2. The ethics of psychologizing

To psychologize an agent is to deny that the reasons the agent takes to be authoritative for them are in fact authoritative for them and to instead claim that there is an alternative psychological/causal explanation for their behavior. Psychologizing someone may well be illuminating: we might discover that the basis for our beliefs and actions is different than what we took them to be in a way that helps us arrive at a better self-understanding, for example. In this paper, I follow Flowerree in focusing on dismissive psychologizing.Footnote ³ When I dismissively psychologize, I reject an agent’s avowed reasons for belief and action as their actual reasons and interpret their actual reasons for belief and action as a product of alternative psychological/causal factors, where the agent would disavow this re-interpretation. When I psychologize in this sense, I endorse explanations of the following form: “‘You really believe that because_______’ or ‘You really did that because ______’ (fill in the blank: ‘you’re racist’, ‘sexist’, ‘chauvinist’, ‘hungry’, ‘homophobic’, ‘hormonal’, ‘biased’…)” etc. (968). As Flowerree’s examples indicate and as Smyth also argues, the alternative explanations need not involve an agent’s personal psychology. They may also involve references to larger social patterns or institutional contexts in which one is embedded. What matters is that they are causal explanations that conflict with the agent’s own view of their reasons. Flowerree is further interested in cases of psychologizing what she calls a “sincere” interlocutor. A sincere interlocutor takes their avowed reasons to be their actual reasons for belief. An insincere interlocutor, by contrast, “knowingly avows a reason for her action/attitude that is not her real reason” – a more Machiavellian figure that raises separate questions (971). For the remainder of this discussion, whenever I refer to “psychologizing,” I have in mind this specific form of dismissive psychologizing.

On Flowerree’s and Smyth’s accounts, psychologizing a sincere interlocutor can wrong and harm them. Smyth considers various problems with psychologizing: it can be ineffective for changing minds, degrade the character of those who reflexively psychologize, and contribute to social fragmentation and warfare. I will be most concerned here with what Flowerree takes to be the underlying harm of premature, dismissive psychologizing – that it fails to treat an individual with the respect they are due “as a rational agent – one whose acts and attitudes are governed by reasons, and who is equal to you as a fellow rational agent” (Flowerree Reference Flowerree2023: 972). This disrespect seems best read as a form of dehumanization – a point both Flowerree and Smyth emphasize. By dismissing the sincere agent’s avowed reasons and reinterpreting their beliefs through a primarily causal or psychological lens that they would disavow, we fail to engage their characteristically human capacity of giving and asking for normative reasons.Footnote ⁴ To fail to engage this capacity when we should therefore means we fail to respect their humanity.

It is important to distinguish between a private-facing judgment we might make that psychologizing another agent is apt versus a public-facing assertion of this psychologizing. Flowerree believes that both private and public psychologizing have the potential to wrong and harm. I do not see the harm of private psychologizing, but I will not litigate this point here. I will focus on what I take to be the more plausible wrong of public, premature dismissive psychologizing.

Flowerree further points out that when we carry out this kind of psychologizing publicly, it “has the effect of shutting the other out of the conversation…. We make it difficult (or impossible) for the other person to have uptake in the conversation” (972). When we psychologize, the agent’s avowed reasons do “not even come up for consideration. Their reasons are silenced from the conversational context” (972). Of course heinous or egregiously misguided reasons should not receive uptake in any conversational context. But that is because they “ought to be rebutted or defeated” – i.e., they ought to be given uptake as reasons, but then repudiated as the bad reasons they are (972). When we psychologize, however, “the reason does not even come up for consideration” (972). Flowerree also emphasizes that this form of psychologizing risks exposing the agent to “criticism, censure, and belittling” (972). Because of the potential for these harms, Flowerree’s view is that “our default should be to accept the reasons our sincere interlocutor offers as genuine for her,” unless certain defeating conditions arise, which I will discuss shortly (970).

The most striking feature of Flowerree’s account is that we should not psychologize interlocutors when we merely suspect them of being self-deceived. Instead, we should psychologize only if we are in a position to know that they are self-deceived; otherwise, we enact the panoply of harms listed above. Flowerree asks us to consider a particularly salient case for our current political moment called “Xenophobia.” A friend of ours is an avid supporter of a politician we despise because of their xenophobic views and policies. Our friend says they support this politician because of their economic and education platforms and dismisses the civil rights reasons we cite: “‘That’s just talk,’ she says, ‘he doesn’t mean that anyway’. As the conversation continues, [we] grow suspicious. Could it be that, despite her sincere assertions, she actually supports the politician because she is actually racist and xenophobic?” (976).

For Flowerree, while we nonetheless “suspect [our] friend is xenophobic, [we] ought to set that aside and engage with her reasons directly,” or else we harm them (980). In fact, because this case concerns political issues, Flowerree warns that the harms of psychologizing are potentially far worse: “[P]sychologizing explanations at work in political discourse are insulting in the extreme. They involve explanations that make the opponent a moral monster…[C]harging another with being in bad faith – if the opponent is in fact not in bad faith – is disrespectful, dehumanizing, and wrong” (980). Smyth makes a similar point: “Interpersonal psychological debunking is usually morally problematic because it is a kind of shaming or verbal humiliation. It is painful to think that we are bad people or unreliable moralizers, or even just to realize that others think this way of us” (Reference Smyth2022, 259). It feeds directly into the drive to “demonize and dehumanize” political opponents (267–8). It is not hard to see the more wide-ranging implications of such arguments. Dismissive psychologizing in political contexts has been ubiquitous post-2016, whether carried out by politicians, political parties, media, academics, and of course interpersonally in our everyday lives. Yet if Flowerree and Smyth are right, much of this psychologizing is morally wrong and harmful.

