%20Commentary.png)
Introduction
Herbert Simon (1916-2001) is one of the most celebrated intellectuals of the twentieth century. His work bridged many disciplines and he was awarded two of the world’s highest academic honors: the Turing Award in 1975 and the Nobel Prize in economics in 1978. He is widely regarded as one of the fathers of artificial intelligence, making his work more relevant than ever. Even more remarkably, his work bridged the divide between the academic world and organizations functioning in everyday settings, as the title of his first book, Administrative Behavior, attests (Simon 1947/1997).
Simon and other twentieth-century pioneers have led to a new paradigm for thought and action in the twenty-first century, based on a combination of complex systems science and evolutionary science (Beinhocker 2006; Wilson and Kirman 2016; Wilson and Snower 2024). The word paradigm is overused and notoriously fuzzy. According to Masterman (1970), Thomas Kuhn used the word in 21 different senses all by himself. The substantive meaning employed here is “an interlocking set of ideas that makes sense of the world but has difficulty escaping its own assumptions.”
A comparison of our culturally evolved paradigms with our genetically evolved organs of perception is informative. We do not sense the world as it really is. Instead, we sense only what was relevant to our survival and reproduction in our ancestral past. In the case of vision, we sense only a narrow band of the electromagnetic spectrum, even though other species and instruments invented by us can sense the rest of the spectrum with ease. Even the portion of the spectrum that we can see, which exists as a continuum of wavelengths, is perceived as discrete colors to enhance our ability to forage, detect predators, and so on (Hubel 1988).
It is humbling to contemplate that our culturally evolved mental constructions, which we describe with words such as “worldview,” “meaning system,” “religion,” “theory,” or “paradigm” are designed primarily to help us function within our respective environments, not to directly apprehend what’s “out there.” Just as for our vision, to see in some respects requires becoming blind in other respects. Only a cultural system that is explicitly designed to generate objective knowledge (the scientific process) can overcome this limitation (Strevens 2020).
Against this background, a combination of complex systems science and evolutionary science provides a new way of seeing that is described as two “toolkits” in Table 1. Each is an interlocking set of ideas that can be applied to a diversity of seemingly disparate topics.
The complex systems toolkit is more general because it applies to both nonliving and living complex systems. To understand the nature of living complex systems, however, the evolutionary toolkit is essential.
Now we can look back to see how a pioneer such as Herbert Simon contributed to the new paradigm. Other pioneers include Friedrich Hayek, James Buchanan, and Elinor Ostrom, all of whom were awarded Nobel prizes in economics. While their careers overlapped and they might have occasionally cited each other, it is only in retrospect that we can trace how their work became interwoven. Also, no pioneer is clairvoyant and it is important to avoid the temptation of regarding everything they wrote as authoritative. Evolutionary biologists revere Darwin as a pioneer but also know that he was hopelessly muddled about the mechanisms of heredity and that the science that bears his name has gone beyond him in many other respects. This is the same attitude that we should cultivate for the pioneers of the new complexity/evolution paradigm.
I have spent a large part of my career helping to develop the new paradigm, updating its pioneers,1 and—like Simon—operating outside in addition to inside the Ivory Tower. In this article, I will attempt to update Herbert Simon’s corpus of work through the lens of the new paradigm that he did so much to create, using the third edition of his book The Sciences of the Artificial (Simon 1969/2019) and his autobiography (Simon 1991) as my main texts.2
Before proceeding, I will relate my own encounter with Simon in the late 1980s. He had taken an interest in the evolution of altruism at a time when the “selfish gene” perspective was dominant among evolutionary biologists. I was one of the very few proponents of multilevel selection theory, which explains the evolution of altruism as a product of selection among groups in a multigroup population, in contrast to selection among individuals within groups (Wilson and Wilson 2007; Wilson 2015). He invited me to be his house guest so we could discuss an idea of his, which explained the evolution of altruism as a form of docility (Simon 1990). I remember him showing me his collection of puzzles, including the Tower of Hanoi, which he famously wrote a computer program to solve (unlike my own efforts!). We talked late into the night, and he later invited me to co-author his article with him. I have the dubious distinction of declining his offer, for reasons that will be explained below.
In the following sections, I will first briefly describe some of the “tools” in the complexity and evolution “toolkits” and then show how Simon contributed to the development of the tools. I make no attempt to be comprehensive and encourage others who combine knowledge of the new paradigm and familiarity with Simon’s work to offer their own thoughts.
