We looked at psychological studies of experienced utility. The authors take it as obvious that the correct way to measure happiness is to sum the episodes of felt happiness and unhappiness over a period of time. They claim that experimental psychology has shown that people are very poor judges of happiness, measured in this way.
I stuck my chin out and said that they had not shown that one of their cases involves an error.
Some of the experiments show that people’s judgments about their past happiness follow the peak/end rule. For two similar experiences, people think the one that had a higher average of peak and end experiences was better, even if it involved a smaller sum of episodes of felt happiness.
I said that it wasn’t obvious to me that this is an error. Maybe happiness is better measured by the peak/end rule than by the measurement that the authors think is obviously correct. The experimental evidence strongly suggests that most people find this to be true.
As Zach pointed out, there is precedent for denying that happiness is simply the sum of momentary feelings. For instance, John Stuart Mill distinguished between the quality and quantity of pleasures, such that a smaller quantity of a higher pleasure could be more valuable than a larger quantity of a lower pleasure. Hence, Mill could conclude that it is “better to be Socrates dissatisfied than a fool satisfied”.** Utilitarianism, Ch. 2, par. 7. Socrates may not have many pleasures, but they are the high quality pleasures of the mind, and thus worth much more than the lower foolish ones. In other words, Mill denied that you could measure happiness by summing up the momentary episodes of happiness. That would leave out the distinction between higher and lower quality pleasures.
I was playing two roles: the provocateur, who generates discussion with an ununusal claim, and the genuine believer, who articulates what he believes is true, even if it’s boring. I say that because I’m not sure exactly what I believe. You can distinguish two claims:
I started with the Weak claim (which really does seem true to me) and occasionally defended the Strong one.
The Strong claim is subject to an objection that my two colleagues made: you could imagine two experiences with an equal average of peak and end states but very different summed moments of happiness. You could have a ten minute colonoscopy and an ten hour colonoscopy that have the same peak/end average. But isn’t it bizarre to say that the unfortunate subject of the second is just as happy as the first? There was an extra nine hours and fifty minutes of colonoscopy there!
Someone who really wanted to defend the Strong claim could dig in his heels and insist that there really is no difference in how happy these two characters would be. I myself doubt that the experimental evidence would support that conclusion.
A more modest claim might be that both the peak/end rule and the sum of momentary happiness matter in measuring happiness. This would fall between the Weak and Strong claims identified above. How much does each measure matter? I suppose you could say that their relative weight should be determined by further experiments. That’s what got us (well, me) thinking that the peak/end rule might be accurate in the first place.