Agner Fog: Cultural selection 1999

13. Discussion and conclusion

Let us start by recapitulating the cultural selection model. Just like Darwin's theory of biological evolution, the theory of cultural selection has three basic elements: innovation, reproduction, and selection.

Innovation means any new cultural trait or idea, whether it has arisen by random chance, errors, play, experimentation, religious interpretation of dreams etc., or by intelligent problem solving and rational planning.

Reproduction or transmission is the process whereby people acquire the cultural traits of others. Possible mechanisms of cultural reproduction include imitation, socialization and teaching. The transmission does not only go from parents to children, the model allows any person to reproduce the behavior and ideas of anybody else.

The third basic element is selection. This is the most interesting of the three elements to study, because it is the selection that determines the direction of cultural evolution.

Many different selection mechanisms are at work in any society. The most direct form of selection is the conscious choice by each person about which of the known behavior patterns he or she wants to acquire, and which forms he does not want to incorporate into his repertoire. The conscious selection may also take place at higher levels where the leader of a society makes a choice for his subordinates.

Besides the more or less intelligent choices exercised by humans, a lot of 'automatic' selection mechanisms also exist, of which the individual human is not conscious or is unable to influence. Let me mention some of the most important of these mechanisms:

The unconscious and 'automatic' selection mechanisms are particularly interesting to study because they are not planned and their consequences are not consciously intended or foreseen. The study of such mechanisms can throw light on many societal phenomena which hitherto have been inexplicable.


13.1 Cultural r/k-selection

Most of the previously published models for cultural selection have focused on mechanisms. Unfortunately, there are often so many different mechanisms working in parallel and so many unknown parameters that it is impossible to set up a useful mathematical model. I have therefore concentrated on another approach, namely cultural selection criteria. If it is possible to identify a general fitness measure or selection criterion for a specific cultural system, then it will also be possible to determine the direction of the cultural evolution (but not its speed) even if the selection mechanisms are not known in detail. This is the rationale behind a new model for cultural selection which I have introduced. I have called this model cultural r/k-selection because of certain theoretical similarities with the known biological r/K-selection model, which also is based on fitness measures rather than on mechanisms.

The balance between internal and external conflicts in a well-defined society or group is important for the cultural r/k-theory. A society which is marked by severe external conflicts will have to spend much of its resources on strengthening its position in these conflicts. A society which is not capable of this will lose the war to neighboring societies and thereby be deselected in the sense that its cultural characteristics will disappear.

I have defined a regal society as a society which allocates a high proportion of its resources to the strengthening of its position in external conflicts. A regal society gives considerations for the community a higher priority than considerations for the individual. Its policy is characterized by a central hierarchical government, strict discipline, uniformity, and a high birth-rate. The ideology, religion, morals, rituals, and art of a regal society will develop in a direction which supports this policy and form of organization.

The opposite is a kalyptic society where the external conflicts are minimal and unimportant, and where more consideration is given to the solution of internal conflicts and to making the individual members of the society happy. The obvious example is a society on an isolated island. The welfare of the individuals here is given more importance than the safety of the community, and the population will not accept that the society makes demands on such a high proportion of the resources and on the freedom of the individual as would be necessary in the case of war or threat of war. The kalyptic society is characterized by tolerance, individualism, freedom, and little or no population growth. The beliefs, rituals, art, etc. of a kalyptic society will evolve in a direction which supports these values.

The theory says that a society will evolve in the regal direction if it has the possibility of conquering territory from a weaker neighboring society, or if it has a risk of loosing territory to a militant neighboring society. Contrarily, evolution will go in the kalyptic direction if wars or massive migrations are perceived as unlikely.

You may imagine a continuous r/k-scale where the most regal cultures are placed at one end of the scale and the most kalyptic cultures at the opposite end. This scale may be used for classifying entire cultures as well as subcultures and single culture elements, such as an ideology or a piece of art. But there is reason to warn that this scale is nothing but an abstract construction, and it is hardly possible to assign absolute numbers to the points on the scale. For example, it would not be very sensible to compare the r/k-value of a primitive hunter-gatherer culture with a modern urban culture. It makes more sense to compare the r/k-status of, for example, two different political systems, two religions, or two genres of art.

