Abstract
People vary in the extent to which they imbue attitudes with moral conviction, and this variation is consequential. Yet we know relatively little about what makes people’s feelings about a given attitude object transform from a relatively nonmoral preference to a moral conviction. In this article, we review evidence from two experiments and a field study that sheds some light on the processes that lead to attitude moralization. This research explored the roles of incidental and integral affect, cognitive factors such as recognition of harm, and whether attitude-moralization processes can occur outside conscious awareness or require some level of conscious deliberation. The findings present some challenges to contemporary theories that emphasize the roles of intuition and harm and indicate that more research designed to better understand moralization processes is needed.
There was a time when abortion was not as controversial as it is today, even in the United States. There were no protests, and abortion services and drugs were marketed openly. Moreover, the earliest motivations to legally control abortion were not rooted in morality, religion, or politics but were instead based on concerns from the increasingly professionalized (and male) medical community’s desire to stem competition from other health care providers (midwives and homeopaths; Reagan, 1998). In short, one of the most polarizing and moralized topics in American politics—abortion—was not always seen as a moral issue.
There is also variation at the national and individual levels in the extent to which contemporary citizens see abortion in a moral light. The Chinese public, for example, sees abortion as benignly acceptable as other forms of birth control (Osnos, 2012). There is even considerable variation among the American public about whether positions on abortion are rooted in personal moral conviction (Ryan, 2014; Skitka, Bauman, & Sargis, 2005). Some people’s abortion attitudes are rooted in strong preferences—they simply prefer to have a backstop protection against unwanted pregnancy. Others’ positions are rooted more in the stances taken by their faith communities, without any personal moral investment in the issue.
This backdrop of historical, national, and individual differences in the moralization of abortion attitudes raises an important question: What leads to attitude moralization—in other words, what processes lead people to perceive some issues (but not others) as ones that reflect their core moral beliefs and convictions (i.e., to moralize those issues)? Despite the call from Paul Rozin (1999) nearly two decades ago for researchers to dig into the moralization process, very little research has addressed this question. This article reviews recent research (two experiments and a field study) designed to understand moralization and to spur more work on this topic. Before reviewing this research, we will briefly review theoretical perspectives that make different predictions about the likely psychological routes to attitude moralization.
Theoretical Perspectives on Attitude Moralization
Contemporary moral theory provides suggestions about how attitudes may become moralized. One key possibility is the strong association between morality and processes rooted in emotion. Haidt’s (2001) social intuitionist model of moral judgment, for example, posits that intuition (a fast, visceral form of cognition) is the linchpin of recognizing the moral significance of a thought, feeling, or behavior. Most tests of this hypothesis have used emotional cues to arouse moral intuitions (e.g., Wheatley & Haidt, 2005). A defining feature of an intuition is that people do not need to be consciously aware of its source (Haidt, 2001). People might be aware of the outcome of a moral judgment or recognition but not of the process itself. Even intuition aroused by an emotional source completely unrelated to a judgment (e.g., hypnotically induced disgust; Wheatley & Haidt, 2005) should be enough to produce moralization.
Other theorists, however, place more emphasis on conscious awareness of associations between emotional cues and moralization. Discrete emotions and associated cognitive appraisals (e.g., recognition of harm) are thought to increase the salience of moral concern, which will, in turn, affect judgments related only to the source of that concern (e.g., Horberg, Oveis, & Keltner, 2011). This theoretical perspective implies that intuition may be necessary, but not sufficient, to lead to attitude moralization: Recognition of specific attitude-relevant harm may be necessary as well. This position is consistent with agent-patient theories of morality (e.g., Gray, Young, & Waytz, 2012) as well as theories that argue that morality is connected to a desire to avoid moral harms and to approach moral goods (Janoff-Bulman & Carnes, 2013).
Although some research has supported predictions of the social intuitionist model by demonstrating that incidental emotions (and more specifically, incidentally aroused disgust) lead to harsher blame and wrongness judgments (e.g., Schnall, Haidt, Clore, & Jordan, 2008), a recent meta-analysis of these moral-amplification effects suggests that if such effects exist, their size is very small and not robust to corrections for publication bias (Landy & Goodwin, 2015; cf. Schnall, Haidt, Clore, & Jordan, 2015). Regardless, harsher moral judgments may not be the same thing as increased attitude moralization; judging a behavior as wrong is not the same as seeing an attitude as reflecting a personal moral conviction (Bauman & Skitka, 2009).
