In Evolution in Four Dimensions (E4D; 2005), Jablonka & Lamb (J&L) present an interesting synthesis of evolution that takes a stance against purely gene-centered approaches. By providing evidence of the importance of not only gene-based mechanisms, but also epigenetic, behavioral, and symbolic mechanisms in the establishment of inheritable traits, the authors propose to reconsider evolution as being dependent and effective along these four dimensions. This way, the authors seek to reconcile genetic, developmental, behavioral, and socio-cultural studies under a systemic, comprehensive framework for evolutionary theory. Through interpretative mutations as another mechanism of variation, both Darwinian and Lamarkian aspects find their place in this proposal; we believe it represents a much needed, challenging, and serious attempt at moving forward our understanding of evolution. However, while the authors dispute the gene-centered notion and consider evolution as a systemic multilayered phenomenon, we believe they fall short in one critical aspect: J&L rely heavily on the notion of “information transmission” in a rather loose manner. Their approach is liable to the argument that in order to have any such thing, one needs a transmitter, a message, and a receiver – something that is not easily found when dealing with biological phenomena.
What distinguishes living systems from the rest is a difficult question that can have non-trivial consequences for our understanding of evolution. One influential hypothesis states that living systems are those that maintain organizational closure: they are constituted by networks of self-sustaining processes, regardless of the materials used to instantiate such loops; that is, they are autonomous systems (Maturana & Varela Reference Maturana and Varela1973; Varela Reference Varela1979). When one understands organisms this way, the notion of information transmission becomes less appealing: a closed system cannot “have” information in itself. As E4D describes, information transmission depends on the existence of a source, a receiver, and a functional relationship between them. However, although J&L recognize the modulation of source organization and dynamics by the receiver's processes through feedback loops (first-order cybernetics), they fail to incorporate the role of the observer (second-order cybernetics). This step is fundamental if one is to distinguish both ends of the information transmission process. Moreover, the authors state that the “receiver's functional state is changed in a way that is related to the form and organization of the source. There is nothing intentional about the receiver's reaction and interpretation” (Jablonka & Lamb Reference Jablonka and Lamb2005, p. 54). This description is incompatible with systems biology, in which organisms are fundamentally intentional and interpretative, thereby reacting to perturbations precisely according to their internal state. There is nothing informative or valuable in a sucrose gradient until a hungry cell interprets it.
There are, however, several statements in the book that lead us to think that the authors are somehow aware of these aspects. Take the following assertions: “the receiving animal is not just a vessel into which information is poured” (p. 172), “Animals must therefore possess some kind of internal filter – some set of principles or rules” (p. 175), and “what is information for them (animals) may not seem like information to us, and vice versa” (p. 334). All of these quotations point to autonomy in a very direct manner. In fact, the last claim begs the question for the utility of using the lens of information when understanding the properties of the interaction even in communicative actions between animals.
We understand that the use of the concept of information attempts to support the existence of the three non-gene-based systems of inheritance, as a consequence of the discussion about units and levels of selection. However, taking the above considerations into account, and in the context of the evolutionary problem, we believe that inheritance can be understood as the subsistence of an adapted (and adaptive) organization across generations. Considering that DNA, self-sustained cytoplasmic loops, behavior, and so on, transmit information among organisms, has no explanatory value when trying to formulate an evolutionary theory. This process is more easily conceptualized as the reconstruction of a “parent” organization by an “offspring,” using multilevel templates (genetic, epigenetic, behavioral, and symbolic), in the context of unpredictable environmental influences. No information metaphor is thus needed. Indeed, when J&L define inheritance as a bias in the reconstruction of an activity or state in next generations, they are emphasizing the active and autonomous role of the offspring in inheriting, without using an information metaphor.
In a quite similar line, Aboitiz (Reference Aboitiz1992) discussed the relevance of development and morphogenesis for a better understanding of adaptive evolution, considering in his proposal Darwinian and Lamarckian mechanisms, as well as behavioral and social levels of organization. We consider that the proposals of E4D and Aboitiz are consistent in this respect. Nevertheless, Aboitiz did not use the information transmission analogy, remaining in a purely biological framework. We want to briefly elaborate on this view by looking at the evolution of the nervous system and language.
Perspectives such as generativist grammar, sociobiology, and evolutionary psychology share a common tendency to explain human behavior and brain operations relying heavily on genetic mechanisms. E4D presents an alternative view where genetic factors are considered to have relevant, but not exclusive, roles in development. Considering this scenario, we agree with J&L that when talking about a whole-organism operation such as language, it cannot be understood as a genetically determined function, but as a consequence of genetics, development, and culture. Faster language learners evolved thanks to many social, ecological, and morphological conditions, some of which are precisely described by J&L. Aboitiz (Reference Aboitiz1988) also considers genetic factors, but calls attention to the epigenesis of the human brain through the regulation of nerve cell proliferation and the selective stabilization of synapses in a specific context of interactions.
Genetic mutations may contribute to sharpening neural networks involved in specific cognitive processes such as phonological perception and certain aspects of syntactical processing (Aboitiz Reference Aboitiz1995; Aboitiz et al. Reference Aboitiz, Garcia, Bosman and Brunetti2006), but a large amount of the connectional repertoire may also be shaped by epigenetic processes such as the selective stabilization of neural networks and biomechanical structures within the context of social interaction. In this context, imitative capacities, specified by highly elaborate neural networks already present in nonhuman primates, begin to work as a mode of cultural propagation of adaptive behavioral and symbolic patterns. More specifically, in early proto-humans, the development of a short-term phonological memory system within the framework of an elaborate imitative capacity permitted the acquisition and subsequent elaboration of highly sophisticated proto-linguistic utterances. This short-term memory network eventually expanded, contributing to the establishment of increasingly complex networks generating higher levels of linguistic interaction. Through tight social coupling and sophisticated imitative mechanisms such as the mirror-neuron system, language can evolve without any information being “passed on”.
We can see how genetic, epigenetic, behavioral, and symbolic patterns of organization evolved into language through an interplay of conservation and transformation of their properties along generations. Using this brief example, and in the context of our discussion of biological autonomy, we pose the question: Is the analogy of information transmission truly necessary in a new evolutionary framework?
