Abstract
Muhammad Ali Khalidi asks how natural kinds differ from nonnatural kinds. This seems to be the focal question that Khalidi wants to answer. Other questions about whether there are natural kinds in sciences such as fluid mechanics and chemistry, in the biological sciences, in the social sciences, in medicine, and in psychiatry are addressed only as part of discovering a general answer to the focal question of how natural kinds differ from nonnatural kinds. Not to keep you in suspense, Khalidi’s answer to the question of how natural kinds differ from nonnatural kinds is that natural kinds are those entities whose properties are projectible and enmeshed in causal networks, whereas nonnatural kinds have no projectible properties and are outside chains of causality. You may wonder what is behind this concern with natural kinds. As will be seen in the discussion that follows, all this has to do with the kind of naturalism and the kind of realism that Khalidi embraces.
Chapter 1 dispenses with “metaphysical Realism” or the doctrine of universals and “essentialism,” of the Aristotelian kind, as revived in the theories and discussions of some current philosophers of science, to clear the pathway to presenting an updated version of naturalism more in tune with a naturalism based on a close study of the theories of the modern sciences. As Khalidi says in chapter 2, “In the previous chapter, I argued that metaphysical Realism is powerless to distinguish natural from nonnatural kinds, and the criteria put forward by essentialism are not suitable for doing so. This leaves us with a philosophical problem: What is the difference between natural kinds and nonnatural kinds?” (p. 42).
Before going on further with outlining the book, I want to make a few general comments about its style and methodology.
The book has meticulous discussions of some of the literature relevant to its topic. The book has meticulous arguments for the theses proposed and meticulous considerations of the possible counterarguments and objections. Moreover, the book is concisely and densely written. The user-friendly Preface contains an overview of the argument in each of the six chapters. Moreover, the Preface places the book in its wider philosophical context as an extended argument for naturalism that includes a briefer argument for a novel viewpoint on scientific realism. Khalidi says, “ . . . to be real, a kind need not be independent of human beings or their minds; it must simply be manifested in the world (a world that includes the human mind.) The surest way to ensure that our categories identify real kinds is to pursue a scientific method that serves epistemic purposes” (p. xiv). Because Khalidi takes “scientific method” and “epistemic purposes” to be central to the understanding of how science discovers reality (or “natural kinds”), the reader might expect a section, or at least a subsection, on scientific method. But the reader will not find it, and this, I think, is the Achilles’ heel of the book. However, I suggest that the reader overlook this Achilles heel in the book’s argument so as to appreciate the book for what it does accomplish: an interesting and intellectually important contribution to a nonreductionist naturalism.
For the purpose of acknowledging and going beyond the Achilles’ heel of the book, let me discuss the book’s weakness very briefly. Khalidi avoids stating in any extended way let alone arguing for, any view of the methods of the sciences. The closest to a developed statement that I could find was not until the final sixth chapter. This statement is the fulcrum point of the presentation of his theory of scientific realism, which is the punch line of the book. But, I must, at this time, present a disclaimer. The following quote will be a plot-spoiler in revealing the truly novel idea in this book and might lead you to skip immediately to chapter 6 as a shortcut to reading the entire book to find out how Khalidi comes up with this idea (I interject a comment in italics enclosed in square brackets): . . . Objectivity when it comes to natural kinds is a matter of being guided by epistemic purposes; mind-independence is just a diversion . . . philosophers of science have made substantial contributions to these questions throughout the modern era from Bacon’s “Four Idols” onwards. But an account of scientific methodology is well beyond the scope of this particular inquiry, and I have taken it for granted in this account of natural kinds. If that is granted, then one way to formulate realism about natural kinds is as follows: Natural kinds divide the world into individuals that share causal properties, enter into the same or similar causal relationships, and give rise to the same or similar causal processes. Scientific categories that correspond to natural kinds are therefore projectible, enabling us to infer the presence of some properties from other properties on the basis of the causal relations between them. [I have italicized the part of the sentence that shows Khalidi is an inductivist.] Natural kinds need not be mind-independent, since some of those causal processes may involve humans and their mental states, but in order to identify them we must be guided by epistemic purposes and not be deflected by non-epistemic interests. (pp. 222-3)
Actually, Khalidi takes the property of “projectibility,” which is one of the code words used by inductivists, just as “paradigm-shift” is one of the code words for Kuhnians, “falsifiability” is one of the code words for Popperians, and “socially constructed” is one of the code words for the strong program in science studies, as a key to the scientific reality of a proposed natural kind in a scientific discipline. In this sixth concluding chapter, Khalidi comes out as an inductivist kind of philosopher of science in stating: What characterizes natural kinds is that, even when one or a few properties are central to a kind, there are a number of other properties associated with that kind that are causally related to them. It is this network of properties that seems to distinguish natural kinds from nonnatural kinds. The causal relations between the properties in the network ensure that natural kinds are projectible and play a central role in inductive inference . . .. (p. 204)
Khalidi takes “projectibility” to be the epistemological property that allows a hypothetical natural kind to play “ . . . a central role in inductive inference”. Also, Khalidi takes “projectibility” to be the key identifier or criterion of whether a proposed natural scientific kind is actually a real natural kind. Therefore, Khalidi assumes the truth of his inductivist approach to scientific methodology; and, hence, begs the question of how to distinguish natural kinds from nonnatural kind. Therefore, Khalidi assumes the truth of his inductivist approach to scientific methodology; and, hence, begs the question of how to distinguish natural kinds from nonnatural kinds. In other words, the general assumption of the existence of natural kinds as nodes in causal networks of projectible properties allows for induction, and induction over projectible properties in hypothesized causal networks of natural kinds as nodes, is used to pinpoint the existence of hypothesized natural kinds. Although Khalidi briefly considers whether we can do without the theory of natural kinds and without the theory of causality to understand science, he tacitly assumes the truth of inductivism: . . . we need to ask how the concept of natural kind can help us achieve our philosophical and scientific goals. Here, the context that is most pertinent to the identification of natural kinds is that of scientific inquiry, since scientific disciplines aim to locate projectible categories that enable them to predict and explain the phenomena of interest. (p. 202) . . . one important principle that guides our investigation into the characteristics of natural kinds concerns their projectibility, and hence their centrality to causal patterns or causal networks . . .. (p. 203)
So, Khalidi begs his main question about how to distinguish between natural and nonnatural kinds by assuming inductivism and the related ideas of “projectibility” and “causality” as indicators of natural kinds. The only way not to beg the question would have been to launch into an extended discussion of scientific methodology, inductivism, and its related theories of causality and projectibility. Still, the book is of great value in providing a novel approach to naturalism and a theory of scientific realism that involves enlightening discussions of various theories and cases in the sciences. Moreover, those who have alternative accounts of naturalism that are reductionist, that reject natural kinds, or that reject causality need to take seriously the arguments for natural kinds, causation, and scientific realism that are presented in this book. A brief overview of those arguments is to come.
