Tuesday, October 23, 2007

Pride and Preferences – Or How We Live With Our Decisions

No one likes to admit to having made a mistake. Just look at all the politicians and business people who, with a mess on their hands and owing the public or shareholders an explanation, have uttered the famously weak cop out “Mistakes were made” – a rhetorical device that political consultant William Schneider has suggested we call the "past exonerative” tense. While acknowledging an error, the passive voice of the past exonerative distances the speaker from any causal role in their execution.

Realising — and, worse still, publicly admitting — that we have made an error of judgement, a bad call, or acted in a way we are less than proud of is frequently a painful experience. Whether it’s our choice of job, which political party we voted in, the stereo we bought, or how we responded to the homeless guy panhandling at the ATM, most us like to think that we’re intelligent, competent decision-makers and, in general, morally worthy people. When we’re confronted with evidence to the contrary, we feel a mental strain and discomfort that psychologists call cognitive dissonance – two dissonant cognitions, such as “I’m a smart consumer” and “I’ve paid my hard-earned cash for this crappy stereo”, are in conflict, and something has to give. Typically, the preferred cognition is preserved and the other discarded (1).

Since the notion was first put forward some 50 years ago, psychologists have made cognitive dissonance one of the most-studied mental phenomena around. And one thing is abundantly clear from this research: humans are equipped with a variety of dissonance-reducing mechanisms that enable us to live with our decisions, our actions and, ultimately, ourselves.

The study of cognitive dissonance has thrown up some paradoxical results. For example, people tend to prefer an outcome if they endure more hardship, pain or suffering to achieve that end. In one study, participants were more satisfied with a fraternity they joined the harsher the initiation into the fraternity, all else being equal. From the behaviourist perspective dominant when the idea of cognitive dissonance was first mooted, this makes no sense: why should an outcome associated with pain or suffering be deemed more rewarding than one reached through a less unpleasant route?

Seen through the lens of dissonance reduction, however, it makes more sense. As an intelligent, sensible person, we wouldn’t go through a painful or humiliating experience if it wasn’t worth the outcome – in this case joining the fraternity. The dissonance produced by the two cognitions “I am not an idiot who would suffer pointlessly” and “I underwent a severe hazing to join this fraternity” is resolved by declaring the fraternity to be worth joining – and the harsher the hazing, the better the decision (2).

One of the ugliest sides to cognitive dissonance comes to light in the self-serving rhetoric we use to justify prejudices. It is depressingly common that persecuted individuals and groups are dehumanised and made to appear as animals — by being kept cramped and naked and filthy in concentration camps, for instance. The victimisers then respond with disgust at the debased and depraved creatures they have created: “Look at these revolting people! How justified I am in treating them as animals!”.

Yet for all the importance of cognitive dissonance, the precise mechanisms by which we deal with discordant thoughts and feelings, and the ultimate purpose these mechanisms serve, are not well understood. One way to approach these issues to look at the origins and evolutionary roots if dissonance reduction in human children and nonhuman primates. And in a recent study published in Psychological Science, Louisa Egan, Laurie Santos and Paul Bloom have taken just such a comparative, developmental tack to the problem of cognitive dissonance (3).

Egan and colleagues, based at Yale University in Connecticut, devised two tests, one for children, the other for capuchin monkeys, each designed to reveal the reduction of cognitive dissonance in action. The specific kind of cognitive dissonance the authors explored in this study arises when an individual, usually an adult in most studies to date, is forced to choose one item from a set of equally desirable items. Prior to choosing, you have no strong preference for any particular item. Yet being made to choose an item – being forced to create a preference – is discordant with your feelings about the decision you faced. And so after the event, this dissonance is reduced by updating your preferences to reflect the decision you actually made. In the future, the preference generated by the forced choice will stick if the dissonance-reduction machinery has done its job effectively (that is, your new preference is for the selected item, explaining satisfactorily to yourself why you chose it).

In the children’s test, each child had to rate the desirability of animal stickers, which kids seem quite keen on, using a scale of increasingly smiley faces (a few kids were eliminated because they had difficulty with the rating system). The researchers then selected sets of three stickers that a given child had rated as equally desirable, and randomly labelled them as A. B or C. A and B were then presented to the child, who was asked to pick one to take home. Then, the unchosen option was offered up against C: so if A was initially picked out of A and B, then B would subsequently be offered alongside C, and vice versa.

The idea behind this test is as follows. The three stickers in each triad tested were all rated as equally desirable by the child, so there was no preference for one over another. Then the child is made to choose between two stickers of previously equal desirability (A and B). This sets up dissonance between ordinarily selecting things with the greatest utility (the most preferred) and, in this case, making a choice without a preference - dissonance between “Ordinarily picking according to preferences” and “Picking without preferences in this case”. The tension is resolved by unconsciously updating the preference to match the choice actually made, which enhances the perceived value of the chosen sticker and derogates the value of the deselected sticker. So if A is picked first, B is, after the fact, deemed to be a worse choice, thus explaining and justifying the decision just made: “I picked sticker A because sticker B is rubbish”. No dissonance there. Then when B is offered against C (stickers that were previously seen as equals), C will seem relatively more attractive. So when Egan and colleagues saw this pattern of choice, they took this as evidence of cognitive dissonance, and its resolution, in operation.

