The Importance of Context in Moral Judgments

Contextualizing Actors

Did the employees of Theranos have a moral obligation to report their company to regulators? When I posed this question to students in my Introduction to Business Ethics class, they had a heated debate about the difference between actions and omissions, shareholders and stakeholders, and internal and external reporting. My students debated concerns that could apply to any employee in any organization. However, when my students posed this question to Cheung, her answer was quite different. Cheung hesitated to uniformly condemn the complicity of her previous colleagues. Although her moral courage cost her socially, psychologically, and financially, as an American citizen, Cheung knew she was not at risk of being deported or harming dependents, unlike many of her colleagues. Underlying Cheung’s response was a belief that an employee’s specific identity can reasonably affect moral evaluations of their behaviors.

Although we can debate the morality of actions performed by nameless, genderless, ageless, family-less, race-less actors, individuating factors shape our moral evaluations. In Bad Blood, John Carreryou (2018) profiled two Theranos whistleblowers—Erika Cheung and Tyler Schultz. Cheung is the daughter of Asian immigrants who grew up in a working class family. When her name was leaked to the press, she was forced into hiding, sleeping on friends’ couches, unable to hire top lawyers. Schultz is an upper-class White man and the grandson of former Secretary of State George Schultz—who was a trustee at Theranos and a close adviser to Theranos’s founder. His family paid more than $400,000 to secure top legal representation. Although both whistleblowers performed similar actions, the evaluations of those actions might differ in theoretically informative why.

By examining specific identities, scholars can uncover how factors such as power, status, race, gender, and personal relationships each affect moral judgments. Researchers can examine whether low-power or -status individuals, such as Cheung, are perceived as having less responsibility for speaking-up compared with high-power individuals in organizations (Wellman, Mayer, Ong, & DeRue, 2016). Researchers can also examine whether disloyalty is judged more harshly for Asian American women who might be uniquely penalized for violating stereotypes of submissiveness or for violating the model-minority stereotype. Part of our identity as individuals is also interpersonal: We are children, parents, coworkers, countrymen, neighbors. Researchers can examine whether (a) someone like Tyler Schultz is viewed as having a special obligation to intervene because of his privileged relationship with people in power or (b) his whistleblowing is viewed more harshly because his betrayal also violated family obligations. Examining the impact of personal relationships can advance theorizing about whistleblowing (Waytz et al., 2013; Weidman et al., 2019).

In this issue, Hester and Gray systematically examine the limits of previous approaches in moral psychology with regard to the social identity of the actors and observers and provide concrete suggestions for future scholarship. In lieu of further discussion of identity here, I implore all scholars interested in moral judgments to carefully read their article.

Contextualizing Actions

The first eight words in Watkins’s (2020) abstract, “what is judged as morally right and wrong,” epitomize the structure of decontextualized morality research. Scholars have sought to uncover general truths about morality, examining what differentiates moral from nonmoral evaluations (Nichols, 2002; Turiel, 1983), what process (or processes) underlie moral judgments (Greene, 2007), and how the moral mind is structured (Schein & Gray, 2018). To draw conclusions that apply across a variety of moral situations, they often present participants with vignettes that fail to specify the larger context of the actions. For example, in my own research, I examine perceptions of brief, context-less vignettes (e.g., “having sex in exchange for money”; Schein, Ritter, & Gray, 2016) or the morality of one-word actions (e.g., “murder,” “pornography”; Schein & Gray, 2015). With trolley problems, even if a context is provided, it is often one that is far beyond believability (Bauman et al., 2014).

The next two words in Watkins’s (2020) abstract, “in war,” contextualize her research question in a way that advances both the psychology of war and moral psychology. Although murder might typically be labeled immoral, Watkins’s research tests principles that affect people’s moral condemnation of this otherwise grave moral violation (Watkins & Brandt, 2019; Watkins & Goodwin, 2020). Following Watkins’s example, moral psychologists can benefit from critically considering morality of unique settings, such as the morality of artificial intelligence (Bigman & Gray, 2018) or the morality of suicide (Rottman & Kelemen, 2014).