Cultures placed around the middle of the cultural r/k-scale may be called solidaric. In a solidaric society, much importance is attached to collaboration and community, but the collaboration is based more on voluntariness and mutual advantage than on compulsion and central government. Of course, it is a simplification to claim that solidarity is always highest at the middle of the r/k-scale. If you want a more precise description, then you may define solidarity versus individualism as an independent cultural dimension.

The cultural r/k-scale is not a universal measure for cultural evolution, but only one among many dimensions in the social structure, controlled by cultural selection. You may choose to study other dimensions, such as ecological niche, flexibility versus specific adaptation, conservativism versus innovativism, size of political units, technological complexity, economic structure, division of labor, the relationship between the sexes, etc. I have chosen to concentrate on the r/k-dimension because this factor seems to permeate almost all aspects of cultural life. This makes it possible to disclose connections between different social phenomena that hitherto have been regarded as independent of each other. The cultural r/k-theory turns out to be useful for explaining many phenomena within politics, ideology, religion, art, and sexual behavior - including phenomena which hitherto have been difficult to explain. It is also possible that the theory can be used for predicting future political developments or for steering the social evolution in a certain direction (more about this in chapter 14).

Among the cultural phenomena I have studied, the one that depends least on the r/k-factors is sport, and the ones that most clearly reflect the regal or kalyptic values in a society are art forms such as music and pictorial art. A number of characteristics of regal and kalyptic cultural phenomena are summarily presented in table 2 page 102.


13.2 Cultural transmission media

On page 78 I have described why the cultural evolution is much faster than the genetic evolution, and how the cultural mechanism makes possible the evolution of much more complex structures than that which the genetic evolution alone can achieve. The human capacity for culture is therefore a highly effective metaadaptation, i.e. a trait which facilitates the evolution of other traits. The evolutionary advantage that the human race has gained over other animals, thanks to this metaadaptation, is so enormous that it more than compensates for the considerable amount of resources spent on the establishment of an effective cultural transmission mechanism - or rather, several cultural transmission mechanisms. As explained in the preceding chapters, cultural traits are transmitted through several parallel channels including imitation, play, socialization, education, religious rituals, and most surprisingly: through art.

I have assumed that art is a communication medium for social instructions and that this communication is mainly unconscious to the sender as well as to the receiver. This is the most important function of art and the ultimate reason why humans have evolved a propensity for producing and consuming art. I am defining art to include music, song, dance, tales, theater, pictures, architecture, body decoration, etc. It may occasionally be difficult to distinguish between art and rituals, but this distinction is not important here since these two cultural phenomena often have the same function.

Art is a markedly social phenomenon, and the instructions transmitted through art are first and foremost instructions about the social structure. The position of the society on the cultural r/k-scale is a particularly important piece of information in this respect and is therefore reflected in almost all branches of art.

It is of vital importance for the survival of a social group in competition with neighboring groups that it is able to adjust itself to the optimal r/k-value as fast as possible (see chapt. 4.7). It is therefore necessary to have a mechanism which makes it possible to find the optimal r/k-value and transmit this value to all members of the group. The optimal r/k-value cannot be decided by the leader alone because he would have an egoistic interest in making the group more regal. What is needed here is a kind of compromise-seeking negotiation process where every member of the society gives his contribution. I have proposed that the exchange of art is indeed such a negotiation process, and that every individual expresses his or her opinion about the social structure through the personal aesthetic taste. The communal singing, dancing, etc. in a primitive tribe makes up a compromise between the aesthetic tastes of the individual members, and thereby represents the best possible estimate of the optimal social structure. The r/k-value is probably only one of the many dimensions of social structure which may be determined in this way. This negotiation mechanism is comparable to a decision process which Kummer has observed in a group of baboons when 'discussing' by means of movements where to go on a foraging trip (see chapt. 11.1).

The reason why I am drawing a comparison with apes here is that I regard this art-mechanism as evolutionary very old. At least older than spoken language. Chimpanzees and other animals do make dance-like scenes and drumming sounds by beating their belly or other objects (see chapt. 11.1). Observations of baboons indicate that they have a mechanism for regulating their social structure in order to adapt to changing ecological conditions (see chapt. 7). Similar mechanisms in humans may originally have evolved as a means for adapting to the genetic r/K-dimension, or other parameters of importance for the social structure. This mechanism may later have been further developed to also regulate the cultural r/k-dimension.

