Two Ways to Answer “Why”

The most influential methodological handbook in political science asserts that attempting to infer from something that is directly observed to something that is not is the hallmark of all science (King, Keohane, and Verba Reference King, Keohane and Verba1994, 8). What the authors (widely known as “KKV”) call “descriptive inference” involves inferring something about the characteristics of phenomena that are not observed from those that are. When the phenomenon whose properties are inferred will only occur after the relevant inference has taken place, descriptive inference takes the form of prediction. But although descriptive and predictive inferences can be very valuable, mere description or prediction often leaves scholars dissatisfied.Footnote ¹ They also want to know why something is as it is or will be as it will be. For this we need to move beyond description and prediction to explanation. As Bas Van Frassen’s classic analysis makes clear, “an explanation is an answer to a why-question” (Reference Van Frassen1980, 134).

One means of explaining why some phenomenon is as it is to discover what caused it to be that way. KKV see what they call “causal inference” as the goal of most social inquiry. Their official view of causation is a counterfactual one, in which X causes Y to the extent that the value of Y would have been different had X not occurred (KKV Reference King, Keohane and Verba1994, 76–82). Although this definition of causality does not itself involve the need to identify specific causal mechanisms, the identification of these mechanisms is, for KKV, nonetheless the central task of social science (85–86).

There are three well-established methods for identifying causal mechanisms. First is qualitative process-tracing. Second is controlled experimentation. Third are specific statistical techniques that have been developed to identify causal mechanisms using nonexperimental data, with old-fashioned regression using nonexperimental data now properly seen to be merely descriptive. Most of these methods depend on an assumption of unit homogeneity: causal mechanisms are expected to produce the same effects in all instances (91). In practice, however, even the best causal explanation never accounts for every single instance of a given phenomenon. The hope is that outliers left unexplained by one causal explanation can be accounted for by another one, which is typically left to future research. In the meanwhile, they remain unexplained.

When it comes to human actions, however, simply because no causal explanation is available does not mean that the action in question cannot be explained in other ways. A particle that fails to behave like other particles of its kind probably cannot be explained by current physics. By contrast, a person who fails to vote for the candidate usually preferred by members of the demographic categories to which she belongs can usually explain why she acted in this unusual way. Unlike most other natural entities, human beings do much of what we do intentionally—that is, for reasons. Even if we are not actively conscious of our reasons for action, we may be able to identify subconscious reasons upon later reflection, or empathetic others may help identify them for us.

Intentional action is fully liable to explanation, albeit not explanation of a causal sort. We are familiar with this sort of explanation from everyday discourse. When someone asks us why we did what we did, we are unlikely to respond in terms of causal mechanisms. When asked why we voted for our chosen candidate, for example, it would be very strange to trace a causal chain beginning with how our parents disciplined us as children, which caused us to have an (anti-)authoritarian personality, which then caused us to support (anti-)authoritarian candidates. Instead, we would defend our vote in terms of the reasons we have for voting as we did—in terms of subjectively appreciated factors that we hope will justify our actions and that we expect our interlocutor to be able to understand intersubjectively.

In some cases, these reasons will take the form of subjective analogues to the kind of objective mechanisms that might form a part of a causal explanation. Both a causal social scientist and a voter herself might refer to the voter’s demographic characteristics in explaining her choice of candidate. When the voter refers to these characteristics, however, it is likely to be in terms of how her race, religion, and gender shaped her values and interests and how these values and interests make voting for a certain candidate the morally and prudentially right thing to do. While it is possible to see objective demographic characteristics as a cause of our behavior, our subjective understanding of these characteristics transforms them into reasons that make our choices both understandable to others and at least potentially justifiable according to at least one relevant domain of normativity (ethics, aesthetics, prudence and instrumental rationality, etc.).

Rather than a matter of tracing a mechanism, an interpretive explanation is a matter of providing a meaningful narrative—in more colloquial terms, telling a story—in which protagonists act for understandable and potentially justifiable reasons. When we explain ourselves to others, we tell a story about our beliefs, motivations, and values with ourselves as the protagonists. Since it relies on the empathetic understanding of these sorts of narratives, the mode of explanation in interpretive social science resembles that familiar from literary criticism and other humanities disciplines, while causal social explanation more closely resembles explanation in the natural sciences.

It is important, however, not to confuse the distinction between causal and interpretive explanation with a host of related dichotomies: scientific versus humanistic research, quantitative versus qualitative methods, positivist versus anti-positivist worldviews, “thin” versus “thick” description (Geertz Reference Geertz1973), and so on. While it is true that interpretive explanation has an affinity with humanistic and qualitative forms of social science—and causal explanation has an affinity with quantitative, experimental, and other forms of social science modeled on the natural sciences—in both cases this is just an affinity. It is entirely possible to offer causal explanations in qualitative research, to offer interpretive explanations in quantitative research, or to offer mere description or prediction without explanation in both. While there may be important ethical considerations at stake in the much-discussed Methodenstreit between “quants” and “quals,” they are not my subject here.