In addition to examining unique moral scenarios, moral psychology can grow from a greater focus on everyday morality (Hofmann, Wisneski, Brandt, & Skitka, 2014) and applied moral questions (Inbar, 2018). Moral psychology is uniquely situated to shed light on some of today’s most pressing challenges. In the #metoo age, companies have both a moral and financial incentive to understanding the barriers to whistleblowing (Weidman et al., 2019). When faced with the imminent threat of climate change and global food insecurity, understanding the nature of people’s moral aversion to genetically modified food has immediate policy implication (Royzman, Cusimano, Metas, & Leeman, 2020; this issue, pp. 250–272). How these applied question will advance theorizing is an open empirical question. Regardless of their findings, psychological science and society at large can benefit tremendously when editors continue to publish these research programs in flagship, high-impact outlets and do not relegate these cases to specialty or applied journals.

Contextualizing Judges

Even when identical actors perform identical actions in identical situations, moral evaluations can differ widely depending on who is doing the judging. Compared with the previous two contextual factors, there has been ample research in moral psychology on how individual differences or environmental pressures on the observer affect moral judgments (Hester & Gray, 2020). Previous research has documented the impact of incidental emotions (Gawronski, Conway, Armstrong, Friesdorf, & Hütter, 2018; Landy & Goodwin, 2015; Schnall, Benton, & Harvey, 2008), cognitive load (Wright & Baril, 2011), feeling of powerfulness (Lammers, Galinsky, Dubois, & Rucker, 2015), political ideology (Graham, Haidt, & Nosek, 2009), religion (Cohen, 2015), gender (Ward & King, 2018), and cultural differences (Buchtel et al., 2015; McNamara, Willard, Norenzayan, & Henrich, 2019) on moral judgments. Given the breadth of previous scholarship, here I will focus on one specific feature that has received relatively less attention in moral psychology research: the subjective experience of the observers.

Imagine that you are a participant in a moral-psychology study and are asked to rate the morality of consensual sibling incest. You tell the researcher that incest is immoral; it harms the siblings, undermines their relationship, and causes future harms. However, every time you bring up one of these concerns, the researcher shoots you down. The researcher claims that the act is harmless, the couple is using protection, they love each other, and they promise not to talk about their night with anyone else. Although you do not accept the researcher’s premise (Royzman, Kim, & Leeman, 2015), you eventually acquiesce and claim that incest is just wrong, even if you cannot articulate why.

In the dumbfounding moral paradigm just described, understanding participants’ subjective experience fundamentally shifts the interpretation of the results (Royzman et al., 2015). Haidt and colleagues interpret this response as evidence of the primacy of emotions and intuitions over reasons in moral judgments (Haidt, 2001). After all, participants insist that incest is immoral even after all their moral reasons are invalidated. However, this conclusion fails to adequately consider the perspective of the participant, who consistently denies the researcher’s claims that the act is harmless. Accepting that participants have consistent reasons for their beliefs, which researchers failed to acknowledge, undermines claims in favor of social intuitionism (Guglielmo, 2018; Royzman et al., 2015). Taking participants’ experiences seriously totally shifts the interpretation of the findings.

At its core, the study of social psychology is the study of human experience (Wegner & Gilbert, 2000). Without a direct window into other people’s mental states, researchers are constantly faced with the challenge of inferring participants’ internal experiences from survey responses and behavioral choices. At times, the relationship between people’s external responses and their latent beliefs is fairly straightforward. It is uncontroversial to infer disapproval of GMOs from the endorsement of statements such as “GMOs should be prohibited.” Inferences are trickier to make from questions that have multiple parts or require participants to imagine a counterfactual state such as “GMOs should be prohibited no matter how great the benefits and minor the risks for allowing it” (Scott, Inbar, & Rozin, 2016).

