Using Memetics to Grow Memetics

1. Crises in Memetics and Psychology

As sciences, psychology predates memetics by a century. But both are facing crises. For memetics, this issue marks the end of the first incarnation of the Journal of Memetics. In psychology the United States government is reorganizing its National Institutes of Mental Health. It will now focus on helping people with mental and behavioral disorders while substantially narrowing its support for basic behavioral research in social psychology, personality, learning and memory and animal studies. The February issue of American Psychological Society Observer devoted both its feature article and its presidential column to the crisis. APS Fellow Daniel Gilbert of Harvard was quoted saying “This is an absolutely devastating blow to the infrastructure of basic science.”

The two crises are of fundamentally different sorts. The crisis in memetics stems from a dearth of empirical studies. In an age when governments, universities and foundations support empirical studies, a science without a track record has a difficult time raising funds. Established sciences, like psychology, are stronger competitors. Memetics may be a future oak tree, but currently the forest of established social sciences is using up its resources.

The crisis in psychology actually stems from a problem identified more than a half century ago by one of the greatest of early psychological scientists, Karl Lashley (1951). Psychology has no true unit of analysis for spontaneous and complex human behavior. Reflexes (including neural responses), conditioned responses, test and questionnaire responses all restrict behavioral possibilities so much that researchers are often unable to distinguish their own biases from the behaviors being studied. When compared to other sciences, these restricted-possibilities units have caused psychology to be more fraught with unresolved conflicts and less clear about scientific community directions for addressing societal problems related to its topic.

De Jong, Konstantions and Mamadouh (2002) found convincing evidence that system upheaval or severe performance crises are more fertile breeding grounds for pervasive change than other occasions. They conclude that “reformers do best when they are prepared for oncoming windows of opportunity and save their energies to act at such moments.” Though a harrowing time for the psychological research community in the United States, psychology’s crisis is a window of opportunity for memeticists to begin demonstrating how its evolutionary approach can bring practical results that people hoped psychology would address more effectively than it has.

Key to realizing the opportunity for memetics is establishing the meme as the unit of analysis for spontaneous and complex human behavior. The view that memes are methodologically complex hampers their growth as such a unit. There are two aspects of this supposed complexity: the research triad and evolutionary data. The research triad refers to the fact that identifying memes requires three people: a model to exhibit the meme, a learner to replicate it and an experienced judge to reliably determine whether the model and learner are producing the same meme.

Replacing models with artifacts, such as videos, texts or diagrams will reduce some of the methodological complexity in the research triad. The fundamental aspect of this complexity, however, stems from the beginnings of psychology. The community trying to establish psychology as a science made an early mistake in attempting to simplify the research setting. Since then it has practiced that mistake so long, that it may as difficult to undo as to start over as a science. Danziger (1994) documented how psychology eliminated the judge from the experimental paradigm. Memetics theory, however, implies that reliable judges are not the problem. The very foundation of memetics is the fact that humanity evolved a unique capacity to reproduce the actions of other humans. The ability to judge the similarity of memes, therefore, is an evolutionarily maximized quality of humanity.

In support of the memetics implication that reliable judges are not the problem in psychological research is the fact that even the mighty computer is no competition for the human in the realm of judging the similarity of memes. The poverty of computerized speech recognition and translation compared to recognition and translation even by experienced children proves the poverty of the “eliminate-the-judge” meme for psychological research. If the computer cannot assess behavioral similarity, surely the lowly test and questionnaire are not up to the task. Memes are not methodologically over-complex; rather psychology as a science is suffering from an artificial assumption of methodological simplicity.

The second aspect of methodological complexity of the meme as unit for spontaneous, complex human behavior is its emphasis on evolutionary data. Dirlam, Gamble and Lloyd (1999) showed that developmental psychology systematically collected data on gradual changes over time only with respect to physical growth. Worse, the field gradually eliminated such studies a half century ago in favor of simplifying to one, two or a few developmental data points. Other social sciences have addressed the distribution of behaviors in space, but psychology does not even consider such distributions an issue. With such a methodological bias, it is no wonder that psychology has missed the memetic evolutionary aspects of human behavior. Again memetics is not methodologically too complex; rather, psychology is methodologically too simple.

For spontaneous and complex human behavior, the meme is a better unit of analysis than any that psychology has constructed. The problem, however, is how to grow usage of the meme in the context of overwhelming hegemony of those previously constructed units.

