Short Synopsis
The proposed title ('Are Species Real?') is the question my book attempts to answer. My answer (spoiler alert!): yes; or more precisely: yes and no. I'll explain presently why this is not a hollow cop-out. First, it's worth noting the attraction to the view that species are indeed real features of the world. Species are deeply embedded in both scientific investigation and everyday enthusiasm. Our species goes to considerable effort and expense to watch, document, count, breed, cultivate, hunt, and conserve other species. Scientists often report discovering new species, estimating that there are ten million distinct species presently on earth (and many more that have disappeared).
But what are species? Do everyday intuitions that species are real stand up to philosophical and scientific scrutiny? Two rival accounts of species' reality have dominated the discussion: that species are natural kinds defined by essential properties and that species are individuals. I contend that neither fully accommodates biological practice or require uncontroversial metaphysical premises. One of the central motivations for the species-as-individuals view has been that the view that species are natural kinds saddles us with an expensive ontology of abstract objects (sets) which cannot accommodate the changing natural world. Ironically, my primary criticism of the species-as-individuals thesis is that, by treating species as objects, its defenders incur demands for metaphysical precision that are seriously out of step with biological practice.
I shall argue for an alternative approach to the metaphysics of species. Species are real, I say, not in virtue of there being particular sets or individuals in the world, but in virtue of there being some organisms bearing certain relations to one another. Sentences like 'There are six species of giraffe' are true, not because this sentence refers to six distinct things, but in virtue of some animals sharing, say, particular breeding patterns. There are six populations of organisms that are natural kinds (understood in a non-essentialist way). Thus, I propose to separate the question of whether something is a natural kind from the question of whether it is a set or an individual. A natural kind, in my view, is not an ontological category. Natural kinds can crosscut ontological categories like sets, individuals, properties, relations, events, processes, and so on. Much confusion in the debate about the metaphysics of species has devolved from failing to separate questions about species' metaphysical category from the question of whether they are natural kinds. I shall show that we can treat species names as plural referring expressions — referring not to individual sets (abstract objects), but to collections of things — and yet nevertheless treat these expressions as names of natural kinds.
This raises a key question: what is a natural kind? By refusing to associate natural kinds with a particular ontological category, I free myself up make this a partly epistemic question. In something like the spirit of Goodman and Boyd, I suggest that the natural kinds are those things (be them individuals, sets, collections, whatever) which play characteristic roles in our inductive and explanatory practices. But while there is no doubt a pragmatic element to these practices, it is something about the metaphysics of these collections that suits them to this role. I develop a novel conception of natural kinds closely allied with Boyd's Homeostatic Property Cluster (HPC) conception. But whereas Boyd says that the relevant clusters of properties must be maintained by homeostatic causal mechanisms, I require only that the properties be stable across various counterfactual perturbations. Which perturbations? This is where the pragmatics of explanation and induction reenter the picture. The last third of the book will be devoted to articulating the account — what I call the "Stable Property Cluster" (SPC) account of natural kinds — testing it against case studies, and exploring several of its biological and metaphysical consequences.
Chapter Sketches
Chapter 1: Introduction
This chapter will motivate the intuitions that species are real features of the world, by briefly surveying the some inferential and explanatory uses species play in modern science. It will sketch the basic positions philosophers have taken on how to understand species' reality and sketch the argument of the book. (5,900 words)
Chapter 2: Natural Kinds
This chapter surveys different theoretical approaches to the general notion that our classification schemes aim and succeed in accurately describing the world. I introduce the notion of a natural kind as traditionally understood and describe how Kripke and Putnam attempted to reinvigorate the notion of an essence, describing the various empirical and philosophical worries their version of essentialism faces. (16,200 words)
Chapter 3: Species as Natural Kinds
I begin this chapter by explaining in more detail why essentialist views of natural kinds appear to be unable to accommodate facts about species. Recently, however, essentialism has been going through something of a rebirth. A first rebirth (propelled by Griffiths [1999] and Okasha [2002]) suggested that species might have historical essences; a second rebirth at the hands of both philosophers (Walsh [2006] and Devitt [2008]) and biologist enthusiasts of "genetic fingerprinting" maintained the idea that species have intrinsic essences. I argue that none of these novel forms of essentialism are plausible. (~9,000 words)
Chapter 4: Monism and Pluralism
