Underappreciated features of cultural evolution

(a). Storage

The way in which information is stored differs fundamentally between cultural and biological evolution. In biological evolution, information (genetic traits) is encoded in and stored as the sequence of physical molecular structures, the genome. By contrast, in cultural evolution, cultural information within individuals (cultural traits) takes the form of mental representations, which are encoded as neuronal connections in the brain. This difference in the nature of trait storage means that an individual cannot change its genome, but can make changes to the composition of its cultural traits. For example, while genes can be frequently and accurately accessed and copied, there are very limited possibilities of passing information to the genome that can be inherited by offspring. By contrast, mental representations cannot be directly observed or copied [47], but due to the high flexibility of the nervous system, mental representations can be continuously altered before they are exhibited and transmitted to others. These differences have consequences for the evolutionary dynamics and concerns several of the other items in the table, which we will address in §3.

Since artefacts (cultural products) reside in the external world, all the information for producing them often need not be contained within individuals and can be easily transmitted and accumulated over generations. Furthermore, in the light of our ability to externalize information storage, computation and filtering (and in part creation), an increasingly important question is how information technology changes cultural evolution as well as our mental abilities (e.g. [29]).

(b). Acquisition

Based on the difference in the nature of trait storage in cultural and biological evolution, the acquisition of traits is also fundamentally different. Genetic traits are in place at the beginning of an individual’s life, whereas cultural traits are acquired throughout the entire life. An important consequence of continuous acquisition is that culture is acquired sequentially through a combination of episodes of social learning (learning from others), and individual practice (trial and error, insight, etc.). The importance of individual practice cannot be understated. It has been demonstrated in numerous studies in developmental psychology and educational science that many cultural traits require substantial individual efforts to be successfully learned (e.g. [48,49]). For example, hunting skills require extensive social learning to acquire the fundamental behaviours and techniques but they also rely on lifelong improvement through individual learning [50]. This difference between cultural and biological evolution has received rather limited attention, but it may be one of the most important ones influencing evolutionary dynamics [51,52]. As we will discuss below, the fact that humans are born without culture and acquire culture sequentially, making the individual an active force in the transmission, has several consequences for the dynamics of cultural evolution, including, for instance, the fact that previously acquired traits can facilitate or even preclude the subsequent acquisition of other traits, and that an individual may transmit different trait variants during different stages of its life. The fact that culture is acquired sequentially does not exclude the importance of inborn predispositions or abilities in the acquisition of culture.

(c). Transmission

(d). Sorting processes

Sorting processes are processes that lead to a non-random change in the frequency of traits. In biological evolution, the main sorting process is natural selection. How it operates is usually clear and well understood, and it brings a general understanding of biological adaptations. In cultural evolution, the situation is considerably more complex, which is in part due to the active role each individual takes in acquiring, exhibiting and transmitting cultural traits [17]. The most explored sorting mechanism that involves the individual in this way, operates in the receiver during social learning, in which content, model or conformity biases favour the acquisition of certain traits while rejecting others [67]. A trait may also be favoured or rejected subsequently when individuals get experiences of using the trait ([66], this issue). Sorting will also occur if individuals selectively display traits to potential learners, and when the generation of new trait variants is incentive-driven. Individuals may also exercise social control by, for instance, trying to influence what others do, or communicate by selectively punishing or rewarding such acts [68–70]. In addition, there is natural selection in cultural evolution, since cultural traits can influence survival and reproduction of their carriers and so affect trait frequencies [15,17]. For instance, a habit that increases survival may spread because it will make carriers of the trait more common among demonstrators. For the same reason, a habit that decreases survival will gradually disappear.

In summary, sorting is possibly the most complex part of the evolutionary dynamics of cultural evolution, as it involves a number of different processes, and several of these occur in combination with other key processes such as transmission and generation of new culture.

