Procrastination and weakness of the will (as when people continue to smoke and to eat a lot of red meat even while they understand the health risks and want to lead a long healthy life) are obvious problems for the view you are entertaining. Another problem is moral and altruistic conduct. You are kind to a stranger, or generous to a rival, at some cost to yourself -- are you acting in your own (perceived) best interest? Not in any ordinary sense. Agents themselves will often deny that they decided on the basis of what was in their own best interest: "Here I tried to act in his best interest, not my own."

Now you can simply always overrule such agents. You might say that an agent's conscious conduct necessarily is conclusive evidence that she must be taking herself to have some interest that she takes this conduct to promote -- perhaps an interest in being regarded (by others or at least by herself) to be kind or generous, or a strong interest in smoking, eating red meat, or procrastinating. But if you say this, your point is in danger of disengaging from empirical reality. It is then no longer a proposed insight into human behavior (which can be empirically investigated and supported or refuted with evidence), but becomes a stipulation or axiom of your thinking which is wholly immune to empirical assessment (and indeed, you once call your point a premise).

Insofar as you are interested in understanding human behavior, you want to avoid interpreting your point as a stipulation. You want to interpret it instead as an empirical hypothesis. You then need some empirical criteria for the predicate "believes at time t that doing X is in her own best interest". And you can then empirically investigate whether indeed, whenever a person decides at time t to do X she also believes at time t that doing X is in her own best interest.

You ask for an accessible read for a busy salaryman. I would recommend Derek Parfit's Reasons and Persons, and there especially the long discussions of the self-interest theory (attractively named "S") in Parts I and II. This is not an empirical theory about how people decide, but a normative theory about how they have most reason to decide. Nonetheless, you will learn a lot from the differentiations and difficulties Parfit discusses in regard to what it means to act self-interestedly. You will see how very hard it is to give clear meaning to your phrase "acts in her perceived best interest".

This is an interesting question, in fact, so interesting that I am writing a whole book about it (Rediscovering Free Will). As Miriam says, much depends on how you define free will. Let's not begin with the problematic assumption that free will requires a non-natural power to transcend the causal interactions in the natural world, though I think we can begin with the idea that free will involves our powers to control our actions in light of our deliberations about what to do, such that we can be properly held responsible for our actions. In that case, we should not begin with the assumption made by some scientists writing about free will: that increasingly complete scientific (naturalistic) explanations of human decision-making thereby rule out any role for free will. Rather, it may be that neuroscientific and psychological explanations of human decision-making can help to explain--rather than explain away--our capacities to deliberate about our reasons and to control our actions accordingly. Here is a very brief summary of what I think science might say about human free will:

1. Scientific evidence for determinism would not prove that free will is an illusion. This is because determinism does not properly entail most of the things people take to be intuitively threatening to free will, including the things that scientists have recently been talking about when they say they are showing free will is an illusion, such as the idea that our "conscious will" plays no role in our actions (e.g., Libet, Wegner, etc.). Determinism, properly construed, is consistent with our conscious deliberations and intentions influencing our actions. And these purported scientific threats are also consistent with indeterminism. (By the way, neither neuroscience nor psychology is going to show determinism is true--i.e., that given certain causal antecedents, certain effects necessarily follow. Only fundamental physics has a hope of showing this to be true, and the current interpretation of quantum physics suggests determinism is false.)

2. The scientific evidence for epiphenomenalism--i.e., the causal irrelevance of conscious mental states--is not there. The claims that non-conscious processes are sufficient to cause our actions, while our conscious awareness of our intentions comes too late to play a causal role, are not supported by the evidence, especially if one considers conscious deliberation, planning, and intention formation that occurs well before action, which is the sort that seems most relevant for free will. It's more important that my thoughts today about what I want to do tomorrow (or with my life!) affect what I end up doing than that my thought about which finger to move in a second affects which finger I move.

3. However, this doesn't mean we are out of the woods yet. There is some worrying evidence from psychology that we often do not know why we do what we do and are influenced by factors we would not want to influence us if we knew about them (advertisers certainly know this!). To the extent that is true, it seems we have diminished free will, because we are unable to act on the reasons we have accepted in prior deliberation--or even that we would accept if we did deliberate about it. I don't think the evidence goes as far as some suggest, but I think it may suggest we have less free will than we tend to think (I take free will to be a set of capacities we possess and exercise to varying degrees, rather than an all-or-nothing thing).

So, in my view, once we work out a clear conception of free will--preferably one that is amenable to empirical investigation--then the sciences of the mind have the potential to inform us about how it works (e.g., how the brain subserves it) but also to show how it is limited.

Hope this helps!

Let me turn the tables on you and ask, "What is free will?" When people use this concept, they may have any of several different ideas in mind. Some people think of free will as freedom from external factors such as bribery or threats, some think of it as freedom from acting in accordance with one's own baser urges, some think of it as lack of determination (by the laws of nature/brain processes etc.) I think different sciences are relevant to each of these questions and that we can have evidence supporting or disconfirming claims about free will.

Another way for you to think about your question: if there is no free will, what would you have lost?