Generalized Darwinism
Generalized Darwinism includes any process combining the three ingredients of variation, selection, and replication. It is both old and new. Old, because Darwin and his contemporaries knew nothing about the mechanisms of variation and replication, so all their theorizing had to be mechanism-free. New, because once Mendelian genetics was discovered as one mechanism of variation and replication, it quickly became treated as the only mechanism (the so-called Modern Synthesis), as if the only way offspring can resemble their parents is by sharing the same genes. The broader study of Darwinian evolution didn’t resume until the 1960s (e.g., Campbell 1960, 1974) and didn’t gather steam until the turn of the twenty-first century. Today it is in full swing (e.g., Henrich 2015; Laland 2017; Wilson 2019; Muthukrishna 2023). Darwinian processes other than genetic evolution include epigenetic evolution (changes in gene expression rather than gene frequency), forms of social learning found in many species, and forms of symbolic thought that are distinctively human (Jablonka and Lamb 2006). Also, Darwinian processes can be intragenerational (e.g., the adaptive component of the immune system) in addition to intergenerational. An important contribution was B.F. Skinner’s (1981) article “Selection by Consequences,” where he pointed out the similarities between genetic evolution, human cultural evolution, and the operant conditioning that shapes the behavior of individuals.
Generalized Darwinism is in the process of transforming our understanding of human origins and history, leading up to the present. We can begin to see cooperation as the signature adaptation of our species and nearly everything that sets us apart from other species, including our capacity for symbolic thought, as a form of cooperation. We can also interpret human history as an increase in the scale of cooperation, with many reversals along the way (Turchin 2015). Finally, the same theoretical framework that can help us understand our past can help us work in the present to evolve a more cooperative future, ultimately at the global scale (Wilson et al. 2024).
Part of going beyond the study of genetic evolution includes rethinking the possibility that evolution can have an intentional component. The architects of the Modern Synthesis worked hard to assert that evolution has no purpose. Organisms simply vary and only the immediate environment does the selecting. In retrospect, this is an extremely restrictive view, not even consistent with the early history, such as Darwin’s reliance on artificial selection to explain natural selection and what became known as the Baldwin effect in the early twentieth century, whereby the selection pressures acting on genes are shaped by the learning abilities of individuals (Schneider 2014).
Today, it is obvious in retrospect that all three ingredients of a Darwinian process—variation, selection, and replication—can have intentional components and remain a Darwinian process. This goes without saying for genetic algorithms on computers, including John H. Holland and his 1975 book Adaptation in Natural and Artificial Systems. Today, artificial intelligence (AI) could just as well be called artificial evolution (AE).
Against this background, Simon’s book The Sciences of the Artificial, first published in 1969, seems to start out on the wrong foot by contrasting natural science (“most unequivocally” physical and biological science) and everything constructed by humans (the artificial). This might seem to imply that the artificial must be explained by a different set of laws than the natural, when the whole thrust of generalized Darwinism is to explain all things human, from our origin as a species to our artifacts, as subject to the same set of laws.
Soon enough, however, Simon acknowledges the similarity between adaptations that evolve by natural selection and human-made artifacts. Both are functionally organized in relation to a given environment (p. 6).
Notice that this way of viewing artifacts applies equally well to many things that are not man-made—to all things in fact that can be regarded as adapted to some situation, and in particular it applies to the living systems that have evolved through the forces of organic evolution.
Of course, the specific mechanisms whereby adaptations evolve by natural selection, compared to artifacts created by humans, can differ. The human process can be called rational, but only boundedly rational. Optimization was typically not possible, so satisficing would have to do. These are Simon’s signature concepts, critiquing the “Olympian” perspective of neoclassical economics and its unrealistic assumptions about human rationality.
The more bounded human rationality gets, the more the design of artifacts must be attributed to a variation/selection/replication process, which can range from conscious experimentation to a blind process of cultural evolution—many inadvertent social experiments, most that fall apart and a few that hang together. In this regard, Simon cited another pioneer of the new paradigm, Friedrich Hayek, quoting this passage from Hayek (1945):
The most significant fact about this system [markets] is the economy of knowledge with which it operates, or how little the individual participants need to know in order to be able to take the right action.
To summarize, Simon was a pioneer in seeing all forms of functional design, natural and artificial, as ultimately a product of variation/selection/replication processes. One contemporary author developing the same theme is Joseph Henrich and his 2015 book The Secret of Our Success: How Culture is Driving Human Evolution, Domesticating Our Species, and Making Us Smarter. The secret of our success is not our individual intelligence, but our capacity for cultural evolution.
The Ultimate/Proximate Distinction and Tinbergen's Four Questions
As stated earlier, the revolutionary nature of Darwin’s theory was appreciated and discussed for decades without any knowledge of the mechanisms of heritable variation. How was this possible? Because the novelty of the theory was to call attention to the shaping influence of the environment whenever organisms vary in a heritable fashion. To the extent that the physical makeup of organisms results in heritable variation, that is the extent to which it can be ignored.
For example, we can confidently predict that in desert environments, many species will evolve to be sandy colored to avoid being detected by their predators and prey. We can make this prediction for any color of sand (e.g., tan, black, white) and any type of species (e.g,. insects, reptiles, birds, mammals), even though the species have different genes and physical exteriors.
