetic stance, for interpreting, understanding, explaining and predicting human behaviour. Just as it is possible to interpret human behaviour as if it emanates from a homuncular centre of intentionality, our propensity to copy behavioural strategies, and the behavioural clustering that results, means that it is sometimes useful to understand non-genetic information transmission as if it were a contagion. By this, all I mean is that our production of behaviour is sometimes dependent on the extensity and intensity of our exposure to that behaviour, or at least some representation of that behaviour.

By understanding the social agent as differentially selecting and replicating the information it is exposed to, models such as the Bass equation (Bass 1969) can be used to predict and model the spread of behaviours and products through a population without invoking or relying on homuncular folk psychology. Such an approach does not ignore cognitive processing, it simply shifts the focus to the description of how one subset of non-genetically transmitted input information is selected for output over another.

Of course, social agents do not blindly and directly imitate all the strategies they are exposed to, as in a simple game of Tit For Tat. Replication is altogether more sophisticated, temporally and contextually cued, and critically, involving the selection of one strategy instead of another. We are highly discriminating and discerning imitators. Indeed, the evolved structure of our brains has produced a remarkable intentional system out of our learning capacities such that we can usually view our behaviour as if it were produced by a principle of utility maximisation. Social agents are pretty good control systems, selecting various strategies to attain antecedently fixed ends.

However, the frame problem precludes optimality in this utility maximisation process, and this means that the intentional stance sometimes breaks down - we sometimes seem to behave irrationally. This, I suggest, is where memetics may be useful. Just as evolutionary psychologists may focus on how biases in strategy selection might result from the evolved cognitive architecture of our brains, memetics might inform us of how the differential exposure to culture influences strategy selection. Just as the evolved architecture of the brain may produce an intentional system with certain biases towards inclusive genetic fitness enhancement (within the environment of evolutionary adaptedness), a memetic stance might legitimately invoke an analogous heuristic of inclusive memetic fitness enhancement (within a particular context) to explain non-rational strategy selection. Because differential exposure to stimuli impacts upon how we respond to situations (by increasing the likelihood of replicating that stimuli), the trajectory and evolution of cultural information through a cultural environment will appear as if the information "wants to get itself" replicated.

The principle of inclusive memetic fitness enhancement is simply a heuristic (as opposed to an ontological) claim for interpreting the production of culture as if it were an attempt to enhance the survival chances of that culture, that is, as if the culture itself had a rudimentary intentionality. Crudely speaking, rather than always behave selfishly for what I personally may want (whatever that may mean), I will fight for my ideas and help others who share my ideas. Just as I may lay down my life for more than 2 brothers or more than 8 first cousins, I might also lay down my life for more than 2 brothers in arms, or more than 8 individuals with whom I share more than an eighth of my culture. Note that this memetic stance is neither incompatible with the intentional stance, nor with folk psychology; it simply provides a different stance for analysing human interaction, one that adumbrates certain features of culture and clarifies others. In sum, what I am proposing is that the memetic stance may provide a useful heuristic for interpreting non-rational behaviour, i.e. when the intentional stance breaks down. This memetic stance of course needs to be tested, but if it turns out to be useful in modelling and predicting behaviour, integrating it with its sociobiological counterpart may one day provide a real alternative to homuncular folk psychology.

Theorists often speak in terms of architecture. They describe their work in terms of constructing theoretical edifices and laying theoretical foundations. This is useful language, it highlights the laborious step by step nature of theory building, where each move forward, whilst a tentative and speculative move into the dark, is both theoretically and empirically informed by what has gone before it, and is designed so as to be open to testing. If Aaron Lynch's intention in writing Thought Contagion was to put memetics on more serious footing in the sciences, in this he must be seen to have failed quite spectacularly. Fortunately, memetics has somewhat more secure foundations, as exemplified in the recent publication of Sue Blackmore's The Meme Machine (Blackmore 1999), which I commend to you as a useful contribution to ECT.

* Appendix: Lynch's Seven Modes of Memetic Transmission

* References

ALLISON P. 1992. The Cultural Evolution of Beneficent Norms, Social Forces, 71:279-301.

BAGEHOT W. 1887 [1869]. Physics and Politics or Thoughts on the Application of the Principles of 'Natural Selection' and 'Inheritance' to Political Society, eighth edition, Kegan Paul, Trench and Company, London.

BALDWIN J. M. 1894. Imitation: A Chapter in the Natural History of Consciousness, Mind, 3:25-55.

BASS F. M. 1969. A New Product Growth Model for Consumer Durables, Management Science, 15:215-227.

BLACKMORE S. 1999. The Meme Machine, Oxford University Press, Oxford. [JASSS review]

BOYD S. and P. J. Richerson 1985. Culture and the Evolutionary Process, University of Chicago Press, Chicago, IL.

BURNS T. R. and T. Dietz 1992. Cultural Evolution: Social Rule Systems, Selection, and Human Agency, International Sociology, 7:259-283.

CAMPBELL D. T. 1974. Evolutionary Epistemology. In P. A. Schilpp, editor, The Philosophy of Karl Popper, The Library of Living Philosophers Volume 14, Book 1, Open Court, La Salle, IL, 413-463. [Reprinted in Plotkin 1982.]

CAVALLI-SFORZA L. L. and M. W. Feldman 1981. Sociocultural Transmission and Evolution: A Quantitative Approach, Princeton University Press, Princeton, NJ.

CLOAK F. T. 1975. Is a Cultural Ethology Possible? Human Ecology, 3:161-182.