Turning to Normative Ethics to Choose Explanatory Strategies

Once we are clear on the differences between them, we see that social scientists must choose whether to use causal or interpretive explanations. While some have claimed that philosophers can dictate which choice they should make, others have insisted that there is no objectively best form of social explanation. I will be adopting a middle ground between these two positions. My stance is moderate in two distinct ways. First, in this subsection, I will argue that philosophy, while not wholly determinative, nonetheless has much to contribute to social scientists’ choice of an approach to explanation. This becomes evident as soon as we turn our attention away from the metaphysics of reasons and causes and instead toward normative ethics. Later, once we see the relevant normative arguments, we will come to the similarly moderate conclusion that both causal and interpretive explanations can be ethically valuable.

While I will be defending this middle ground philosophically, many social scientists stake a middle ground in practice, combining interpretive and causal explanations throughout their work. They explain certain phenomena causally, while explaining others interpretively, and then examine how these two kinds of phenomena interact.Footnote ² In addition to combining different modes of explanation for different phenomena, it is also possible to offer both causal and interpretive explanations of the same phenomena. Subjects’ reasons for action can rightly be said cause them to act as they do in the world. These actions in turn set off further causal chains, which can only be explained when we interpretively understand the reasons that first set them in motion.Footnote ³

Even though other phenomena can thus be explained as caused by reasons, many argue that reasons cannot be caused by other phenomena. Under this view, reasons only qualify as such when they are unmoved movers, the uncaused first links in a subsequent series of causes. This position can then be used to defend the further claim that action-for-reasons cannot be explained by appeal to the sorts of causal mechanisms familiar from the natural sciences. The conclusion is that, since the reasons at the root of all social phenomena are genuinely uncaused, all social explanation must take a fundamentally interpretive form (see, e.g., Bevir and Blakely Reference Bevir and Blakely2018; Hutchinson, Read, and Sharrock Reference Hutcheson, Read and Sharrock2008; Winch Reference Winch1958/1990). Social scientists thus fall under the rule of philosopher-kings, who can declare a priori which of their explanatory practices are intellectually legitimate.

Here is hardly the place to settle these metaphysical issues. It seems reasonable to assume, if only for purposes of argument, that nothing in the fundamental nature of reality precludes our explaining social phenomena causally, interpretively, in both ways, or in neither. Given the plausibility of this view, it might be tempting to think that there can be no compelling answer to the question of social explanation at all, only irresolvable struggles between competing schools. This is not to say that the contest at hand would then just be a petty fight among professors for control of scarce academic resources. It might instead be a grand battle between competing regimes of knowledge-power, each seeking ideological hegemony as an integral element of its quest for political domination.

My attempt to address the conflict between causal and interpretive explanations stakes a middle ground between the view of explanation as enacted metaphysics, on one hand, and the view of explanation as a weapon of the powerful on the other. Explanation remains a form of enacted philosophy, but the branch of philosophy in question is now normative ethics. Choosing different modes of social explanation will reflect different stances towards others, stances that can then be evaluated ethically. This article will focus on one important consideration at stake: respect for explanatory subjects as persons.

The Relevant Form of Respect for Persons

This is no more the place to defend a general theory of respect than it is the place to defend a general theory of reasons and causes. Instead, this subsection will simply identify the specific form of respect that will be relevant to the normative argument later in the article. The relevant form of respect is what I will call recognition-respect for others as persons.

Most analyses of the concept begin with the observation that respect is relational: A person respects an object of respect, establishing a respectful relationship between them. The relationship need not be symmetrical. A person’s object of respect may not respect her back. It may not even be the sort of being capable of respect, since it is possible to respect a physical thing (e.g., a bible) or an abstraction (e.g., the Bible, defined in terms of its semantic content). In order to show respect, a being must be a person—that is, an agent, an entity capable of action for reasons that are interpretively understandable in the manner described in the previous section. Symmetrical respect relationships can only occur when the object of respect is also a person. Self-respect occurs when the agent and the object of respect are the same person, while mutual respect occurs when two persons respect each other. Not all forms of respect, however, are capable of such symmetry.

The respect relationship is also asymmetrical in a second way. While the agent of respect needs to be a person to be capable of respect, none of her other features are important to the respect relationship. Instead, the reasons for respect are grounded solely in features of the object. When a person responds to these features respectfully, there is an implicit claim being made that any other person in her position has reason to do the same.

There is an ambiguity about what it means for a person to orient herself respectfully toward an object of respect. This orientation can be understood as a belief, a feeling, a set of actions, a disposition to perform such actions, or some combination of these (Green Reference Green2010, 219). One might worry that, since feelings and beliefs cannot be produced at will, respect must be understood as a matter of action if it is to be obligatory. This is not a problem, however, if respectfulness is understood as an Aristotelian virtue—that is, as a complex of right beliefs, right feelings, right habits, and right actions that cannot be demanded of any person at any moment, but must be cultivated over the course of a lifetime. A respectful person will show respect as warranted, but not everyone is a respectful person, nor can a disrespectful person become respectful through a sheer act of will.

