Reasoning is like driving. We all think we're above average, and woe betide anyone who suggests otherwise. We all want to be reasonable, or at least to appear reasonable (often less work and as effective as actually being reasonable). No one likes to admit their reasons are poor and their arguments more fudge than forensic. Our only hope is that, just as few of us would claim to be infallible drivers, few of us would claim to be cognitively perfect (and those that do can be discouraged from flying planes and becoming surgeons). For most of us, there's room for improvement, and Nigel Warburton is the perfect mind mechanic and this book the perfect pocket-sized manual.

A glance at the first few entries - peppered with Latin phrases like "ad hoc" and "ad hominen" and odd-sounding activities like "affirming the antecedent" - might make you wonder whether this is all going to be a little too technical, pretentious even, hardly useful outside of the classroom. As for more familiar words like "argument", "assertion" and "assumption", surely we already know what they mean? (Actually, for many of them, we don't, at least not precisely.) Warburton's style is clear and straightforward, always beginning with a crisp definition and following up with a short discussion and simple examples from everyday life. In fact, there are very few arcane entries - the vast majority are ordinary phrases that crop up all over the place, often without the slightest regard to their philosophical meaning.

For example, consider the common or garden "argument". Not an altercation in a pub car park but "Reasons supporting a conclusion". (Just as the word itself has two very different meanings, arguments themselves are often marred by "equivocation" - where the meaning of a word is surreptitiously switched midway through.) Terms that have their own entries elsewhere are in bold, and so from this one simple word can be assembled a more complex set of concepts, including "deduction", "induction", "validity", "sound argument", and so on. (These last two terms are fascinating, and good examples of the importance of precision: a valid argument with true premises guarantees a true conclusion. A valid argument may also contain one or more false premises, in which case the conclusion is not necessarily true, although false premises may sometimes lead to a true conclusion. Only a sound argument guarantees a true conclusion. "Validity should not be confused with truth. Validity is always a quality of the structure of arguments; statements are true or false. Arguments can never be true or false, statements never valid or invalid...")

One of my favourite entries is "humptydumptying" - not an illegal pastime, but giving "private meanings to words in common use". Some people think winning an argument involves confusing their opponent into silence, and humptydumptying along with ambiguity and equivocation are very often their weapons of choice. Clarity, of course, is no guarantee that sense is being made. There may be no argument at all, just a string of assertions (unsupported statements of belief). As Warburton drily comments, "Merely asserting something, no matter how loudly, doesn't make it true."

The "no hypotheticals move" is a clumsy phrase describing a very common tactic, used all the time by politicians who don't want to answer a question. If elected, will you raise the rate of VAT? We're not in government yet, so let's wait and see what the blah blah blah. Such evasiveness is given short shrift by the likes of Paxman but the more cognitively vulnerable will be lulled by its superficial plausibility.

In light of the message of this marvellous little book - Be reasonable! - one surprising absence is "reason" itself. Warburton does use the word, but not as often as you might think in a book that could have been called, less catchily and more pompously, "Reasoning from A to Z". In an important sense, however, the whole book is an account of reasoning and all its ramifications, with the crucial fact always in mind that human reason is fallible. We are often motivated to connect the ideas we want to be true in any way we can. Creationists expend huge resources and energy in putting forward reasons and arguments for their beliefs (think of the millions of dollars behind the Answers in Genesis Creation Museum and Harun Yahya's "Atlas of Creation"), it's just that their reasons and arguments are not very good. And there's the rub. What is a good reason? What makes a good argument? Who says? Fortunately, logic and experience provide the basis for an objective standard of truth, and we don't have to rely on any single person. Having said that, Nigel Warburton and this book are a reliable starting point. While most of us get by on our intuitive reasoning, with little idea how bad our arguments sometimes are, he's rather more rigorous and professional. Whether you think human reason is a gift from the gods (unreasonable) or a result of a billion years of evolution (reasonable), it's still something rather special, and well worth tuning up.