Rare is the philosopher who doesn’t give happiness a significant place in the good life. Even Kant, often caricatured as making no room for happiness, still says both that it is a duty to secure one’s own happiness in this world, and that one needs to hope for happiness in the afterlife. Happiness, then, is a topic of key philosophical importance, whether by “happiness” we mean the pleasant mental state aimed at by Bentham or the broader conception of human flourishing in Aristotle’s eudaimonia; and most accounts of the latter include some element of the former.

We would do well, then, to pay attention to the burgeoning field of psychologists’ empirical research on happiness. The field faces a number of methodological problems, but comes to interesting insights in spite of these. One deservedly popular book in the field is Daniel Gilbert’s Stumbling on Happiness, a well written and engaging summary of current research. Gilbert does a good job of summarizing many psychologists’ counterintuitive findings about happiness.

The problem is that some of Gilbert’s conclusions contradict not only common sense – which isn’t a problem, because contradicting common sense is the point – but each other. He concludes at the end that we are not as different from other people as we think we are, and that therefore in order to be happy we should ask other people what makes them happy. Yet elsewhere in the book he acknowledges that people don’t themselves know what makes them happy. The most obvious example is children: ask anyone who has children and they will tell you their children are their key source of joy, yet every study on the subject concludes we get less happy when children are born, and happier again when they leave. Which is to say that according to Gilbert’s own data, other people’s self-report is not the best place to find out what will make you happy.