One cannot deny the central role the concept of natural kind plays for understanding and examining classification practices in science. In recent years, there has been an increasing literature on philosophical approaches to natural kinds. Adding to the recent contributions on this topic by Richard Boyd, Ian Hacking, John Dupré, Brian Ellis, and others, there is a new monograph by Muhammad Ali Khalidi. In addition to a fine-grained examination of metaphysical and epistemological dimensions of scientific classifications, he expounds his own view of natural kinds.
Natural Categories and Human Kinds is divided into six chapters. In the first, the author analyses the essentialist notion of natural kind. According to this notion, ‘natural’ classifications are based on the identification of a set of necessary and sufficient properties that are also intrinsic and reflect in some way the micro-structure of the categorized phenomena, which is consequently discovered through scientific research. In Khalidi’s opinion, these criteria are problematic, since contemporary scientific theory takes as paradigmatic natural kinds cases in which these aspects are not included, especially in fields such as the biological or social sciences. Following Khalidi, this makes the essentialist view unsatisfactory and justifies the development of an alternative proposal.
This alternative is articulated in the second chapter. In Khalidi’s proposal, discoverability by science is the only essentialist requirement that is not controversial. So, he takes it that natural kinds are discoverable by science and are considered to be revisable in light of new empirical evidence. However, this characteristic is not as uncontroversial as Khalidi claims. In fact, there is an extensive discussion about whether these kinds are discovered or stipulated (see, LaPorte; Bird). ‘Discoverability by science’ is an aspect with relevant metaphysical implications, as well as epistemic ones, and the current philosophical approaches to this issue are far from general agreement.
In a similar vein, the ‘revisability’ of natural kinds that Khalidi takes for granted is also debatable. When does a kind that is revised, that is, that changes its configuration (extension, reference, membership criterion), become a different natural kind? The relevant question, which Khalidi’s proposal eludes is how far we can revise these kinds that are actually ‘carving nature at its joints’ and reflect real divisions.
In any case, Khalidi’s approach offers a new and promising theoretical account for natural kinds. For instance, natural kinds according to Khalidi can overlap; that is, a particular individual can belong to different kinds. That allows us to account for many problematic cases for which essentialism is too limited, and points to a new theoretical framework better suited to scientific classificatory practices, without abandoning a realist stance. Many scientific kinds are polythetic and have fuzzy boundaries. Khalidi claims that natural kinds are those that are fertile from a predictive and explanatory point of view.
In the third chapter, Khalidi addresses causal relations between kinds of different sciences, revising classical issues in the philosophy of science such as multiple realizability, the distinction between empirical generalizations and scientific laws, and inter-level causation. Khalidi defends the view that science tries to identify properties and kinds based on causal relations detected in their own areas and concludes that projectibility and predictive reliability reflect accommodation of these kinds to causal structures, as claimed by Richard Boyd. Footnote 1 In fact, Khalidi advocates a definition of ‘naturalness’ based on what Carl Craver has labelled a ‘simple causal theory.’ According to this theory, a natural kind is related to an underlying causal network in which the presence of some properties has a causal influence for the generation of other properties.
The fourth chapter is dedicated to answering some arguments against considering kinds in the biological and social sciences as natural kinds. Khalidi shows that these kinds can be quite different (etiological, historical, interactive, mind dependent, etc.), but all of them are projectable and can be defined according to the ‘simple causal theory.’ In other words, Khalidi shows how these apparently problematic biological and social kinds are perfectly suited to his own notion of natural kind. Accordingly, the fifth chapter focuses on some examples and analyses how they are interpreted and used in scientific practice. Khalidi analyses cases such as lithium, virus, and polymer and concludes that in all these cases scientific research suggests that these are genuine natural kinds. The criterion for ruling that two entities belong to a kind is not that they share a number of properties but that they participate in the same causal relations.
In conclusion, this book advocates an integrative theory of natural kinds and defends an interpretation of ‘naturalness,’ based on a ‘simple causal theory,’ which is sufficiently broad to account for common kinds, both in the basic and in the special sciences. Additionally, Khalidi considers his definition restrictive enough for preserving the adjective ‘natural.’ Despite the theoretical charge of the realist position that the author takes up, this book makes a remarkable contribution to the current debate. It analyses in a rigorous and clear way ideas implied in natural kinds (especially in relation to the essentialist perspective) and it addresses many complex aspects related to the causal dimension commonly associated with natural kinds.
