Lamarckism and epigenetic inheritance: a clarification

Since the 1990s, the terms “Lamarckism” and “Lamarckian” have seen a significant resurgence in biological publications. The discovery of new molecular mechanisms (DNA methylation, histone modifications, RNA interference, etc.) have been interpreted as evidence supporting the reality and efficiency of the inheritance of acquired characters, and thus the revival of Lamarckism. The present paper aims at giving a critical evaluation of such interpretations. I argue that two types of arguments allow to draw a clear distinction between the genuine Lamarckian concept of inheritance of acquired characters and transgenerational epigenetic inheritance. The first concerns the explanandum of the processes under consideration: molecular mechanisms of transgenerational epigenetic inheritance are understood as evolved products of natural selection. This means that the kind of inheritance of acquired characters they might be responsible for is an obligatory emergent feature of evolution, whereas traditional Lamarckisms conceived the inheritance of acquired characters as a property inherent in living matter itself. The second argument concerns the explanans of the inheritance of acquired characters: in light of current knowledge, epigenetic mechanisms are not able to drive adaptive evolution by themselves. Emergent Lamarckian phenomena would be possible if and only if individual epigenetic variation allowed the inheritance of acquired characters to be a factor of unlimited change. This implies specific requirements for epigenetic variation, which I explicitly define and expand upon. I then show that given current knowledge, these requirements are not empirically grounded.