DAWKINS R. 1976. The Selfish Gene, Oxford University Press, Oxford.

DAWKINS R. 1993. Viruses of the Mind. In B. Dahlbohm, editor, Dennett and His Critics, Basil Blackwell, Oxford, 13-27.

DENNETT D. C. 1990. Memes and the Exploitation of Imagination, Journal of Aesthetics and Art Criticism, 48:127-35.

DENNETT D. C. 1991. Consciousness Explained, Penguin, Harmondsworth.

DENNETT D. C. 1995. Darwin's Dangerous Idea: Evolution and the Meanings of Life, Allen Lane, London.

DURHAM W. H. 1991. Coevolution: Genes, Culture and Human Diversity, Stanford University Press, Stanford, CA.

DURKHEIM E. 1952 [1897]. Suicide: A Study in Sociology, Routledge, London.

FLINN M. V. 1997. Culture and the Evolution of Social Learning, Evolution and Human Behavior, 18:23-67.

GATHERER D. 1998. Why the Thought Contagion Metaphor is Retarding the Progress of Memetics, Journal of Memetics - Evolutionary Models of Information Transmission, 2, <http://www.cpm.mmu.ac.uk/jom-emit/1998/vol2/gatherer_d.html>.

HEYES S. C., L. J. Heyes, H. W. Reese and T. R. Sarbin, editors, 1993. Varieties of Scientific Contextualism, Context Press, Reno, NV.

HULL D. L. 1988. Science as a Process: An Evolutionary Account of the Social and Conceptual Development of Science, University of Chicago Press, Chicago, IL.

JAMES W. 1880. Great Men, Great Thoughts, and the Environment, The Atlantic Monthly, October, 46(267):441-459.

LANGTON J. 1979. Darwinism and a Behavioural Theory of Sociocultural Evolution: An Analysis, American Journal of Sociology, 85:288-309.

LA PIERE R. T. 1934. Attitudes Versus Actions, Social Forces, 13:230-237.

LE BON G. 1903 [1895]. The Crowd: A Study of the Popular Mind, T. Fisher Unwin, London.

LEVY D. A. and P. R. Nail 1993. Contagion: A Theoretical and Empirical Review and Reconceptualization, Genetic, Social and General Psychology Monographs, 119:235-183.

LOVEJOY A. O. 1936. The Great Chain of Being: A Study of the History of an Idea, Harvard University Press, Harvard, MA.

LUMSDEN C. J. and E. O. Wilson 1981. Genes, Mind and Culture, Harvard University Press, Harvard, MA.

LYNCH A. 1996. Thought Contagion: How Belief Spreads through Society: The New Science of Memes, Basic Books, New York, NY.

MARSDEN P. S. 1998. Memetics and Social Contagion: Two Sides of the Same Coin? Journal of Memetics - Evolutionary Models of Information Transmission, 2, <http://www.cpm .mmu.ac.uk/jom-emit/1998/vol2/marsden_p.html>.

MARSDEN P. S. 1999. A Strategy for Memetics: Memes as Strategies, Journal of Memetics - Evolutionary Models of Information Transmission, 3, <http://www.cpm.mmu.ac.uk/jom-emit/1999/vol3/marsden_p.html>.

MEDAWAR P. 1959. The Future of Man, The BBC Reith Lectures. [Transcript ]

MONOD J. 1971. Chance and Necessity, Alfred E. Knopf, New York, NY.

PETTY R. E. and J. T. Cacioppo 1981. Attitudes and Persuasion, William C. Brown Company, IA.

PLOTKIN H. C., editor, 1982. Learning, Development, and Culture: Essays in Evolutionary Epistemology, John Wiley and Sons, Chichester.

PLOTKIN H. C. 1994. Darwin Machines and the Nature of Knowledge, Penguin, Harmondsworth.

POPPER K. R. 1979 [1974]. Objective Knowledge: An Evolutionary Approach, revised edition, Clarendon Press, Oxford.

RINDOS D. 1985. Darwinian Selection, Symbolic Variation and the Evolution of Culture, Current Anthropology, 26:65-88.

RINDOS D. 1986. The Evolution of the Capacity for Culture: Sociobiology, Structuralism, and Sociocultural Selectionism, Current Anthropology, 27:315-332.

ROGERS E. M. 1995. Diffusion of Innovations, fourth edition, The Free Press, New York, NY.

RUNCIMAN W. G. 1998. The Selectionist Paradigm and Its Implications for Sociology, Sociology, 32:163-189.

SIEGFRIED A. 1960. Germs and Ideas: Routes of Epidemics and Ideologies. [Partial transcript]

SPERBER D. 1994. The Modularity of Thought and the Epidemiology of Representations. In L. Hirschfied and S. Gelman, editors, Mapping the Mind, Domain Specificity in Cognition and Culture, Cambridge University Press, Cambridge, 39-68.

SPERBER D. 1996. Explaining Culture: A Naturalistic Approach, Basil Blackwell, Oxford.

SPERRY R. 1965. Mind, Brain and Humanist Values. In R. Platt, editor, New Views on the Nature of Man, University of Chicago Press, Chicago, IL, 71-92.

TARDE G. 1963 [1903]. The Laws of Imitation, Peter Smith, Clouchester, MA.

ZIPF G. K. 1965 [1935]. Psychobiology of Languages, The M.I.T. Press, Cambridge, MA.

ButtonReturn to Contents of this issue

© Copyright Journal of Artificial Societies and Social Simulation, 1999