So far, all of these are purely formal features of respect. In order to add content to this form, we need to specify what an object of respect is being respected as. Think about what it means to respect the Bible as literature but not as scripture. Respecting an object as something situates it in an evaluative domain, highlighting certain of its features over others. While the contours of a given evaluative domain may be contingent on the surrounding culture, the intersection of natural and social facts that makes a certain object of respect fit a particular domain in a given cultural context should be recognizable by any competent agent within it (Green Reference Green2010, 228).

There are two ways an object can deserve respect as a member of an evaluative domain, which Stephen Darwall (Reference Darwall1977) calls appraisal respect and recognition respect. Appraisal respect involves evaluating the object positively within a certain normative domain. I show the Bible appraisal-respect as a work of literature if I rate its literary merits highly, but do not respect it in this way if I think it is literarily mediocre. Recognition respect, by contrast, involves appreciating the status of an object as a member of a certain normative domain without necessarily evaluating it. Entities may solicit our recognition respect, and all it entails, regardless of how high they rank within the relevant domain. All works of literature, regardless of their quality, may, according to our local customs, deserve a place in our national depository library. Recognition respect of this sort is a matter of status, not of merit.

Deciding not to procure a copy of a book for the national depository library is to deny that book recognition as literature. When the book is in fact literature, such action is disrespectful; it implicitly or explicitly denies its object the status it deserves. Procuring a copy of the book for the library, by contrast, is respectful, expressing recognition of the book as the work of literature that it is. These two categories are far from exhaustive, however. There are many ways to interact with a book that neither affirm nor deny its status as literature: one can count its pages, mail it to one’s aunt, place it gently on a clean table, etc. Such actions, and the orientation toward an object that they express, can be called “nonrespectful.” Nonrespect can then be supplemented with either respect or disrespect, which will then determine whether an agent is being respectful overall. When mailing the book to one’s aunt, one could include a respectful note suggesting that she write back with her thoughts on its literary merits, or a disrespectful note that she tear out the pages and use them as a solution to the current toilet paper shortage.

The form of respect that will be most important later in this article is a form of recognition respect. It will be important to my argument, however, to understand that recognition and appraisal respect do not operate independently. Evaluating something within a normative domain by either granting or denying it appraisal respect is to give it recognition respect within that domain. If my aunt reads and dislikes the book I send her, she may deny it appraisal respect as literature, but in so doing she is showing it recognition respect. In ordinary language, it is thus perfectly sensible to say that she both (appraisal-)disrespected the book when we contrast it to the works of the literature that she prefers and (recognition-)respected the book when we contrast it to the non-literature that she uses as toilet paper. Recognition respect, as assignment to a normative domain, always carries with it the risk that something will be found wanting when judged within that domain. It does not necessitate evaluation of its object, but it does make the object liable to evaluation by assigning it criteria by which it could be properly evaluated.

The form of recognition respect relevant for this article is respect for subjects of social explanation as persons. Darwall argues that those who confuse appraisal and recognition respect (such as Cranor Reference Cranor1975) are particularly likely to go awry in their analysis of respect for others as persons. In ordinary language, to respect someone as a person is to assign them to the evaluative domain of ethics or morality (terms I will be using here interchangeably). Everyday discourse, however, is often ambiguous as to whether it refers to appraisal respect or recognition respect (Hudson Reference Hudson1980, 73). When we say that we respect Gaugin as an artist, but not as a person, we probably mean that we appraise his aesthetic achievements highly, but do not think much of his moral character. In saying this, however, we are at the same time showing recognition respect for Gaugin, acknowledging him to have the status of a person liable to ethical evaluation. If I were to say that I do not respect a robot as a person, however, I might instead be talking about recognition respect, denying that the robot qualifies as a person at all.

Determining who deserves appraisal respect as a person can be understood as the task of much of the discipline of normative ethics. Determining who (or what) deserves recognition respect as a person is not quite as monumental a task, but it too is no easy matter. There is widespread agreement that “normal” adult human beings deserve to be recognized as persons, but since there is no agreement about which of their features solicit this recognition respect, there is no such agreement for a host of other cases: children, the incapacitated, nonhuman animals, collectivities, robots, etc. I will avoid these thorny issues by stipulating that the subjects of social explanation being discussed are the sort of beings properly recognized as persons; the reader can feel free to include or exclude any controversial cases.Footnote ⁴

The Importance of This Form of Respect

Once we have identified the form of recognition respect at issue in this article, the case must be made that showing such respect is morally important. Respect in its most general sense—regardless of whether it is appraisal or recognition respect, and regardless of whether it is respect for entities as persons or as something else—seems inherently valuable. Proper recognition respect involves situating something within a normative domain to which it truly belongs, and proper appraisal respect involves evaluating it accurately within that domain. To fail to show proper respect would be to fail to see the properly respect-soliciting features of the object, and hence to fail to orient oneself properly to the world as it is. In this way, the value of respect is connected to the value of truth (Frankfurt Reference Frankfurt1997, 12).

The world, however, is filled with trivial truths. There may be some truth of the matter as to whether a certain long-forgotten mystery novel deserves either appraisal or recognition respect as literature, but the issue is hardly an important one. Yet there is widespread agreement that respecting the status of persons as such is important in a way that respecting the status or merit of many other entities is not.