Various sciences have produced theories about community spirit, national character, collective conscience, the collective unconscious, or other expressions that the members of a society have certain mental structures in common. Such collective psychological phenomena imply the existence of a transmission medium which can communicate mental structures from one human to another. If we reject the possibility that the genes alone can be responsible for this transmission29, then there is reason to assume that art and rituals contribute to such a cultural transmission.


13.3 The pleasure principle

Having attributed a social function to art, we also have found an ethological explanation for the aesthetic taste of humans. Any human who exercises his aesthetic taste by producing art or by choosing between existing alternative pieces of art, thereby unconsciously expresses a message about the social structure as he perceives it or as he wants it to be. This innate ability is part of the human capacity for culture.

It is characteristic of drives or instincts30, that they express themselves by means of pleasure and pain. You may say that the various pleasure-feelings are the psychological manifestations or driving forces of the genes or instincts. Various instincts are expressed in various kinds of pleasure-feelings which we have given different names. The pleasure of eating something nutritious is expressed as: "It tastes good". The unpleasure of eating something poisonous is called: "It tastes bad". The pain of bodily injury is called: "It hurts". The pleasure of sex is called: "love" or "lust". The discomfort of missing social company is experienced as "loneliness". Children's desire to play is expressed as "It is fun", etc. etc. Many of the things that we do every day are controlled by inner desires or feelings of pleasure and pain. And behind all these desires and feelings lie the genetic instincts which impel us to do exactly these things.

Recent investigations indicate that the aesthetic preference of humans for beautiful landscapes also has an evolutionary biology explanation. Those landscapes which are perceived as most beautiful are exactly those areas which are most suited as habitats for primitive man (Orians & Heerwagen 1992; Kaplan, S. 1992). Beauty is not an objective characteristic of an object, but lies in the eye of the beholder. Or, as evolutionary psychologists say: beauty is in the adaptation of the beholder. The same is the case when a potential sexual partner or a piece of music is perceived as beautiful. The perception of beauty is a pleasure-feeling which represents an instinctive preference, the ultimate function of which we may not be conscious about.


13.4 Are we slaves of the culture?

Some biologists have provocatively characterized genes as selfish beings having as their sole aim to reproduce themselves, and the body as a machine which only serves as a means for the reproduction of its genes (Dawkins 1976). Using the same argument you may regard cultural phenomena, or their information carriers (memes), as parasites or viruses using the humans as helpless tools for their own reproduction. This metaphor has been criticized with the argument that humans control the cultural evolution by rational exercise of their free will. As I have made evident throughout this book, neither of these opposite standpoints represents the whole truth. The interesting thing is that these two opposing statements, when pinned down to mathematics, lead to exactly the same formula. In fact it is the same phenomenon viewed from different angles: a reductionist biological and an idealist anthropocentric viewing angle, respectively. Many people have a psychological resistance against seeing human phenomena from the biological angle because, in their worldview, humans are conceived as sublime beings exercising a free will, rather than robots controlled by biological processes and random external influences (see chapt. 1).

There is no doubt that humans often make intelligent and advantageous choices, and the progress of science has made the possibilities for intelligent choices still bigger. On the other hand, we have seen many examples of selections which are unconscious or which have unintended consequences. Unconscious choices are very often rationalized so that people believe that they have made a rational choice (Visser 1994).

Humans are very susceptible to indoctrination and have a pronounced ability to internalize a religion or ideology without being conscious of the choices they make or the consequences thereof. If you ask a deeply religious person why he believes in religion A and not religion B, then he will probably answer that A is the only true religion, or he may say that his parents have taught him so. It is very unlikely that you will get the answer that A is the religion which is best adapted to the surroundings in which he lives. The selection of religious dogma is seldom based on a conscious evaluation of the consequences for the individual or for the society as a whole.

