In this essay I shall use the word "grog" to mean such-and-such.
As long as such stipulations are clear and consistent and the author consistently holds to them, there is no objection.
If a philosopher asks a question like "What is death?" on the other hand, he's not just after some stipulative answer. He wants to know what death really is. He wants to know what we're thinking and talking about when we think and talk about death. He's seeking an analysis of our pre-existing concept of death.
Thought-Experiments and Counter-Examples
One way we test analyses is by trying to come up with counter-examples.
Say for instance that Professor Smith analyzes death as: having the biological processes of your body stop.
To test his analysis, we try to imagine a case where some creature has died but the biological processes of his body continue, or a case where the creature's biological processes have stopped but the creature is not yet dead. To do this is to engage in a thought-experiment. A thought-experiment is sort of like an imaginary test case. We're trying to see whether we can conceive of some situation that's incompatible with the proposed analysis.
Philosophical thought-experiments often involve pretty far-out science fiction. For instance, this term we'll be discussing brain transplants, teletransportation, and time-travel. Newcomers to philosophy tend to find all this science fiction bewildering. What relevance can science fiction cases have to real life?
To answer this question, you have to understand the nature of philosophical claims and what's required to produce a counter-example to them.
Professor Smith, for instance, is trying to tell us what death is. He's not just making a claim about actually existing creatures on the planet Earth, and what happens when they die. He's making a claim which purports to be true of any imaginable creatures anywhere, no matter how bizarre and science-fiction-y they may be.
Hence, Professor Smith's claim about what death is seems vulnerable to the following counter-example. Suppose Charles is put into suspended animation, and his body is frozen to near absolute zero. One week later, he is thawed out and revived. Now, during the period where he was frozen, all biological processes in his body had stopped. But it does not seem correct to say that Charles was dead during this period. Hence, Professor Smith's analysis of death is incorrect. Charles' biological processes had stopped but he was not dead.
Perhaps it is not in fact technologically possible to freeze a person and revive him again. This is not important. Professor Smith's claim purports to be true of any imaginable creatures anywhere. So if it's possible even in principle for someone to be frozen, and for his biological processes to stop, without his thereby dying, then Professor Smith's claim is false. This is what our counter-example purports to show.
First, Professor Smith might have given us a perfectly good biological test for death: a way of checking whether actual creatures of the sort we're likely to come across have died. It's just not a good analysis of death.
In general, we want to distinguish between questions about what it is to be X and questions about how we find out that something is X. In the same way, defining the difference between two things--say, hydrogen and helium--is different from finding a practical way to tell hydrogen and helium apart.
Second, in offering our counter-example, we appealed to certain intuitions about whether Charles had died in the imagined scenario. We often do this in assessing philosophical claims.
It's important to acknowledge that our intuitions aren't sacrosanct. Sometimes they're wrong. For example, modern physics forces us to revise many of our intuitions about time, space, and probability.
So it can sometimes happen that we ought to accept a philosophical claim that conflicts with our intuition, and throw out the intuition. But in general, there is a presumption that our pre-philosophical intuitions are true, and we should throw them out only if we have very good reasons for doing so.
Sometimes we can say "Imagine a situation in which..." and go on to describe a situation which is incoherent or contradictory or otherwise impossible. For instance, if a philosopher says "Every square has four corners," and you say "Not so! Imagine a round square," you haven't in fact described a coherent possibility, and so you haven't succeeded in offering a genuine counter-example to his claim.
Sometimes it's hard to tell whether you've described a coherent possibility or not. That's a big part of what makes philosophy so difficult.