The same reasoning, called “Adaptationism” or “Natural selection thinking” can be used to form hypotheses about any other property of an organism relevant to its survival and reproduction. These hypotheses are not invariably correct, but they are excellent starting points for scientific inquiry. Amazing!
Of course, every adaptation has a physical basis that is also important to know about. The evolutionary biologist Ernst Mayr coined the terms “ultimate causation” and “proximate causation” for these two modes of explanation (Mayr 1961). At about the same time, the pioneering ethologist Niko Tinbergen observed that every product of evolution requires four explanations, concerning its function (if it is an adaptation), its history, its mechanistic basis, and its development during the lifetime of the organism (Tinbergen 1963). Tinbergen’s function and history questions correspond to Mayr’s ultimate causation, while the mechanism and development questions correspond to Mayr’s proximate causation.
Ultimate and proximate causation stand in a one-to-many relationship. For example, Richard Lenski (2011) selected for the ability to digest glucose in twelve isolated lines of the bacterium E. coli. The lines were initiated with E. coli from the same clone and therefore started out genetically identical. The environment of each line was also identical—ten millimeters of well-shaken laboratory medium with glucose as the only energy source, refreshed daily. All twelve lines evolved an improved ability to digest glucose (ultimate causation), but the particular mutations arose by chance and therefore differed between lines. Hence, each line had a different proximate mechanism for the same functional adaptation. More generally, both genetic evolution and human cultural evolution are replete with examples of convergent evolution, which have the same explanation in terms of ultimate causation but different explanations in terms of proximate causation.
Scholars can decide whether Simon was directly influenced by Mayr or Tinbergen (neither are cited in his autobiography or The Sciences of the Artificial), but he had a fine sense of what they were pointing out, which might well itself be a case of convergent evolution. In a section titled “The Environment as Mold” (p. 5) he stressed that functionally designed human artifacts, no less than products of natural selection, must be adapted to certain environments. The function of a clock is to tell time, but a nautical clock, a sundial, and a pocket watch are subject to different design constraints based on the environment that they are adapted to inhabit. In this passage, he makes the fundamental distinction between ultimate and proximate causation, beginning with an example similar to my desert example (p. 7).
Many animals in the Arctic have white fur. We usually explain this by saying that white is the best color for the Arctic environment, for white creatures escape detection more easily than do others. This is not of course a natural science explanation; it is an explanation by reference to purpose or function. It simply says that these are the kind of creatures that will “work,” that is, survive, in this kind of environment. To turn the statement into an explanation, we must add to it a notion of natural selection, or some equivalent mechanism.
An important fact about this kind of explanation is that it demands an understanding mainly of the outer environment. Looking at our snowy surroundings, we can predict the predominant color of the creatures we are likely to encounter; we need know little about the biology of the creatures themselves, beyond the facts that they are often mutually hostile, use visual clues to guide their behavior, and are adaptive (through selection or some other mechanism).
Analogous to the role played by natural selection in evolutionary biology is the role played by rationality in the sciences of human behavior.
Simon’s appreciation of ultimate causation did not deter him from paying equal attention to proximate causation. He was among the first to realize that proximate mechanisms of human cognition can be formulated and tested with computer simulation models. In this respect, he differed from B.F. Skinner, who also appreciated ultimate causation in his “Selection by Consequences” article, but stayed away from “opening the black box” of proximate mechanisms.
Dual Inheritance Theory
An important part of generalized Darwinism is Dual Inheritance Theory, which posits two streams of inheritance in our species: a genetic stream found in all species and a cultural stream that is distinctively human (Richerson 2017). While many other species have cultural traditions (Whiten 2021), we are distinctive in our capacity for symbolic thought, capable of conveying large amounts of learned information across generations. Symbolic thought, in turn, would be impossible without a high degree of cooperation. Cooperation came first in our origin as a species, and our capacity for symbolic thought evolved as a type of cooperation. A landmark book on this subject is The Symbolic Species by the evolutionary neuroscientist Terrence Deacon, published in 1998. For such a foundational book to appear less than 30 years ago illustrates the recency of the complexity/evolution paradigm, which I stressed earlier.
Thinking of our symbolic systems as the cultural equivalent of our genes has tremendous metaphorical transfer that is only beginning to be explored. Wilson et al. (2014) coined the word “symbotype” to stress the comparison with our genotypes, as shown in Figure 1. Each person has a collection of genes (their genotype), which influences just about everything that can be measured about them (their phenotype). Each person also has a symbolic system (their symbotype) that also influences just about everything that can be measured about them (their very same phenotype). A person’s genotype and phenotype interact with each other during the lifetime of the person (which is where epigenetics, or changes in gene expression, comes in). And the two streams of inheritance have been co-evolving with each other during our history as a species.