I don't specifically recognise this argument as having been put forward by anyone in particular: but I'm having trouble seeing why it's supposed to be an argument against determinism. If anything, the notion that it is "necessary" to "make free choices" seems to be tending more towards compatibilism: that is, the theory that determinism and free choices are both real, and that they can comfortably coexist together.

Determinism doesn't imply that you should "just sit there expecting all the events leading up to this situation to make this decision for you". Rather, it implies that those past events will cause you to make a certain decision. The decision itself might be predetermined, but that doesn't take away the fact that you are the person who is formulating it. By contrast, if things are indeterministic, wouldn't the right attitude be to say: there's no way of predicting what I'm about to do, because my behaviour doesn't fit into the normal causal structure of events, so I'm just going to have to sit back and wait and see what spontaneously happens?

You likely have in mind something we might call "covert coercion." I might hypnotize you to vote for McCain-Palin but do it in such a way that you feel like you really want to vote for them (and if I ask you why, you'll come up with lots of reasons--they're "mavericky!" Assume you would otherwise have voted for Obama). Or maybe Professor Black gives you a sleeping pill and while you're down he does futuristic brain surgery on you to make you feel like you really want to divorce your spouse (whom Black wants to seduce) and you do it. It's very scary to think that you could be manipulated not only to leave your lover but to feel like doing so (to fall out of love), and some might argue that this is simply impossible. But we know that people change their minds (and change how they feel), so these processes just seem to be doing it faster and in a different way.

The difficulty then is figuring out how these types of manipulation, which seem like they make you unfree (and not responsible for what you do), are different from, on the one hand, slower processes that you may be aware of, such as someone offering your arguments that convince you to vote for McCain, and on the other hand, the general processes that lead you to believe what you do and feel what you feel, processes that ultimately are beyond your control. Indeed, some philosophers argue that covert coercion or manipulation is no different in principle from determinism (or, what is different, complete causation of everything that happens, including our thoughts and feelings). Hence, they think that if everything that happens is determined (or completely caused by events in the past and the laws of nature), we do not have free will.

I think these arguments are mistaken, but it is very hard to see exactly why. That is, I think that agents manipulated in these ways lack free will (to answer your question) and I think that determinism or complete causation need not rule out free will. What these cases clearly suggest to me is that free will requires that we somehow "participate in" the process of coming to think, feel, and choose as we do, that we are conscious of some of the forces influencing us (such as people offering us arguments and our considering their merits) and our being aware of what's going on has a causal influence on how we turn out (but this does not require that determinism is false or that we are somehow--impossibly--outside of the causal stream).

I hope this helps!

Try this out: We cannot blame John morally for the particular behaviours he exhibits while under the spell of the lovely Joyce; after all, he is not in control of himself. Nor can we blame John morally for being the kind of person who -- because of his 'biological makeup' -- is prey for Joyces. However, we CAN blame John for being the kind of person who falls heavily for unsuitable Joyces, IF we believe that the cultivation of the kind of person we are is in some measure in our control and the object of particular trends in our choices. John should have learned from his mistakes with the last Joyce; John should understand his own weaknesses and find ways to compensate for them, perhaps seeking counselling; John should learn to trust his friends advice; John should learn to read people; John should get out more. It doesn't seem unreasonable to ask these things of John's character, in general. Although, first, we have to get rid of Joyce.

I am not aware of a conventional way in which philosophers standardly draw this distinction. However, if a particular author distinguishes between "causing" and "bringing about", she might have in mind any one of several possible distinctions. Here are three candidates:

(i) causing versus being part of the causal background: The alarm clock's ringing causes me to awaken. That I am not deaf, that I was asleep to begin with, that there was air to conduct the sound from the alarm bell to my ear, etc., were all needed for me to awaken; without them, I would not have awakened when the alarm clock rang. So they, too, are causes -- at least, broadly speaking. But we might well want to privilege the alarm clock's ringing from among all of the other, background causes, and say that it was the alarm clock's ringing that brought about my awakening.

(ii) causing versus preventing a potential preventer. Dick, pilot of a bomber, bombs a city. His actions cause the city to be bombed. Jane, pilot of a fighter, shoots down an enemy plane that would otherwise have shot down Dick before Dick could have released his bomb. Jane, then, is also causally responsible for the bombing, broadly speaking. But we might want to distinguish Jane, who prevented a potential preventer of the bombing, from Dick, who actually dropped the bomb.

(iii) causing (intentionally) versus merely bringing about (which might be unintentional): A physician, intending to help a sick patient, gives the patient a drug. Unfortunately, unbeknownst to the physician, the patient is allergic to the drug and dies as a result of the reaction coupled with the original illness. Let's even say that no physicians at the time knew that anyone was allergic to the drug; the given physician was not negligent in prescribing the medication. We could even stipulate that there was no other available course of treatment. Under these circumstances, we might be reluctant to say that the physician caused the patient to die; that might suggest that the physician deliberately brought about the patient's death or was somehow negligent. Rather, the physician brought about the patient's death, but inadvertently and without negligence.

