In Part One of this series, we saw that culture doesn’t suffer from the problem that Darwin’s theory of natural selection successfully solved: the problem of how change accumulates in biological lineages despite changes acquired over a lifetime being discarded at the end of that lifetime. We also saw that units of culture don’t possess the algorithmic structure of something that can evolve through natural selection. In this post, we see how culture violates the requirements that make natural selection an applicable model.

A model is an artificial construct; no process adheres to a model exactly. Biological evolution adheres to the conditions for natural selection sufficiently well that it serves as a useful approximation, but cultural evolution does not. A model that captures the skeletal essence of the process increases understanding because it enables us to vary one part of the model and see how other parts are affected. It can form the basis of a theoretical framework that unites disparate bodies of knowledge under one umbrella and shows how they’re related. An inappropriate model can, however, create confusion by suggesting relationships that don’t exist.

Nature Violates the Requirements that Make Natural Selection a Viable Model

A model is an artificial construct; no process adheres to a model exactly. Biological evolution adheres to the conditions for natural selection sufficiently well that it serves as a useful approximation, but cultural evolution does not. A model that captures the skeletal essence of the process increases understanding because it enables us to vary one part of the model and see how other parts are affected. It can form the basis of a theoretical framework that unites disparate bodies of knowledge under one umbrella and shows how they’re related. An inappropriate model can, however, stymie and mislead by suggesting relationships that don’t exist, or misportraying ones that do.

Violation of the Requirement that Variation Be Random

For the framework of natural selection to be applicable, new variations must be random with respect to fitness; there can be nothing biasing them to make the organism better adapted to its environment than due to chance. That’s because, if variations are systematically biased toward improvement, then change over time is due to whatever creates that bias in the first place, not to natural selection weeding out less fit variants after the fact. Biological evolution adheres to this condition sufficiently well that natural selection can serve as a useful approximation; now and then, due to chance mutation, a new beak shape or eye color or irises. However, the greater the extent to which variation deviates from random, the less adequately natural selection can explain the process.

This requirement is violated in cultural evolution; human minds don’t generate cultural novelty randomly. We solve problems using strategy, knowledge, intelligence, creativity, imagination, and intuition, scouring our memory for clues about how to proceed. Moreover, whereas natural selection requires generations to produce an effect, mental operations can generate improvements in minutes. Cultural changes operate on time scales so fast, they drown out any effect of a process that requires generations. Biological evolution is a “breadth-first” process: randomly generate huge numbers of variants, such that by chance a few are fitter than what currently exists. In contrast, cultural evolution is “depth-first”; it uses pattern-matching abilities and content-addressable memory to land on, and iteratively refine, ideas that are more likely than chance to be useful and relevant.

Violation of the Requirement of Discrete Units and High-Fidelity Copying

Natural selection requires clearly identifiable, discrete units that are copied with high fidelity; otherwise, it’s not possible to track changes in their frequencies over generations. Genes fit this bill nicely, but cultural units do not. Ideas are often ill-defined, half-baked, context-sensitive, and constructed on the fly (Scotney, et al., 2020; Veloz et al., 2011). An idea in one person’s mind may be subtly or radically different from the “same” idea in another person’s mind. During transmission, ideas are routinely modified, reinterpreted, or transformed into something quite different.

Even when cultural content appears to be transmitted with perfect fidelity, its meaning may radically change. Consider this example: Ann tells Bill, “Dave asks lots of questions,” meaning Dave seems engaged and interested. Bill then says the exact same words to Cathy—“Dave asks lots of questions”—but the subtext of Bill’s message is that Dave is suspicious, and may be onto them. The words were copied perfectly, but the content changed it completely. Bill’s message acquired a trait (expressing danger) that was passed to Cathy. Even with perfect fidelity information transfer, we have transmission of an acquired trait, the very phenomenon Darwin’s theory was designed to work around.

Competition

Natural selection requires that different variants of a trait compete for representation, with some being selected and others excluded. In biological evolution, different versions of genes (alleles) do compete; an allele that predisposes organisms to disease typically loses out to healthier variants.

However, most cultural phenomena can’t be neatly divided into competing variants like biological alleles. Concepts and ideas often coexist, mutually support each other, and sometimes enable one another rather than competing. A new artistic style doesn’t necessarily eliminate earlier styles; a new technology often depends on and incorporates earlier ones.

Horizontal Transfer

Moreover, biological evolution assumes a tree-like pattern of descent in which lineages can split and diverge but cannot rejoin. Because there is no rejoining of branches, similarity is due to inheritance; it reflects shared ancestry through vertical descent (or convergent evolution, in which case they independently evolved the similar trait). However, in cultural evolution, similarity can reflect the horizontal transmission of acquired traits. Moreover, horizontal transfer is ubiquitous in culture; it is present whenever people share ideas, fuse cuisines or musical styles, or combine ideas from different lineages, as in the piano stairs that play notes as you walk on them. Cultural descent diagrams form reticulated networks. Forcing such patterns into a branching tree-like model misleadingly portrays similarity due to horizontal idea transfer as similarity due to common ancestry.

In short, natural selection explains cumulative, adaptive change through differential reproduction of randomly generated heritable variations. Cultural evolution violates the necessary conditions for natural selection to be a viable theory: random variation, discrete units that compete and are copied with high fidelity, and vertical transmission of inherited traits (not horizontal transmission of acquired traits). We can only maintain a Darwinian perspective on culture by viewing humans as passive imitators and transmitters of pre-packaged cultural units, ignoring exactly those abilities that define our humanness.