Why do memes die? Paul Bouissac paul.bouissac@utoronto.ca University of Toronto "We have come to the edge of a world of which we have no experience, and where all our pre- conceptions must be recast." D'Arcy W. Thompson (1942:77) Introduction: What is a Meme? It is being increasingly recognized that the notion of "meme" (Dawkins 1976) as a self-replicating information pattern (cultural unit) whose biotic environment is formed by human, and possibly primate[1], brains is one of the most intriguing, even upsetting hypotheses put forward in recent years. Like the more palatable -- because of its inherent functionalism -- notion of "culturgens" proposed by Lumsden and Wilson (1981), this new way of framing semiosis and cultural dynamism in an evolutionary, Darwinian, perspective should be of prime interest to semioticians in as much as it provides a novel approach to the conceptualization of signs and their action. In spite of some efforts to trace the idea back to earlier metaphoric attempts "to define a cultural analogue of the gene" (Costall 1991: 323) or to show that traditional Western philosophy has always tried to account for the dynamism of signs or ideas (Moritz 1990: 5-9), Dawkins' tentative proposal that cultural units of replication may have some features in common with genes has struck many as an entirely novel, counterintuitive perspective deserving serious consideration, mainly in view of the theoretical context provided by The Selfish Gene (Dawkins 1976). Delius (1989), Dennett (1990), Hull (1988), Koch (1986), Moritz (1990), for instance, have devoted considerable efforts to showing how this hypothesis may be much more than a metaphor and may indeed bring about a revolution in the way we conceptualize the dynamics of cultural, mental and intellectual life. This seems all the more interesting since the theoretical link between this hypothesis and earlier, seminal definitions of semiotics can be made without conceptual strain, albeit somewhat unexpectedly: on the one hand, the meme hypothesis matches some important aspects of Peirce's accounts of semiosis and gives a surprising turn to his "perfusion of signs" metaphor; on the other hand, Saussure's definition of semiology as the study of the life of signs in social life (my emphasis) takes up a new meaning if "life" is understood literally, i.e., in its full biological sense, rather than metaphorically as some kind of abstract dynamism. Let us recall the initial, somewhat awkward formulation of the meme hypothesis: "Just as genes propagate themselves in the gene pool by leaping from body to body via sperms or eggs, so memes propagate themselves in the meme pool by leaping from brain to brain via a process which, in the broad sense can be called imitation. If a scientist hears, or reads about a good idea, he passes it on to his colleagues and students" (Dawkins 1976: 206). The term "meme" was somewhat playfully coined after the French "même" for "same" and with phonetic allusions both to "mimetic" or "memory" and to "gene." In the paragraph following this introduction of the new concept, Dawkins quoted the summary that one of his colleagues, N.K. Humphrey, produced in reaction to the reading of an earlier draft of this chapter: memes should be regarded as living structures, not just metaphorically but technically. When you plant a fertile meme in my mind you literally parasitize my brain, turning it into a vehicle for the meme's propagation in just the way a virus may parasite the genetic mechanism of a host cell. And this isn't just a way of talking -- the meme for, say, "belief in life after death" is actually realized physically, millions of times over, as a structure of the nervous systems of individual men the world over (as quoted in Dawkins 1976: 207). In his subsequent publications Dawkins returns to the concept of "meme" (1982: 97-117, 1986: 158, 1989: 322) and indicates attempts made by others to develop his hypothesis (e.g., Delius 1990). Until now, the idea of entities endowed with both genotypic and phenotypic structures evolving in a parasitic relation to the organism whose brains provide them with a vital environment, remains an intriguing and fascinating metaphor which shatters our worldview as if it were an aftershock of Darwinism but it may seem to be too "odd" to be taken at face value. Dawkins himself has been cautious in his subsequent and brief treatments of the "meme" idea, as if he were daunted by the full implications of this emergent theory. However, in the context of the changing epistemological landscape of the 1990s, this reconceptualization is being given new dimensions by a range of recently developed concepts such as computer viruses (e.g., Cohen 1990; Delius 1990) and artificial life (Langton 1992) which provide a quasi-operational context for the speculative manipulation of this new idea (Bouissac 1992). For example, advances in