Thursday, September 19, 2013

Distilling the Essence of an Evolutionary Process: Implications for a Formal Description of Culture


From arXiv.org's division of quantitative biology, a production of the Cornell University Library, this is an older paper just now posted at this site outlining a model for understanding cultural evolution by looking at a specific process of biological evolution - selection.

Their model echoes the argument in favor of the human brain's open architecture (posted here yesterday), that the unique ability of the human brain are responsible for culture:
To invent in the strategic, intuitive manner characteristic of humans requires a cognitive architecture that supports the capacity to spontaneously adapt concepts to new circumstances and merge them together to conceptualize new situations.
This article is more than 10 years old, presented in 2000 at a conference and then published in 2005 in a book. Still, this is a cutting edge topic right now as we begin to embrace the idea that cultural evolution is an emergent property of biological and consciousness evolution.

Oh yeah, a note on the image above. In terms of variation, biological evolution is much less random than we imagined (and certainly less so than Darwinians, like Richard Dawkins, will currently admit). Likewise, cultural evolution is much less directed than originally believed. In a marketplace of ideas, so to speak, new memes (like "twerking") are generally random in their emergence, not orchestrated and directed.

Distilling the Essence of an Evolutionary Process and Implications for a Formal Description of Culture 

Liane Gabora, Diederik Aerts

(Submitted on 18 Sep 2013)

It has been proposed that, since the origin of life and the ensuing evolution of biological species, a second evolutionary process has appeared on our planet. It is the evolution of culture-e.g., ideas, beliefs, and artifacts. Does culture evolve in the same genuine sense as biological life? And if so, does it evolve through natural selection, or by some other means? Why does no other species remotely approach the degree of cultural complexity of humans? These questions lie at the foundation of who we are and what makes our lives meaningful. Although much research has been done on how selective pressures operating at the biological level affect cognition and culture, little research has focused on culture as an evolutionary process in its own right. Like biological forms, cultural forms - ideas, attitudes, artifacts, mannerisms, etc. - incrementally adapt to the constraints and affordances of their environment through descent with modification. In some respects culture appears to be Darwinian, i.e., a process of differential replication and selection amongst randomly generated variants. This suggests that knowledge of biological evolution can be put to use to gain insight into culture. However, attempts to apply Darwinian theory to culture have not yielded the kind of unifying framework for the social sciences that it provided for the biological sciences, largely because of the nonrandom manner in which the mind - the hub of cultural change - generates and assimilates novelty. This paper investigates how and when humans became capable of supporting culture, and what previously held it back, focusing on how we attained the creative powers we now possess. To invent in the strategic, intuitive manner characteristic of humans requires a cognitive architecture that supports the capacity to spontaneously adapt concepts to new circumstances and merge them together to conceptualize new situations.

Journal Reference: 
Gabora, L. & Aerts, D. (2005). In (W. Kistler, Ed.) Proceedings of Center for Human Evolution Workshop #4: Cultural Evolution, May 18-19, 2000. Bellevue, WA: Foundation for the Future.

Cite as: arXiv:1309.4712 [q-bio.PE]
(or arXiv:1309.4712v1 [q-bio.PE] for this version)

CONTENTS

1 Do Evolutionary Models Capture the Dynamics of Culture? ................. 3

1.1 Memes .............................................................................................. 4
1.2 Mathematical Approaches ................................................................ 4
1.3 Computer Models ............................................................................. 4
1.4 Where Do We Stand? ........................................................................ 6
2 Background from Cognitive Science .......................................................... 6
2.1 Conceptual Space and the Distributed Nature of Memory................. 6
2.2 Conceptual Integration....................................................................... 7
2.3 Focusing and Defocusing................................................................... 7
3 Evolution of the Culture-evolving Mind ..................................................... 8
3.1 What Sparked the Origin of Culture?................................................. 8
3.2 The Earliest Modern Minds and the ‘Cultural Revolution’................ 8
4 Rethinking Evolution .................................................................................... 9
4.1 The Cultural Replicator: Minds Not Memes....................................... 9
4.2 Creative Thought is Not a Darwinian Process .................................. 10
4.3 Evolution as Context-driven Actualization of Potential.................... 11
5 Concepts: An Enigma at the Heart of the Problem .................................. 12
5.1 The SCOP Representation of a Concept ........................................... 12
5.2 Embedding the SCOP in Complex Hilbert Space ............................. 13
5.3 Concept Combination ........................................................................ 14
6 Summary and Conclusions .......................................................................... 14