(e). Generation of variation

The continuous generation of variation is a central aspect of evolution. A generally recognized difference between biological and cultural evolution is that genetic variation is random and unrelated to any need or purpose. By contrast, in cultural evolution the generation of new variants can also be incentive-driven. This means that it is directed towards solving a concrete problem or reaching a certain goal. Important mechanisms include cognitive processing [71], deliberate experimentation [72–74], as well as trial-and-error and associative learning processes [71,75]. However, cultural variants may also occur because of mistakes, transmission errors, by-products of trying to solve other problems, or other circumstances that produce results that the individuals may learn from and subsequently transmit to other individuals [72]. Note that successful learning from unintended or unexpected outcomes is not free from incentive-driven aspects as it requires that the individual recognizes how such observations can be used.

(a). Potential for fast evolution and complex culture

Cultural evolution can proceed much faster than biological evolution, both in rate of change and in terms of trait complexity. This has been pointed out previously, and several reasons are generally given for why this is the case [1,11,15,17,67,76,77]. Firstly, acquisition and transmission of cultural traits occur along many pathways and throughout an individual’s entire lifetime. Secondly, novel or modified traits can be transmitted directly to other individuals (which is why cultural evolution is sometimes described as Lamarckian rather than Darwinian). Thirdly, the generation of novel variants can be incentive-driven. This may have a dramatic effect on the rate of evolution, since cultural traits will be much more likely to be productive once they ‘enter’ the real world. And, finally, language can be an important tool for efficient cultural transmission. It is necessary for transmitting abstract ideas, beliefs and values, and is a tool for the transmission of complex cultural traits, such as opaque processes, inaccessible observations, practices, algorithms and other mental skills.

A full understanding of evolutionary rates and complex culture also requires a detailed understanding of cumulative cultural evolution. Cumulative culture is not only the accumulation of traits but also the innovation of new traits by combining or recombining existing traits [78–80]. This can result in a gradual increase in complexity, as combinations can be recombined continuously, and may offer a better theory of how cultural complexity and cultural systems emerge (see §3d). Also, it could explain, through the power of combinatorics, how cultural evolution can generate a never-ending stream of new variation.

For a better understanding of the role of language and recombination in cumulative cultural evolution, future research needs to address several important questions, such as:

• i.

  What are the combinatorial properties that allow a seemingly unbounded generation of diversity and complexity in human cumulative cultural evolution?

• ii.

  What is the role of language, symbolism, and cognitive abilities in cumulative cultural evolution?

• iii.

  Can all types of cultural change be understood in evolutionary terms (listed in table 1), or do we need additional theoretical frameworks?

(b). Developmental processes are directly involved in acquisition and transmission

Development here refers broadly to the processes that shape and modify the individual during its lifetime. In biological evolution, development is usually separated from the processes that transmit genetic traits and its evolution is studied in life-history theory, which emphasizes the role of trade-offs [81]. By contrast, in cultural evolution, development and transmission processes tend to be more intertwined, as they occur on the same time scale. On the one hand, an individual’s experience is shaped by sequential encounters with its surrounding cultural environment. On the other hand, the individual itself contributes to the makeup of that cultural environment.

Relevant progress in our understanding of developmental processes in humans has been made in a variety of disciplines, such as developmental psychology, educational science and linguistics [48,49]. Several general findings from empirical research in these fields are relevant to cultural evolution. First and foremost, development is generally incremental and time-consuming, starting with learning simple features and elements that are later integrated into a functional practical or mental skill. This is obvious for advanced skills like mental arithmetic [82–84], and reading and writing, but also required for more basic skills (e.g. [82,85,86]).

A second key finding is that traits are typically acquired through a combination of social interactions and individual efforts [48,49,87,88]. Thus, social and individual learning are operating together rather than being alternative strategies. Furthermore, research on development has also clearly established that there is an extensive behavioural and mental flexibility in humans. This allows humans to acquire a large range of skills, many of recent origin and not part of our prehistory, like reading and writing, agricultural practices, playing chess, and computer programming. Performance is generally strongly correlated to the amount of training, for both domain-specific and domain-general skills [89]. Thus, psychology and educational science suggest that when we see faithful copying over generations, this depends on substantial learning efforts among individuals. The situation is complicated by findings showing that learning and memory abilities themselves respond to training, and are thus partly cultural and subject to cultural evolution [48,89,90].