Our political, moral, and religious views are controlled to a significant degree by factors of which we are not conscious. One such factor, which is particularly important for the r/k-theory, is the phenomenon that people become authoritarian when their society is in crisis (see chapt. 4.7). One of the most irrational products of authoritarianism is witch hunts (see chapt. 8.4). No matter how irrational a witch hunt is, it may still be functional in the sense that it contributes to the preservation of the existing social structure. Whether this structure is also the optimal for the population (measured in life-quality, fitness, or whatever criterion) is still an open question. Undeniably, a witch hunt has considerable costs in the way that innocent people are persecuted and punished. The witch hunts in early modern Europe lasted for several hundred years, and similar phenomena are still evident today (see chapt. 8.4 and chapt. 10.4). This indicates that humans are not always as rational as they believe themselves.

Even a perfectly conscious choice may have negative consequences. Consider, for example, when fear of war between two countries leads to an arms race which drains both countries of resources and threatens with total destruction. Even though both countries understand the mechanism behind this arms race and its negative consequences, they are unable to stop it as long as mutual trust cannot be established.

Drug addiction is another well known example of how wide a parasitic phenomenon may spread and how difficult it is to stop it despite the most intelligent efforts.

Of course these examples belong to the extremes, and humans are generally becoming better and better at making intelligent and beneficial choices. But we are still far less rational than we want to believe, and the irrational or unconscious choices are the most interesting to study because of their unintended consequences. This is where the cultural selection theory has its greatest utility because it is able to throw light on unplanned phenomena which hitherto have been inexplicable.


13.5 Testability and sources of error

In this book I have proposed several theories and hypotheses, but how
tenable are they? Can they be verified?

The fundamental theory is that cultural evolution involves innovation, reproduction, and selection. There are no problems with this formulation: It is obvious; it has been known for more than a hundred years; and it has been examined and verified by numerous examples throughout this time. There may be questions about the relative importance of various forms of selection, but it is beyond doubt that selection is taking place.

Next we have to evaluate the theory of cultural r/k-selection, i.e. the claim that external conflicts can influence the social organization, ideology, religion, etc. of a society in a particular direction, which I have called regal, and that the absence of such conflicts or threats will make the evolution go in the opposite direction, which I have defined as kalyptic. The possibility of proving such a theory is limited by the difficulties of making experiments with humans. Experimenting with individuals would not suffice. We are talking about groups comprising hundreds or thousands of people and the experiments have to span several generations in order to prove the cultural r/k-theory. Such a giant experiment is totally impossible to carry out, not only for practical and economic reasons, but obviously also for ethical reasons. Animal experiments are a theoretical possibility, but those animals which might be suitable, for example baboons, have such a low degree of culture that the results would tell very little about human cultural evolution.

The only realistic possibility for supporting the cultural r/k-theory is therefore natural experiments, i.e. the study of events which already have taken place or which take place whether we study them or not. Human history contains an immensely rich source of events suitable for this purpose, and I have already mentioned several historical courses of events which confirm the r/k-theory. Numerous other historical events and epochs may be taken up and investigated in order to evaluate the r/k-theory. Here is enough work for keeping a bunch of historians occupied for many years. The biggest problem with natural experiments is that there always will be confounding factors for which it is difficult to correct. This is a general problem encountered with any social theory. But in the present case, the correlations are so strong and the quantity of data so big that the possibilities for testing this theory are much better than for many other social theories. The reader may already have relevant examples in mind from his historical and social knowledge. Another way of testing the cultural r/k theory is to make predictions about the future based on the theory and see if the predictions come true. In fact, the attractiveness of this theory lies exactly in its usefulness for making predictions.

While the cultural r/k-theory is reasonably well founded and testable, we have more serious problems with the most daring part of my theory: the claim that art has a social function which for most people is unconscious. It is obviously not possible to provide a definitive proof or disproof for the existence of a phenomenon which takes place in the unconscious, which is hypothesized to have arisen by biological evolution millions of years ago, and which today has been partially superseded by more effective mechanisms. We have to be satisfied with statistical indications and let the hypothesis stand until somebody else comes up with a better one.

It is a frequently used argument in ethological research that if a phenomenon exists then, according to darwinism, it must have an adaptive function. The weakness of this argument is that non-adaptive characters will always appear with a certain low frequency; that fitness-neutral characters may arise purely by random; that the evolution of adaptive traits may have side effects in the form of non-functional epiphenomena; and that traits which earlier in evolutionary history have been adaptive may survive for a very long time although altered conditions have caused them no longer to be functional. The more elaborate and complex a phenomenon seems, and the more resources it makes demands upon, the more difficult it is to explain it away as a random dysfunction, and the higher is the likelihood that the phenomenon has arisen by a selection process.