At any moment, a person’s combined genotype and phenotype result in a repertoire of behavior in response to their environment (phenotypic plasticity). Going beyond this repertoire requires changing the symbotype, genotype, or both. Apart from mutations, a person’s genotype does not change during their lifetime, at least until the advent of genetic engineering. In contrast, a person’s symbotype is much more responsive to change during their lifetime. I have already stressed that Darwinian evolution can be an intragenerational process in addition to an intergenerational process.
The academic literature identified by search terms such as “dual inheritance theory” and “gene-culture co-evolution” is small, much of it is theoretical, and much of the empirical research is macro in scale, such as long-term human history (e.g., Turchin 2015, Henrich 2020) and comparisons of modern societies at a national scale (e.g., Muthukrishna et al. 2021). Largely lacking, so far, are studies of human symbolic systems interacting with their environments at the micro-evolutionary scale. On this front, Simon was a true pioneer and still ahead of his time. For those who might assume that human artifacts are restricted to physical objects, he stresses the importance of symbols at the very beginning of The Sciences of the Artificial (p. 2).
Moreover for most of us…the significant part of the environment consists mostly of strings or artifacts called “symbols” that we receive through eyes and ears in the form of written and spoken language and that we pour out into the environment—as I am now doing—by mouth and hand. The laws that govern these strings of symbols, the laws that govern the occasions on which we emit and receive them, the determinants of their content are all consequences of our collective artifice.
In a section titled “Symbol Systems: Rational Artifacts,” he begins to explore the fertile ground by comparing the computer and the human (p. 22).
The computer is a member of an important family of artifacts called symbol systems, or more explicitly, physical symbol systems. Another important member of the family (some of us think, anthropomorphically, it is the most important) is the human mind and brain. It is with this family of artifacts, and particularly the human version of it, that we will be primarily concerned in this book. Symbol systems are almost the quintessential artifacts, for adaptivity to an environment is their whole raison d’etre. They are goal-seeking, information-processing systems, usually enlisted in the service of the larger systems in which they are incorporated.
Simon’s distinction between “inner” and “outer” corresponds to the symbotype-phenotype description outlined above (p. 6):
An artifact can be thought of as a meeting point—an “interface”, in today’s terms—between an “inner” environment, the substance and organization of the environment itself, and an ”outer” environment, the surroundings in which it operates.
Here is how he describes the interface as a computer programming task (p. 116; passages in brackets are mine):
The logic of optimization methods can be sketched as follows: The “inner environment” of the design problem is represented by a set of given alternatives of action [phenotypic plasticity]. The alternatives may be given in extenso: more commonly they are specified in terms of command variables that have defined domains. The “outer environment” [where the agent acts and is selected upon] is represented by a set of parameters, which may be known with certainty or only in terms of a probability distribution. The goals for adaptation of inner to outer environment are defined by a utility function—a function, usually scalar, of the command variables and environmental parameters—perhaps supplemented by a number of constraints (inequalities, say, between the functions of the command variables and environmental parameters). The optimization problem is to find an admissible set of values of the command variables, compatible with the constraints, that maximize the utility function for the given values of the environmental parameters.
It is an open question whether the symbotype-phenotype relationship of a computer program corresponds mechanistically to human symbotype-phenotype relationships. If they differ, they will need to be functionally equivalent to adapt the agent to its environment. This returns us to the one-to-many relationship between ultimate and proximate causation described earlier. To summarize, Simon’s work is foundational for studying Dual Inheritance Theory at the microevolutionary scale.
Multilevel Selection Theory
Naïve portrayals of natural selection often assume that adaptations evolve at the level of groups and multi-group societies as easily as at the level of individual organisms. The actual situation is more complex because selection is based on relative fitness. It doesn’t matter how well an individual survives and reproduces in absolute terms—only that it does so better than neighboring individuals. This places individuals who behave “for the good of the group” at a relative fitness disadvantage compared to more self-serving individuals who accept social benefits without reciprocating. It might seem that we can solve this problem by rewarding public good providers and punishing cheats, but rewarding and punishing are themselves public goods (called higher-order public goods) that place the enforcers at relative fitness disadvantage, compared to non-enforcers within the same groups.
Based on the perverse logic of relative fitness, we can conclude that adaptation at the level of groups requires a process of selection among groups in a multi-group population and tends to be undermined by selection within these same groups. The same goes for selection at the level of multi-group societies. The general rule is “Adaptation at a given level of a nested hierarchy of groups requires a process of selection at that level and tends to be undermined by selection at lower levels” (Wilson 2024). Self-preservation—a good thing—easily becomes self-dealing. Helping kith and kin—a good thing—easily becomes cronyism and nepotism. My nation first—a good thing—easily becomes an international conflict. Growing our economies—a good thing—easily leads to overheating the Earth.