While the value of respect in general is connected to the value of truth, the unique value of recognition respect for persons is also connected to the value of reciprocity (see Sennett Reference Sennett2003, 219). This is because recognition respect for others as persons should take the form of mutual respect. To recognize someone as a person is to acknowledge her to be the kind of entity that is capable of respecting as well as being respected. It is to see in her the very features in me that allow me to show her respect. If I am a properly self-respecting person, I recognize that we have these features in common. Since any instance of respect includes an implicit claim on all relevantly situated persons to respect its object, a self-respecting respecter of others expects them to respect her as a person as well.

The value of creating a community of mutual respect becomes clearer when we consider the alternatives. Disrespectfulness for others as persons can take many forms, not all of which can be listed here, but when we consider a few of them we can see why they are nontrivially objectionable. Recognition respect for others as persons is, first, incompatible with infantilization or paternalism (see Sennett Reference Sennett2003, 102–107). Regardless of the status of older children, infants and young children are at most proto-persons, beings with interests but without full agency. It is disrespectful to treat full-fledged persons paternalistically, as though they had interests but no agency.

Second, respect for persons is also incompatible with the sort of instrumentalization or objectification of which Aristotle is guilty in his defense of slavery (c.Reference Aristotle, Sinclair and Saunders330BC/1992, 1253b1–1256a1, 62–75). Objectification in this sense (though not in all senses in which the term is used) involves treating someone as a non-person, something without interests of its own but which is nonetheless valuable insofar as it can be used to advance someone else’s interests (Lucas Reference Lucas2011; Nussbaum Reference Nussbaum1995). Both infantilizing paternalism, and instrumentalizing objectification may involve forms of manipulation—that is, strategies of control designed to bypass agency. As objectionable as manipulation may be, worse still are dehumanization and demonization. These involve an attitude that the disrespected lack even instrumental value—indeed, their very existence may be of negative value—and are hence fit for destruction (Smith Reference Smith2011).

Disrespect of any of these types is objectionable regardless of whether it is directed at individuals, groups, or humanity as such. A sociopath obviously wrongs the individuals that he manipulates. We also condemn a demagogue who manipulates anonymous crowds or whole societies. We even condemn a sociopathic hermit who cultivates a belief that humanity is a collection of objects to be manipulated, and we do so even if social isolation prevents him from ever manipulating them.

Yet while there is widespread agreement that recognizing and respecting others as persons is morally important, there is less agreement as to exactly how important it is. Many philosophers—typically taking inspiration from Kant (Reference Kant and Gregor1785/1996, 8:428–429, 77–80) or Hegel (Reference Hegel and Knox1821/1967, ¶35–36, 37)—have argued that interpersonal recognition respect cannot be outweighed by competing moral values, that it is a categorical imperative. Some have gone so far as to identify it as the categorical imperative, or at least one of its possible formulations, and hence the grounds for all moral duties as such (Donagan Reference Donagan1977; Downie and Tefler Reference Downie and Tefler1969). There is no need, however, to adopt such a strong Kantian position here.Footnote ⁵ It is also possible that this form of respect is merely one moral good to be weighed against others in a consequentialist calculus. It might be inherently good—perhaps as an instantiation of a more abstract inherent good like truth or reciprocity—or it might merely be instrumentally valuable in the pursuit of something else, like equality or happiness. The precise moral importance of respect will thus depend on one’s comprehensive ethical views.

While I do not wish to tie myself to any single ethical theory here, the kind of respect that I am discussing, while often associated with Kant or Hegel, also fits with Aristotelian virtue ethics, its notable absence from Aristotle’s own philosophy notwithstanding (Weber Reference Weber, Giorgini and Irrera2018). One advantage of a virtue-based over a consequentialist approach is that, like deontology, it can make sense of our intuition that disrespect is wrong regardless of its consequences. Unlike deontology, however, it can also make sense of the intuition that respect might not always be an absolute obligation. Respectfulness might be one virtue among many, just as consequentialists may believe it to be one good among many.

Virtue ethics also allows us to deal with the worry that, while respect for persons may be important in general, its role in choosing between modes of social explanation is morally trivial. One might think that, since most social science has little effect on anyone or anything, the consequences of engaging in either respectful or disrespectful research will be unimportant. Yet even if the consequences of a researchers’ choices are negligible—and there is no reason to believe that they always will be—social scientists are still human beings, and therefore have reason to be good human beings. The virtues or vices instantiated in social explanations are important parts of a researcher’s psyche regardless of whether the character traits developed in the explanatory stage of research spill over into other stages of research, other spheres of researchers’ lives, or the wider society around them.

Interpretive Explanations Are Respectful

Before arguing for the thesis that interpretive explanations are inherently respectful in the relevant sense, it is necessary to distinguish my argument for this thesis from a different, less successful one. This alternative argument begins with the premise that, since recognition respect for persons ought to take the form of mutual respect, the most respectful way for me to explain others might be to explain them as I explain myself. Self-explanation, as we saw earlier, typically takes an interpretive form. We might then conclude that, in order to be respectful, explanation of others must also take an interpretive form.

