Marc asked:

My question concerns real vs. nominal definitions.

In brief: is it possible for real definitions to be either true or false?

For example, let’s assume I fix the denotation of the term ‘tiger’ (as I point to a large, four-legged cat). Then, I give a real definition of ‘tiger’: an eight-legged invertebrate.

Would it be reasonable to say that the real definition of ‘tiger’ I have given is false? Assuming the earlier denotation of ‘tiger’ I gave by pointing to actual large, four-legged cats?

Answer by Helier Robinson

I assume that you mean a real definition to be a definition of something in reality and a nominal definition to be a definition in language. A real definition is then a special case of a nominal definition (since it can hardly be a definition if it is not put into words) while a nominal definition may have no reference to reality but still be a definition (such as a mermaid being half woman, half fish).

We can then say that all real definitions are true, and all false definitions are nominal definitions. Ostensive definitions (definitions by pointing) become real definitions by assigning them a word, such as tiger, or a phrase such large four-legged cat; but the phrase must be a true description: eight-legged invertebrate is a false description and so a merely nominal definition.


Answer by Geoffrey Klempner

Your question relates to the analysis of ‘natural kind’ terms, which was revived by Hilary Putnam and Saul Kripke in the 1970s. Originally, a distinction coined by Locke in his Essay Concerning Human Understanding the notion of a ‘real’ as opposed to a merely ‘nominal’ definition fell into disrepute — it had become associated with the supposedly ‘bad’ Aristotelian metaphysics of ‘substance’ and ‘essence’, notions that looked suspect when viewed from the perspective of post logical-positivist analytic philosophy.

In his landmark paper, ‘Naming and Necessity’ (published in book form in 1980) Kripke showed a way to make sense of ‘real definitions’ of natural kind terms. To take an often quoted example, Gold, defined nominally as a ‘yellow metal, etc.’ is in fact not yellow in its pure form, but white. Iron Pyrites (fools gold) is not gold even though it is yellow and looks metallic.

What is Gold? Anything with the same atomic structure of that (pointing to a sample of gold). We identify samples of elements, say (it could equally be samples of animal species, such as a Tiger) and define the term as anything ‘similar in the theoretically relevant way’ to that.

This procedure, however, is not assumption free. You have to assume that the story about the Periodic Table of elements is true. It may be difficult to see how we could be wrong about that, but there will be other cases of ‘real definitions’ of supposedly ‘natural kinds’ which get going only because of a theory which is in fact false. We thought we knew what ‘kind’ we were pointing to but we were wrong.

With the perspective of a few decades, it now looks to me that the whole issue of ‘real essences’ and the revival of the Aristotelian/ Lockean notion of ‘essence’ was a trifle over-stated. Yes, there is an interesting distinction to be made, in relation to the practice of science — the way things are grouped into theoretically significant kinds. But really a ‘real definition’ is just a more sophisticated ‘kind’ (from the perspective of logic) of ‘nominal definition’. The idea of a real ‘real definition’ — which gets right down to the metaphysical nitty gritty of things — looks just as suspect now as it did in the heyday of logical positivism.