the understanding of self-replication as an emergent property arising from local interactions in fairly simple systems (Reggia et al. 1993) suggest a possible avenue for a formal, canonic description of memes. However, strong resistance to this approach, ranging from amused disbelief to philosophical arguments -- mostly drawing from the long standing controversy concerning intentionality (e.g., Costall 1991) -- has transformed a seemingly playful suggestion with which eccentric minds would toy into a potentially dangerous frontier of science, as if the Darwinian theory of evolution were bracing for a renewed confrontation with the last defense line of a long established worldview order dominated by dualism (Bouissac 1991). This paper proposes to explore further the "meme" hypothesis by raising two heuristic questions: "Why and how do memes die?" and "What can we learn from the way they die about the way they live?" Thus, it will focus on some of the features and processes which may account for the successful replication (conservation) or the extinction (dissipation) of "memes," a phenomenon which requires to be examined from a multi-disciplinary perspective. Along the line of Dawkins, "memes" will be taken here as informational structures, or sets of instructions, endowed with their own dynamism (autopoiesis) and constrained by the same evolutionary laws as genes, but their biological status will be more specifically ascribed to parasitism both in the predatory sense -- as N.K. Humphrey first visualized the idea (see quotation above) -- and the symbiotic sense, along the line of the developing paradigm of symbiogenesis (Khakhina 1992). But a note of caution is in order. In dealing with such problems, it is obvious that the heuristic status of the meme-as-parasite hypothesis should be emphasized even if some scientifically established data and currently accepted theories are brought into the argument. All too often, the humanities and the social sciences have borrowed ideas out of context from the natural and empirical sciences and have generalized them abusively in the form of all-purpose metaphors of little descriptive value, such as the notion of social darwinism or the right/left dichotomization of cerebral functions. This contribution to the exploration of the meme hypothesis tries to avoid such pitfalls by a constant concern for the falsifiability of the propositions it may come to formulate. At the very least the relevance to semiotics and to the social sciences of the issues potentially addressed by the meme hypothesis must be acknowledged, and even if the conceptual exploration developed in this paper proves to be misguided it might help better define the limits and pertinence of this challenge to established wisdom. Memes' Ways of Dying It can be hypothesized that there are at least four causes which can bring about the demise of a meme. These causes should be critically examined in relation to case studies such as the extinction of a language, the discarding of a fashion, the refutation of a theory (in the range of macro-memes), and the phasing out of a word, a habit, a belief (in the range of micro-memes). However, as a first step in this direction, these causes will be considered here ina purely speculative manner. The first cause is a direct result of the parasitic mode of survival of memes; the other three are not specific to memes but apply to memes as to any other organism submitted to the laws of evolution by selection, although they can be expressed in a meme-specific manner. These various models of extinction will now be examined serially. In as much as memes depend on other organisms for sustenance and reproduction (replication), their survival is bounded by the life-cycle and reproductive success of their hosts. As bearers of brains which afford the conditions necessary for memes to spread through replication, Homo sapiens is a vital resource which is itself dependent upon predation and other environmental commodities for the species' evolution and maintenance. Moreover, Homo sapiens' biological mode of survival involves a high degree of motility and dispersal. Therefore, memes can become extinct for the same reasons as parasites. An examination of the life conditions of parasites thus sheds some heuristic light on memes' modalities of existence. Typically, parasites are dependent upon patchy and ephemeral resources because the other organisms -- plants or animals -- which are their hosts are themselves scattered over large areas and are not constant over time. The feeding specialization of the hosts compounds with the feeding specialization of the parasites, which is often extremely narrow.