The problem with this argument is that, while self-explanation most often takes an interpretive form, it does not always do so. When someone asks me why I did something, my immediate response is usually to give a narrative of my intentions and their relationship with my broader goals. If my action was a rational means of pursuing my intended end, my response can end there. In some cases, however, I may realize that my action was not a rational means of pursuing my intended end. Alternately, my intended end might be one that I never had any good reason to pursue in the first place. I may even realize that I did not behave with any reason or intention at all, but merely out of unthinking habit or reflex. If I were familiar with the jargon of social science, in these sorts of regrettable circumstances I might turn to talk about my social circumstances or psychological compulsions. In other words, in situations of self-acknowledged error I readily move from an interpretive explanation of my meaningful action to a causal explanation of my errant behavior (Dennet Reference Dennett1987, 86).

If causal forces such as individual psychological pathologies or large-scale social structures are sufficiently strong to entirely determine my behavior, overpowering my agency, they may serve to excuse my error. If society provides me with only one inescapable course of action rather than a menu of choices, or if mental illness robs me of the power to choose, then I am not to blame for what I did. If these sorts of causal forces merely influence my behavior without determining it, I may still be culpable for my error, but at least I can trace these causal mechanisms to help explain why I went wrong, hence lessening the degree of my culpability. Perhaps my willpower was weakened by strong emotions, or society offered me a very limited menu of choices, all of them bad. Only on those rare but precious instances when we refuse to offer causal explanations of this sort do we take full moral responsibility for our errors. A true apology involves acknowledging the wrongness of my decision qua decision, not merely lamenting an unfortunate chain of causally determined events.

A scholar seeking an interpretive explanation of a subject’s action is in an analogous situation. Rational action in pursuit of ends that the scholar believes his subject has good reason to pursue is easy to explain interpretively. “We obviously ‘understand’ without further question a person’s solution of a certain problem in a manner which we ourselves regard as normatively correct,” Max Weber observes (Reference Weber, Shills and Finch1917/2011, 40–41). Things become more difficult when a subject’s standards of what counts as rational action or what counts as a legitimate reason to adopt a certain end differ from those of the researcher. This, Weber says, requires drawing on the investigator’s “capacity to ‘feel himself’ empathetically into a mode of thought which deviates from his own and which is normatively ‘false’ according to his own habits of thought” (Reference Weber, Shills and Finch1917/2011, 41). Things become more difficult still when subjects fail to abide by their own standard of normative correctness. A subject’s self-acknowledged failure to act rationally or meaningfully cannot be explained through the usual interpretive appeal to meanings and reasons. Since, in these situations, subjects would themselves offer causal explanations for their behavior, a researcher attempting to see things from their point of view may be forced to offer a causal explanation as well.

Weber is thus not entirely correct when he claims that interpretation does not involve value judgments, that it merely involves “possible relationships of objects to values” (Reference Weber, Shills and Finch1906/2011, 143). It is true that, by offering interpretations of subjects’ actions, scholars need not imply that the subjects are behaving correctly according to the scholars’ values or standards of rationality. They do, however, inescapably affirm that the subjects’ actions make sense in terms of the subjects’ values and standards of rationality. The distinctively agential categories used in interpretive explanation—meanings, intentions, reasons, and so on—are by their nature normative. We can succeed or fail in carrying out our intentions, respond or fail to respond to relevant reasons for action; we can make sense or be entirely senseless.

Coherence may not be the most admirable of virtues, but it is a virtue nonetheless, as can be seen when it is contrasted with the vice of utter incoherence. Interpretive charity therefore really is a form of charity—a willingness to give subjects the benefit of the doubt and assume most of what they say and do is coherent most of the time. The insistence that this is not a value judgment leads interpretive scholars today into such absurdities as claiming that since the goal of their work is “not to correct, instruct, or pass judgment,” the subject must always be treated as “an expert about the topic at hand” (Schaffer Reference Schaffer, Yanow and Shwartz-Shea2006, 159). To deem someone an expert is to pass judgment about them; it is to judge them approvingly.

Yet even if interpretive social scientists are mistaken to believe that they are not judging their subjects, they are not to be condemned for the judgments that they pass. All else being equal, charity probably is a virtue, and a willingness to err on the side of excessive rather than insufficient appraisal respect is well worth cultivating. Admittedly, the first maxim of virtue ethics is “moderation in all things.” When charity is highly excessive, it can become a sort of vice.Footnote ⁶ Caught up in the charitable attitude that their methods require, interpretive social scientists do have an unfortunate tendency to forget the fact that part of what it means to recognize someone as a moral agent is to admit that she is capable of failing to live up to her own standards.

More importantly for purposes of our larger argument, recall that showing others appraisal respect as persons entails showing them recognition respect as persons. This holds true regardless of whether the appraisal respect is warranted. Even when we over- or underestimate others’ virtues, when we evaluate them according to distinctively human standards (such as morality and value rationality or prudence and instrumental rationality), we recognize them to be persons. While their tendency toward excessive, unacknowledged appraisal respect of their subjects’ choices may have its ethical drawbacks, interpretive researchers are to be credited for the recognition respect that they show to their subjects in this way.