Introduction


It has been proposed that, since the origin of life and the ensuing evolution of biological species, a second evolutionary process has appeared on our planet. It is the evolution of culture—e.g. ideas, beliefs, and artifacts—and the creative minds that invent them, adapt them to new situations, and play with them for artistic expression and fun. But does culture evolve in the same genuine sense as biological life? And if so, does it evolve through natural selection, or by some other means? Why does no other species remotely approach the degree of cultural complexity of humans? These are questions that must be addressed because they lie at the foundation of who we are and what makes our lives meaningful.

Although much research has been done on how selective pressures operating at the biological level affect cognition and culture, little research has focused on culture as an evolutionary process in its own right. Nonetheless, culture does appear to evolve. Like biological forms, cultural forms—ideas, attitudes, artifacts, mannerisms, etc.—incrementally adapt to the constraints and affordances of their environment through descent with modification. Agricultural techniques become more efficient, computers get faster, scientific theories predict and account for more observations, new designs are often artistic spin-offs of those that preceded them. And in some respects culture appears to be Darwinian, that is, a process of differential replication and selection amongst randomly generated variants. For example, different brands of peanut butter may be said to compete to be ‘selected’ by consumers. This suggests that knowledge of biological evolution can be put to use to gain insight into cultural patterns. However, the attempt to straightforwardly apply Darwinian theory to culture has not been overwhelmingly fruitful. It certainly hasn’t provided the kind of unifying framework for the social sciences that Darwin’s idea of natural selection provided for the biological sciences. This is largely because the underlying substrate of the process—human beings—are notoriously complex and unpredictable! For example, natural selection cannot tell us much about how someone came up with the idea for turning peanuts into a spreadable substance in the first place!


The difficulty applying evolutionary theory as it has been developed in biology to culture arises largely because of the highly nonrandom manner in which the mind—the hub of cultural change—generates and assimilates novelty. To understand how, when, and why the human mind became capable of supporting culture, and what may have previously held it back, we need to know something about how we attained the creative powers we now possess, and how creative processes actually work, in groups as well as individuals. To invent in the strategic, intuitive manner characteristic of the human mind requires a cognitive architecture that supports the capacity to spontaneously adapt concepts to new circumstances and merge them together to conceptualize new situations. Thus we find that at the heart of the puzzle of how culture evolves lies the problem of concepts, not so much just how we use them to identify and classify objects in the world, but their contextuality and compositionality, and the creative processes thereby enabled.


We will see that the change-of-state a mind undergoes as it develops an idea is not a natural selection process, and indeed it may be that culture evolves, but only in small part through Darwinian mechanisms. We suggest that its basic mode of evolving turns out to be a more general process referred to as context-driven actualization of potential. Thus the story of how ideas are born and bred in one mind after another leads us to another story, that of what it means to evolve, and how an evolutionary process could work. Finally, this paper will touch on how an evolutionary perspective on culture can shed light on questions of a philosophical or spiritual nature that have been with us since the first fledgling creative insights glimmered in our ancestors’ brains.
 

1. Do Evolutionary Models Capture the Dynamics of Culture?


Let us consider how well attempts to formally or informally describe culture as an evolutionary process do at capturing the cultural dynamic.
 

1.1 Memes

Perhaps the most well known attempt to apply Darwinism to culture is the meme approach (Aunger 2000; Blackmore 1999, 2000; Dawkins 1976). It simplifies things by restricting what counts as ‘culturally transmitted’ to things that are passed from one person to another relatively intact, such as eye-catching fashions, or belief in God. This approach quickly runs into problems. First because ideas and stories are not simply stored, outputted, and copied by others as discreet chunks, complete unto themselves. They are dynamically influenced by the context in which they appear, and we process and re-process them in ways that reflect our unique experiences and unique style of weaving them into an internal model of the world, or worldview. Furthermore, the meme perspective leads us to view ourselves as ‘meme hosts’, passive imitators and transmitters of memes. Although some authors have capitalized on the shock value of the ensuing dismal view of the human condition, clearly we are not merely passive hosts but active evolvers of culture.
 