Other research yields findings that are more consistent with the idea that emotions are integral to the attitude itself or that recognition of harm can be moralized. People who are more disgusted by smoking tobacco and who see the practice as more harmful, for example, are more likely to see smoking as a moral issue (Rozin & Singh, 1999). Importantly, recognition of harm is generally based on more deliberative rather than emotional forms of reasoning (Lazarus & Folkman, 1984). If recognition of harm is the key to attitude moralization (in addition to or instead of emotions), intuitions alone are unlikely to lead to attitude moralization; more deliberative forms of cognition will be needed as well. So is attitude moralization based in intuition and emotion? Is it based on recognitions of harm? Our recent research provides initial answers to these questions.
Examining Attitude Moralization in the Lab
We first tested competing attitude-moralization hypotheses by manipulating harm and emotions in the lab. We recruited participants for a short, computerized recognition task that involved displaying six pictures and six words one at a time in random order (Wisneski & Skitka, 2017). Participants were told to identify whether a picture or a word appeared on the screen as quickly and accurately as they could. Participants were exposed to the images and words at speeds that either did or did not allow for conscious awareness (14 ms vs. 250 ms). We exposed participants to disgust-inducing pictures of harm relevant to the issue of abortion (photos of aborted fetuses), disgust-inducing and similarly bloody depictions of harm irrelevant to abortion (depictions of animal abuse), disgust-inducing pictures irrelevant to abortion depicting nonharmful stimuli (e.g., toilets overflowing with feces), or control images (pictures of everyday objects, such as tables and chairs). 1 This allowed us to compare emotionally relevant stimuli with harmful and emotionally irrelevant stimuli as well as nonharmful but emotionally irrelevant stimuli.
After the recognition task, participants completed a second, “unrelated study” that measured their position on abortion and the degree of moral conviction they had about this issue. Participants’ moral conviction about abortion increased only if they were exposed to the abortion-relevant images at speeds that allowed them to be consciously aware of what they were seeing; attitude importance was unaffected. These findings were replicated in a second study that found that increased moralization in the aborted-fetus condition was mediated by disgust and not by harm or anger. This research suggests that attitude moralization is neither intuitive nor rooted primarily in perceptions of harm.
Attitude Moralization in the Wild
Although studies run in the lab allow for carefully controlled tests of hypotheses, they can also be artificial. To further investigate attitude moralization, we tested the role of beliefs and emotions as predictors of attitude moralization in a real-world context: political candidates over the course of the 2012 U.S. presidential election (Brandt, Wisneski, & Skitka, 2015). A longitudinal design allowed us to test whether changes in moral conviction associated with preferred and nonpreferred major-party candidates were predicted by participants’ beliefs about the harmful or beneficial consequences of either candidate becoming president, their emotional reactions to the candidates, or some combination of both.
We surveyed a large U.S. sample of participants in early September 2012 (shortly after the Republican and Democratic national conventions) and again the week prior to election day. Participants rated a range of emotions they felt when thinking about the two major-party candidates, including how much the candidates made them feel hostile (e.g., angry, disgusted), fearful (e.g., afraid, frightened), or enthusiastic (e.g., enthusiastic, excited; Watson & Clark, 1999). Participants wrote down the consequences of each candidate winning the election and rated how harmful or beneficial they thought those consequences would be.
Results from an analysis of latent difference scores (McArdle, 2009) found that only participants’ emotional reactions to the candidates predicted changes in moral conviction. Enthusiasm for people’s preferred candidate predicted increased moral conviction about that candidate. Conversely, hostility toward people’s nonpreferred candidate predicted increased moral conviction about that candidate. Changes in moral conviction were unaffected by perceived harms, perceived benefits, or fear associated with preferred and nonpreferred candidates. Moreover, emotions tied only to a specific candidate (not emotions associated with the other candidate) predicted attitude moralization, a finding at odds with the prediction of the social intuitionist model that intuition triggered by less integral emotions can moralize. Just as our laboratory study found that integral emotion is needed to moralize people’s abortion attitudes, integral emotions—and not perceptions of harm—are also needed to moralize attitudes about political candidates.
