Brian Earp's ‘Some Writing Tips for Philosophy’ (Reference Earp2021) is useful for someone who wishes to do well in analytic philosophy as currently practised. It is also useful in casting doubt on why such analytic philosophy would have any interest for someone who wants to learn something new. It contains many good tips, such as: early in your paper explain clearly its structure; use short, simple sentences; avoid including irrelevant material; explain connections clearly; maintain a consistent terminology; make claims that are clear rather than vague; eliminate ambiguities; try to strengthen a position before criticizing it; acknowledge, at least implicitly, one's fallibility. However, two bad tips, if followed, as they generally are by analytic philosophers, will tend to produce a paper that says next to nothing. That is a problem because philosophers are supposed to be contributing to the growth of knowledge and they often used to do so. I list the two faulty tips, show how the practices of great philosophers and scientists contradict them, then set out some contrary good tips for philosophers who aim to write a paper that makes a contribution to our knowledge. I consider a possible objection before concluding.
Bad tip 1. Be very careful about making any universal claims (involving words such as ‘every’, ‘never’, ‘always’). Such a claim can be refuted by just a single counterexample. Do not leave yourself open to such refutation. Make a universal claim only if you are sure that there are no counterexamples.
Bad tip 2. Pick a smaller, more narrow thesis and argue for it thoroughly rather than a more ambitious thesis for which you argue less thoroughly.
Traditionally, philosophers often made bold, ambitious, universal claims to which there were many accepted counterexamples. But they then set out to explain why the counterexamples should be rejected. For example, Parmenides proposed that the world is an undivided homogeneous whole, without parts and without motion. He presented his bold claim as the conclusion of a deductive argument from the single premise ‘what is not is not’. He was fully aware that the world seems to be made up of myriad things which change and divide. But he dismissed such seeming counterexamples as illusion. The senses were not to be trusted; intellect is the only route to knowledge. The ancient Greek atomists, having more faith in the senses, regarded Parmenides’ claim as empirically refuted. They rejected Parmenides’ single premise by distinguishing the void from the non-existent. Empty space is nothing (void), and in that sense it is not; but empty space nevertheless exists, and in that sense it is. Parmenides’ argument depended on the conflation of the two senses of ‘is not’. The atomists’ bold claim was that the world is made up of atoms and the void. These ancient philosophers were not engaged in idle speculation. Atomism eventually became part of fundamental physics; and Parmenides’ ‘block universe’ has been likened to the universe posited by general relativity theory (Popper Reference Popper1972: 79–81, 145–6).
Rationalist philosophers, such as Plato, René Descartes, and Gottfried Leibniz, had maintained that some or all of the concepts or ideas that we use to understand the world are part of the cognitive equipment that belongs to our mind, rather than being derived from sensory experience. The concepts are innate, but not in the sense that we are explicitly aware of them when we are born; rather, in the sense that they are prior to our sensory experiences but are called forth by them. Kant provided a bold and complex a priori argument for the existence of a specific set of such concepts. In outline, he argued that we become aware of something by classifying it in some way. Classification involves a judgement, such as ‘this is of kind K’. But every judgement has a logical form which is linked with non-sensory concepts, or ‘categories’, such as substance–attribute, one–many, conditional dependence, necessity–possibility, and so on. Those categories have metaphysical import. So, when we obtain sense-experience of things, we obtain experience as of a world that conforms to the metaphysical principles of our conceptual scheme. The idea was taken up by empirical psychologists who discovered many different ways in which what we perceive (or think we perceive) depends upon our antecedent theories or expectations (for example, Piaget Reference Piaget1929: chs. 5–7; Kalat Reference Kalat2011: 123–40).
In the mid-nineteenth century, Urbain Leverrier was aware that the observed motions of Uranus did not agree with the predictions of that planet's motions that had been derived from Newton's theory. He claimed that there was an eighth planet which had just the right size and orbit to exert a gravitational pull on Uranus that would account for its observed motions according to Newton's theory. That was a bold claim because it was an accepted fact that there were seven planets. It was also a risky claim because, once he worked out the supposed eighth planet's size and orbit, astronomers could use their telescopes to check whether there was a planet in the places and at the times that Leverrier claimed. However, Leverrier's claim survived that empirical test: Johann Galle discovered the new planet. It was named ‘Neptune’.