Royzman and colleagues’ body of research consistently demonstrates how to conduct studies that accurately gauge human experience (Royzman, Cassidy, & Baron, 2003; Royzman et al., 2015). In this issue, Royzman, Cusimano, Metas, & Leeman (2020) demonstrate the importance of carefully considering how participants interpret research questions. They lay out two prerequisites for making valid inferences from survey questions: the comprehension requirement (do participants understand the question) and the task-acceptance requirement (do participants have “the cognitive aptitude . . . to execute the task in question”; p. 254). In a series of studies, Royzman et al. (2020) found that participants do not fully grasp the intended meaning of Scott et al.’s (2016) initial question. By carefully taking participants’ understanding of the question into consideration and examining what “no matter the costs” really means, they were able to simultaneously replicate and undermine Scott et al.’s (2016) initial claims. More broadly, they provide an illustration of the importance of researchers’ careful examination of how exactly participants understand and experience the research question before making broad theoretical inferences.

Contextualizing Values

Although moral questionnaires frequently ask participants to endorse abstract moral statements, or consider moral violations in isolation, real-life moral decisions involve values situated in specific contexts. Sometimes moral situations involve value dilemmas. Here, I am not simply referring to sacrificial dilemmas (Cushman & Greene, 2012; Foot, 1967) that involve trade-offs between deontological principles (e.g., do no harm) and utilitarian consequences (e.g., killing one to save five). These specific dilemmas fail to capture the full breadth of moral conflicts and lack ecological validity (Bauman et al., 2014). Rather, I am referring to realistic scenarios that require navigating competing moral and nonmoral values. For example, in the whistleblowing case, loyalty to employers conflicts with protecting consumers from harm (Dungan, Young, & Waytz, 2019; Waytz et al., 2013). In free-speech debates, liberty and autonomy conflict with safety and social order (Gibson, 2013). In debates over corporate social responsibility, concerns for environmentalism, patriotism, or ethical supply chains often conflict with short-term profitability and shareholder value.

Endorsements of abstract moral principles such as loyalty, liberty, environmentalism, or patriotism are likely to be poor predictors of judgments when values conflict. To best understand whistleblowing decisions, it may not be enough to ask about one’s general concern for loyalty or fairness in isolation. What might better predict moral evaluations is how one prioritizes obligation to one’s own group relative to obligations to the broader community (Dungan et al., 2019). Political science research on civil liberties clearly demonstrates how decontextualized, abstract moral questions are poor predictors of concrete concerns. Although American participants overwhelmingly endorse decontextualized statements such as “it is important to extend political rights to all people, regardless of their beliefs,” when this freedom has historically conflicted with other values, such as moral disapproval for homosexuality or communism, the endorsement of specific civil liberties falls substantially (Gibson, 2013; McClosky & Brill, 1983). Placing abstract values such as loyalty or liberty into concrete situations shifts moral judgments in meaningful ways.

Measurements of abstract values not only fail to capture moral judgments when values conflict but also fail to adequately capture how people perceive value in the world. With regard to civil liberties, although almost all Americans believe at an abstract level that civil rights should exist, perceptions of specific civil liberties depends on who’s perceived to have these inherent rights (Gibson, 2013). In 1972, 48% of survey respondents claimed that “a man who admits that he is a homosexual” should not be allowed to teach in a college or university. That percent was down to 11% in 2018 (Smith, Davern, Freese, & Morgan, 2019). Questions measuring broad endorsement of civil liberties would be unable to detect this change. Likewise, moral judgments of disobedience depend on perceptions of legitimate sources of authority (Frimer, Gaucher, & Schaefer, 2014), and moral judgments of harm depend on perceptions of vulnerability and moral rights (Schein & Gray, 2018). The same action, such as destroying a fetus, can be construed in vastly different ways depending on whether a fetus is perceived as a mass of mindless cells or a human capable of experiencing pain, and worthy of moral concern.

