The aims of philosophical and non-philosophical writing are not identical. Good philosophy essays often break rules recommended for other kinds of writing, e.g. that essays have substantial introductions and conclusions, or exhibit copious reading, or eschew the first-person pronoun. Knowing what's required for a philosophy essay can therefore be difficult. The following might help. (So might an excellent book on writing and doing philosophy, The Practice of Philosophy, by an ex-teacher of mine, Jay Rosenberg.)
Before writing, leave enough time to read, think, and plan. These overlap. Good, critical reading isn't passive; it involves thinking, interrogating the text, developing questions and objections. As for planning, I use big A2 sheets of paper, criss-crossed with arrows and diagrams aimed at laying bare the structure of issues and arguments. Only then do I start to get a sense of exactly what I really think about it. Even when writing, thinking hard about the issues remains crucial. It might lead to a change of mind and some re-writing. Despite the overlap, each element -- reading, thinking, planning -- is crucial. Don't, for instance, spend so much time reading that you have no time to think about the issues for yourself or to plan what you want to say.
The main point
Argue for a thesis; address an issue; answer the question. It's not enough just to string together some interesting remarks about some theme, e.g. "Freewill". You must set up a clear issue to be addressed, a puzzle to be solved, a question to be answered. And then, of course, you must get stuck in, engage with it; take a view and defend it. This might require some context-setting, some comparison of the standard views and arguments, but you must go beyond this. Description isn't sufficient. Explain your position and, crucially, make a case for it. Answer the question, and make it clear you've done so. If the question contains the phrase "sense data", say, don't just write down everything you know about sense data; look at precisely what is being asked about sense data, and speak to that.
Don’t merely survey the literature. Locating your point within an existing debate can be important, but don't only do that. And while it's fine to make points found in handouts and lecture notes, slavish regurgitation of teaching materials is easily spotted and frowned upon.
Avoid long, aimless introductions and conclusions. A snappy introduction can usefully sketch the issue and gesture at the shape of what is to follow. But keep it simple and brief; avoid the grandiose (e.g. "Since the beginning of time, man has wondered ..."). If you have a concluding paragraph, use it to succinctly tie the threads together. But then stop. In philosophy, it's okay to jump straight in and out, more-or-less.
Have a strong authorial voice. You should be attempting to persuade your reader of your thesis, so the first-person pronoun ("I") will often be useful. Its use is not discouraged.
But be modest and avoid polemics. Having a strong authorial voice doesn't mean, for instance, not acknowledging how finely balanced a particular issue is. Doing so can be laudable. Don't give the impression that the answers are obvious unless they really are.
Striking originality is neither necessary nor sufficient. That your argument or thesis can be found in the literature is fine, provided you make it yours -- not by failing to acknowledge your sources (you must acknowledge them), but by endorsing it, clarifying it, and trying rationally to persuade your readers of it. Originality can sometimes just be a matter of making a point in a better or clearer way than others, perhaps drawing a distinction or effecting a connection that others miss or misdescribe. It can be particularly impressive if your point hasn't been made elsewhere, of course, but this is exceedingly rare, and impressive only if the point is plausible, clearly articulated, and cogently defended.
Don’t feel obliged to mention everything you have read. Reading lots is crucial, but as a means to knowledge and understanding, not as an end in itself. The mere fact that you've read a lot will not be rewarded.
Don't pack too much in. Lazy students' essays are often too thin; conscientious students sometimes pack so much in that they lack space in which to explain it or develop it. Packing too much in is extremely tempting if you want to show that you've worked hard, but the temptation should be resisted, since it prevents depth. Of course breadth too is important; some points are so crucial that space needs to be found for them somewhere. Getting the breadth/depth balance right is, of course, a matter of judgement.
Avoid excessive quotation. It reduces your contribution. Avoid, in particular, the following sort of thing: "According to the Oxford English Dictionary, ‘a priori’ means …”. Dictionaries of philosophy will help you think about what a term of art term means. But if the term needs explaining (which will depend on the context), explain it in your own words. Quotations play a greater role in more historical essays, e.g. in a discussion of which of two interpretations of Descartes is correct.
Clarity and Structure
Be clear. Read your essay aloud to yourself or others. Even if it seems to make sense to you, think whether it would if you hadn't written it or hadn't taken the course. Put yourself in the position of an intelligent reader not already immersed in the issues. Don't think, "My tutor's marking this and will know what I'm getting at". Try different ways of explaining difficult points. Read them to yourself and use the clearest. Ensure the central point stands out and is not lost amongst the detail. And, of course, have a final check of spelling and grammar before submitting the essay.
