Moral Judgments and Happiness part 2

In an earlier post I described an experiment that I believe provided some evidence for the following hypothesis:

That the folk concept of happiness involves some normative or evaluative element such that, of the things that determine whether we take somebody to be happy, one is whether we think that this person is living a good life or not.

This means that, if we think that somebody is living a bad life---perhaps by being a morally bad person---then we might not attribute happiness to her even if we believe that she likes and wants to continue living in the way that she does. See this link for a description of the earlier experiment:

Moral Judgments and Happiness

Now on to a new experiment that I believe gives further support for this partly normative theory of the folk concept of happiness. In this new experiment, the subjects (71 Michigan undergraduates) were all given one and the same vignette, namely the following one:

Bruce was brought up in a religious home and used to go to church every Sunday. However, since coming out as gay, Bruce has stopped going to church. He now meets new people at nightclubs, parties, and bars, instead of in church as he did before.

One of Bruce’s strongest desires is to find a man to have a long-term relationship with. He thinks that Andrew, whom he just recently met, might be what he is looking for. One of the reasons Bruce thinks this is that they are very compatible, sexually speaking.

Occasionally Bruce feels a little guilty about being in a relationship with a man. It goes against his upbringing; he was raised to think that one of the most important things in life for a man is to find a wife. But, most of the time Bruce does not feel these kinds of guilty feelings. And, when he does, Bruce reminds himself that he thinks he really has no reason to. Bruce then also reminds himself that he really likes Andrew and the people they socialize with. And so he doesn’t want to change anything about his lifestyle.

Subjects were then asked three questions (which were in different orders on different surveys): (1) Would you agree with the statement “Bruce is happy”?, (2) How much distress do you think Bruce feels? and (3) Do you think Bruce’s lifestyle is immoral? Subjects recorded their answers on a scale from 1 to 7 where 1 meant “disagree”, 4 “in between”, and 7 “agree” in the happiness question, 1 meant “none at all”, 4 “in between”, and 7 “a lot” in the distress question, and 1 meant “No”, 4 “neutral”, and 7 “Yes” in the immorality question.

Subjects who believed that Bruce experienced distress judged that Bruce was not happy. But, their distress ratings could not fully account for their low happiness ratings. And, differences in immorality ratings could explain the extra difference in happiness ratings. In other words, if subjects judged that Bruce’s lifestyle is immoral, then this had a negative effect on their happiness-judgments. The results were highly statistically significant.

The distress question was used to show that what happens is not just that some people attribute further mental states to the people in the vignettes and then base their happiness judgments on these fuller pictures of the imagined person’s mental life. As this experiment suggests, happiness judgments are also based on some sort of normative or moral evaluation of those about whose happiness level we are forming a view.

To sum up: these two experiments, the one described in my earlier post and the one described above, provide evidence, I think, for the thesis that the folk concept happiness has some normative or evaluative component. This is interesting since most research on happiness in psychology uses definitions of happiness that have no such component and that are wholly psychological or non-evaluative. This means that when a psychologist concludes that somebody is happy, perhaps the folk would not.

This raises the question of whether the kind of happiness that psychologists talk about is better, as good as, or less valuable than the happiness that the folk desire. Another interesting question is whether psychologists are using a concept of happiness that is better, as good as, or less good than the folk concept of happiness. Providing a philosophical account of happiness, I tentatively conclude, is thus a normative project: we have to take a stand as to whether we should favor some version of the folk's normative concept or the psychologist's non-normative, wholly psychological concept of happiness.

I’d be very interested to hear people’s thoughts on this new experiment and the conclusions I draw based on it.

Comments

[jonathan weinberg]

Very interesting line of research. Could you perhaps say a bit more about how you analyzed your data?

Unrelatedly: I would worry about whether the "distress" question would sufficiently cover all the relevant ways in which someone might have filled in Bruce's psychology. "Distress", as a mental state, might seem to be something rather acute & generally conscious. But perhaps some of the subjects giving high answers to your third question, thought that Bruce would have some sort of low-grade sadness or something like that, which wouldn't count as distress but which would still undercut his happiness.

Also, out of curiosity, did you look at your data for order effects? This isn't meant as a possible confound -- you say you balanced the orders -- but I'd love to know whether the different orders did indeed produce changes. And it might be relevant to your claim, if it turned out that high-moralizing subjects had a greater impact of their moral views when the morality question came first.