2. A Memetics Solution

Memetics can contribute to the growth of the meme as a unit of analysis in psychology. First, reflexes, responses, tests and questionnaires are all memes. Dirlam (2003) showed that the growth of memes obeys ecology’s Lotka-Volterra model of growth. Thus, four general factors control the growth of memes: initial strength, characteristic growth rate, resource availability and competitive strength. In current strength, psychology’s restricted-possibilities units are several orders of magnitude stronger than the meme. The problems with psychology’s oversimplification of the research setting strongly suggest that the meme’s characteristic growth rate and competitive strength are greater than psychology’s restricted-possibilities units. This means that those interested in growing the meme must focus on the resources.

Certainly the scientific apparatus of journals, university departments, funding agencies and societies are resources. But memetics allows us to look at a more general level of the resource problem. We have a thorough conception of the general resources required by species, such as temperature, food, water etc. But at this general level, what are the resources that lead to meme growth? The psychologist, Albert Bandura (1977) devoted his career to the study of “observational learning” or imitation. In his analysis, there were four essentials: attention, retention, reproduction and motivation. This is a good starting point for identifying resources for meme growth. If the “meme-as-a-unit-of-analysis-for-psychology” meme is to spread, memeticists need to provide it with attention, retention, reproduction and motivation. Memetics history of self critique suggests that motivation is the central problem. Therefore, Bandura’s criteria will be discussed in the reverse order.

2.1. Motivation

Judging from the philosophical articles on memes in both the Journal of Memetics - Evolutionary Models of Information Transmission and in collections of readings such as Aunger (2000), the central resource for growing memetics is motivation. Each of the authors had clearly attended to, retained and reproduced memetic thinking. But the question of long-term support seemed ever present. In Aunger’s words “The ultimate test—which would pre-empt theoretical objections—is whether memetics can produce novel empirical work or insightful interpretations of pervious results. It has not yet done so, but must do so in the near future. Otherwise, it is likely that memetics will be perceived to be a misguided enterprise. The clock is ticking.” The impression is that there is a plethora of theoretical critiques and a scarcity of data. But what would motivate a change to memetic data collection both within and outside of the memetics community?

The most insightful article on motivation for change that I have encountered was the brilliant analysis of dietary change by economist/public policy analyst, Carl V. Phillips (1999). When we eat, we start with a strategy that will satisfy our needs for good nutrition, good taste, ease of obtaining and familiarity (four key resources for eating). The high fat, high sugar diet that we get at fast food chains is an example. As we practice, we get better and better at satisfying our needs with this food. In time, we get so good that when we try to change, we are usually less satisfied. From a nonlinear dynamics point of view, we have found a relative maximum. Trying to cut back within our favorite diet leaves us feeling hungry, resulting in a lower overall utility value. We easily backslide into our old habit patterns.

Eventually for some people, the strategy fails. They take a step back and consider making a radical change. They might try a vegetarian diet. This whole new strategy results in a whole new set of experiences, at first not as satisfying as the old diet. As they keep trying, they gradually get better at satisfying their four needs with the new approach. While they are working at satisfying their needs with the new approach, they frequently backslide into the old one. When they get as good at satisfying their needs with vegetarian food as they did with fast food, however, they rarely backslide. A radical change in strategy that needs practice to get right might be called a “memeplex replacement.” Evidence presented in the next section shows that it applies not just to eating, but to all the life strategies used in complex human activities.

For memetics to succeed in psychology there needs to be a radical change in the activities of psychological researchers. To summarize the dietary analogy, people are dying from the “fast food” diet. That is motivation to try the leap to the vegan diet. Trying to maximize the four utilities without radical change will not work. It helps to have models who have shown that they can successfully maximize their utilities with the new approach, but each individual must commit to the change and then learn to make it work. To apply these concepts to psychology, maximizing the utilities of psychological research will not motivate people to change. A crisis, such as the US NIMH (the National Institutes of Mental Health) precipitated, must motivate researchers first to commit to change memeplexes and then follow up by learning how to make the new approach work.

People have died from fast food for decades. That fact has not been sufficient to create change. A rich set of data has confirmed that fast food impairs people’s health and lifestyle. A public critique of fast foods is growing. The food industry is creating new foods in response and there is a growing list of healthy food cookbooks coming from publishers. There is less data about the failings of psychology, but a public critique is beginning to emerge. Several core concepts in psychology lack hard empirical support. These include developmental stages, intelligence, the genetic transmission of psychological traits, and psychological disease. Also, if there really was a science of mind that worked, social scientists would be as much in demand in corporations and governments as biologists and engineers. Instead, at least in the U.S. policy makers seem committed to seeking biological solutions to social problems, no matter how weak the solutions may be.

