Can it feel good to feel bad?

We assume that, given the choice, a sane and healthy person will always prefer feeling good over feeling bad – but recent studies indicate that sometimes we may actually  seek to feel unpleasant emotions.

Do normal, healthy people ever really want to feel angry, frightened or sad?

Before you answer that question, keep in mind that you can want something either consciously or unconsciously. In other words, you may seek to feel a specific emotion whether you realize it or not.

Historically, most research on emotion has focused on how people (and other animals) modify their feelings, not why they do it. Maya Tamir bucked this trend in a 2009 study, What Do People Want to Feel and Why? Pleasure and Utility in Emotion Regulation  where she wrote:

People want to maximize immediate pleasure. Therefore, they want to feel pleasant emotions and avoid unpleasant ones. The emphasis on short-term pleasure has dominated research on emotion regulation. However, people also want to maximize utility. Therefore, they may also want to feel emotions that are useful (not merely pleasurable) and avoid harmful ones. The approach that views emotion regulation as instrumental proposes that what people want to feel depends on both pleasure and utility. When immediate benefits (i.e., immediate pleasure) outweigh long-term benefits (i.e., delayed pleasure derived from successful goal pursuit), people should prefer pleasant emotions. When long-term benefits outweigh immediate ones, people should prefer useful emotions.

If this is true, when we’re in pursuit of a short-term goal we should just unconsciously seek to do the thing that makes us feel good in the moment. But when we’re chasing a goal that represents delayed reward we should seek the emotional state that is best suited to achieving that goal, even if it is an unpleasant emotion.

Tamir decided that situations where it would be most adaptive for people to seek to feel unpleasant emotions would provide the strongest test of her theory, because in these cases the immediate reward offered by pleasure would diametrically oppose the negative emotion favored by utility. She decided to test whether it was adaptive to feel angry when pursuing confrontational goals. Participants were asked to engage in anger-inducing activities before they played a confrontational game. Tamir found that this group who did this performed measurably better at the game than a control group who did not. She also observed that engaging in the same anger-inducing activities did not enhance participants’ performance in a separate nonconfrontational game.

After she found that anger could help a person achieve a confrontational goal she conducted a second test to find out whether people would seek to feel angry when they were told that they were preparing to engage in confrontational activities. Even though most participants stated that they expected anger-inducing activities to be unpleasant, they still actively preferred to engage in them when they were told that they were going to participate  in the confrontational game. Participants did not seek to engage in the same anger-inducing activities when they were told that they’d participate in a nonconfrontational game.

Feelings of fear should enhance performance in escape and avoidance activities, so another group of participants were told they’d be playing a computer game where their goal would be to avoid various threats. These subjects preferred to engage in fear-inducing activities before playing, and according to Tamir, “the more participants expected an activity to make them afraid, the more they wanted to engage in it before playing the threatening game.”

Though others have stated that negative emotions are the only ones that drive behavior, Tamir proposed that happiness should make us feel more cooperative. She tested this by telling some participants that they’d be engaging in negotiation activities with collaborative goals. These subjects displayed a marked preference to engage in  happiness-inducing activities before negotiating.

Tamir’s results are fascinating but we don’t (and can’t) always consciously choose how we want to feel because feelings related to expectations operate largely within unconscious mental processes. So how do we know which emotion will be the most useful in helping us achieve a given goal? Tamir thinks that we learn this with experience and she implies that context is important in this type of learning.

No one knows exactly how emotion affects behavior. The most widely accepted view is that emotion directly causes behavior, but some recent studies have proposed that rather than triggering behavior directly, conscious and unconscious emotions drive behavior through positive and negative feedback loops.

In an article published in Personality and Social Psychology Review in 2007, Roy F. Baumeister, Kathleen D. Vohs, C. Nathan DeWall and Liqing Zhang state that:

Conscious emotion commands attention and stimulates analysis, learning, and adaptation, often occurring in the aftermath of behavior and its outcomes. It may occasionally have a direct effect on behavior (for good or ill), but directly driving behavior is not its main function. Automatic affective responses can preserve the lessons and information from previous emotional experiences. The combination of previous emotional outcomes and current affect also contributes to making people start anticipating emotional outcomes — and to choose their actions according to the emotions they expect will ensue.

They go on to state that if positive and negative feedback loops are important in the emotional regulation of behavior, the anticipation of an emotion may be more important than the actual emotion. When we feel the way we expected to feel about an event our past learning is reinforced by the correlation. When we don’t feel the way we expected to, the emotional contrast causes us to update our mental database. If this is true, emotional expectations play a vital part of the learning process.

Baumeister et al.’s ideas on the role of our expectation of emotions in learning ties in well with Tamir’s theory of emotional priming. If we seek an unpleasant emotional state like anger or fear with the expectation that it will help us achieve a delayed goal, we may either be rewarded by a feeling of success when we achieve the goal or punished by the feeling of frustration when we don’t. While most of this expecting, comparing, rewarding and punishing goes on underneath our conscious intellectual radar, it still plays a key role in learning.

If emotional expectancies play a foundational role in how we perceive and understand the world, experiencing unpleasant emotions is a vital – and unavoidable – part of learning. It’s not just good to feel bad, it is absolutely and completely necessary to.

With the best of intentions, people often seek to protect their children, pets and other loved ones from any experience associated with negative emotions. We should, perhaps, reassess our goals in this area. If “negative” emotions are vital in controlling how complex animals (like people and dogs) achieve our goals, these “negative” emotions aren’t just necessary and adaptive – they are also (at least in some cases), the emotions that we prefer to feel in a given situation.

This is new, cutting edge, work so it may or may not hold up under further scrutiny, but it is certainly food for thought.