Colin asked:

What is the definition of ‘definition’?

Answer by Geoffrey Klempner

Ah, what a great question! I don’t recall anyone asking this (going back to when Ask a Philosopher was launched in 1999). However, earlier this year in an answer to a question on the difference between ‘real’ and ‘nominal’ definitions, Hubertus Fremerey states:

“My definition of ‘definition’ would be: A method of bringing some order into the boundless chaos of experiences — sensual and intellectual.”

This is a nice idea, but to my ear falls short of the requirements of a definition. Fremerey goes on to say:

“The whole concept of ‘real definition’ is a misnomer, and even ‘nominal definition’ is. What you have as primary givens are experiences, and then you first attach labels to them and if needed you re-define the (sensual or rational) objects in the context of a theory.”

This is Fremerey’s ‘take’ on the question. A take isn’t a definition, although the way you take something can be relevant to formulating a definition. A case could be made that when analytic philosophers ‘define’ concepts (the concept of a person, or causation, or event, or etc.) what they are really doing is offering takes or theories. What they are looking for is an insightful view of the concept in question and how it fits in to our conceptual scheme.

A notorious case would be the fruitless attempt to define ‘knowledge’, with ever more elaborate sets of conditions, designed to cope with every possible counterexample. A theory of knowledge is what one is after, but a theory doesn’t necessarily require that you give a set of conditions that uniquely identify the concept in question.

So let’s narrow the question and concentrate on definitions rather than theories. A definition of a term should give you sufficient information to be able to use that term successfully and correctly, provided only that you understand the terms used in the definition. The Oxford English Dictionary (‘on Historical Principles’) is a model of this approach, which offers examples of the use of the word in question, especially early or first uses, as well as a sketch of its etymology.

A popular question in English-speaking philosophy in the 50s or 60s would have been, ‘What is the difference between a philosophical analysis and a dictionary definition?’ Now that I know how, e.g., the English word ‘person’ is used in normal conversation, its derivation from the Latin ‘persona’, what more do I need to grasp the concept of a person?

Let’s stick with this example. Suppose that the predicted breakthroughs in AI come to pass, and the first intelligent creatures with artificial brains roll of the production line. Are they persons? Whom should you ask, a philosopher or a lexicographer?

Imagine a future dystopian society where AI creatures are exploited and abused because they are not regarded as ‘persons’ by the general public despite the objections of philosophers. Or, alternatively, a society which happily embraces these mechanical beings into the ‘human race’ despite the objections of philosophers. A lexicographer is interested in how a thing is regarded, without questioning too hard the basis for the belief in question. The philosopher is the one who asks, ‘But is it really?’

As stated above, the philosopher may only be able to offer a take, not a definition. But a take can be sufficient to defeat a false definition.

What is a definition of ‘definition’? There is more than one kind of definition. I have given two, for the sake of contrast, but there are probably more (consider, e.g. the use of definition in mathematics). If there are two kinds of definition, then there are two (or possibly four!) definitions of ‘definition’, and if there are three, then etc. The point, however, is to decide what kind of definition we (as philosophers) are primarily interested in.