[Sven Nyholm]

Thank you, Jonathan, for these comments and questions.

My fellow Michigan grad student Warren Herold helped me out with the statistical analysis. He will post something about what, exactly, he did sometime later today.

I take your point that the "distress" rating question might not cover all the different ways in which people might have further filled in Bruce's psychology. My thought was that "distress" would be a broad enough mental state to cover many of the different further mental states that people might add on their own. I'd be happy to hear suggestions as to which further kinds of questions I might ask if I do further experiments that could help to take care of this worry.

As for order-effects, there were some very small, but not significant ones, unless my memory fails me. Like I said, Warren (Herold) will post something about these things later today.

Thanks again!

[Dan Haybron]

Clever strategy, Sven! I'll set aside my general qualms about the folk concept and what these sorts of intuitions tell us about it, since you've already heard them. (In short: if there is a "the" folk concept of happiness, it's rather a mess, plus I'm not sure how to individuate folk concepts.) A few things specific to this study:

First, following up on Jonathan's point, I could imagine 2 further types of state missed by "distress":
1. Unconscious states--eg, repressed distress, trouble in the boiler room, which only arises at times. Seems happy but really isn't. (In my work, this falls under the heading of "mood propensity.")

2. Life satisfaction: he might not truly be satisfied with his life and hence, by the lights of some, not happy.

I suspect that you'll get the effect even controlling for such states (which may be hard!), since I think there is an evaluative "sense" of the word that isn't clearly distinguished from the psychological "sense" (a mongrel of mongrel concepts?). I bet the effect gets stronger if you ask if Bruce has a happy *life*!

Religious students, btw, may well use 'happy' differently from others, tending more toward the 'well-being' sense of the term, which is predominant in religious thought, than others.

I think you're right that philosophical work on happiness has a normative aspect, but there's a gap btwn these kinds of results and the provocative thought that psychologists might not "really" be talking about happiness (as the folk know it): are the folk *attached* to thinking of happiness in this (semi-evaluative) way, or is it just a product of sloppy thinking, not really distinguishing various things they care about? If the latter, then psychologists might ascribe H where the folk would tend not to, yet not meet with any objection from the folk, who might on reflection conclude they were confused.

I've been meaning to post my own results, which are coming out in my book but need to be followed up more formally, so I'll try to do it soon. One of the instruments may be of use to you in sorting out possible diffs in concept btwn religious and others....

[Dustin]

In my Christian upbringing, I was raised to believe that people who are "detached from God" are never really happy, regardless of how happy they think they are. Of course, I was also taught to believe that being detached from God was immoral. But the reason for the unhappiness is simply the detachment, independent of the fact that the detachment is immmoral. So let me suggest this as alternative hypothesis: those of your subjects that judge Bruce's behavior to be immoral probably had a similar Christian upbringing to my own, and thus, probably judge that Bruce is unhappy, not because he is immoral, but because he is detached from God.

To rule-out this hypothesis, you might want to try including in the prompt that Bruce currently has a "very strong relationship with God" or something like that. Of course, that probably wouldn't work, since many of your students will either reject the case as impossible (how can one be gay *and* have a strong relationship with God?!) or else understand you as saying that he *thinks* he has a strong relationship with God.

[Meryn Stol]

"if we think that somebody is living a bad life---perhaps by being a morally bad person---then we might not attribute happiness to her even if we believe that she likes and wants to continue living in the way that she does."

I think that "she likes and wants to continue living in the way she always does" can be true on one level of consciousness, and false on another. It might have to do with mild dissociation.

I'm not a psychologist, but I think that people deep, deep down inside truly want to behave moral (what is moral obviously depends on situation), so even when people say they like behaving immoral, some part of them (a part they dissociate from) doesn't like it. This may not turn up in surveys at all, because people have different parts of their brain active depending on situation, i.e. they may answer different to a psychologist than to their pastor.

That's my lay-man's theory. Can you confirm anything of it?

My point is that if more people share my theory, then they'll think someone can't be truly (i.e. without partial dissociation) happy like that.

In this particular case, I guess that some people think that the person in question for one part beliefs that homosexuality is ok, but deep down, another part of him thinks it not. He might not even be conscious of that thought because it conflicts with the logical arguments he based his justification on. It remains only a gut-feeling.

