Localización y características del proyecto

The replicator view has many critics. Jablonka and Lamb (2005) criticize it for failing to individuate replicators satisfactorily (375) and for establishing a barrier

between replication and development that distorts our view of epigenetic and behavioral inheritance systems (102, 189). Godfrey-Smith (2000, 413; 2009, 31-39; 2015, 1) has argued that replicators are not necessary for evolution by natural selection, and that there are cases in which the replicator view provides a poor description of the biological facts. Griesemer et al. (2005) worry that the replicator view cannot offer insights into the evolution of replicators. Proponents of niche construction theory (NCT) (Laland 2004) and developmental systems theory (DST) (Griffiths and Gray 2001) have articulated similar dissatisfactions with the replicator view.

8 See also Hull 2001, 57-58 for an endorsement of the replicator commitment.

9 For example, in Blount et al's 2012 paper on an experimentally induced evolutionary novelty we read that, “A complete understanding of the evolution of a novel trait requires explanation of its ecological function, its physiological basis, the underlying mutations and the history of the accumulated changes.” See also Wagner and Lynch (2010).

Though the details of these criticisms differ, they have a common structure. The strategy is to call attention to specific examples in which the replicator view performs badly, show that it is possible to conceive of inheritance without making reference to replicators, and then abandon the replicator view altogether. There is a diverse set of alternatives to the replicator view, but the common commitment is that any processes that produce heritable variation may be inheritance systems, regardless of whether those processes involve the transmission of replicators from parents to offspring. In this respect, they all hearken back to the “classical” (Godfrey-Smith 2009) descriptions of evolution by natural selection (see, for example, Lewontin 1970).

The classical commitment: any developmental resource that produces heritable variation is hereditary material.

Of course, such an abstract characterization needs to be supplemented by an explanation of what it is that disposes a developmental resource to produce heritable variation. This is where the classical views diverge. Jablonka and Lamb (2005) individuate inheritance systems mechanistically. Godfrey-Smith (2009) prefers the statistical approach embodied by the Price Equation. DST (Griffiths and Gray 2001) offers a holistic take on inheritance which declines to distinguish between inheritance mechanisms except for pragmatic purposes. Mameli (2004) distinguishes between

genotypically and envirotypically produced variation and suggests a way of incorporating the latter category into traditional population-genetic models.

Despite their differences, these accounts all emphasize the connection between inheritance and the generation of variation. Jablonka and Lamb even define evolvability as “ the capacity to generate heritable phenotypic variation” (Kirschner and Gerhart 1998,

8420). While the replicator view privileges the transmission of variation over its

generation, classical views are much more explicitly interested in the ways in which non- replicators produce truly novel evolutionary outcomes (Laland et al. 2015). Rather than the power of cumulative selection, they are focused on identifying the sources of the variation on which this selection acts, and on pointing out that the replicator view, carried away by its enthusiasm for high-fidelity transmission mechanisms, fails to see these sources of variation as hereditary material. Thus, classical views emphasize a different aspect of evolvability than the replicator view does.

Unfortunately, classical views shift the emphasis from transmission to generation without offering a satisfying response to the claim that a general account of inheritance should explicitly acknowledge the connection between replicators and cumulative selection. This is a significant oversight, because this is the feature that defenders of the replicator view see as the chief advantage of their position. Stereleny and other

proponents of the replicator view agree with Jablonka and Godfrey-Smith and others that it is possible to give a general account of inheritance that does not reference replicators, but the reason they have sought to extend, rather than reject, the concept of replication is that they believe that replicators have played a crucial role in the evolutionary history of this planet, and that we compromise our ability to explain how lineages produce complex, adaptive phenotypes unless we emphasize replicators.

Proponents of the classical view do not deny this claim, but they do not take it as important enough to justify retaining the replicator category in their own accounts of inheritance. Jablonka and Lamb, for example, allow that the genetic inheritance system is an important form of inheritance, but they resist classifying genetic and non-genetic

replicators together, as a type of hereditary material with a distinctive causal and explanatory role. They write that “it is best to avoid this concept [of the extended replicator] and concentrate on the way variation is constructed and transmitted through developmental processes” (2005, 376).

In the remainder of this chapter, I argue that replicators do have a distinctive causal and explanatory role, that there are disadvantages to abandoning the replicator concept in our theoretical articulation of inheritance, and that retaining it does not compromise the classical view's insight that there are many non-replicator sources of heritable variation. My alternative offers an increase in explanatory power over both the replicator and classical views. This is important, not only because explanatory power is a consideration in choosing between accounts of inheritance, but also because rejecting the replicator commitment is central to two emerging and influential perspectives in biology: Eco-evo-devo (Gilbert et al. 2015, 616, 619) and the extended evolutionary synthesis (EES) (Laland et al. 2015, 2). If rejecting the replicator commitment carries with it a loss of explanatory power, then this is a problem for both Eco-evo-devo and EES. My

proposal can help these two perspectives respond to criticisms and open up new avenues for research.