1.2 Mathematical Approaches

Others have drawn from mathematical models of population genetics and epidemiology to model the spread of ideas (Cavalli-Sforza & Feldman, 1981; Schuster & Sigmund, 1983; Boyd & Richerson, 1985). They examine the conditions under which mutated units of culture pass vertically via family, or horizontally through a community by imitation within an age cohort, and proliferate. The limitations of this approach are expressed succinctly by Kauffman (1999):
True, but impoverished. Why impoverished? Because the concept of meme, and its descent with modification is taken as a, or perhaps ‘the’ central conceptual contribution to the evolution of human culture. But the conceptual framework is so limited as to be nearly trivial. Like NeoDarwinism, it suffers from the inability to account for the source of new forms, new memes. Further, mere descent with modification is a vastly oversimplified image.

Consider the new concepts, artifacts, legal systems, modes of governance, modes of coevolving organizations at different levels that have come into existence in the past three million years. Our understanding of these and other aspects of culture transforms every day. Take, for instance, the Wright brother’s airplane. It is a recombination of four technological facts: an airfoil, a light gas engine, bicycle wheels, and a propeller. The more diversity that exists in a technological community, the more diversity of novel combinations of existing elements are present that might later prove useful in some context. Thus, 200,000 years ago, the diversity of the economic web of goods and services was severely limited. Today it is vast. 200,000 years ago, finding a technological novelty with the stone and bone implements available was hard. Today, with millions of artifacts already in existence, the generation of novel ones is easy.
 

In short, memes do not just descend with modification. A rich web of conceptual interactions is at work as humans happen upon, design, and implement a combinatorially exploding diversity of new goods and services. This WEB structure of technological and cultural evolution is far richer, and far closer to the truth, than mere meme descent with modification. Indeed, this broader view helps us begin to understand how and why memes recombine and diversify. It is a more generative picture, undoubtedly still inadequate, but far better than a naïve copying of neoDarwinism.

1.3 Computer Models

To what extent we can computationally abstract the underlying skeleton of the cultural process and actually evolve something with it? If culture, like biology, is a form of evolution, it should be possible to develop a minimal model of it analogous to the genetic algorithm, a biologically inspired search tool that evolves solutions to complex problems through a reiterated process of randomly mutating information patterns and selectively replicating those that come closest to a solution (Holland 1975). Meme and Variations (or MAV for short) is to our knowledge the first computer model of the process by which culture evolves in a society of interacting individuals. It is discussed only briefly here since it is presented in detail elsewhere (Gabora 1995). MAV consists of an artificial society of neural network-based agents that don’t have genomes, and neither die nor have offspring, but that can invent, assess, imitate, and implement ideas, and thereby gradually increase the fitness of their actions. Agents have an unsophisticated but functional capacity to mentally simulate or assess the relative fitness of an action before actually implementing it (and this capacity can be turned off). They are also able to invent strategically and intuitively, as opposed to randomly, building up ‘hunches’ based on trends that worked in the past (and this too can be turned off). This was possible because of the integrated structure of the neural network. All the agents’ concepts are connected, if indirectly, to one another, and thus each can influence, if only weakly, each other. The architecture of MAV is also such that it implements a cultural version of epistasis. In biological epistasis, the fitness conferred by one gene depends on which allele is present at another gene. In MAV, the fitness conferred by the locus determining the movement of one limb depends on what the other limbs are doing.

Initially all agents are immobile. Every iteration, each agent has the opportunity to acquire a new idea for some action, either through 1) innovation, by strategically modifying a previously learned idea, or 2) imitation, by copying an action performed by a neighbor. Quickly some agent invents an action that has a higher fitness than doing nothing, and this action gets imitated by others. As ideas continue to be invented, assessed, implemented as actions, and spread through imitation, the diversity of actions increases. Diversity then decreases as the society evolves toward implementing only those actions that are most fit.