Previous theoretical models of cultural evolution typically use simplified assumptions about individual development and rely on simple copying mechanisms and strategies for social learning [91–94], or treat social and individual learning as opposing alternatives [17,95], whose evolution is generally studied in the absence of a learning–reproduction trade-off (see [96] for how classical results are critically changed once the learning–reproduction trade-off is taken into account).

It is thus becoming increasingly clear that future models should incorporate developmental processes, especially for traits that cannot be acquired in a single learning episode (because they are combinations of simpler traits or because they require individual improvement). These models should further be embedded into a proper life-history context so as to make all traits pertaining to learning co-evolutionary [97].

Challenges and questions for future research:

• i.

  How can developmental psychology and related fields inform cultural evolution research? For example, in the debate between the ‘California’ and the ‘Paris’ school on why cultural learning can be faithful to its source and maintained over generations, while single transmission events are error-prone [53,54,98,99], developmental psychology might explain faithful transmission in terms of sequential learning including efficient corrective feedback loops.

• ii.

  What is the scope of cultural evolution in terms of mental abilities? To what extent are mental abilities and psychological mechanisms learnt from or influenced by the cultural environment [100,101], and thus subject to cultural evolution themselves [30]? Future models that combine developmental processes and cultural evolution may help to distinguish the roles of inborn factors, acquired traits and individual experience.

• iii.

  How does learning depend on an individual’s lifetime and how has it co-evolved with life history?

(c). Different evolutionary outcomes and cultural fitness

In biological evolution, the fate of a trait can be understood by considering the average reproductive success of all carriers of that trait [102,103]. It is common to use optimization and game theoretic approaches to study evolutionary endpoints (i.e. adaptations). By contrast, in cultural evolution there is no generally accepted formulation of cultural fitness [104] as the relationship between the individual and its traits is more complicated. Here, transmission is not tied to biological reproduction, can occur between arbitrary individuals, and the individual is not fixed with respect to the cultural traits it possesses and exhibits to others during its life. These differences have implications both for the outcome of evolution and for how cultural evolution should be theoretically explored [11,15,17,52]. In principle, anything that influences transmission, retention and use of cultural traits needs to be considered.

For instance, traits that are contagious enough can spread and be maintained in a population even if they reduce the survival and/or biological reproduction of individuals [105–107]. Suppose lifestyle A produces more children than lifestyle B and that parental lifestyles are often transmitted to the children. However, children are also exposed to the alternative lifestyle B, which is more attractive (e.g. because it is easier to learn). This may be driven by a genetic or cultural bias for this lifestyle, and could result in B increasing in frequency compared to A. Figure 1 provides a numerical example. A complete analysis of cultural evolution scenarios of this kind requires exploration of both the evolution of lifestyles and the evolution of preferences for lifestyles [108].

In any human society, we find a range of both material and non-material culture serving the well-being, survival and reproduction of individuals or the group. However, we can also observe phenomena that are more difficult to interpret as adaptive in this way. We spend significant amounts of time learning to play musical instruments and chess. Often, our actions and beliefs do not even promote the goals they are aimed at, like when anti-vaxxers refuse medically motivated vaccinations in order to promote health [109].

It is quite possible that concepts like adaptation and fitness in cultural evolution will not have equivalent and as straightforward meanings as they do in biological evolution. The complication is that individuals are subject to change during their lives, which can be illustrated by comparing a trait that transmits rather slowly to others, but tends to stay with an individual once it is acquired, with a trait that transmits quickly to other individuals, but is easily abandoned or replaced. The latter trait may be easy to learn, or has some immediate attraction, but is less productive. Which trait spreads can depend on trait content and thus may not be predictable by considering an elementary set of rules of transmission [110].

It has been suggested that the fitness of cultural traits can be based on just the traits or free-floating memes [104]. While this can be a useful approach for certain questions, a general theory of cultural evolution cannot ignore the individual because of its active roll in acquiring, processing and exhibiting cultural traits [104]. This is one of the reasons that make it so tricky to find a useful concept of units of evolution in cultural evolution (a unit might be the single cultural trait or a set of traits, but also the carrier of a trait, etc.).