In this issue, Beal (2020; pp. 273–290) compellingly demonstrates the benefit of shifting from questions that gauge abstract moral values to questions that more accurately measure participant’s perceptions of the inherent value of entities involved in a scenario. His example of slavery is particularly illustrative:

If you could ask an early White American one question to discover their moral feelings about slavery, what would it be—a question probing how much they value justice, welfare, or liberty, or a question probing to what extent they view non-Whites as being equal to Whites? It seems obvious to me that the question about values would lead one astray. (p. 282)

I share Beal’s intuition that the context-specific question would be a better predictor than questions about general values. In lieu of questions about abstract values, Beal proposes that scholars ask about specific “ontological frames.”

Although the term “ontological framing” is new to the field of moral psychology, there is a long history of similar concepts in the literature. Since the 1930s, psychologists have challenged anthropological assertions of ethical relativism by appealing to differences in how people understand and perceive meaning in the world (Duncker, 1939). For example, in his 1952 textbook, Solomon Asch appeals to “situational meaning” (p. 377) to explain why the existence of cultural differences in infanticide is consistent with the existence of psychological universals. According to Asch, what matters for moral values is how people construe meaning in specific contexts. More recently, scholars have argued that moral diversity can emerge from differences in “informational assumptions” (Turiel, Killen, & Helwig, 1987), differences in where people perceive vulnerability (Schein & Gray, 2018) and humanity (Haslam, 2006), differences in who is included as part of the moral community or moral sphere (Crimston, Bain, Hornsey, & Bastian, 2016; Waytz, Iyer, Young, Haidt, & Graham, 2019), and differences in who or what is seen as possessing sacred value (Fasoli, Saunders, & Andrade, 2018). Although some of the scholars just cited disagree over the structure of morality, they generally embrace moral diversity by accepting some version of “universalism without the uniformity” (Shweder, 2012; p. 88). Beal’s criticism of previous moral theorizing suffers from overlooking important historical and contemporary scholarship that actually bolster his larger claim.

What I appreciate about Beal’s article is that it places these subjective perceptions at the forefront of research. He suggests that ontological frames are not simply a way of capturing moral diversity but rather a starting point for theorizing. Although I find the term “ontological frames,” somewhat unwieldy, Beal’s term has the benefit of avoiding the linguistic ambiguity of “value,” which can refer both to a person’s abstract moral values and to subjective perceptions of who has moral value. The term is also agnostic about what constitutes worth, allowing for moral rights to stem from harm, sanctity, or any source of value. Psychology as a field can grow by following Beal’s suggestions to move beyond broad, abstract moral-values questions. Instead of asking participants to endorse abstract harm principles, researchers can examine who is vulnerable to harm (Schein & Gray, 2018). Instead of asking about the importance of sanctity, researchers can examine where sanctity can manifest in the world, examining who and what is seen as having sacred value (Fasoli et al., 2018).

Additional Benefits of Contextualization

Throughout this article, I have highlighted research questions that emerge from contextualizing moral psychology research. Next, I briefly focus on two specific predictions for how contextualized morality might outperform decontextualized moral questions (a) by improving predictions of moral behaviors and (b) by improving predictions of moral change, two areas of moral psychology research in need of further research.

Conclusion

Of all the topics debated in courtrooms, dining rooms, boardrooms, and bedrooms, few issues effect people more profoundly than questions of morality. Moral psychology is uniquely situated to offer concrete, theory-informed insights into topics that are compelling, relevant, and deeply personal. However, by studying judgments of nameless, genderless, race-less, relationship-less actors, performing actions in unspecified or highly unrealistic settings, moral psychologists have overlooked broad areas of important moral concern. When scholars start leaving moral scenarios embedded in their typical contexts, they open up a world of compelling moral questions.