Impose a sensible and sign-posted structure. Sections can be useful. Sensibly sized and themed paragraphs are essential. Avoid overly long paragraphs. The machinery of many good novels is hidden, but not so the machinery of many good philosophy papers. To make it explicit, use such sign-posting as “First”, “Secondly”, “Thirdly”; “This shows that…”, “I have shown that…”, “I will show that …”, “I argued in the preceding two paragraphs that p, but now I turn to an objection to p…”. Ensure the important points are centre stage, and less crucial points remain in the background.
Pay attention to how others write. You’ll read authors you find hopelessly obscure and others who make everything click into place. Think about how the latter manage to be so clear. Think how to emulate them.
Avoid both the book report and the polemic. Book reports look (a little!) like this:
A. The Book Report
There’s a philosophical problem X, in response to which philosopher A says that p, B says that q, C says that r. (Footnote: see how much I’ve read.) The End.
Although they often aren’t, book reports can be good, if they are well structured and express clearly a knotty philosophical issue and some treatments of it. But they are unlikely to be excellent, simply because they fail to take a view; they don't endorse or argue for a thesis. They are merely descriptive, perhaps historical. They are about philosophy but they don't do philosophy.
With this criticism ringing in their ears, some students produce the following:
B. The Polemic
There’s a philosophical problem X, in response to which philosopher A says that p, which is obvious rubbish. The truth, obviously, is q. The End.
Or, to fill out the structure a little:
“Do Descartes’s Arguments for Dualism Succeed? Since the dawn of time, man has worried about what sort of thing he is. Descartes says that each of us is a soul. But how could somebody hold such an unscientific view? Descartes only believed this because of his religious feelings and his fear of death. Rather, it’s quite obvious that we are connected to our brains; our brains cause our mental states. For my part, I used to believe I was a soul, but now I've read some neuroscience, I find it incredible. Many people reject my view, but only because they cannot stomach the unromantic, scientific truth. The End.”
I too am no fan of Cartesian souls but this isn’t the way to make a case against their existence. Admittedly, the essay has a thesis, which is a step in the right direction, but it doesn't contain an argument for it.
It's just dogmatism and rhetoric. You should avoid rhetorical flourishes like “since the dawn of time man has worried about …”. (Man hasn't worried about anything since the dawn of time.) Also avoid rhetorical questions (“How could somebody hold such a unscientific view?”) and autobiography (“I used to think that p, but then I read so-and-so, who made me see the light”). And, while I am at it, avoid psychological speculation ("Descartes only believed this because ..."). The issue is not what explains Descartes’s beliefs, but what if anything justifies them--that is, whether his or others' arguments are any good.
The above essay also has far too little by way of explanation or clarification. What is dualism? What are souls? What are Descartes’s arguments for them? What does it mean to say we’re “connected” to our brains? Is the idea that our brains cause our mental states really incompatible with dualism? Something clear and helpful needs to be said on all these points.
So what do good essays look like? They'll usually contain at least the following three ingredients:
(i) clear, accurate, and illuminating exposition of a question or puzzle or a position of some philosopher;
(ii) a clear and plausible thesis;
(iii) and some cogent argumentation for that thesis.
There are countless shapes that essays with these virtues might have. Here is just one promising example:
C. A promising structure
There’s a philosophical problem X. It can at first sight seem spurious because … But actually it is a deep and perplexing difficulty because … In response to it, A says that p. What p amounts to is … By contrast, B says that q, in other words that …. I think B is wrong because his q faces the following difficulties … Now, B’s argument for q is … but this argument doesn’t work because … . Turning to A, she has two arguments for p, namely … . B criticises both arguments. Actually, B is surely right about the first, which doesn't work because … . But A’s second argument is more promising, especially when re-formulated as follows … Thus formulated, it escapes B’s objection because … . Hence p is an attractive way of dealing with problem X.
Needless to say, you shouldn't crow-bar every essay into this template, which is only one of countlessly many. And whether the resulting essay really is any good depends on how the shape is fleshed out, i.e. whether the exposition is clear and accurate, whether the author advances a well articulated and plausible thesis, and whether the author's arguments are lucid and cogent.
So that's it. Get to work; and good luck!
(There is more essay-writing advice here. But keep in mind that, as helpful as all this advice can be, there's a point at which what's needed is not more advice but more time and effort reading, writing, and doing philosophy!)