Challenges and questions for future research:

• i.

  Is it possible to provide a general theoretical account for the observed diversity of outcomes of human cultural evolution?

• ii.

  How can we define and recognize cultural adaptations, at what level are they manifested, and how can we use them to identify useful units of cultural evolution?

• iii.

  What could be a useful concept of fitness in cultural evolution, and, if it exists, how can it be used to study cultural evolution?

(d). Trait interactions, cultural systems and evolving evolutionary dynamics

The facts that individuals acquire traits sequentially and that traits can interact may have important organizing effects. The resulting developmental path dependence (see §3b) can limit which traits an individual acquires, and so promote the evolution of cultural systems [22,55,66,111,112]. For instance, an individual might acquire a belief A about the world, which subsequently facilitates the acquisition of other beliefs that are consistent with A but leads to the rejection of beliefs that are not [113]. If, instead, the individual would first acquire an alternative belief B, then this could trigger the eventual acquisition of an entirely different belief system. Similar reasoning can apply to other categories of traits. In fact, there are many ways in which traits can interact and influence evolution. Traits may or may not operate well together (within or between individuals), they might express preferences towards other traits, and define goals the individual strives for that are better served by certain traits than others. Traits can also be clustered in space (e.g. co-occurrence of traits relevant for fishing near bodies of water) or time (e.g. seasonal occurrence of sets of traits), and may even influence social network structure, and thus lead to exposure to different sets of traits [62]. The fact that culture exists in the form of systems, contexts or structures of interrelated parts, manifested both in individuals and in groups of individuals, is generally recognized in the humanities and social sciences [22]. It is possible that many cultural phenomena can only be understood from an explicit systems perspective.

Trait interactions could also have even more pronounced effects, and bring about changes to the evolutionary process itself, that is, important aspects of cultural evolution dynamics may themselves be results of cultural evolution rather than determined by genetic factors or external circumstances [22,114,115]. Examples of such changes in biological evolution are migration, mutation and sexual reproduction, which are called modifier traits. These traits have important consequences for the dynamics of biological evolution, but they are also products of that same evolution, thus resulting in counterintuitive feedbacks [116]. All items in table 1 are potentially subject to cultural evolution, and it has been suggested that cultural evolution itself can influence social learning abilities [1], efforts by senders’ and receivers’ willingness to change [114,117], and transmission pathways [62]. Culturally established goals could have a dramatic effect on the outcome of cultural evolution. Acerbi et al. [114] introduced the term regulatory traits for cultural traits that influence the dynamics of the evolutionary process and suggested that their evolution depends on their ability to directly or indirectly influence their own evolution.

The organizing effects of interactions and feedbacks can also be seen in the broader perspective of evolutionary transitions. These are fundamental modifications to the evolutionary process, caused by evolution itself. Maynard Smith & Szathmáry [21] recognized several properties common to such transitions, which we can also find in human cultural evolution. The first property is that new ways of transmitting information are formed. Humans have invented a series of information technologies, like reading and writing, radio and television, databases and pocket calculators, which can store, as well as communicate or broadcast, cultural traits, and also help us think and make decisions. A second property of transitions is that smaller entities aggregate to form larger ones and that the smaller entities tend to differentiate. Humans form a diversity of societies that are conceivably different from those that existed at the dawn of human culture, with for instance complex political and social organizations, division of labour, individual specializations and long production chains, and organized transmission of cultural traits via institutions like schools (e.g. [5,118–120]).

Challenges and questions for future research:

• i.

  What new understanding can be gained from a systems approach that cannot be gained from considerations of simpler representations of culture?

• ii.

  How can the evolution of cultural systems be explored theoretically? This requires formal descriptions of cultural systems, an understanding of trait relationships, and evolutionary modelling approaches that take into account that cultural systems exist and operate within and between individuals ([66], this issue).

• iii.

  How do externalized information storage and processing affect cultural evolution?