Discussion
Our laboratory and field investigations of attitude moralization converged on a common conclusion using very different methods. Attitude moralization results from integral emotion. Neither intuition nor harm were sufficient to produce moralization in our studies.
Attitude moralization may not be intuitive
If attitude moralization is an intuitive process, then people should (a) not need to recognize or be aware of the source of an emotional association with a given attitude object or (b) not require cognitive awareness for emotion to have an effect. In other words, incidental emotions aroused outside of conscious awareness should be sufficient to produce moralization via moral intuition, but we found no evidence to support this hypothesis. Our laboratory experiments found that strong but incidental disgust cues did not have the power to moralize abortion attitudes—only abortion disgust cues presented at conscious levels of awareness work—a result that was replicated across two studies. Similarly, emotions tied only to specific candidates predicted changes in candidate-based moral convictions. Changes in enthusiasm for preferred candidates and hostility for nonpreferred candidates did not spill over to predict changes in moral conviction about the other candidate. Just as we found in the lab, the emotions that moralize are the ones tied specifically to the target of moralization.
Attitude moralization may not be based on harm
If harm is a necessary and not just a sufficient basis for attitude moralization, then (a) recognition of harms and benefits should mediate any effects of emotionally relevant cues on changes in attitude moralization and (b) perceptions of harms and benefits should predict changes in attitude moralization. Our studies clearly demonstrated that attitude-specific emotions, and not perceptions of harm (or benefits), were the driving force that predicted moralization. Disgust and not harm mediated the experimental effects we found in the lab.
Where to go?
The findings we share in this article are not intended to be the final word on attitude moralization or whether moralization processes are intuitive or harm based. It is simply not possible to draw definitive claims from two experiments and one study about how moralization occurs under all circumstances and for all people. That said, the research reviewed here provides clear evidence that moralization does not always result from either intuition or appraisals of harm. This claim should not be controversial. Indeed, enormous amounts of data would be needed to support any argument claiming to have found a single explanation for how an important aspect of morality functions for all (or even most) situations or individuals, a requirement that should apply not only to moralization but also to similar concepts such as moral judgment. With that in mind, we hope this article helps motivate more research into the important question of how various phenomena come to be imbued with moral relevance. In the remainder of the article, we point to a few possible directions that this research could go.
First, future research could leverage research on moralization to test other theories of moral judgment. For example, do the contents of moral foundations (Graham, Haidt, & Nosek, 2009) cause people to moralize particular issues? Although prior work has linked moral foundations with support for political issues (Koleva, Graham, Iyer, Ditto, & Haidt, 2012), they have not been linked with changes in moral relevance. In addition, some perspectives suggest that harm intuitions (rather than appraisals such as those tested in the studies reviewed here) are particularly relevant for moral judgments (Gray, Schein, & Ward, 2014; Schein & Gray, 2017), and it is possible they are also relevant for moralization.
Second, more research is needed to test the generalizability of these effects across a wider array of attitude objects. For example, it is possible that intuitions (be they grounded in harm or otherwise) come into play only very early in the moralization process and draw one’s attention to morally relevant aspects of the situation, and that attitude-relevant emotion is then needed to carry the moralization process forward. Such a possibility would imply that moralization happens in stages (e.g., Lovett & Jordan, 2010) and that it may be helpful to distinguish between instances when an attitude acquires an initial moral association and those where that association becomes strengthened.
Finally, theoretical claims about moral intuition have been widely embraced without a track record of strong empirical support. Empirical tests about moral intuition find that incidental affect or hypnotic associations with disgust sometimes lead to harsher moral judgments (e.g., Wheatley & Haidt, 2005). A meta-analysis, however, found that these effects are collectively not robust (Landy & Goodwin, 2015), and few, if any, studies have tested the cognitive implications of social intuitionism directly—that is, tested the degree to which moralization requires conscious versus unconscious awareness and deliberative thought. More research is needed to directly test just how many controlled, deliberative processes are required to produce moralization. Is simply the recognition of the association between an attitude and an emotion enough? Or are predictions from theories based in moral reasoning closer to the truth, and people, in fact, often think their way into or out of moralization (e.g., Narvaez, 2010)? Major theories in moral psychology should make bold and risky claims, but those claims should also be thoroughly tested before they receive widespread acceptance.
