How to do Philosophy
We are confronted by a world we don’t fully understand. We try by various methods to gain understanding of it, sometimes successfully. Philosophy is one such effort. Suppose we want to understand X: we talk about X, have thoughts about X, but we don’t know the nature of X, or the complete nature of X. We are in a state of ignorance about X, though we know enough about it to make it an object of contemplation. We then seek a method that will reveal the nature of X—a source of data, information, insight. Let’s suppose our interest is of the kind we call philosophical (as opposed to scientific, historical, or practical); it isn’t easy to say what this kind of interest consists in, but let’s leave that question aside. As an initial example, suppose X is pain: we refer to pain, feel pain, remember pain, expect and dread pain—but we don’t know what pain is. Is it a state of the nervous system, a functional state, a state of an immaterial substance, or something else entirely? That is, we are considering the mind-body problem. Presumably there was a time when the first human did this: he reflected philosophically on the nature of pain. He wasn’t thinking about the causes and effects of pain, possible cures for it, its prevalence in the population, or which pains were the worst; he was considering what we might describe as the very nature of pain. He was doing metaphysics. Evidently, his current knowledge of pain didn’t settle this question, so he needs to appeal to a further source of information. It is not obvious what this would be: he is ignorant of the answer, but nothing immediately suggest itself as the route to enlightenment. That is one of the first problems of philosophy—how is it to be done? We know there is something we don’t know, but we also don’t know how to remedy this lack. Should we just concentrate hard on the phenomenon and try to discern its nature? But that doesn’t seem to help—the real nature of pain is not given to us, not directly transparent. We therefore need to find an alternative source of data: introspection alone won’t cut it, so we need to look elsewhere—we need an indirect source of data. Then we can infer the nature of pain from the data. It is the same with many topics of philosophical interest—knowledge, meaning, causation, necessity, truth, time, goodness, etc.: we don’t get far by attempts at direct inspection, by looking and seeing. What would this even be in many cases—how would you look at and see truth, for example? The mind’s eye turns up nothing. I suggest that there are four possible ways of trying to get at the nature of the thing that puzzles us, all deployed at one time or another: language, consciousness, modal intuition, and knowledge. How is X represented in language? How is X presented to consciousness? What are the possibilities and necessities attaching to X? How do we know about X? All four are directed at elucidating the nature of X philosophically. As far as I can see, there are no other ways. I will call these the Four Ways, echoing Eastern modes of expression; and my first claim is that the Four Ways exhaust the ways. They don’t exhaust all the ways of knowing about the world available to us as human beings, but they do exhaust the ways of answering philosophical questions available to us (asking God’s opinion is not an available way). So, the way to do philosophy is to employ one or more of the Four Ways. We can do philosophy linguistically, phenomenologically, modally, or epistemologically. I don’t need to say much about each of these, as they are well known, but it will be useful to set them out in relation to each other. My general position is that all four are legitimate and potentially fruitful; I am not a philosophical sectarian. Thus, the linguistic way involves us in questions of meaning, speech acts, logical form, grammar, and lexical analysis. The phenomenological way involves us in acts of consciousness, intentionality, introspective intuition, perceptual and intellectual apprehension. The modal way invites us to use our modal intuition to decide what is essential to the thing we are investigating and what is contingent: what is conceivably true of X and what is inconceivable (for example, false knowledge is inconceivable but knowledge in the womb is conceivable). This method enables us to separate what belongs to the very nature of X from what is extraneous to that nature, and hence provides useful information. It is where counterexamples to claims of analysis come in. The epistemological way asks how we know the thing in question—a priori or a posteriori, directly or by inference, with certainty or without, in virtue of knowing something else or primitively. One application of this method is the so-called knowledge argument: that is, whether it is possible to know the nature of X by knowing facts seemingly distinct from facts about X—as in knowing the nature of the mind by knowing the nature of the brain, or knowing the nature of the physical world by knowing facts about the phenomenal world. This enables us to determine whether some proposed analysis of a thing’s nature is really successful. In general, the idea is that the nature of X will be (partially?) revealed by examining how X is known, since the nature of a thing fixes the way in which it is known (e.g., mathematics). The picture is that X has a determinate nature N and N is revealed in the aspects of X that show up in its relations to language, consciousness, modality, and knowledge. The Four Ways tap into the nature of X, each in their different way, so that we can gain a fuller picture of this nature by seeing how they each reflect it. They each point to this nature, though they don’t exactly contain it (they are not identical to it); they provide us with evidence about the thing that interests us. In some cases, we can be confident that we have identified the correct nature by employing these methods, as with simple analytic truths: each method will certify that “Bachelors are unmarried males” states the essence of bachelorhood (I leave this as an exercise for the reader). In other cases, this won’t be obvious but can be ascertained by diligent conceptual analysis (e.g., Bernard Suits’s definition of a game). In yet other cases, we can provide a partial analysis but perhaps not a complete analysis (e.g., the analysis of knowledge as true justified belief). Perhaps in rare cases no account of X’s nature is forthcoming, possibly because of conceptual poverty. Convergence of the Four Ways will be taken as confirmatory. Is any primary? Probably not: each has its strengths and weaknesses. But it’s not a competition—we can use all four ways to conduct our philosophical investigations. Each of the Four Ways is fallible, of course, but that is just part of rational inquiry. At least we have a method, or several methods; we are not saddled with blank staring and unsullied ignorance. It is actually possible to do philosophy! The real nature of things can be accessed by the examination of related areas—language, consciousness, possibility, and knowledge. These all contain clues, indications, suggestions, even if they don’t provide transparent revelations of the nature of things. We shouldn’t really expect more; after all, philosophy only exists as a non-trivial subject because our minds don’t already contain knowledge of essences (philosophical knowledge is not present in the genes). In this regard, philosophy is very like empirical science, i.e., not possible by simple unaided intuition. In fact, I think philosophical methods are less naturally prone to error than scientific methods, because of the existence in the latter case of perceptual illusions, experimental mistakes, and the prevalence of induction (not to mention the necessity of sheer speculation). In philosophy, at least, what we are interested in is right under our noses, unlike the physical universe. Still, if it weren’t for the existence of language, consciousness, modal intuition, and knowledge, philosophy would be well-nigh impossible; we would remain in a state of utter ignorance about the nature of many things. We don’t possess a dedicated organ for obtaining philosophical insight, so we have to rely on these indirect sources of information, imperfect as they are.
 It is a strange thing that many philosophers have shown a tendency to restrict their sources of data to one class of data–as it might be, language (often further restricted to communicative use). They then argue with each other about what method is best. One would think that philosophy is difficult enough without imposing such restrictions on its methods. Better to adopt a more pluralist approach.
 A common error here is to suppose that because we rely on these sources of information philosophy must be about them. But it is not, any more than physics is about meter readings; it is about the nature of things—reality itself. So-called linguistic philosophy (rightly understood) was about reality via language.