Multilevel selection (MLS) theory, as this framework is called, began with Darwin and explains both the presence and absence of functional organization at various levels in biological and human systems (Wilson 2015). During the mid-twentieth century period that Simon was working, however, it was largely rejected by evolutionary biologists in favor of more reductionistic accounts that interpreted all adaptations as forms of individual and genetic self-interest (Williams 1966, Dawkins 1976). This individualistic perspective in evolutionary biology coincided with similar perspectives in economics (the rational actor model) and the social sciences (methodological individualism; Hodgson 2007). Individualism during this period was like a tide that lifted all disciplinary boats.
Simon was somewhat immune to these trends because he chose human organizations as his object of study, starting with his PhD thesis and first book, Administrative Behavior (Simon 1947/1997). He easily rejected neoclassical economics’ idealized notion of the firm in favor of the empirical study of real organizations. This made him part of the “New Institutional Economics”, along with other pioneers such as Douglas North, Oliver Williamson, Vincent Ostrom, and Elinor Ostrom. Ostrom’s work has become foundational for MLS theory applied to human institutions and will be briefly described before returning to Simon’s contributions.
The two key concepts from Ostrom’s work are core design principles (CDPs) for the governance of single groups and polycentric governance at the scale of multi-group cultural ecosystems. In her study of groups that attempt to manage common-pool resources, Elinor Ostrom discovered that some groups (not all!) were able to effectively self-govern, avoiding the famous “tragedy of the commons” (Hardin 1968), by implementing eight core design principles (Ostrom 1990). My collaboration with her and her (then) postdoctoral associate, Michael Cox (Wilson, Ostrom, and Cox 2013) generalized the core design principles from an MLS perspective to include nearly all cooperative endeavors. Ostrom’s wording of the CDPs, the generalized wording, and the connection with MLS theory are shown in Table 2. CDP1 defines the boundaries of the group, its membership, and its purpose. CDP2-6 regulates social interactions within the group, coordinating prosocial activities and suppressing the potential for disruptive self-serving behaviors, making the group the primary unit of selection. CDP7-8 ensure appropriate autonomy for the group and apply the same principles to between-group interactions. In other words, the CDPs are scale-independent, applying to all levels of a multi-tier social hierarchy, which is a tremendous conceptual simplification.
The concept of polycentric governance notes that: 1) life consists of many spheres of activity; 2) each sphere has an optimal scale; and 3) good governance requires finding the right scale for each sphere and appropriately coordinating among the spheres (McGinnis 1999; Ostrom 2010). Stated so simply, it can scarcely be otherwise, but that’s not how most polycentric systems are governed!
One implication of polycentric governance that appears obvious in retrospect is that many spheres of activity are neither markets, states, nor business firms. The “none of the above” category includes most of the small groups of our lives—our families, neighborhoods, schools, churches, and volunteer activities. The generalized core design principles can include these social lifeforms, along with those more narrowly associated with economics, within the same theoretical framework (Aligica 2018).
Simon’s approach to organizations, unsurprisingly, was computational. He regarded human society as regulated by many coordination mechanisms, not just markets, with business firms and other organizations as another form of coordination (p. 31):
Most economic units in capitalist societies are mostly business firms, which are themselves hierarchic organizations, some of enormous size that make almost negligible use of markets in their internal functioning. Roughly eight percent of the human economic activity in the American economy, usually regarded as almost the epitome of a “market” economy, takes place in the internal environments of businesses and other organizations and not in the external, between-organization environments of markets. To avoid misunderstanding, it would be appropriate to call such a society an organization & market economy; for in order to give an account of it we have to pay as much attention to organizations as to markets.
Strikingly, Simon identifies loyalty as a key feature of organizations, converging upon Elinor Ostrom’s first core design principle (strong sense of identity and purpose).
Brief mention was made earlier of a crucial reason why so much human activity takes place within organizations: people acquire loyalty, and often a large amount of loyalty, to the groups, including organizations, to which they belong.
With this simple and self-evident observation, Simon goes beyond neoclassical economics’ reliance on individual utility maximization as the sole human preference. The reason he could do this with ease was that he had rejected the entire mathematical edifice of neoclassical economics, which requires such a simplifying assumption.
Loyalty is not only motivationally advantageous for an organization; it is also cognitively advantageous. In other words, if members of a group are loyal to a single collective goal and have subordinated their more narrow self-interests, the task of coordinating their behavior is simpler than if their narrow self-interests are primary. Simon appreciates that this only solves the problem of within-group coordination, leaving open the potential for between-group conflict (p. 44).
Organizational loyalty is perhaps better labeled identification, for it is both motivational and cognitive. The motivational component is an attachment to group goals and a willingness to work for them even at some sacrifice to personal goals (in effect, group goals become personal goals). The ethnic conflict we observe in many parts of the world provides vivid evidence of this attachment to group goals and the differential treatment it generates between “we” and “they.”