Another concern for essentialism is the notorious plurality of legitimate species concepts. Pluralism also apparently poses a problem realism about species. Does not realism require just one "privileged taxonomy" of species? Metaphysical and methodological issues must be disentangled here: epistemological arguments for species pluralism tend to be inconclusive and of unclear relevance to the metaphysical status species. I will return to the question of pluralism after introducing my own conception of natural kinds. (~9,000 words)
Chapter 5: Species as Individuals
Inspired by the apparent failings of essentialism and concern over taxonomic pluralism, some proposed a "radial solution to the species problem": that they are individuals — literally, spatiotemporally extended objects, "hunks of the genealogical nexus". Bolstered by the burgeoning consensus for historical approaches to taxonomy, proponents of the "Species-as-Individuals" (SAI) thesis argue that it makes sense of our pre-theoretic realism and anti-essentialism about species. Likewise, they argue, pluralism can be set aside as a purely epistemic artifact. Unfortunately, proponents of SAI have conspicuously failed to establish their thesis, either begging the question against deniers of SAI or taking as explananda contentious theses. This chapter surveys both the original attempts at establishing SAI (Ghiselin 1974, Hull 1978) and recent arguments for SAI (Coleman and Wiley 2001, Brogaard 2004, Crane 2004, Richards 2010) which draw on different, incompatible, and contentious metaphysical presuppositions. (~15,000 words)
Chapter 6: Against Species as Individuals
These worries notwithstanding, the Species-as-Individuals thesis might still be true. Indeed, many philosophers of biology seem to have adopted a conciliatory approach to the issue, regarding the metaphysical question as "biologically neutral". This, I think, is the wrong approach. In this chapter, I explore what metaphysical concessions one would have to make in order to accept SAI, arguing that developments in metaphysics over the last twenty years strongly legislate against SAI. (12,900 words)
Chapter 7: Metaphysics of Species for the Commitment-Wary
In the wake of the failure of biological essentialism and individualism, realism about species may appear doomed. I articulate a deflationary stance on species-reference which allows us to refer to species using plural referring terms. Species, on this view, are neither classes nor individuals. They are natural kinds of populations, as clarified in the final three chapters. (11,700 words)
Chapter 8: Natural Kinds as Stable Property Clusters
But an anti-essentialist conception of natural kinds has emerged as a strong contender: Boyd's Homeostatic Property Cluster (HPC) approach to natural kinds. While the HPC account has much going for it (I defend it against several criticisms), I argue that it imposes both too vague and too strict requirements on "homeostasis". I identify cases in chemistry and biology which apparently fit into the inferential role of HPC kinds but which do not seem to feature the right kind of "homeostatic mechanisms". Fortunately, a relatively minor modification of the HPC account addresses the above worries. On the view I favor — what could be called a "Stable Property Cluster" (SPC) approach — natural kinds are characterized by the existence of a cluster of properties that are collectively stable. What matters for natural kinds' role in our inferential and explanatory is the mere fact of a cluster of properties' stability. Whereas Boyd holds that stability must be secured by productive causal relationships, I hold that (especially in the biological world), stability is often maintained by the absence of certain kinds of causal pathways. And whereas Boyd views the HPC account of natural kinds as a kind of natural kind, I commend it as a general account, able to accommodate kinds with essences, kinds whose stability is secured by homeostatic mechanisms, and non-causal means (e.g., phylogenetic inertia). (11,000 words)
Chapter 9: Many Species are SPC Natural Kinds
This chapter will put the SPC account to work, showing that it nicely accommodates a variety of case studies of species discovery, inference, and explanation, evading some of the problems that have dogged the HPC approach (such as those of development and polymorphism). Unlike many adherents to SAI or essentialism, I do not wish to say that all recognized species are natural kinds — there may be some groups of organisms that do not have what it takes to satisfy the SPC account — but I shall show that these groups will typically not play nearly as robust a role in our inferential and explanatory practices. This chapter and the next will articulate many of the details of the SPC account in context of actual scientific practice. (~12,000 words)
Chapter 10: Pluralism and Realism Reconsidered
The final chapter returns to the question of realism about species in earnest, resuming the thread from Chapter 4: the SPC account's flexibility invites a robust kind of pluralism (one that cannot be deflected nearly as easily as the pluralism I discuss early in the book). But this raises a tricky question: Can a minimalist, pluralist view of species count as realism about species? A frustrating answer seems inevitable: "yes and no". Rather than attempt to parlay the SPC account of biological taxa into one of the slippery realist/antirealist categories, I argue for a comparative claim: species are as real as many other significant scientific categories. Insofar as the SPC approach to natural kinds explains the role of biological kinds in explanation and induction, it offers a sense in which our pre-theoretic intuitions about the reality of species are correct. (~12,000 words)