The above books are great, but unless you are really into philosophy I don't think you'd gain much from reading primary sources. (With the exception of Hume, he's brief and brilliant.) I think someone casually interested in philosophy would be better off reading articles from the Stanford Encyclopedia of Philosophy (plato.stanford.edu); and if a subject is particularly interesting then they can always go read some related primary sources.

If you do start reading primary sources, particularly the early modern philosophers, I recommend visiting www.earlymoderntexts.com. Jonathan Bennett has done a great job "translating" the works of Hume, Kant, Locke, Berkley, etc. from old-English (or German, French...) into modern-English so that their philosophical arguments can be understood without being obfuscated by language quirks.

I respectfully disagree. In many cases, reading secondary sources is pointless without also reading the originals. (It's a bit like reading a review of a meal, rather than eating any of it.)

That said, the quality of secondary works varies a lot. Some things, like Mackie's Problems from Locke, are now primary sources in their own right, albeit in a secondary way.

But you can't really pick up most primary sources and actually get anything out of them without understanding the historical/philosophical context they were written in. If you can find someone with little to no philosophical background who can pick up A Critique of Pure Reason or the Tractatus and actually get anything out of it then you've probably found Kant or Wittgenstein reborn.

That's why the parent commenter said you need a guide/instructor to properly understand what is going on in these books, and I agree with that. But for people without such a luxury, secondary sources and articles are a good method for gaining some rudimentary understanding of the philosophical topics.

I don't disagree with some of your larger point. Most people will get the fullest benefit out of a classroom setting, with an engaging and well-read class leader and an engaged and thoughtful group of students to argue with.

However, I still believe that a determined outsider can get real benefit out of philosophy on his or her own. (Not the fullest benefit, but significant benefit.) I completely agree with you about cultural/historical/philosophical background, which is why in my answer I made a point of recommending not just titles but specific editions for classical works. Maybe we could call this the best of both worlds: good primary sources, filled out with enriching secondary material, all in one edition? (It does tend to make those editions more costly than some, and also heavier.)

Yes, this is why I gave a long caveat, and tried to picked only the most accessible primary sources. If there's some work in particular in the list that you think is inappropriate, I could try to defend it.

I think that the Stanford Encyclopedia of Philosophy is very mixed. Often (not always) the articles spend a lot of time making distinctions among many views, without adequately explaining any of them at an introductory level, and at the expense of any sort of broad view of the issues. So like most philosophy reading it's most useful if you're already in a philosophy course or something like that.

This has always puzzled me. Why does it seem that the vast majority of philosophers love to stray away from plain, clear, specific definitions and step-by-step, verifiable logic? Why is it that people can't be more like Hume or Feynman? Maybe they think that complex ideas can only be explained in a nebulous, vague manner? Every time I read an overgeneralization such as "subject <A> is <x>", especially when the author does not seem to be a scholar in the subject, I cringe. (I'm talking about the obvious need for "guide books". In some cases it's helpful to know the background and such, but many of these books are almost a paragraph by paragraph analysis.)

Actually, a lot of contemporary philosophy -- at least as it is practiced in the English-speaking world -- strives for exactly the "plain, clear, specific definitions" and "verifiable logic" you're looking for. I think the overgeneralizations and nebulous explanations you refer to are more typical of "Continental" philosophy and related disciplines like "critical theory."

The former style of philosophy does produce work that is clearer and easier to read, but it has its own limitations: in the worst cases, philosophers get so hung up on being clear that they shy away from discussing the difficult problems that bring people to philosophy in the first place, and instead spend their careers arguing over minutiae or practicing pseudo-science. Clear definitions aren't going to provide you with the final word on how to treat others, how to organize society, what really exists, or how we know about it.

Then again, the latter style of philosophy doesn't necessarily do much better on addressing these "big questions." Sometimes it is better at keeping these questions in view, but (as you point out) it is often extremely esoteric.

The best philosophers, I think, understand that clear definitions are a starting place, a model that needs to be developed. They see that model as a tool for addressing the thorny, complicated, and tangled questions that philosophers are asked to answer, not as a truth that must be defended in its own right. The goal is to find real philosophical truths. Where a rigorous, methodical approach helps us attain that goal, it should be welcomed; but where that approach becomes a hindrance, it should be left behind.

No Hegel?