Group members need not only a primary identity for the whole group, but also secondary identities for the roles that they play within the group. They also need primary and secondary identities for all the other groups in their lives. Our capacity for cooperative social behavior is so great that we can participate in many groups simultaneously, recognizing the appropriate contexts. The entire concept of social identity can be interpreted as a cognitive adaptation for coordinating behavior in a given context. This appreciation is so native to Simon’s way of thinking, yet so incompatible with the strictures of neoclassical economics, that contemporary economists such as Akerlof and Kranton (2011) and Hoff and Stiglitz (2016) still must struggle to introduce the concept of social identity to the economics profession.
To summarize, Simon’s work on single organizations and multigroup cultural ecosystems is as pioneering and foundational for modern MLS theory as the work of Elinor and Vincent Ostrom. This brings us to my interaction with Simon in the late 1980s, his invitation to co-author his article, and my polite refusal. As previously mentioned, I was part of a small minority working to revive MLS theory, which explains the evolution of altruism and other forms of prosociality at face value. Later, I was to combine forces with the evolutionary biologist Edward O. Wilson in a 2007 review article titled “Rethinking the Theoretical Foundation of Sociobiology,” which concluded with the passage “Selfishness beats altruism within groups. Altruistic groups beat selfish groups. Everything else is commentary.”
Back in the late 1980s, Simon was familiar with my work reviving MLS theory but had another idea about how altruism might evolve within a single group, based on a combination of docility and bounded rationality, which he eventually published in the journal Science (Simon 1990). By docile, Simon meant a tendency to accept the social influence of others, which he regarded as individually advantageous. Docile individuals might still be selfish or altruistic, so Simon invoked bounded rationality to explain why selfish docility is not an option.
Under these assumptions, an individual who is docile, enjoying the advantage (d) of that docility, will consequently also accept society’s instructions to be altruistic as part of proper behavior. Because of bounded rationality, the docile individual will often be unable to distinguish socially prescribed behavior that contributes to fitness from altruistic behavior. In fact, docility will reduce the inclination to evaluate independently the contributions of behavior to fitness. Moreover, guilt and shame will tend to enforce even behavior that is perceived as altruistic. Hence, the docile individual will necessarily incur the cost, c, of altruism.
I regarded Simon’s idea as intriguing in some respects but also full of unstated assumptions that would need to be unpacked, including the invocation of shame and guilt, which themselves require an explanation from a MLS perspective. The invocation of bounded rationality to exclude a behavioral strategy (selfish docility) that would otherwise be advantageous struck me as especially far-fetched. Every human group is vulnerable to passive free-riding and more active forms of self-serving behavior that are not good for the group! Every group is also vulnerable to coordination failures, even if all members have prosocial intentions. Simon appreciated these points in other contexts, so his docility model seemed inconsistent with other aspects of his own work. There was so much to work through that I elected not to become associated with his model.
In recent years, the concepts of “social selection” and “self-domestication” have become hot topics for other species in addition to humans (Boehm 2011; Wrangham 2019; Hare and Woods 2020). While these share some aspects of Simon’s docility model, they are much more attentive to the need for social control mechanisms as public goods that require higher-level selection to evolve.
Complex Systems Science
Up to now, I have focused on Simon’s contributions to the “tools” of the evolutionary science “toolkit,” which is my own primary area of expertise. His contributions to the “tools” of the complex systems “toolkit” are so numerous that only the briefest summary will be provided here. He pioneered the concept of simple interactions leading to complex outcomes, most memorably in his fanciful example of an ant walking on a beach (p 51). The ant’s path is complex as it navigates the little hills and valleys of its landscape, but the cognitive rules that guide the ant on its path are simple. The complexity of the path is due to the interaction between the simple rules of the ant and the complexity of the environment. Simon’s daring hypothesis was that human cognitive rules are not much different than the ant’s.
In Chapter 7 of the third edition of The Sciences of the Artificial, Simon describes the history of complex systems science as recurrent “bursts of interest” during the twentieth century (p. 169).
This century has seen recurrent bursts of interest in complexity and complex systems. An early eruption, after World War I, gave birth to the term “holism” and to interest in “Gestalts” and “creative evolution.” In a second major eruption, after World War II, the favorite terms were “information”, “feedback”, “cybernetics”, and “general systems”. In the current eruption, complexity is often associated with “chaos,” adaptive systems,” “genetic algorithms,” and “cellular automata.”
With respect to holism, Simon distinguishes versions that deny reductionism altogether from versions that are compatible with reductionism. While agreeing with his account, I think it is important to appreciate how far ultimate causation goes toward holism without requiring a reductionistic account, even though a reductionistic account is available in every case. The holistic statement “the properties of the parts permit, but do not cause, the properties of the whole” is truly the case when the parts result in heritable variation and the whole is shaped by selection (Wilson 1988). In Lenski’s experiment with E. coli described earlier, the proximate mechanisms that evolved in any given line were not necessary and indeed did not evolve in the other lines. If that’s not a whole that cannot be reduced to its parts, what would be?