[Meryn Stol]

The comment from Dustin crossed mine. "detachment from God" could have a parallel with mental dissociation.

[Meryn Stol]

I think the most important assumption people make is that their judgement of what's moral is right, in absolute terms, and that they know and feel that deep down inside. Based on their belief in absolute morality, or maybe of relative morality but with the knowledge that Bruce lives in a circumstances comparable to them, they conclude Bruce must feels the same, deep down inside, even if he is not aware of it.

[Sven Nyholm]

Thanks for all these helpful comments!

I think that I could get similar results if I ran a study that wasn't about a person that some might judge to be immoral, but instead about somebody who was doing things people feel are completely meaningless/not worthwhile ways of spending one's time. Suppose the vignette was about somebody who spends all her time playing computer games and eating potato chips. Suppose further that it was stated in the vignette that this person gets pleasure from this and prefers doing these things to developing her talents. I have a hunch that people would be disinclined to think that she is happy. But, if somebody was described as working to develop her talents, perhaps for violin playing, and the vignette said that this gives her pleasure and she prefers doing this to playing video games, then people would be inclined to agree that the person is happy, I think. And, I predict that they would not give very high distress ratings. So, there are ways of testing the normative hypothesis that don't appeal to people moral or religious beliefs, I think.

The evaluative component of our concept of happiness is not wholly moral, I think, but normative in some more general sense such that, if we think somebody is living a bad or meaningless life, then we are likely to think that she is not happy even if we think she is in many of the mental states usually associated with happiness. This intuition would be especially strong in people if we asked them if the person is "truly happy", I think.

[Sven Nyholm]

Dan, thanks for the feedback. I'm inclined to agree with most of what you say. Also, it'd be great if you posted some of your results on this blog!

[Sven Nyholm]

Some of these responses seem to suggest that happiness has to do with how one feels deep down. And, it has to do with how one feels so deep down that one might not even know that one has these feelings. This sort of unconscious unhappiness is a fascinating idea. It, too, would be a different idea than the kind of happiness that much happiness research in psychology is about. Much of it takes, if I am not mistaken, happiness to be a matter of how one's mood is from moment to moment, or whether one is satisfied with one's life and how it is going, both of which are more conscious things.

[Warren Herold]

I helped Sven out with the statistical analysis, so I'm probably in the best position to explain what we did.

The analysis was very straightforward. I ran a simple multivariate linear regression of 'happiness' (H) on the two independent variables, 'immorality' (I) and 'distress' (D), and then estimated the coefficients using ordinary least squares (OLS). In other words, I estimated the coefficients (B0, B1, and B2) for the following equation:
H = B0 + B1*I + B2*D + u

One would expect B2 to be negative (if people think Bruce is more distressed, they would likely think he is less happy). If happiness is partly evaluative, then one would expect B1 to be negative as well (more immoral, less happy).

This is exactly what we found. We obtained the following coefficient estimates (standard error in parentheses):
B0: 7.03 (0.44)
B1: -0.22 (0.06)
B2: -0.43 (0.10)
All coefficient estimates are significant at the 1% level. The interesting variable is B1: it suggests that higher immorality evaluations lead to lower happiness evaluations, holding distress evaluations constant.

I also ran various regressions that include a dummy variable equal to 1 for surveys in which the immorality question appeared before the happiness question. In all cases, B1 and B2 remained negative and statistically significant (at either the 1% or 10% level, depending on the regression). The dummy variable was negative and statistically significant in each case, indicating that asking the immorality question before the happiness question tended to reduce the subjects' assessments of Bruce's happiness. I can provide more details on these regressions if anyone is interested.

These are very rough, first attempts to analyze the data. If anyone has any suggestions of how to improve the analysis, we're all ears.

[Meryn Stol]

I'm not deep into positive psychology, but this might be a good entrance for further research:

"Seligman, who is the figurehead of the positive psychology movement, goes further than suggesting people learn to think positively. He has worked out what he sees as a blueprint for happiness that people can use to set them on the path to a fulfilling and satisfying life. He believes there are three routes to happiness, which he calls the "pleasant life", the "good life" and the "meaningful life". Some are better than others, although a mix of all three is ideal."

http://www.guardian.co.uk/society/2003/nov/19/1

I hope to read more on this in the future. :)

[Sven Nyholm]

Thanks for the quote and the link, Meryn.