There are a host of other subtle distinctions in the neighborhood as well. But these are the three that most readily come to mind.

Interestingly, one of the more well-known statements of your premise -- that belief in most cases is a matter of accidents of birth and circumstance -- was offered by a well-known defender of religion, the British philosopher John Hick. But we'll get to that.

Most people don't think very hard about their religious beliefs. And when we get to the level of specifics (that Jesus was God incarnate, that the Koran was delivered to Mohammed by an Angel, that the Amida Buddha built the Western Paradise...), it's guaranteed that most people are wrong, because there are no majority beliefs at this level of detail. But what to make of this is harder to say.

After all, something like both of these points (beliefs held by custom and habit and no majority view in any case) may be true for political beliefs, and for views on certain controversial ethical matters. It's likely true even for certain sorts of scientific beliefs, and ceretainly for various broad background "philosophical" or "metaphysical" commitments. So the first point is that it may be a bit harder than it seems to single religion out. But there are a couple of other points.

Even if my commitment to liberal democray, or libertarianism, or communism or socialism or whatnot isn't well-thought-out, it doesn't follow that such commitments are rotten by nature. After all, some people hold their views thoughtfully. And this goes for religious views as much as for any other sort. But we can add that there are many ways of holding religious beliefs. There are plenty of believers who realize that they really don't know a lot about ultimate things. The specifics of their traditions give them ways of conducting their religious lives, and that could be valuable for a variety of reasons, not all of which have to do with getting the details right. Hick, by the way, thinks that religious views are partial attempts to grasp a reality that we can't fully grasp, but that many religious traditions can put people in touch with ultimate reality, even if the believer's account of the matter is confused. (Compare: I don't need to understand the active ingredients in the medicine I've been given for it to do its job.)

All this is consistent with thinking that there's far too much thoughtless religion in the world, and that fair bit of evil that can be traced to unthinking but zealous acceptance of bad dogma. For all that, however, the fact remains: it's not quite as easy as it seems to dismiss religious belief by the sort of argument you offer.

I think this is a very interesting question, one that has inspired some of my recent research. It has been said that it is just obvious that determinism rules out free will. Here is Robert Kane:

In my experience, most ordinary persons start out as natural incompatibilists. They believe there is some kind of conflict between freedom and determinism; and the idea that freedom and responsibility might be compatible with determinism looks to them at first like a ‘quagmire of evasion’ (William James) or ‘a wretched subterfuge’ (Immanuel Kant). Ordinary persons have to be talked out of this natural incompatibilism by the clever arguments of philosophers. (1999: 217)

Like you, I have been curious why philosophers have taken incompatibilism to be the commonsense view. So, the first thing I did, along with my co-authors, was to test whether non-philosophers actually take determinism to rule out free will and moral responsibility. Our studies suggested that most people (between 2/3 and 3/4) do not think they are incompatible. Here is a paper discussing the motivation for these studies, the results, and the implications: http://www2.gsu.edu/~phlean/papers/Is_Incompatibilism_Intuitive.pdf

Assuming it is not so intuitive that determinism rules out free will, the next question to ask is why has it been taken to be intuitive. I think your explanation for why determinism need not rule out free will also offers an explanation for why it seems to. My view is that many people interpret determinism to suggest a sort of reductionistic, mechanistic view that says the physical world pushes around our decisions and actions and makes our minds (our selves, our desires) irrelevant. If we start with some non-physicalist intuitions about our minds, then it's easy to think that a complete explanation at the physical level means there's nothing left for the non-physical mind to do. Hence, I think the free will question is motivated largely by the mind-body problem.

Having offered this error theory for incompatibilist intuitions, I should point out that the traditional way of motivating incompatibilism is this basic sort of argument (van Inwagen 1983):

If determinism is true, then everything that happens is a consequence of the state of the universe in the distant past and the laws of nature. But it's not up to us what happened in the distant past or the laws of nature. So nothing that happens is up to us.

There's a lot to say about this sort of argument and a lot of it has been said. But I think that whether free will is compatible with determinism depends on how one understands determinism and its consequences, as well as free will itself.

It is not entirely straightforward to come up with a cogent statement of determinism. But perhaps something along the following lines will do: our world is a deterministic one if the laws of nature are such that, given the past and current state of the world, there is only one possible way its future state can evolve. If you prefer that in "possible world" talk, then the idea is that the actual laws are such that any other possible world which shares these same laws, and whose past and present duplicates that of the actual world, will also be a future duplicate.

Thus understood, the claim that our world is a deterministic one is a claim about the shape of the laws of nature governing the world. Do they, so to speak, uniquely fix what will happen next (given the past and current state of the world); or do they allow e.g. for irreducibly chancy events?

And that is surely an empirical question. We can't settle from the armchair whether (i) we live in a "classical" world where the laws make the world run like complicated clockwork, or (ii) we live in a "chancy" world where e.g. the laws governing fundamental particle interactions only settle the chances of various outcomes, or indeed (iii) whether the whole assumption that the world is totally subject to laws (whether classical or chancy) is badly wrong. We have to go and do the science.