Simon appreciated that these “eruptions” of interest were dependent upon the tools for studying them. The current eruption simply could not have taken place without the advent of widespread computing ability. This is why the new complexity/evolution paradigm is genuinely new in the history of human thought as a way of seeing and acting upon the world.
Having rejected the “Olympian” aspirations of neoclassical economic theory, Simon also rejects similar aspirations for complex systems science (p. 173):
During these postwar years, a number of proposals were advanced for the development of “general systems theory,” that, abstracting from the special properties of physical, biological, or social systems, would apply to all of them. We might well feel that, while the goal is laudable, systems of these diverse kinds could hardly be expected to have any nontrivial properties in common….If a general systems theory is too ambitious a goal, it might still not be vain to search for common properties among broad classes of complex systems.
In other words, Simon is arguing for a relatively humble and pragmatic “toolkit” approach, very much in keeping with what I have presented in Table 1.
Conclusions
Human history, including the history of ideas, is the fossil record of human cultural evolution. It is important for it to be understood in as much detail as possible, with the same care—and much the same toolkit—that paleontologists bring to the study of the biological fossil record. A careful accounting of the history of ideas goes beyond academic interest and is integral to the current use of the ideas to translate thought into action. My reflection on Herbert Simon is written in this spirit.
Notes:
1. See Wilson (2020) for updating Friedrich Hayek, Wilson (2023) for updating Pierre Teilhard de Chardin, and Wilson et al. (2013) for updating Elinor Ostrom.
2. Throughout this article, page numbers refer to Simon (1969/2019).
References:
Akerlof, G. A., & Kranton, R. E. (2010). Identity Economics: How Our Identities Shape Our Work, Wages, and Well-Being. In Identity Economics. Princeton University Press. https://doi.org/10.1515/9781400834181
Aligica, P. D. (2018). Public Entrepreneurship, Citizenship, and Self-Governance. Cambridge University Press. https://doi.org/10.1017/9781316888728
Beinhocker, E. D. (2006). Origin of Wealth: Evolution, Complexity, and the Radical Remaking of Economics. Harvard Business School Press.
Boehm, C. (2011). Moral Origins: The Evolution of Virtue, Altruism, and Shame. Basic Books.
Campbell, D.T. (1960). Blind variation and selective retention in creative thought and other knowledge processes. Psychological Review, 67, 380–400.
Campbell, D. T. (1974). Evolutionary epistemology. In P. A. Schilpp (Ed.), The Philosophy of Karl Popper (pp. 413–463). Open court publishing.
Dawkins, R. (1976). The Selfish gene (1st ed.). Oxford University Press.
Deacon, T. W. (1998). The Symbolic Species: The Co-evolution of Language and the Brain. W. W. Norton & Company.
Hardin, G. (1968). The Tragedy of the Commons. Science, 162, 1243–1248.
Hare, B., & Woods, V. (2020). Survival of the Friendliest: Understanding Our Origins and Rediscovering Our Common Humanity. Random House.
Hayek, F. A. (n.d.). The Use of Knowledge in Society. Ameican Economic Review, 35, 519–530.
Henrich, J. (2015). The Secret of Our Success: How culture is driving human evolution, domesticating our species, and making us smarter. Princeton University Press.
Henrich. (2020). The WEIRDEST People in the World: How the West Became Psychologically Peculiar and Particularly Prosperous. Farrar, Straus & Giroux.
Hodgson, G. M. (2007). Meanings of methodological individualism. Journal of Economic Methodology, 14(2), 211–226. https://doi.org/10.1080/13501780701394094
Hoff, K., & Stiglitz, J. E. (2016). Striving for balance in economics: Towards a theory of the social determination of behavior. Journal of Economic Behavior & Organization, 126, 25–57. https://doi.org/10.1016/j.jebo.2016.01.005
Holland, J. H. (1975/1992). Adaptation in Natural and Artificial Systems: An Introductory Analysis with Applications to Biology, Control, and Artificial Intelligence. Bradford Books.
Hubel, D. H. (1988). Eye, Brain, and Vision. W.H. Freeman. http://www.amazon.com/Brain-Vision-Scientific-American-Library/dp/0716750201/ref=pd_sim_b_1?ie=UTF8&refRID=1A84WEWA12V0DNABHTMP
Jablonka, E., & Lamb, M. J. (2006). Evolution in Four Dimension: Genetic, Epigenetic, Behavioral, and Symbolic Variation in the History of Life. MIT Press.
Laland, K. N. (2017). Darwin’s unfinished symphony: How culture made the human mind (p. 450). Princeton.
Lenski, R. E. (2011). Evolution in Action: A 50,000-generation salute to Charles Darwin. Microbe, 6, 30–33.
Masterman, M. (1970). The Nature of a Paradigm. In A. Musgrave & I. Lakatos (Eds.), Criticism and the Growth of Knowledge: Proceedings of the International Colloquium in the Philosophy of Science, London, 1965 (Vol. 4, pp. 59–90). Cambridge University Press. https://doi.org/10.1017/CBO9781139171434.008
Mayr, E. (1961). Cause and Effect in Biology. Science, 134(3489), 1501–1506.