Much of the early writing about memetics has involved philosophical critiques. If someone wants to help some friends change their diets, it is certainly useful to avoid suggesting a diet that would make matters worse. But the critiques of memetics have focused too much on the problems of memetics and not enough on the problems of the alternatives.

Clearly many memetics solutions are not going to be immanent, but some are. For example, the developmental stage is an almost mystical concept that at best prevents people from trying to teach concepts to children without adequate preparation. If psychologists replaced the concept of “stage” with one of memeplex change, however, it would cause educators to look for the utilities involved in the activity and to create “recipe” books that might help learners to maximize these utilities.  

Likewise, psychological “disease” is a mystical concept that causes people to look for medications that alleviate the symptoms. That is roughly equivalent to solving addictions by giving drugs that block the effects of the addicted drug. A perspective that leads to some interesting possibilities for memetics is to view psychological disorders as self maintained addictions – people learning to manipulate their own neurotransmitters until their brains begin to malfunction. Such a view would motivate clinical researchers to look at the utilities of the neurotransmitter manipulation and at how to create the commitment to change. Ultimately it might lead to developing “recipe books” for how to achieve those utilities without manipulating the neurotransmitter. As an example consider that television watching produces attentional problems (Christakis, Zimmerman, DiGiuseppe, and McCarty, 2004). In the U.S. more and more children are being given drugs for attentional problems. From a memetics point of view, a more lasting and less damaging solution would be to discover ways to serve children’s entertainment needs other than television.

Many psychologists are already struggling against the current diet of such “fluff” concepts as developmental stages and psychological diseases. They are working in areas such as naturalistic observation, qualitative data analysis, ecological psychology, observational learning, etc. The memetics community should invite such psychologists to contribute to memetics methodology.

2.2. Reproduction

From an empiricist’s point of view, empirical research in memetics is currently difficult to reproduce. No one has yet written the methodological recipe book for memetics research. A good starting point is to make analogies with ecology. Before attempting ecology’s Lotka-Volterra model, I spent a frustrating decade looking for a mathematical model that fit data from over 1,200 drawings made by 5 to 18 year olds and coded for nearly 30 drawing qualities by three independent judges. Probability theory offered a reasonably good fit using the generalized gamma law developed in the context of radioactive decay from one element to another. But why development would proceed by decay was a mystery. The ecology equation, however, fit the data like a statistical skin-tight glove. When the same equation fit data on the historical development of developmental research methodologies (as coded by two independent judges in nearly 1000 journal articles written from 1930 to 1992), it opened up several powerful theoretical insights:

1. A competing-agent action of the qualities in both settings fits well with the meme concept.
2. An outside-of-the individual status of the qualities also fits the meme concept.
3. Decay happens because later memes compete with earlier ones.
4.  “Development” refers to changes in the frequency of use of competing memes described above as memeplex replacement.
5. An easy to study developmental succession of memeplexes replaces the fluffy stage concept.
6. The easily researched process of commitment to radical change followed by relearning readily replaces the inferred process of developmental stage acquisition.
7. Controlling factors of meme populations are initial frequency, growth rate, resources and competitive strength.
8. Given constant resources, four growth patterns identify four categories of memes:
• Default – high initial strength, low growth rate and competitive strength.
• Pioneering – high growth rate (enough to exhaust all resources in the absence of competition), low initial and competitive strengths.
• Niche – high competitive strength, low initial strength and growth rate.
• Dominant – high competitive strength, low initial strength, moderate growth rate ((not enough to exhaust all resources in the absence of competition).

Other analogies from ecology should produce similarly powerful insights not only for human development but for memetics research related to all the social sciences.

2.3. Retention

Key factors that influence retention are repetition, organization, context similarity and distinctiveness. Repetition, organization and context similarity factors depend on growth. Based on Google searches “Memetics” currently appears on the internet about 0.1% as often as the term “internet” itself (controlling for growth in the internet as a whole), as often as “psychotherapeutic”, 3% as often as anthropology or sociology and 1% of psychology. It’s ranking among such terms grew rapidly last fall, but has declined this spring. The crisis represented by the journal change may be a key reason for the recent decline. Since there are none of the other traditional academic resources mentioned above for memetics, a large impact of a change in the only journal might be expected. In this context, closing the journal would be roughly akin to people aspiring to change from a fast-food to a vegan diet by throwing out their only vegan recipe book.