MAV exhibits many phenomena observed in biology, such as drift—changes in the relative frequencies of different alleles (forms of a gene) as a statistical byproduct of randomly sampling from a finite population. Second, as in biology we find that epistasis increases the amount of time it takes to evolve. Third, although in the absence of variation-generating operations culture does not evolve, increasing innovation much beyond the minimum necessary causes average fitness to decrease, just as in biology.


MAV also addresses the evolutionary consequences of phenomena unique to culture. Imitation, mental simulation, and strategic (as opposed to random) generation of variation all increase the rate at which fitter actions evolve. The higher the ratio of innovation to imitation, the greater the diversity,  and the higher the fitness of the fittest action. Interestingly however, for the society as a whole, the optimal innovation-to-imitation ratio was approximately 2:1 (but diversity is compromised). For the agent with the fittest behavior, the less it imitated (i.e. the more effort reserved for innovation), the better. This suggests if you’re the smartest one around, don’t waste time copying what others are doing!
 

Thus it is possible to genuinely evolve information using a computer algorithm that mimics the mechanics of culture [1]. More recent computer models of cultural evolution (e.g. Spector & Luke, 1996a, b; Baldassarre, 2001) embed the cultural dynamic in a genetic algorithm. Thus agents not only exchange ideas but bear offspring and die. Although these models have unearthed interesting results concerning the interaction between biological and cultural evolution, we believe the first priority is to first learn what we can through computer simulations of culture alone before combining the two. After all, culture is not merely an extension of biology. Biology does not provide adequate explanatory power to account for the existence of widgets (just as physics cannot explain the existence of worms). Culture is spectacularly unlike anything else biological processes have given rise to. Indeed there is much left to do with such a culture-only modeling approach. Everyday experience suggests that human culture exhibits other phenomena observed in biological evolution that could be investigated with this kind of computer model, such as Founder Effect (stabilization in a  closed-off social group) and altruism (being especially nice to those who are related to you). In fact one could argue that humans feel more altruistic toward their ‘cultural kin’ than their biological kin. (For example, who would you go out of your way for the most: someone who has the same eye color or blood type as you, or someone who shares your interests?)

1.4 Where Do We Stand?
 

How well have we done at capturing what really happens in cultural evolution? At best, invention and imitation are modeled as single-step processes, in no way coming close to what really happens as a novel idea is churned through. There is a saying, ‘you never step into the same stream twice’, and it applies to streams of thought as well as streams of water. Units of culture are not retrieved whole and discreet from memory like apples from a box. Humans not only have the ability to blend and adapt ideas to new situations and see them in new perspectives, we are compelled to. And we are compelled to entice others to see things our way too, or to bat ideas around with one another, using each other as a mental scaffold. Moreover, just about anything is food for thought, and thus food for culturally transmittable behavior. Some items in memory, such as a recipe for goulash, may be straightforwardly transmitted through imitation. Others, such as, say, an attitude of racial prejudice, appear to be culturally transmitted, but it is impossible to point to any particular phrase or gesture through which this transmission is mediated. Still others partake in the cultural dynamic in even subtler ways, as when a composer releases the painful experience of his daughter’s death in a piece of music.
 

As an idea passes from one individual to another, it assimilates into the various minds it encounters, and these minds are altered to accommodate not only the idea but also what it may, perhaps only subtly, imply or suggest. An idea has a different impact on different individuals, depending on the beliefs and preconceptions already in place. Furthermore, individuals differ in the extent to which they process it, and thus the extent to which their worldview is affected by it and by its ‘halo’ of implications. They also differ in the extent to which their processing of the idea takes place alone or through interaction with others. There are individuals who are never directly exposed to the idea, but indirectly altered by it nevertheless, through exposure to others who are directly exposed. In short, the evolution of the ideas, stories, and artifacts that constitute culture is a subtle matter.

Notes:
1. MAV will be elaborated such that agents have a more realistic method of generating novelty, and multiple drives that are satisfied to different degrees by different actions, and the fitness function for the evaluation of an idea emerges from the drive strengths.
Go read the whole article.
Post a Comment