Few sentences have irritated me more in recent weeks than this, from the jacket of The Procrastination Equation by the psychologist Piers Steel: "If you think you procrastinate because you're a perfectionist, you're wrong." Singling out Steel is harsh, I admit: he probably didn't write that line, and the book itself is great. Even the argument behind that sentence is persuasive: perfectionists are more likely to seek help for procrastination, he argues, because it offends their perfectionism, so researchers mistakenly imagine there's a causal link. But the jacket's tone encapsulates a position that's increasingly prevalent today – in David Cameron's "happiness index", in much coverage surrounding the launch of the (broadly excellent) Action for Happiness movement, and in 100 books on "the surprising science of" this or that. The position is this: that if scientific studies have reached a conclusion on some aspect of psychology, the argument's over. (And, yes, I know I rely on such studies in this column. "Hypocrisy is the homage vice pays to virtue" – La Rouchefoucauld.)

For all the near-unchallenged popularity of the science of happiness, though, there are some awkward problems. One has to do with unrepresentative samples: as mentioned here before, too many experiments rely on citizens of white, educated, industrialised, rich and democratic (Weird) countries. Then there's the intrinsically provisional nature of science: any conclusion reached might be overturned next week. But there's a more fundamental philosophical problem, too: even the best scientific studies can't fully penetrate the experience of being you.

This is because, as Daniel Dennett puts it, science is by definition "third person": its whole point is dealing with things others can observe, and in procedures they can try to replicate. The place where happiness and sadness happen – your conscious experience of being yourself – is intrinsically first person. You can never perfectly communicate, even to your closest friend, your experience of subjectivity. That doesn't mean scientists can't study consciousness – they do so all the time – but it does mean that to talk about things like happiness, science must translate what you mean when you say, "I feel happy" into something more objective: your responses to a questionnaire, say. (According to one head-spinning philosophical position, the whole external world is ultimately ungraspable, not just other people's minds. After all, everything we know about it comes through our senses – so how could we ever be certain our senses weren't fooling us?)

None of this is to suggest happiness science is worthless: going with the scientific evidence is usually advisable, especially if you're spending public money on policies to promote wellbeing. Nor should it provide solace for pseudoscientists, who peddle what looks like scientific evidence but isn't. Rather, it highlights the possibility of strategies for happiness that are neither science nor pseudoscience – a case made recently for psychoanalysis by the writer Robert Rowland Smith, who suggests that its focus on the unique properties of the client-analyst relationship might place it beyond meaningful experiment. And it shows that while the science of happiness may come close to understanding your mind, we shouldn't forget that it can never quite get all the way. In the final analysis, when it comes to your inner experience, no study (or dust-jacket) can conclusively tell you you're wrong.

oliver.burkeman@theguardian.com
twitter.com/oliverburkeman