[2] In addition, a parasite's life-cycle often requires more than one host, thus increasing the complexity of its reproduction, mainly if stages in the parasite's development must match stages in the hosts' lives. Since many parasites themselves use other parasites as hosts, it should be obvious that trophic and reproductive constraints make parasites' survival more complex than is the case for free-living predatory organisms, which depend on a much larger array of feeding resources and whose reproductive process is more flexible. As a result, parasitic species can never reach a state of equilibrium with their environment (i.e., their hosts) unless their parasitic mode of life evolves toward mutualism or even symbiosis. Another remarkable result of the parasitic mode of life is that ecological and genetic factors constantly interact to form within a given species "a patchwork quilt of genetic incompatibilities between closely adjacent populations" (Harper 1977: 412). Low probability of colonization and high probability of extinction in patches are among the consequences of the above properties of parasitic species (Price 1980: 77). In response to such ecological constraints these species have evolved a variety of reproductive strategies such as an extreme longevity, including in some cases extended resting stages -- up to several decades -- which can combine with various sexual and non-sexual modes of reproduction such as mating before dispersal (colonization), self-compatible hermaphroditism or parthenogenesis. There are obvious homologies between some of these strategies and what can be observed in the case of memes, namely the clonal mode of reproduction and the reliance on extended resting stages in the form of inscriptions -- using a variety of semiotic systems -- stored into non-organismic supports. These periods of latency, which enable memes to bridge the gaps occurring in their vital resources, do not constitute however an absolute warranty of survival since, as in the case of parasites, they depend upon coming across an appropriate host at least within the period of time during which the physical support ensures their conservation. Another important factor to be considered is that memes can not be confused with evolved wired-in adaptive behavior. As any parasitic organisms, their biological dynamic exclusively concerns their own replication -- in this respect they are strictly "selfish" to use the metaphor that Dawkins applied to genes -- and consequently are neither detrimental nor beneficial to their hosts by necessity. However, some may afford an adaptive advantage which improve the reproductive success of Homo sapiens genes.[3] The reverse can naturally occur and some memes can become extinct because their own success of replication causes the extinction of their hosts. It must be remembered here that their hosts form a patchy and ephemeral resource and that resting stages are not immune to total destruction (viz. the burning of ancient libraries). If memes are indeed a crucial factor in the emergence and evolution of cultures, the latter's patterns of patchy development and eventual sudden or slow extinction are compatible with an important consequence of the parasitic mode of life: namely the impossibility of reaching a state of population equilibrium (Price 1980: 44-52). Beside the death of the host -- or the extinction of its species -- caused either by the ill-adaptive effects of a meme or a combination of memes, or by other factors, memes can die for reasons other than those relating to their parasitic mode of survival. At least three such causes can be considered: changes in the environment, intrinsic fragility, and elimination by other strains of memes. These are general evolutionary conditions applying to all forms of life -- let it be organic or not (Langton 1991). As any organism, memes can not be conceptualized independently of their environment. Adaptation by selection upon both genetic and memetic variations operates with respect to environmental factors. Catastrophic changes in memes' environment, that is the brain or a particular anatomical, physiological or chemical component of it, may eradicate a meme population or trap it in a deadend without possibilities of further replications. Local disruptions causing amnesia, general degenerative pathologies, such as Alzheimer's disease, chemical imbalance occurring naturally or being induced by external factors may seal the fate of memes. Since memes have the peculiarity of replicating via modifications of their hosts' behavior, those which are replicated through verbal means are jeopardized by aphasic conditions, those which spread through observational learning can be threatened by various forms of motor impairments. Such deleterious modifications of the neurological milieu are, from the point of view of memes' evolution, essentially similar to the environmental constraints on any living organism which can become extinct when confronted to sudden and drastic changes. Another factor which must be taken into consideration when assessing the conditions of survival of