Abstract:
Natural Kinds
A natural kind is a kind which has essential properties where those properties establish the identity of the kind and are important for the scientific explanation of the way it behaves. Although properties are essential relative to a conceptual scheme, theoretical identifications such as 'water is H2O' are genuinely explanatory and are compatible with causal theories of reference.
There are important similarities and differences between Aristotle's account of substances which have real definitions and Locke's account of real essences. Locke is more of a sceptic than a nominalist regarding the basic constituents of matter. A rather narrow account of natural kinds as Lockean real essences is accepted. Terms such as 'river' and 'shrub' are consigned the status of real but superficial kinds.
Some of the roles which natural kind terms have been called upon to play are reviewed: firstly, the Kripke/Putnam thesis that the essence of a natural kind is a variety of metaphysical necessity (this is denied); secondly, the thesis that scientific theories are true or that their terms refer to real objects (the latter is affirmed); thirdly, the thesis that natural kinds are a source of natural necessities from which a necessitarian account of the laws of nature can be derived (the discussion is deferred until part two).
Some problematical aspects of the notion of a natural kind are examined. The notion is defended against the charge of mind-dependence and restricted to structural rather than functional kinds. Biological taxa such as species and genera do not have non-trivial essential features and are therefore not natural kinds in the narrow sense defended here.
Laws of Nature
If natural kinds are important to science, can they be used to provide an account of the laws of nature? The naive regularity theory that laws are true universal generalizations is inadequate as it stands. Necessitarian accounts, (including an attempt to argue that the world itself is a natural kind and the laws of nature are its essential properties), are rejected as too mysterious and too expensive ontologically. An account of necessity is sketched which makes it an irreducibly epistemic notion, thus invalidating any account of the laws of nature which depends on an exclusively natural source of necessity. Various sophistications of the regularity theory are discussed, including Lewis's systematic account and van Fraassen's semantic approach. Matters of realism aside, these are the only accounts which appear viable.
Entropy
The second law of thermodynamics states that entropy is increasing in the universe. The claim is examined that this law leads to an alternative paradigm for science where the emphasis is on irreversible processes rather than the time-reversible trajectories of particles.
The nature of entropy and limitations to its significance are discussed. Even though the term is often interpreted too liberally, the direction of time should still be treated as an objective feature of the world.
The thermodynamic perspective threatens to diminish the explanatory power of natural kinds and its implications for laws of nature are even more serious. The combined effects of the reality of time, the primacy of dissipative structures in explanation, and the promotion of indeterminate effects to the macroscopic level would mean that regularity and necessitarian accounts of laws, even if they could be made to work, would be of little relevance to science. Some version of the systematic account, where a law of nature is a generalization which plays a fundamental explanatory role within a theory, seems to be the only contender to survive this change of perspective.