Of course, unless it is cultivated as a robust character trait rather than an isolated feature of one’s explanatory methodology, recognition respect may be incomplete or insincere, perhaps preceded or followed by moments of disrespect. One virtuous act does not a virtue make, and my argument here should certainly not be taken to imply that all interpretive social explanation is ethically unobjectionable, let alone that all interpretive social scientists are good people. Incorrect interpretation of others may be a particularly dangerous evil, especially when it reinforces unjust stereotypes or distorts subjects’ authentic self-understanding (Lucas Reference Lucas2011). Even an entirely correct interpretive explanation, while itself respectful, may be part of a larger project of disrespect. Sociopaths, after all, are famously adept at understanding others’ worldviews and values. What is morally objectionable is that they then use their correct understanding of others for purposes of manipulating them. Similarly, an academic interpretation of others’ actions and reasons could certainly be used to manipulate them more effectively in the future. Anthropologists, for example, are still grappling with the use of their discipline as a means of advancing imperialist domination through the interpretive understanding of indigenous cultures (see Tilley and Gordon, Reference Tilley and Gordon2010). There remains, however, something genuinely respectful about interpretive social explanation as such. What precedes or follows an interpretive explanation may be disrespectful or otherwise unethical, but the explanation itself is not.

Causal Explanations Are Not Necessarily Disrespectful

Now that we realize that interpretive explanations inherently respect their subjects as persons, we might be tempted to jump to the conclusion that causal explanations are necessarily disrespectful. There are many who see something inherently manipulative in explaining human action using the same kind of causal mechanisms that scientists use to explain nonhuman phenomena. The goal of natural science, after all, is not merely to explain the world; it is also to control it (see Snow Reference Snow1959/1964/2012, 67). If scientists explain their fellow human beings as causally determined elements in the natural world, it is an understandable worry that they will also attempt to control them (Bevir and Blakely Reference Bevir and Blakely2018, 185–186).

This concern becomes even stronger when we realize that the two motives for science are not separable from each other in practice. The scientific method for explaining the world involves controlling it, as in a controlled experiment. Some interpretive scholars take pride in the fact that their preferred methods—such as conversational interviews and observer participation—involve less control over others than do standard scientific techniques such as experiments or surveys (see, e.g., Yanow Reference Yanow, Yanow and Shwartz-Shea2006, 70).

It is certainly true that interpretive researchers do not need to control the actions being interpreted in order to interpret them. To the contrary, in order to understand how these actions are grounded in an agent’s own beliefs and values, they must refrain from controlling the actions that they are interpreting. Recall, however, that the respect shown via interpretive explanation can be incomplete or insincere, perhaps preceded or followed by manipulation or other forms of disrespectful control. Conversely, while a controlled experiment will necessarily involve some short-term manipulation of others, the long-term goal may have nothing to do with projects of manipulation or other forms of disrespect.

Yet even when a given piece of causal explanation has no direct relationship with larger projects of control of some by others, it could nonetheless be objected that causal social science trains researchers to overcome the ordinary virtues of interpersonal relationships, including mutual respectfulness. In this, it is akin to the other depersonalizing systems that so many complain have utterly deformed modern life, turning our society into a series of structures of disrespect. For the causal social scientist, as for the bureaucrat described by Marx, “the world is a mere object to be manipulated” (Reference Marx and Tucker1852/1978, 25).

While an undeniably disrespectful and manipulative stance toward subjects is sometimes found among causal social scientists, however, this is not a consequence of causal explanation per se. As we have already seen, nothing precludes self-respecting individuals from providing causal explanations of their own behavior. As a result, there is nothing in causal explanation as such that prevents the kind of reciprocity characteristic of a community of mutual respect.

Those who insist that causal explanation is inherently disrespectful might counter this argument by pointing out that in ordinary life we mostly offer causal explanations of our behavior when we fail to live up to our own standards. If social scientists were to explain subjects as they explain themselves, this would lead them to use interpretive explanation when subjects largely live up to their own standards, turning to causal explanations when they fail to do so. If this were the case, it would be natural to suppose that while interpretive explanations express an attitude of charity and at least mild approval, and hence respect, causal explanations might involve an attitude of suspicion and at least mild disapproval, and hence disrespect.

There are two significant problems with this argument. First, it involves confusion between appraisal and recognition respect. While many ethical theories demand that we show recognition respect to all persons, none require universal moral appraisal respect. In fact, denying others appraisal respect involves respecting their status as persons by situating them within distinctively human domains of normativity. If causal explanation genuinely involved moral or prudential disapproval of its subjects, this would be evidence that it is respectful of them as persons in precisely the recognitional way relevant here.