memes is their intrinsic structure and degree of complexity and specialization. Complexity entails fragility not only because of a closer dependency on environmental conditions but also because of a higher risk of replication errors. These two factors may compound to drive memes to extinction. An ad-hoc example of the latter could be some secret formulas of ancient medicine or medieval alchemy. Selective secrecy and exclusively oral replication -- i.e., absence of resting stage -- can be intrinsic components of a meme's set of instructions and thus account for its built-in fragility. It can be hypothesized that some memes do not survive beyond a small number of replicative generations. The argument of Dawkins according to which longevity is a criterial feature of memes seems to be hardly defensible. With respect to evolutionary time, how can longevity be measured? Even a pattern of information which would replicate only once would qualify as a meme because of this very property which has been thus demonstrated. A meme is no more necessarily a successful meme than a gene is. Unsuccessful memes are just short-lived ones. But there is more: if memes indeed conform to some characteristics of parasitic species -- that is, incidentally more than fifty percent of all extant species according to conservative estimates (Price 1980: 8) -- they may share their general property of "evolving slowly" and representing "deadends in any phylogeny (Price 1980: 11). Although Price qualifies this view, he quotes other parasitologists who thus express current theoretical views: "parasites as a wholeare worthy examples of the inexorable march of evolution into blind alleys" (Noble and Noble 1976: 525). Therefore, memes would die, that is cease to replicate, not only selectively because of an intrinsic structural fragility but generally, as a result of a proclivity attached to their specific mode of survival as parasites, not necessarily because of their host's death. But this is not all. Within a given segment of evolution memes are bound to compete with each other for brain territories and resources which are essential to their existence and reproduction. It can be hypothesized that strains of memes have no choice but strive to eliminate each other directly through structural disintegration or assimilation -- that is by neutralizing the information value of their competitors -- or indirectly by instructing their own host to destroy the hosts of other competing memes. Admittedly, such a formulation make memes' warfare sound and look like a video game, but let us point out that the analogy might not be as far-fetched as it seems given the fact that on the one hand confrontational video games explicitly mimic human affairs and that on the other hand they are based upon the mutual manipulation of information patterns or sets of instruction, and organisms through the medium of a tool which itself would easily qualify as a meme. The hierarchical depth of parasitism commonly involves several levels of embeddings, and there is no reason why this feature could not be heuristically extended to memes. Conclusion: A Program for Evolutionary Semiosis The above speculative exploration of the memes as parasites hypothesis can generate falsifiable hypotheses. For instance, Cavalli-Sforza (1971), Cavalli-Sforza and Feldman (1981) have shown how quantitative approaches to the study of socio-cultural evolution can be modeled with reference to biological evolution. Nevertheless, the meme hypothesis remains a distant, albeit fascinating frontier. On the one hand, it fits to some degree a long standing intuitive knowledge; countless observations have indeed led to a tradition of human anxiety regarding the power of ideas and the alienating effects of ideologies, the taking over of human behavior by directly or indirectly self-destructive programs; "habits" have been scrutinized with some puzzlement by philosophers and psychologists and have even been characterized as "a second nature," a metaphor which would become literal if the meme hypothesis were substantiated; but, on the other hand, it is deeply counter-intuitive and extremely difficult to operationalize; the conditions which would be required to test this hypothesis are still far from being afforded by current knowledge and means of investigation. Computer viruses and parasitic models at best offer intriguing analogies. Like the "sign" of semiotics, the "meme" of evolutionary biology or memetic is a hypothetical entity. It raises many questions and solves none. It is however the contention of this paper that it improves the quality of the questions raised through delineating what could be called a field of relevance and through shifting, on the epistomological spectrum, the heuristic speculation further toward theoretical and empirical determination of cultural evolution. Further advances in the exploration