McGinnis, M. D. (1999). Polycentric Governance and Development: Readings from the Workshop in Political Theory and Policy Analysis.
Muthukrishna, M. (2023). A Theory of Everyone: The New Science of Who We Are, How We Got Here, and Where We’re Going. The MIT Press.
Ostrom, E. (1990). Governing the commons: The Evolution of institutions for collective Action. Cambridge University Press.
Ostrom, E. (2010). Beyond Markets and States: Polycentric Governance of Complex Economic Systems. American Economic Review, 100, 1–33.
Peter J. Richerson. (2017). Recent Critiques of Dual Inheritance Theory. Evolutionary Studies in Imaginative Culture, 1(1), 203–203. https://doi.org/10.26613/esic.1.1.27
Scheiner, S. M. (2014). The Baldwin Effect: Neglected and Misunderstood. The American Naturalist, 184(4), 130–132. https://doi.org/10.1086/677944
Simon, H. A. (1947/1997). Administrative Behavior, 4th Edition. Free Press.
Simon, H. A., & Laird, J. E. (1969/2019). The Sciences of the Artificial, reissue of the third edition with a new introduction by John Laird. The MIT Press.
Simon, H. A. (1990). A mechanism for social selection and successful altruism. Science, 250, 1665–1668.
Simon, H. A. (1991). Models of My Life: The Remarkable Autobiography of the Nobel Prize Winning Social Scientist and Father of Artificial Intelligence. Basic Books.
Skinner, B. F. (1981). Selection by Consequences. Science, 213, 501–504.
Strevens, M. (2020). The Knowledge Machine: How Irrationality Created Modern Science. Liveright.
Tinbergen, N. (1963). On aims and methods of ethology. Zeitschrift Für Tierpsychologie, 20, 410–433.
Turchin, P. (2015). Ultrasociety: How 10,000 years of war made humans the greatest cooperators on earth. Baresta Books.
Whiten, A. (2021). The burgeoning reach of animal culture. Science (New York, N.Y.), 372(6537). https://doi.org/10.1126/science.abe6514
Williams, G. C. (1966). Adaptation and Natural Selection: A critique of some current evolutionary thought. Princeton University Press.
Wilson, D. S. (1988). Holism and reductionism in evolutionary biology. Oikos, 53(269–273).
Wilson, D. S. (2015). Does Altruism Exist? Culture, Genes, and the Welfare of Others. Yale University Press.
Wilson, D. S. (2019). This View of Life: Completing the Darwinian Revolution. Pantheon/Random House.
Wilson, D. S. (2020). Updating Hayek on Cultural Multilevel Selection. In S. Haeffele, S. M. Stein, & V. H. Storr (Eds.), Hayek’s Tensions: Reexamining the Political Economy and Philosophy of F.A. Hayek (pp. 151–172). Mercatus Center.
Wilson, D. S. (2023). Reintroducing Pierre Teilhard de Chardin to modern evolutionary science. Religion, Brain & Behavior, 1–15. https://doi.org/10.1080/2153599X.2022.2143399
Wilson, D. S. (2024). Williams’ Rule and Its Relevance for Positive Change Efforts. This View of Life. https://doi.org/10.5281/ZENODO.10826558
Wilson, D. S., Hayes, S. C., Biglan, A., & Embry, D. (2014). Evolving the Future: Toward a Science of Intentional Change. Behavioral and Brain Sciences, 37, 395–460.
Wilson, D. S., & Kirman, A. (2016). Complexity and Evolution: Toward a New Synthesis for Economics (D. S. Wilson & A. Kirman, Eds.). MIT Press.
Wilson, D. S., Madhavan, G., Gelfand, M. J., Hayes, S. C., Atkins, P. W. B., & Colwell, R. R. (2023). Multilevel cultural evolution: From new theory to practical applications. Proceedings of the National Academy of Sciences, 120(16), e2218222120. https://doi.org/10.1073/pnas.2218222120
Wilson, D. S., Ostrom, E., & Cox, M. E. (2013). Generalizing the core design principles for the efficacy of groups. Journal of Economic Behavior & Organization, 90, S21–S32. https://doi.org/10.1016/j.jebo.2012.12.010
Wilson, D. S., & Snower, D. J. (2024). Rethinking the Theoretical Foundation of Economics I: The Multilevel Paradigm. Economics, 18(1). https://doi.org/10.1515/econ-2022-0070
Wilson, D. S., & Wilson, E. O. (2007). Rethinking the theoretical foundation of sociobiology. Quarterly Review of Biology, 82, 327–348.
Wrangham, R. (2019). The Goodness Paradox: The Strange relationship between virtue and violence in human evolution. Pantheon.