And this is exactly the case with the human production of music, dance, and other forms of art. The philosophical claim that art is entertained for its own sake does not make evolutionary sense because the peculiar human affection for art would never have evolved or would have been deselected long ago if it were nothing but a dysfunctional waste of resources. The various forms of art are so elaborate and well developed and claim such a high proportion of human time and energy in all known societies that it cannot possibly be rejected as a non-functional epiphenomenon. Art must therefore necessarily have, or have had, an adaptive function. That this function has something to do with communication seems obvious, and most scholars agree that art is a form of communication. But there is far from agreement about what the communicated message is. The claim that art contains information about the social structure is based on statistical analyses of the correlation between social structure and artistic style. These statistics prove beyond doubt that there is a connection between art and social structure, but statistics cannot distinguish between cause and effect. So in principle we cannot know whether it is art that influences society or society that shapes the art. Most probably both factors influence each other to a certain degree, but my hypothesis that art may have functioned as a medium for 'negotiating' the social structure cannot be verified by these kinds of statistics. The comparison with the way baboons 'negotiate' where to go for foraging (see chapt. 11) only proves that such a function is biologically possible, not that it exists in humans.

Above (chapt. 13.3) I have described the pleasure principle which says that any feeling of pleasure or pain is related to an instinct or genetically determined function. Since art gives rise to an aesthetic pleasure it must, according to this principle, be guided at least partially by genetically determined inclinations. But we must be careful to avoid a circular argument here: even though the connection between pleasure and instinct is well-known, I would not have formulated it as categorically as here if I did not already believe that art, play, and other pleasurable activities are functional.

Another critical problem is the question of the phylogenetic age of art. I have assumed that basic forms of art - especially music and dance - are considerably older than spoken language. In other words, that art is a more primitive form of communication which has survived beside the spoken language. Of course, such a theory is very difficult to prove. Our ancestors may have danced to primitive rhythm instruments made of wood millions of years ago without leaving any archaeological traces. The theory is based on comparison with the communication forms of other animals, and on the fact that it is very improbable that the primitive form of communication that art comprises would have arisen when a much more advanced and less energy-consuming form of communication already existed, i.e. the spoken language. On the other hand, it is evident that art has developed further and still develops. Such a further evolution has been possible with little or no genetic changes and does therefore not contradict the assumption that art basically is a survival from a more primitive evolutionary stage. You may discuss whether art is a peculiar appendix which has totally lost its function, or whether it still has an importance in modern society. Undeniably, many modern people do attach much importance to art, and when we consider how many resources modern man spends on music, dance, film, theater, painting, monuments, architecture, body adornment, etc. etc., then there can be little doubt that art still is an important communication form, although disproportionately energy-demanding.

If we want a more precise description of what is communicated through a particular piece of art, then we are really on shaky ground, because most of the communication is unconscious to sender as well as to receiver. The study of unconscious phenomena is the domain of psychoanalysis which, as a science, often has to resort to intuition and guesswork. Rigorous testing is not possible within this discipline. Many psychoanalysts therefore have totally ignored the need for testing their theories, which has immensely harmed the entire science of psychoanalysis. In my opinion, this problem has not been taken seriously enough by the psychoanalytical school, and it is therefore with great regret that, in this book, I have had to contribute to the intuitive guessing. Since psychoanalytical science has not developed any general and reliable methods of verification, I have had to resort to other disciplines - most importantly statistical analysis of the connection between social structure and artistic style.

The r/k-status of a culture has influence on many areas of social and private life, including religion, ideology, art, sexual behavior, etc. Changes in the r/k-status of a society are sometimes reflected in these areas faster than you would expect from the immediately obvious selection mechanisms. I have therefore assumed the existence of vicarious selection mechanisms of a psychological nature (see chapt. 4.7 and chapt. 10.1). The effects of these psychological mechanisms are that a person who perceives his life situation, and in particular the situation of his group, as threatened and insecure will have a tendency to submit to a strong leader and strict rules of living. In other words, he will develop what social psychologists call an authoritarian personality. According to my theory, this psychological condition not only influences the political attitude of a person, but also his preferences in art and his sexual morals. It is possible to produce statistics which show a connection between the political situation in which people live and their aesthetic taste, sexual behavior, etc., but even if such a statistic can prove that there is a connection, it cannot say anything about the mechanisms behind this connection. A formal statistical proof would require a controlled double-blind experiment as is common in medical research, but of course such an experiment is as impossible in this connection as it is in other areas of social psychology.