The second problem with this argument is that causal explanation does not actually express disapproval. We must not be led astray here by the fact that self-explanation most often takes a causal form when we fail to live up to our own standards. It does not always do so, and the extent to which it does is culturally variable. Although the scenario is admittedly all too rare in a boastful America, self-effacing over-achievers in Britain often seek to evade others’ praise by pointing to the causal forces responsible for their achievements—be they a privileged social position, fortunate economic conditions, or a fortuitous alignment of political power in their favor. In this way, just as causal self-explanation can save us from others’ disapproval, it can also save us from their approval. Causal explanation defuses all ethical evaluation, positive or negative. As such, it removes subjects from the evaluative domain of morality.

Recall that even in cases of self-professed error we do not offer causal explanations as expressions of regret, remorse, or other attitudes of self-disapproval. To the contrary, explaining our errant behavior causally can often excuse this behavior, or at least reduce the degree of our culpability for it. A kind social scientist who helped us formulate these causal explanations would simply be aiding us in our search for an excuse. Just think of the imagined psychoanalysts and social workers who try to help the protagonists find excuses for their juvenile delinquency over the course of the song “Gee, Officer Krupke” from West Side Story (Sondheim Reference Sondheim1957/2010).

Respectful and Disrespectful Uses of Nonrespectful Causal Explanation

By offering us causal explanations of our behavior when we freely request them to pursue our chosen goals—in the West Side Story case, the goal of excusing our juvenile delinquency—social scientists recognize us as entities who act for reasons and in pursuit of goals. That is, they respect us as persons. This respect, however, is not part of the causal explanation itself. To the contrary, the same causal explanations can either help us better pursue our own goals or help others better manipulate us for purposes that we do not share. The same causal explanations of juvenile delinquency that were used to provide excuses for the miscreants in West Side Story could instead be used to develop brainwashing techniques to control the similar miscreants in A Clockwork Orange (Burgess Reference Burgess1962). In this way, a single causal explanation is capable of being used in both respectful and disrespectful ways.

While interpretive explanations can be used to manipulate others, their humanity is affirmed in the moment of interpretation even if it is denied later in the moment of manipulation. While the manipulation that follows may be morally objectionable, the interpretation itself is not. A causal explanation, by contrast, is silent on the personhood of those being explained, making its moral status depend entirely on how it is used. As such, causal explanations themselves are nonrespectful.

The fact that nonrespectful causal explanations can be used to demonstrate both respect and disrespect for subjects as persons depends on a certain duality in human existence. Even while acknowledging that we are parts of nature and hence liable to causal explanation, Kant famously argues that we must think about the actions and choices for which we hold ourselves responsible “under the idea of freedom” (Kant Reference Kant and Gregor1785/1996, 4:448, 95). While we may occasionally try to escape evaluation by thinking of ourselves as causally determined, this is not an attitude we can maintain consistently while still thinking of our decisions as our own, while still taking our agency and personhood seriously.

While we cannot consistently think of ourselves in this way, we can consistently choose to think of others this way. Reporting on discussions with social scientists committed to the exclusive use of causal explanations, Howard Becker reports that when these researchers switched from discussing the lives of others to discussing their own lives, their modes of explanation switched from structures and causes to values and reasons (Becker Reference Becker1998, 30). This asymmetry between researchers’ attitudes to themselves and their attitudes toward their subjects suggests a lack of reciprocity and mutual respect between them.

While mutual respect prohibits us from explaining others in exclusively causal ways, however, it does not require us to explain them in exclusively interpretive ways. Although we cannot think of ourselves except under the idea of freedom, we also cannot think of ourselves only under the idea of freedom, denying that we are, at least in part, material objects enmeshed in chains of causation linking us to other such objects. If we wish to avoid being overwhelmed by the causal mechanisms acting on these objects, we had best learn how to bring them under our agential control. In order to do so, we must alternate between causal and interpretive modes of self-explanation, learning how to manipulate our own matter causally to better achieve our own freely chosen goals. Reciprocity then requires us to help others do the same.

One of the best ways to recognize and respect the status of persons as such is to help them understand the causal mechanisms to which they are liable so that they can do something about them. In the Marxist tradition, for example, respect for the subjects of structural or other forms of causal explanation is shown, not in the moment when one formulates causal explanations, but when one later uses these explanations for purposes of consciousness raising, empowering subjects to seize control of the forces that would otherwise overwhelm their agency. The idea here is not to advance the well-being of those subject to causal forces paternalistically, but to recruit them in the effort to understand and control these forces themselves. Casual explanations are used respectfully when they are part of a larger project designed to foster self-determination.

The quest for self-determination explains the current fad for using social-scientific techniques to mine data about oneself in order to maximize one’s own health and productivity (see Tucker Reference Tucker2014). This practice, while it may be unappealing for many, is not a symptom of a lack of self-respect, let alone a lack of respect for others. More importantly, the demand for self-control also explains the continuing struggle by democratic polities to seize power over forces of domination and manipulation, the struggle to substitute collective self-determination for the control of some by others or by impersonal social structures.

How to Use Causal Explanation Respectfully

The respectful use of causal explanation requires that the explanation be part of a project willingly undertaken by subjects in the pursuit of goals that they freely embrace as their own. This necessitates bringing the goals of social scientists together with the goals of their subjects. Such harmonization of ends can occur through action on the part of the subject, the researcher, or both. Subjects may choose to embrace researchers’ preexisting goals, researchers may choose to embrace subjects’ preexisting goals, or both may modify the goals that they pursue in order to bring them together.

