Natural species in biology

More recently, a "new biological essentialism" has developed, fundamentally relational in nature, with a dual objective: to show that essentialism is not incompatible with Darwinian evolutionary theory, and to insist on the definition of essence as a relational property, emphasizing the omnipresence of relational concepts in biology. The following arguments were put forward: the weakness of phenetic arguments based on the idea of "resemblance" and "difference"; the importance, on the other hand, of a whole series of relational concepts such as "reproductive exchange", "nicheecology" and "phylogenetics". All this would tend not only to show that Darwinism and essentialism are compatible, but also that it is possible to transcribe the line of argument developed by Kripke and Putnam (Okasha [25], Griffiths [26], LaPorte). However, "relational" essentialism in turn encounters difficulties: each of the concepts entails explanatory limitations that should not, however, lead to a return to intrinsic essentialism (contra Devitt [27]). Before testing our own model, we examined another realistic explanatory model (to which Hacking had also drawn attention): that of homeostatic "clusters" of properties (due to R. Boyd [28]), which has several advantages: it takes into account a certain indeterminacy, flexibility and plasticity of taxonomies; it insists on the causal profile of the species; it does indeed constitute a realistic explanatory model. But it also has its problems: it contradicts biological theory; it tends to give a circular explanation of how taxa are identified; it confuses the biologicalspecies as a "kind" and as an "individual"; it does not avoid a certain vagueness in the analysis. Since traditional essentialism has its limitations, and the (weakly essentialist) cluster model, "intrinsic" essentialism and even relational essentialism are themselves unsatisfactory, should we renounce all forms of essentialism, while giving priority to relations? No, and to show this, we began by recalling the principles governing dispositional aliquidditism and the four assumptions on which the model is based: 1. a causal theory of properties; 2. a dispositionalist causal analysis of laws; 3. efficient and teleological causality; 4. aliquidditism (narrow essentialism).

In a second stage, we emphasized the proximity of the model (on the first three points) to scientific essentialism, and the correctives introduced: "essence" is not a static "quiddity" or a substance, or a pure natural species, but rather a fundamentally dispositional " aliquid " something (close, in some respects, to causal or relational structuralism). The real source of intelligibility of a thing is a set of general dispositions affecting the way the object would tend to behave in certain kinds of circumstances: essence is defined not in terms of intrinsic properties, but of conditional and mutual relational or dispositional properties, orclusters of causal powers. Evaluating the model in the light of concepts used in biology provides a rather conclusive test: there are many arguments against defining species by intrinsic criteria [29] : for example, just as Kripke emphasized that individuals can have extrinsic essential properties (notably their origin), the cladistic approach in biological taxonomy considers that biological taxa are defined not by reference to a hidden internal constitution, but by an extrinsic property, the common ancestor. Then, in evolutionary biology, both in an adaptationist approach (emphasizing the principle of natural selection) and in a non-adaptationist approach (insisting on the fact that mutations occur at least as much as a result of genetic drift, which constitutes a variation in genetic frequencies linked to the mode of reproduction as well as to the size of a given population), the concept of " fitness " (or adaptive selective value) plays a key role that cannot be fulfilled by the explanatory schemes of physics: to show the similarities between, for example, the evolution of a bacterial population, that of a population of peas and that of a population of lions, we need to momentarily disregard some of their physical differences (Sachse, 2011 [30]). Yet the concept of " fitness ", which does not have descriptive value (it arises from physical properties that do indeed exert their constraints) but functions as a statistical model of causal explanation (a probability distribution relating to the organism's progeny), is fundamentally relational and dispositional: indeed, an organism's degree of fitness is its disposition to survive and reproduce, which is based on a fairly static property (that hardly changes over time), but whose manifestation may or may not depend on the evolution (and dynamic changes) of the environmental context (of the "niche"). Another important concept is that of "systemic function": this no longer refers to a goal or intention of the Creator, but to the explanatory potential of a causal power that is no longer limited to etiological criteria constituted by reference to the evolutionary past. It integrates the structural organization of causal capacities and the context or environmental conditions, which is also in line with the causal theory of properties inherent in the dispositional aliquidditist model, since the essence of a property is indeed to "produce the effects it produces". For example, a plant has a genetic basis which, under certain environmental conditions, leads to the production of red petals. Following the systems approach, we would say that "this genetic base is a functional property because the disposition to produce red petals itself contributes to a higher capacity of the plant, namely being able to attract insects for pollination" (Sachse, 2011, 98). Thus, "what fundamentally constitutes the function of a biological property in a system depends on the context, which also includes the internal organization of the organism's dispositions if the property in question contributes to the homeostasis and fitness of the entire system"(ibid.,102).

We conclude these analyses by underlining the historical dimension of biology (dealing with events such as speciation) and the difficulties surrounding the concept of biological "law" (non-universality, laws ceteris paribus), which should not be underestimated in the realist perspective of our model, requiring consideration not only of essentialism, but also of the type of causality at work in nature and the laws that hold things together. This calls for all the more attention to be paid to the necessary multiplicity of explanatory models (natural selection, genetic drift, but also epigenetics - which can no longer be understood as a simple return to vitalism [31]). New associations need to be forged, without losing sight of the necessary classification of species if we are to appreciate biodiversity (Hacking [32]; Dupré [33]), better link biology and ecology (cf. the EVO-DEVO perspective) and also reconcile our two "images" between cladistic systematics (of process as much as of pattern) and evolutionary systematics better able to take into account the causal path that species follow. Against a fundamentally constructionist, historical conception or, conversely, an essentialist, intrinsic or purely relational conception of the species, we are more inclined towards a form of extrinsic essentialist or moderate relational realism.