of the meme hypothesis are obviously depending upon advances in the understanding of the brain because it is generally recognized in evolutionary biology that it is impossible to understand an organism independently from an understanding of the environment within which it has evolved as a part of this environment. Let it suffice for the time being to ponder the basic concepts of parasites' evolutionary biology as expressed by Price (1980: 16-24) but by heuristically replacing "parasites" by "memes:" memes are adapted to exploit small, discontinuous environments; memes represent the extreme in specialized resource exploitation; and memes exist in non-equilibrium conditions. This may provide the shifting of perspective which is necessary to reformulate the fundamental questions of semiotics in a novel, more productive manner, not necessarily by equating signs and memes, but by considering the possibility that memes might be parasite's signs, and, as such, an important dimension of evolutionary semiosis. References Bonner, J.T. 1980The Evolution of Culture in Animals. Princeton: Princeton University Press. 1990 "Cultural Evolution: A Biologist's View," Journal of Ideas, I(1), 24-34. Bouissac, P. 1991 "Semiotics and the Gaia Hypothesis: Toward a Restructuring of Western Thought," Philosophy and the Future of Humanity I(2), 168-184. 1992 "The Construction of Ignorance and the Evolution of Knowledge," University of Toronto Quarterly, 61(4), 460-472. Cavalli-Sforza, L.L. 1971. "Simularities and Dissimularities of Social Cultural and Biological Evolution." In: Mathematics in the Archeological and Historical Sciences. F.R. Hodson (ed.). Edinburgh: Edinburgh University Press. Cavalli-Sforza, L.L. and M.W. Feldman 1981 Cultural Transmission and Evolution: A Quantitative Approach. Princeton: Princeton University Press. Cohen, F. 1990 A Short Course in Computer Viruses. Pittsburgh: ASP Press. Costall, A. 1991 "The 'Meme' Meme," Cultural Dynamics IV(3), 321-335. Dawkins, R. 1976 The Selfish Gene. Oxford: Oxford University Press. 1986 The Blind Watchmaker. London: Longman Scientific. 1989 The Selfish Gene. 2nd edition. Oxford: Oxford University Press. Delius, J.D. 1990 "Of Mind Memes and Brain Bugs," In: W.A. Koch (ed.) The Nature of Culture. Bochum: Studienverlag Brockmeyer, 26-79. Dennett, D. 1990 "Memes and the Exploitation of Imagination," Journal of Aesthetics and Art Criticism 48(2), 125-35. Harper, J.L. 1977 Population Biology of Plants. London: Academic Press. Hull, D.L. 1988 Science as a Process. Chicago: Chicago University Press. Khakhina, L.N. 1991 Concepts of Symbiogenesis. Trans. by S. Merkel and R. Coalson. edit. by L. Margulis and M. McMenamin. Yale University Press. Koch, W.A. 1986 Genes vs. Memes: Modes of Integration for Natural and Cultural Evolution in a Holist Model. Bochum: Studienverlag Brockmeyer. Langton, C.G. (ed.) 1991 Artificial life I. Redwood City, CA: Addison-Wesley. Lumsden, C.J. and E.O. Wilson 1981 Genes, Mind, and Culture: The Coevolutionary Process. Cambridge, MA: Harvard University Press. Moritz, E. 1990 "Memetic Science," Journal of Ideas 1(1), 3-23. Noble, E.R. and G.A. Noble 1976 Parasitology: The Biology of Animal Parasites. 4th ed. Philadelphia: Lea and Febiger. Price, P. W. 1980 Evolutionary Biology of Parasites. Princeton: Princeton University Press. Reggia, J.A., S.L. Armentrout, Hui-Hsien Chou, Yua Peng 1993 "Simple Systems that Exhibit Self-directed Replication," Science, Vol. 249 (26 Feb.), 1282-1287. Thompson, D'Arcy W. 1942 On Growth and Form. 2nd ed. London: Cambridge University Press. Notes ENDNOTES******************************** [1]. However, Bonner who defined memes as "any bit or any collection of bits of information passed by behavioral means from one individual to another" (1980), extends the operationality of this concept to social animals well beyond the primates (1990: 28-31). [2]. It must also be kept in mind that parasites influence many aspects of their hosts' qualities, thus altering the hosts' ecological system. "Behavior, reproductive potential, competitive ability, and susceptibility to predation may all be modified and the degrees of change will differ with levels of infection. This variability in the host population, superimposed on the genotypic and phenotypic differences within it, greatly complicates its ecological and evolutionary processes" (Price 1980: 10-11). [3]. In the same way as random genetic variations sometimes turn out to be adaptive, thus causing evolution to occur, memes can both be aleatoric and occasionally generate in their hosts behavioral modifications which are adaptive (Bouissac 1992). However, meme-driven cultural habits may be only relatively adaptive with respect to a temporary subset of conditions, and may happen to be devastating in the long run. Historical time must be always measured against evolutionary time when assessing cultural changes.