A similar test was also devised for capuchin monkeys, using M&M sweets of different colours instead of stickers. In addition, because monkeys cannot follow instructions the way a human child can, a different way of measuring preferences had to be used: how quickly they retrieved an M&M of a given colour from a testing chamber. Although the details are more complicated the logic is the same, and after the preferences for 20 different colours of M&M had been established, the capuchins were similarly presented with triads of M&M colours.

Both human children and capuchins showed sign of cognitive dissonance and resolution, as revealed by the pattern of preferences for stickers and M&Ms, respectively: in both cases, there was a greater-than-chance preference for C over the unselected option from the A-or-B choice. This clearly suggests that the basic machinery underlying cognitive dissonance, and the tools for making it disappear, are evolutionary old, and emerge relatively early in development (at the least, they don’t require extensive experience in weighing up preferences and evaluating decisions).

But a perhaps more interesting are the questions these findings raise about the function of reducing cognitive dissonance - just what does it achieve? An early suggestion was that it was simply the response to two competing cognitions, which might lead to mental paralysis if not resolved. Later researchers proposed that mechanisms for reducing cognitive dissonance exist to preserve our self-image as intelligent, competent, morally upstanding people.

Capuchin monkeys do not have a capacity for language, and human children are generally assumed to be cognitively much less sophisticated than adult humans. Yet as Egan et al. point out, their results suggest that either cognitive-dissonance reduction is mechanistically simpler than recent work on the subject has suggested and is not neccesarily related to preserving a complex self-image. The alternative would be to ascribe this sort of self-conception to both capuchins and children.

I think, however, that there’s a third way between the options of whether cognitive-dissonance reduction is mechanistically simple or the protector of a complex self-image - it can be both. One of evolution’s favourite tricks is to take some trait that evolved for one purpose and sculpt it to new ends. This process of co-option or exaptation occurred in the evolution of feathers (initially evolved for thermoregulation, than later exploited for flight); exaptation also enabled a sense of distaste, which is widespread in animals, to evolve in humans into ‘core disgust’ (elicited by bodily products, rotting meat and so on), which seems to have been built on through subsequent biological and cultural evolution into the complex cognitive state of moral disgust (the feeling you get when you think about Hitler, or a child molester, for example (4)).

Something similar might have happened with cognitive dissonance and strategies for its reduction. Rather than choosing between accepting dissonance-reduction as a simple process or accepting a complex inner life of monkeys, and perhaps we should conclude that monkeys retain a 'simple' dissonance-reductionmechanism that evolved in the primate line. Humans, by contrast, built on this simple mechanism and linked it to other processes, including those regulating our sense of self. And just as children take time to develop a full-blown disgust reaction (which is a cognitively complex reaction), and even longer to feel moral disgust, so too might children, like monkeys, initially make use of simple dissonance -reduction strategies, only later constructing the complex forms of dissonance reduction and ego preservation that we see in adults.

1. See Mistakes Were Made (But Not By Me) by Elliot Aronson and Carol Tavris for an excellent overview of the literature on cognitive dissonance and the strategies and situations that call for the deployment of powerful dissonance-reducing strategies.

2. Cognitive dissonance and its subsequent resolution are not the only, or perhaps even the preferred, explanation for this behaviour. A similar phenomenon has been observed in pigeons: in one study, food that took more effort to obtain from a feeder was preferred over food associated with less effort. This result was explained by ‘relative hedonic contrast effects’ – that is, the difference between the feeling experienced trying to get the food and that of actually devouring the food. When feeders exert more effort to get a given food morsel, they experience a greater shift in their relative hedonic (pleasurable) status, so the same food seems better after a worse experience. The same good explain some or much of the findings on effort-justification in humans.

3. Egan, L. C., Santos, L. R. & Bloom, P. The origins of cognitive dissonance – evidence from children and monkeys. Psychological Science 18(11), 978–983 (2007).

4. Jones, D. The depths of disgust. Nature 447, 768–771 (2007).


Anonymous Anonymous said...

This comment has been removed by a blog administrator.

4:39 pm GMT  
Anonymous Anonymous said...

"The test of a first rate intelligence is the ability to hold two opposed ideas in the mind at the same time, and still retain the ability to function."
F. Scott Fitzgerald

1:36 am GMT  
Blogger Michael said...

Hi Dan,

This is an amazing entry and I have added your blog to my Google Reader. The implications of the dissonance research for business professionals, negotiators, and even the personal relationships we choose would seem to be dramatic (if we can extend the research that far). Nice job and keep up the great work. I look forward to more posts.

I am not certain the above quote belongs to F. Scott. That exact idea just appeared in Harvard Business Review a couple months ago. I will take a look when I get back to the office.

2:14 am GMT  

Post a Comment

<< Home