Sources of error

I want to draw attention to certain pitfalls and sources of error for those who might want to work further with the cultural r/k-theory. When using archaeological and art-historical sources, there is a systematic bias of which it is important to be aware: regal cultures generally produce big, sumptuous and ostentatious artifacts made of durable materials, whereas kalyptic cultures usually produce small and simple artifacts made of perishable materials. Therefore, the relics of regal cultures have always attracted more historical attention, while the products of kalyptic cultures either have perished or been overlooked.

When evaluating the r/k-level of a culture you must never rely on a single indicator. A reliable evaluation requires the examination of several factors, such as religion, political and military systems, social stratification, criminal law, human rights, population growth, education, art, sexual morals, suicide statistics, etc. In case of ambiguity or disagreement between these factors, a closer examination is needed. Each of these factors may have sources of error when used as r/k-indicators:

In the case of religion, two countries may formally have the same religion while the population in one country is much more orthodox than that of the other. The political system of a country may not be in accordance with the mentality of the population if the system has been forced upon the country from outside. As an example I can mention the state of Czechoslovakia which had a fairly regal political system forced upon it by the Soviet Union, while beneath the surface the population proved to be quite kalyptic.

Factors like military structure, social stratification, criminal law, and human rights, not only depend on the r/k-level of a society, but also on its general technological, economic, and political levels of development.

The population growth of a country may be considerably larger than expected from its r/k-level if the political and social situations are chaotic or if economic factors motivate a high production of children.

Using art as an r/k-indicator, one must acknowledge that there may be a considerable discrepancy between the official and state subsidized high culture which a country proudly presents to the outside world, but which in reality is only cultivated by a small snobbish elite, and, on the other side, the popular music and films which are preferred by the majority of the population. Evaluating the r/k-information of a specific piece of art will always be a subjective judgement, and there are many possibilities for misinterpretation and overinterpretation. Speaking of art, there is a peculiar asymmetry between regal and kalyptic cultures to which it is important to pay attention: kalyptic cultures are tolerant of regal art, while regal cultures are intolerant of kalyptic art. Regal art may therefore be found in kalyptic cultures because of inertia and lack of inventiveness (see chapt. 11.3), because of historical and ethnographical interest, or the regal style may be caricatured in a parody with an ironical distance to the style being imitated.

The sexual morals may be a useful r/k-indicator because they can be measured more precisely than art for example, but it is worth noting that the sexual legislation of a country is no reliable measure of the morals that are actually enforced. The legislation in this area is often applied arbitrarily or not followed at all. This is particularly true in former european colonies where the laws in this area are survivals from the colonial rule.


13.6 Explanatory power

While there are aspects of the cultural r/k-theory which are difficult to verify, there is another aspect which makes the theory very attractive, and that is its impressive explanatory power. I have already shown how the r/k-theory may be applied within widely different areas of social science and historical research. This theory can explain such diverse phenomena as the fall of Rome, renaissance witch-hunts, the emergence of rock music in the late Soviet Union, and the resistance against pornography in the USA.

I expect that the theory may be advantageously applied within widely different branches of science, including sociobiology, anthropology, archaeology, political history, art history, sociology of culture, sociology of religion, social psychology, sexology, conflict- and peace research, and futurology. Obviously, it would be quite impossible for me alone to work through all these areas in order to test the applicability of the cultural r/k-theory. And besides, I do not have the necessary expert knowledge. I therefore hope that this book has inspired scientists trained in some of these disciplines to test the applicability of cultural selection theory within their special field of expertise. I have sketched how it can be done and I am now inviting others to work on within this new paradigm.


29. In his theory of the collective unconscious, psychoanalyst C.G. Jung assumed that it was innate (Jung 1969). See chapt. 11.

30. See note 21 chapter 11.