The Preface provides a concise guide to each of the chapters and their role in carrying forward the argument of the book. So, rather than writing a mirror preface through summarizing each chapter, I will highlight and very schematically outline the three key arguments of the book.
The first key argument is that the sciences show that natural kinds are polymorphic (or “polythetic”). The second key argument shows that there are different kinds of natural kinds. The third key argument is that all natural kinds are natural because they participate in causality.
1. The Sciences Show that Natural Kinds Are Polymorphic (Or “Polythetic”)
The upshot of this argument is that “microphysical fundamentalism,” where reduction of the sciences to physics is required for kinds to be natural kinds, is mistaken (p. 37). The point is that natural kinds have large numbers of properties beyond the properties associated with the nuclear or microphysical. However, this does not imply that ordinary common sense views of what are natural kinds are the last word. Rather, the last word for “ . . . such categories [is if and when they] . . . get incorporated in time as part of a scientific discipline or a systematic inquiry” (p. 36). So, then how does a category that may have originated with common sense, such as fish, become “part of a scientific discipline?”
. . . the category fish is a cluster or polythetic kind . . . Despite the fact that it is not a unitary taxon from the evolutionary or phylogenetic point of view, the category fish has undisputed value as an epistemic kind. There are a number of branches of science, such as ichthyology and marine biology, which use this category to explain and predict natural phenomena. (p. 63)
Furthermore, polythetic natural kinds in the different sciences do not form a hierarchy of levels, but rather are “ . . . crosscutting kinds . . . corresponding to scientific categories introduced in order to enable us to make inductive generalizations and to explain naturally occurring phenomena” (p. 72). Thus, when polythetic and even “fuzzy” scientific categories have a “role in inductive inference, and their explanatory and predictive value reflect causal relationships in which they participate . . .” (p. 80), they are recognized in science as having instantiation in reality as natural kinds.
2. There Are Different Kinds of Natural Kinds
The upshot of this argument is that though there are special sciences with different kinds and different kinds of kinds, the unity of science is due to the different special sciences having the same “epistemic purposes.” Among the different kinds of kinds we have, for instance, in biology, geology, and cosmology, are etiological kinds (p. 134) that are linked historically or genetically. Although they are not “causal kinds,” they are still kinds. Also, there are “normative kinds” and “institutional kinds” that may be linked both historically and causally or only historically and not causally or only causally and not historically. However, ultimately, the epistemic goal of the discovery of the variety of kinds of natural kinds is to show how all the kinds of kinds “ . . . are implicated in causal processes . . . [that] . . . track real causal patterns and enable us to make inductive inferences” (p. 165).
3. All Natural Kinds Are Natural because They Participate in Causality
The upshot of this argument is to exclude a category as naming a natural kind when no causal nexus can be found for the category. So, in agreement with the Frankfurt school and with social constructionists, some categories in the social sciences are created for political purposes. However, even such categories can turn out to be natural kinds or real when they are used for “epistemic purposes” and found to participate in causal networks. For instance, “ . . . even some of those philosophers [from the Frankfurt school] who argue that social science ought to have a moral aim or purposes can agree that it is possible to disentangle that purpose from the epistemic one of understanding the causal nature of social phenomena” (p. 164). Also, not all natural entities form natural kinds. For instance, the category of polymer though all members share a primary property of having “large macromolecules with repeating monomeric units” (p. 173), they only constitute a natural kind because, “From this primary property flow various others as a direct causal result. Many of these other properties do not apply strictly to all polymers without fail, but they apply sufficiently widely as to enable us to make reliable inductive inferences” (p. 173).
The bottom line of the three key arguments about different kinds of natural kinds having a plurality of properties that cross disciplines and that are ultimately part of causal networks is twofold: first, the reduction of the sciences to physics is impossible because of the nature of nature as shown by how natural kinds are used in the sciences. Second, though some categories, especially in the social sciences, are socially constructed and mind-dependent, even socially constructed and mind-dependent categories can name natural kinds: “ . . . mind-independence is irrelevant to realism since what we want from our kinds is not that they should be independent of the human mind but that they should be instantiated in the universe (which includes human minds) and participate in real causal patterns” (p. 227).
The novelty of this book lies in the details of its arguments. But in any case, here is my attempt to squeeze the details into a short summary sentence: the kind of thing reality is, is that it is natural and made up of kinds that are only natural kinds when they are part of causal networks, but there are a lot of kinds of kinds, even mind-dependent kinds that are natural or real if and when causally connected.