The remaining factor is distinctiveness. There have been numerous distinctive contributions to memetics. Topics ranging from anti-religion and anti-self to paganism and marketing are reminiscent of the early days of psychology amplified by a century-and-a-half of self preoccupation. Few of these distinctive contributions, however, have an empirical basis. The deeper question is whether there is a distinctive memetics methodology.

Two candidate methods that set memetics apart from the other social sciences have already been mentioned. The first is the model-learner-judge triad and the other is measurement of the frequency distribution of memes in time and space. Again a comparison of ecology and psychology is illustrative. I counted the frequency of tables and charts in a standard textbook of ecology and one of psychology. From beginning to end of the ecology textbook, about 75% of the charts and tables involved distributions in time or space. The same was true for the first portion of the psychology text that dealt with sensation, perception and conditioning. The last portion of the psychology text, however, dealt with personality and social psychology. For that portion, the ratio was reversed–only 25% of the tables and graphs involved distributions in time or space. Because psychology views the mind as being inside the brains of individuals, it does not consider mapping the distribution of complex human behaviors in time and space. Because memetics views the human mind as dependent on relationships with other humans, past and present, distributional data is central to memetics methodology. Distributional data made the theory of evolution both possible and established. It will also yield the distinctive findings needed to establish memetics.

2.4. Attention

Memetics needs empirical studies. Empiricists pay attention to empirical results. Those that attract the most attention solve intractable societal problems. If memetics is to help solve such problems it will be through its emphasis on adaptive growth. Growing a reduction in crime within a city, a more equitable way to distribute food in an underdeveloped country, a reliable reduction in dependency on addictive drugs or self-manipulated neurotransmitter changes in individuals, or an improvement of skill in a school population would all be spectacular results that would certainly attract attention. If memetics researchers tackled the worst problems in their environment by collecting evolutionary data using the model-learner-judge triad, the results would establish memetics as a science.

3. Conclusions

There is a window of opportunity for memetics resulting from the current crisis in U. S. funding for psychology. To realize the potential of this window memetics must provide other social sciences with the four essential resources for imitation: motivation, reproduction, retention and attention. The discussion showed that the following four strategies will grow the use of memes in the social sciences:

(1) Motivate social sciences by critiquing current paradigms, by calling for a commitment to a radical shift of units of analysis to those involving experienced human judges and the collection of evolutionary data and by encouraging methodological articles in memetics by those already using such units.

(2) Help researchers reproduce memetics by providing methodological models which borrow heavily from analogs of approaches found to be successful in ecological and evolutionary research.

(3) Help researchers retain the memetics approach by repeatedly emphasizing the model-learner-judge triad and the charting of changes in frequency across time and space.

(4) Attract researcher’s attention to memetics by tackling spectacular problems.

References

Aunger, R (2000). Conclusions. In R. Aunger (ed.), Darwinizing Culture: The Status of Memetics as a Science, Oxford, Oxford University Press, 205-232.

Bandura, A. (1977). Social Learning Theory. Englewood Cliffs, NJ: Prentice-Hall.

Danziger, K. (1990). Constructing the Subject: Historical Origins of Psychological Research. Cambridge: Cambridge University Press.

De Jong, M., Konstantions, L. and Mamadouh, V.  (2002). The Theory and Practice of Institutional Transplantation: Experiences with the Transfer of Policy Institutions. Dordrecht: Kluwer.

Christakis, D. A. Zimmerman, F. J., DiGiuseppe, D. L., and McCarty, C. A. (2004). Early Television Exposure and Subsequent Attentional Problems in Children. Pediatrics, 113, 708-713

Dirlam, D. K. (2003). Competing Memes Analysis. Journal of Memetics - Evolutionary Models of Information Transmission, 7. <http://cfpm.org/jom-emit/2003/vol7/dirlam_dk.html>

Dirlam, D. K., Gamble, K. L., & Lloyd, H. S. (1999). Modeling historical development: Fitting a competing practices system to coded archival data. Nonlinear Dynamics, Psychology, and Life Sciences, 3, 93-111.

Lashley, K.S., (1951). The problem of serial order in behavior. In L.A. Jeffress (Ed.), Cerebral mechanisms in behavior. New York: Wiley, 112-136.

Phillips, C. V. (1999) Complex systems model of dietary choice with implications for improving diets and promoting vegetarianism. American Journal of Clinical Nutrition, 70(3), 608S-614S.

Back to Issue 1 Volume 9