Despite their differences, these three examples have much in common. First, in each case, a theorist was trying to find a solution to a problem. Parmenides was struggling to understand how change is possible. Kant was striving to reconcile the invalidity of induction with the existence of scientific knowledge, specifically, Newton's theory, which seemed to Kant and his contemporaries to be certainly true. Leverrier's problem was to explain away the apparent inconsistency between Newton's theory and accepted statements of the observed motions of Uranus. Each theorist (flying in the face of bad tips 1 and 2) tried to resolve his problem by proposing a bold, novel solution involving universal or empirically refutable claims. Such solutions run the risk of counterexamples. Parmenides knew of innumerable counterexamples but was not deterred: he tried to explain them away. Kant's audacious ‘transcendental proof’ has been subject to much criticism with regard to its validity; and it also seems to have been mistaken in trying to prove that there is one particular conceptual scheme that is necessary, given that the replacement of Newton's theory by general relativity showed that we can successfully modify our conceptual scheme (Popper Reference Popper1972: 190–2). Leverrier knew his solution would be refuted if astronomers did not find his posited new planet at the place and time he predicted; but if that had happened, he would then have come up with a different bold hypothesis to save Newton's theory (as is evident from his later attempts to explain away inconsistencies between Newton's theory and accepted statements of the observed positions of Mercury).
The reason that bad tips 1 and 2 are bad is that they inhibit the growth of knowledge. The latter depends upon bold new theories that contradict some of what we currently take ourselves to know. If those bold new theories, after criticism and testing, appear to be better than their available rivals, we end up discovering that some of the theories we held previously had been mistaken. We thereby learn something new. A philosophy paper, then, ought to offer a solution to a problem that gives us new insight, thereby increasing our knowledge. It can do that only by making a surprising claim; otherwise it just tells us what we already know. And it will tell us more, the bolder the claim made, provided that the claim survives criticism. The solution offered in a philosophy paper will therefore be better, other things being equal, the bolder and the more surprising the solution is; and thus the more open it is to the risk of refutation. To get substantial progress, we must take risks; many of the risky claims will not survive criticism; but the ones that do will make a substantial contribution to our knowledge. In contrast, Earp says that we should avoid making claims with a significant risk of being refuted (bad tip 1) and that we should make timid or narrow claims that can more easily be protected from refutation (bad tip 2). By following Earp we would guarantee little progress.
I therefore propose discarding Earp's bad tips 1 and 2. I offer instead the following.
Good tip 1. After a brief introduction and explanation of the structure of the paper, turn to clarifying the problem you are trying to solve.
Good tip 2. Try to propose a bold, novel solution to the problem and do not be frightened of making universal claims.
Good tip 3. Identify the main points at which your solution conflicts with currently accepted facts or theories (if there are none, you are probably wasting everyone's time). Investigate ways in which those accepted facts or theories may be mistaken and look for arguments that can show or suggest that the accepted facts or theories should be revised (a reductio ad absurdum argument is particularly powerful).
Good tip 4. Compare your solution to the leading alternative solutions to assess relative strengths and weaknesses and show how your solution comes out best on that comparison. Some important points of comparison are: whether the solution actually solves the problem; how simple the solution is; how illuminating the solution is; and whether the solution also solves other problems that are not solved by any of the rival solutions.
Good tip 5. Consider and rebut possible objections to your solution. Consider also whether your solution may be empirically testable and suggest, if possible, some empirical test.
It might be objected that I am holding Earp to a wrong standard. Earp, in his abstract, says that his tips were aimed at college students, many of them right out of high school, most of whom had never written a philosophy paper before. I, on the contrary have been considering papers that would be written by accomplished philosophers, perhaps leaders in the field.
The objection is mistaken on two counts. First, although Earp originally wrote his advice for new students, he says that philosophers generally may find it helpful for teaching or writing in philosophy, or in any other relevant field or discipline. Second, while students are more in need of guidance than are established philosophers, what they are being guided to do is philosophy. They ought to be taught how to think and write about philosophical problems, including how to evaluate alternative solutions, and they should be encouraged to come up with solutions of their own, since the best way of understanding a problem is to try to solve it. Our judgement of their efforts will be appropriately lenient; but their efforts should be attempts at doing what philosophers do. Further, as in mathematics, we should start students off with the less technical problems.
My aim here has been to offer a solution to the problem of providing advice for writing a philosophy paper. The solution I offer is bold and novel, though its novelty is due largely to Karl Popper (Reference Popper1972: 193–200), since it is a development of some of his neglected ideas. The solution conflicts with accounts of what philosophy should do that are generally accepted by analytic philosophers; but those accounts divorce philosophy from the growth of knowledge whereas, in fact, in its history, philosophy has contributed to epistemic, including scientific, advance. The solution conflicts with Earp's more or less standard solution to the problem. It is not as simple as Earp's; but, unlike Earp's, it explains how philosophy can be valuable in contributing to the growth of knowledge and thus why a person may have an epistemic interest in writing a philosophy paper. It replaces two defective tips given by Earp with five good tips. The solution survives an objection that might be made on Earp's behalf. It is also empirically testable by considering which types of philosophy papers actually make a genuine contribution to the growth of knowledge, though the form of such empirical testing will require careful consideration.
