In philosophy, the problems are stated, outlined and solved from the armchair; we do not need to go out and do research of the sort the scientist does. In order to cast doubt on Plato's notion of man as a furless, featherless biped (etc) one does not need to go out and shave a chicken. It is sufficient to say, `imagine....' Of course we do need to do research in books, etc., but such resources are themselves, qua philosophy, formulated in the armchair. The main method of philosophy is argumentation and critique, consisting of checking validity, offering counterexamples and thought experiments. Given the nature of philosophy, the problem of the preceding essay has been to outline some constraints that thought experiments ought to take into consideration. Generating a thought experiment that does not contain contradictions is a first step. And such examples are the standard fare of what I have termed the usual, or typical, notion of logical possibility. But this does not get us far; as D.L.C. Maclachlan has argued regarding logical possibility, internal "consistency is a necessary condition that a theory must satisfy to be given a hearing, but a theory that merely satisfies this condition has not even begun to prove its case" (p. 71). The main contention of this thesis has been that this consistency is only a first step, for when dealing with entities that have essences more stringent constraints must be taken into account.

On the typical view we can imagine just about anything: as we saw at the beginning of the essay, Hume challenges us to refute his claim that there are no necessary universal synthetic propositions— viz., that such propositions can always be imagined to be false. One instance of this is the law of causality—that like causes must produce in the future like effects. We have never seen a man survive being thrown into a fiery furnace, but can we not imagine a man living through this in the future? We can imagine without contradiction the heat being abstracted away from the fire so that the man is not burned, or the man turning into stone for the duration of the fire (cf. Madden 1971 p. 68). Of course, the issue now comes down to believability. J. S. Mill once pointed out to Herbert Spencer, who claimed that when he felt cold he could not conceive that he was not feeling cold and that when he was looking at the sun that he could conceive that he was looking into darkness, that surely one can conceive of such things but one cannot believe of such things (in Tidman 1994 p. 303). One can conceive while looking into the sun that he is looking at darkness— but he would be hard pressed to claim that he believes such a thing. Of course Mill puts the issue in psychological terms that I would reject; one cannot believe such things because he knows that they are false-- the standard is set by extramental reality. Often the reason for not believing is that the essential natures of the things involved preclude such beliefs.

The adherent of the typical view of logical possibility says that he can conceive of many things happening; of course he never says that he believes that such things will happen. They are physically impossible. And the point of this essay has been to show that many other claims fall somewhere in between what has been thought to be either physically or logically necessary— such things are essentially necessary. So, what is the force of claiming, in the face of physical or essential necessity, that something is logically possible? It would seem to be that, whether the person believes his claim or not, the force of the very ability to imagine it is enough to show contingent some claim. It is the ability to imagine things differently from our armchairs that leads us to claim that we have produced a counterexample. As Arthur Collins argues:

What is imagined serves as a model for the simultaneous truth of a number of propositions just as a model suffices for the proof of consistency of a set of postulates. The imaginary model is offered as a rough proof of the consistency of the assertions made within the story that we find imaginable. What we imagine could be true and hence could not be contradictory.

That the model exists in fact is crucial for model proofs of consistency. The philosopher's model, however, is only imaginary. It seems a sensible caution to wonder whether we are not merely imagining that the story that we tell could be true, merely imagining that this state of affairs is possible and, therefore, merely imagining that it involves no contradictions ( 1967 p. 51).

As with the time travel example outlined in the first chapter, we need to ask whether the story, within its full context, is noncontradictory, or whether our way of specifying it in a narrow manner makes the story only seem noncontradictory.

So, in what sense does an appeal to logical possibility as a counter example give us reason to change our minds? It seems that, while we can conceive that a man can survive a fall into a fiery furnace, we do not believe that a man could survive such a fall. But nor do we find the ability to conceive it important; if the argument of this essay is correct, we need to consider a posteriori necessity (which was anathema to Hume). Such necessity can be found, according to Kripke and Putnam, in scientific theories which uncover the natures of things (such as gold= atomic number 79). Such truths require someone to get out of his chair and observe and interact with the world which is usually the area of the scientist. While we cannot base philosophical theories on scientific ones, since philosophy is prior to science and must itself explain science, we can inform our philosophical theories with scientific truths. Also, there is the need to do with scientific theories what scientists, qua scientists, do not do with their subject: namely, to see what kind of necessary truths their enquiries imply. Essential necessities about natures of things, which are found through empirical investigation, are the guide we need for constructing thought experiments. When these are taken into account, we can be sure that it is a necessary truth even though the opposites are logically conceivable that sold iron bars sink in water and cows will never jump over the moon.