[Dan Haybron]

Sven--I agree that the empirical research focuses on conscious affects, especially "feelings" (vs, eg, background stress or anxiety). It would be v interesting to show that the folk take unconscious states into account, even in unreflective practice. (I think they do, but showing that is not easy!)

Re. LS, I think your example may leave open that bruce could report being dissatisfied with his life using std measures. I bet you could tweak it to make clear that he would sincerely report being satisfied--yet the folk might still deem him not "really" satisfied given the guilt etc. BTW, this could be evidence that LS measures don't fully track the folk notion of LS.

[Sven Nyholm]

Thanks, Dan.

I share your suspicion that the folk might not fully believe what they judge to be immoral people's reports that they are satisfied with their lives. This might be another example where people think that "deep down" these supposedly immoral people are not really satisfied with their lives even if they themselves think so on a conscious level.

[Alberto Masala]

It seems to me that evaluative judgements are routinely used for psychological explanation. This seems to be a very general phenomenon: anecdoctically, I have often realized that people can't help predicting other people's behaviour assuming that their own values correspond to universal aspects of human functioning. I am referring to sheer prediction, not judgement. People who value monogamy cannot help thinking that promiscuous people must feel hollow deep inside, one day will understand what real commitment is etc. Take for example the debate on the polyamory community: most monogamous just can't believe that polyamorous people are happy like that. Idem for religious people with the need for spirituality. There must certainly be some good empirical work on that, I think that a lack of moral imagination could be the key. The upshot could be that the concept of happiness is indeed psychological, but values inform spontaneous theories of human functioning.

[Sven Nyholm]

Thanks, Alberto, for these comments.

I think your hypothesis is a plausible one. I'd be interested in suggestions as to how it might be tested or in references to empirical work that support it.

[Staffan Humlebo]

Most presentations of utilitarianism I've heard are psychological in nature. The state of pleasure or satisfied desire seems to be the happy thing.

I never liked that. People I trust seem to be happy in different ways. Willing and wanting to do the good e.g. seem to be much more relevant than pleasure or some other state, preferred or not.

That's why I don't like buddhism's emphasis on enlightened states or the definition of unhappiness as a psychological or physical state of suffering.

Action rather than psychological state would be a more fruitful path of deliberation on the concept of happiness.

[John A.]

This may have already crossed your mind, but maybe it would be beneficial to run a parallel study with the scernerio not being grounded in having been brought up in a strong religious environment. This way you could test to see if the effect suggested by Dustin and echoed by Maryn and others is present. I think these finding could be very interesting.

[John A.]

This may have already crossed your mind, but maybe it would be beneficial to run a parallel study with the scernerio not being grounded in having been brought up in a strong religious environment. This way you could test to see if the effect suggested by Dustin and echoed by Maryn and others is present. I think these finding could be very interesting.

[Sven Nyholm]

Thanks for these comments, Staffan and John.

John: I did run another experiment that had no mention of somebody's having a religious background. See a description of it here:

http://experimentalphilosophy.typepad.com/experimental_philosophy/2007/07/moral-judgments.html

[John A.]

Sven
I was thinking of a scenerio with Bruce that did not refer to his upbringing. For example:

"One of Bruce’s strongest desires is to find a man to have a long-term relationship with. He thinks that Andrew, whom he just recently met, might be what he is looking for. One of the reasons Bruce thinks this is that they are very compatible, sexually speaking.

Occasionally Bruce feels a little guilty about being in a relationship with a man. But, most of the time Bruce does not feel these kinds of guilty feelings. And, when he does, Bruce reminds himself that he thinks he really has no reason to. Bruce then also reminds himself that he really likes Andrew and the people they socialize with. And so he doesn’t want to change anything about his lifestyle."

This is probably not worded adequately, but I think you get the idea. This way you would be comparing "apples' as the concern was regarding the effect a person's religious attitutes has on how he (she) answers whether Bruce is happy.

[Sven Nyholm]

Thanks for this suggestion, John. I may run some new experiment that does not mention anything about religion but that in other respects are like this most recent experiment of mine.

[Sven Nyholm]

Thanks for this suggestion, John. I may run some new experiment that does not mention anything about religion but that in other respects are like this most recent experiment of mine.