Ereshefsky, M., Species Pluralism and Anti-Realism, Philosophy of Science, Vol. 65, No. 1 (Mar., 1998), pp. 103-120

Monists believe that biologists should settle on a single species concept. Pluralists maintain that a number of species concepts should be accepted as legitimate.
(..)
study the relation between species pluralism and anti-realism (cf Rosenberg,1985 and Stanford, 1995).
(..)
Contemporary biological theory provides ample evidence that the tree of life is segmented by biological forces into different types of species taxa. Consequently, there is no single unitary species category, but a heterogeneous collection of base taxa referred to by the term 'species'.
(..)
The main argument of this paper casts doubt on the reality of the species category but not the existence of species taxa.
(..)
Two approaches to species - the interbreeding and the phylogenetic - carve the tree of life in different ways. Many interbreeding species fail to be phylogenetic species, and many phylogenetic species fail to be interbreeding species.
(..)
Stanford (1995) has adopted Kitcher's (1984) form of species pluralism and argued that it leads to anti-realism: "Kitcher's species therefore lack that property, supervenience on the state of the mind independent material world, which we demand of real objects." (1995, 83) (..) species concepts are only correct relative to a conceptual scheme and not according to the world.
(..)
Stanford needs to give us more than the standard skeptic's argument against species realism. On the one hand, it fails to cite anything particularly problematic about species or pluralism. On the other hand, it begs the question.
(..)
Why did Linnaeus require that an organism belong to only one species? It stems from his adoption of Aristotelian essentialism (Cain 1958, 146ff.; Larson 1971, 146). According to that form of essentialism, each entity belongs to a single least inclusive kind, what might be called a 'fundamental kind'.
(..)
A realist approach to species pluralism allows that an organism can belong to more than one species, while the Linnaean hierarchy requires that an organism belongs to only one species.
(..)
How can a single organism belong to two different real species? Antirealists can argue that this is not a problem because on their view species do not exist (cf Stanford, 1995).
(..)
A pluralist can drop the essentialist requirement that each individual have a single essence. That would allow an organism to belong to multiple real species.
(..)
Kind essentialism asserts that kinds are defined by kind-specific essences. Individuale ssentialismr equirest hat each individualh as a single set of essential properties throughout its existence. Some authors adopt one form of essentialism while rejecting the other.
(..)
ME: the desire for hierarchical classifications does not pose a conceptual roadblock to a realist interpretation of pluralism.
(..)
What is common to the entities referred to by the term 'species'?
(..)
Taxa are genealogical entities - First, the commonality of species taxa lies in the similar type of process that renders species taxa cohesive entities. Second, the common nature of species taxa lies in their containing a similar structure. The first is a claim about process, the second a claim about pattern.
(..)
Interbreeding species are causally integrated entities whereas many phylogenetic species are not. In other words, species pluralism implies that the species category is an ontological mixed bag.
(..)
In general, many interbreeding species lack the structure of phylogenetic ones, and many phylogenetic species lack the structure of interbreeding ones.
(..)
A realistic interpretation of species pluralism- one that accepts the existence of different types of species taxaimplies that there is no unified ontological category called 'species'. It implies that the species category does not exist.
(..)
[Essentialist account (Locke, Putnam): all and only the members of a natural kind share a set of properties that explains the other necessary properties of those members. Cluster accounts of kinds (Mill, Hempel): the members of a kind share a number of common features that allow us to make new inferences about those members. EM's weaker account: one common feature would suffice, and no new inferences are required]
(..)
The biological world consists of a plurality of theoretically significant base lineages. Species pluralism is the result of a fecundity of biological forces rather than a paucity of scientific information.
(..)
The non-existence of the species category does not imply that the taxa we call 'species' are mere artifacts.
(..)
There still is an ontological problem concerning the reality of the species category. Call those base lineages whatever you want, the existence of the species category is dubious.