Natural selection operates at many levels simultaneously; see Depew and Weber 1995, Chapter 14, on the development of this view among biologists. Sober and Wilson (in Katz 2000, 259) summarize multi-level selection theory as follows:
a gene can evolve by increasing its fitness relative to other genes within the same individual, by increasing the fitness of the individual relative to other individuals within a group, or by increasing the fitness of the group, relative to other groups in the total population.
But at any of these levels, since the ‘other’ to which fitness is relative is co-evolving, competition is not necessarily required. What is essential to evolution is a nonlinear dynamic, a feedback process. This entails that the very act of adapting may change its own context as well as the ‘text’ (the specific transform). Besides, natural selection is not necessarily or exactly an adaptive process.
When adaptation is observed, it can be explained by the differential survival and reproduction of variant types being guided and biased by their differential efficiency or resistance to environmental stresses and dangers. But any cause of differential survival and reproduction, even when it has nothing to do with the struggle for existence, will result in some evolution, not just adaptive evolution.… What evolutionary geneticists and developmental biologists have been doing for the last sixty years is to accumulate a knowledge of a variety of forces that cause the frequency of variant types to change, and that do not fall under the rubric of adaptation by natural selection. These include, to name a few: random fixation of nonadaptive or even of anti-adaptive traits because of limitations of population size and the colonization of new areas by small numbers of founders; the acquisition of traits because the genes influencing them are dragged along on the same chromosome as some totally unrelated gene that is being selected; and developmental side effects of genes that have been selected for some quite different reason.
— Lewontin (2001, 56-7)
Lewontin (2001, 52) speaks of a ‘vulgar Darwinism’
which sees all aspects of the shape, function and behavior of all organisms as having been molded in exquisite detail by natural selection – the greater survival and reproduction of those organisms whose traits make them ‘adapted’ for the struggle for existence.… Evolutionary geneticists, on the other hand, … and most epistemologists take a more pluralistic view of the forces driving evolution.
Gerald Edelman (2004) attributes just such a ‘vulgar Darwinism’ to Alfred Wallace, contrasting this view with Darwin’s own:
Wallace, in fact, concluded that natural selection could not explain the origin of our higher intellectual and moral faculties. He claimed that savages and prehistoric humans had brains almost as large as those of Englishmen but, in adapting to an environment that did not require abstract thought, they had no use for such structures and therefore their brains could not have resulted from natural selection. Unlike Wallace, Darwin understood that such an adaptationist view, resting only on natural selection, was not cogent. He understood that properties and attributes not necessarily needed at one time could nevertheless be incorporated during the selection of other evolutionary traits. Moreover, he did not believe that mental faculties were independent of one another. As he explained in his book The Descent of Man, for example, the development of language might have contributed to the process of brain development.
— Edelman (2004, 2)
It would appear, then, that Darwin was not a ‘vulgar Darwinist,’ and in fact anticipated the co-evolutionary theory developed by Deacon in The Symbolic Species (1997).