Standard practices of informed consent use the first of these three approaches. By providing potential subjects with all the information required to understand the risks and benefits of a given study, informed consent ensures that subjects understand the goals that the researcher is pursuing, how the methods of the study are a rational means of pursuing these goals, and why these goals are considered valuable. When this information is presented effectively, potential subjects act almost like interpretive social scientists studying the activity of the researcher, coming to grasp the researcher’s values and standards of rationality. By freely consenting to participate in a piece of research after coming to an interpretive understanding of it, subjects take on the researcher’s reasons as their own. In this way, the study becomes a joint project in pursuit of shared goals and, hence, a community of mutual respect.

With their myopic focus on signatures on informed consent forms, however, current research ethics regulations neglect the other ways in which the goals of researchers and their subjects can be brought together. Informed consent after a study has been designed, but before it commences, is not necessary to bring researchers and subjects together in a community of mutual respect. The work of harmonization can occur either before or after this particular stage of the research process. It can be part of the initial design of a study, when researchers successfully identify the goals of given set of subjects and determine what kinds of causal explanations would be helpful to them in pursuing their chosen ends. Harmonization can also occur over the course of a piece of research, as all those involved come to understand one another better and increasingly share each other’s goals as their own. Finally, harmonization can occur after research is completed and disseminated to its subjects, when it becomes clear exactly how research findings can empower subjects to better pursue their chosen ends.

These alternative paths to respectful causal explanation are a poor fit with current research ethics regulations, structured as they are around the preapproval of predesigned research. First, these rules make it difficult to collaborate with potential subjects in the initial design of social research. If subjects can only be recruited once all the potential risks and benefits of a piece of research have been identified, then these subjects are precluded from actively collaborating in designing the specific kinds of social science that they need to better pursue their chosen goals. Second, by forcing social scientists to commit to a particular research design in advance, existing regulations make it difficult to make significant changes to a study once it is already underway. As a rapport develops between researchers and subjects over the course of a study, it may become evident that the goals of each could be better aligned by undertaking a project quite different from that originally intended by the researcher.

While I am loath to grant more power to overbureaucratized ethics review boards, it would be useful if these boards placed less emphasis on obtaining signed consent forms at the beginning of a study and more of an emphasis on the holistic evaluation of the stance that researchers take toward their subjects throughout their research. Unless such changes are made, some paths to the harmonization of goals between researchers and their subjects will present greater obstacles than others. Under the status quo, ethics regulations are designed in a way that makes it easier for a social scientist to recruit subjects to adopt her goals than it is for her to learn about her subjects’ goals and adopt them as her own. The result is a great deal of research that is ethically unobjectionable, thanks to informed consent, but not particularly valuable for those being studied.

Existing regulations requiring preapproval of research also make it difficult to develop a case for the ethical permissibility of a given study based on knowledge that can only be acquired once the study is completed. Since we can never be entirely certain in advance whether a given piece of research will enable subjects to exercise greater self-control, we may only be able to evaluate the respectfulness of a given piece of social science retrospectively.

It would therefore be useful if ethics boards conducted retrospective as well as prospective reviews of causal research, including a review of the dissemination of a study’s findings after the research is completed. A failure to share findings with subjects is not disrespectful if it is done with their informed consent, but in other cases proper dissemination may make all the difference between a respectful and a disrespectful use of causal social science. A focus on dissemination strategies is therefore particularly important when research subjects are anonymous aggregates or entire polities, for whom standard informed consent requirements are not applicable. Even when informed consent is received, researchers can nonetheless reaffirm their respect by offering their findings to subjects in a form that they can understand and use for their own purposes. On the individual level, this suggests much more rigorous debriefing than most ethics boards currently impose, with significantly greater emphasis on making sure that subjects not only understand the research that has been conducted about them but also have the ability to put their newfound knowledge to use in advancing their self-determined goals. On the collective level, respecting one’s subjects may necessitate engagement with the wider political discourse of the societies one is studying, even blurring the line between pure and applied research where necessary to ensure that one’s findings are always tools for democratic self-control and never tools for nonconsensual manipulation of some by others.

In order to show respect for their subjects in any of these ways, causal social scientists will have to incorporate a degree of interpretive analysis of the values and practices of the societies they study, or at least collaborate with other scholars who do. Interpretive anthropologists, for example, could help causal social scientists studying those in cultures other than their own throughout the research process. Experts on the local culture could participate in research design by helping identify the kind of causal knowledge that those being studied would most value for their own purposes. They could then guide the process of informed consent, making sure that research subjects both understand and freely choose to participate in the research being conducted about them. Experts on cross-cultural understanding could also be present throughout the course of research, available to clear up any misunderstandings that might arise. Finally, such experts could help design culturally sensitive practices for the dissemination of research findings, communicating these findings in ways that relate to subjects’ worldviews and values as best as we can interpret them. If respect for subjects as persons is obligatory, then so too is such interdisciplinary collaboration.