Now, Flowerree or Smyth do not think psychologizing is never apt. Smyth argues that while our “epistemic standards for debunking” are typically far too low – providing “no direct evidence” for our claims and instead relying on an intuitive sense of “obviousness” – such psychologizing can be appropriate when we are “reasonably sure of the real causal history behind” the target agent’s “expressions and of the evaluation [we are] making of that history” (260 and 261). Flowerree sets the bar higher, and I will again focus on her formulation. For Flowerree, it is permissible to psychologize when we are in a position to know that our interlocutor is self-deceived. What does the “position to know” consist in? Flowerree gives two paradigm cases: a therapist who has worked extensively with a patient and a long-term romantic relationship. In these cases, “the key lies in the reliability of the practice of applying the psychologizing explanation” (976). In the long-term relationship, there is “an intimate interpersonal history, one that involves many instances of caring for the other person, and finely tuned knowledge of the partner’s reactions” (976). The therapist similarly “listens carefully, and draws on her knowledge of psychology to come to understand [the patient’s] true motivations,” and the therapist only permissibly psychologizes after many sessions of this rigorous work (974). Being in a position to know that an agent is self-deceived therefore involves a very high bar for knowledge.Footnote ⁵ This is why Flowerree insists that “knowing that someone is self-deceived is rare. More likely we suspect it. And if we merely suspect they are self-deceived, we risk wronging them by getting it wrong” (976). In Xenophobia, for example, it is enough that we “cannot rule…out” that our friend’s reasons are her genuine reasons for supporting the xenophobic politician that we count as lacking knowledge that she is self-deceived. Moreover, we lack the necessary “insight into her that would” give us this knowledge (978). Because there is a great deal of disagreement among epistemologists over how good our evidence needs to be to generate knowledge and to indicate the very high bar Flowerree has in mind, let’s call it therapeutic knowledge that is required for permissible psychologizing. One is in this position when one has long-term, intimate interpersonal contact with the agent involving careful and accurate observation of their behavior.Footnote ⁶

In the next section, I will take for granted the harms that are claimed to result from the public, dismissive psychologizing of a sincere speaker when one is not in a position to secure this knowledge. But I will argue that even if these harms occur, we should reject the notion that it is wrong to dismissively psychologize others. Moreover, I will defend the first plank of my own principle – that it is not just permissible to engage in this psychologizing of a specific class of agents, but that we should in fact default to psychologizing them.

3. Psychologizing and formal authority

In this section, I present two cases where we face a decision about whether to psychologize. If we do psychologize, we might wrong the target, but, in psychologizing, we can also prevent them from harming others. If we instead refrain from psychologizing, we will not wrong the target, but we may well end up harming others.Footnote ⁷ Here is the first case:

Note that Rachel accepts that the CEO is not insincere in Flowerree’s sense – i.e., that the CEO is not lying about his reasons. But Rachel is nonetheless not in a position to know that the CEO is self-deceived given the therapeutic knowledge Flowerree requires. For one, Rachel has far less interpersonal familiarity with the CEO than we have of our friend in Xenophobia. In Xenophobia, despite the interlocutor being an old friend, Flowerree claims that we are still not in a position to know whether they are ignorant or self-deceived and therefore “must wait to psychologize. You must take her reasons seriously and provide rebuttal,” or else we wrong her (980). Rachel is far from having the familiarity a therapist or long-term romantic partner would have with the CEO’s mental and behavioral lives. It follows, then, that on Flowerree’s account, Rachel wrongs the CEO by psychologizing him when she is not in a position to know that he is in fact self-deceived. The CEO is disrespected and dehumanized, his reasons silenced, and he is made vulnerable to criticism.

Grant that the CEO is harmed in precisely these ways. In fact, the situation is even worse. We have seen that Rachel is not alone in her judgment; much of the town reasons in just the same way she does and does so publicly. Any harms the CEO experiences as a result of this psychologizing are therefore multiplied many times over. As Smyth emphasizes, “this form of humiliation acquires more and more power as the debunking audience grows” (Reference Smyth2022: 259). But we should now consider a question that Flowerree does not address: are there harms of not psychologizing the target agent, of making the decision to take their reasons at face value? After all, it is not enough to know that there are harms of psychologizing; we need to know if these harms outweigh any harms that might come from failing to psychologize.

Suppose, then, Rachel and the other townspeople decide not to psychologize the CEO and the CEO is in fact self-deceived. In this world, in eschewing psychologizing, we take the CEO’s avowed reasons at face value. Because we adopt this stance, we focus our attention on rebutting his avowed reasons, just as Flowerree advises. For example, we continue to explain why there is toxic waste being released from the factories and why this waste is harmful, and we keep doing so as long as the CEO argues otherwise (which he continues to do). Because we are treating the CEO’s avowed reasons as his genuine reasons – treating him as if he is in fact in the game of giving and asking for normative reasons – then the fitting response is to show that those reasons are mistaken (if we think they in fact are). Presumably, because the CEO’s beliefs and actions trace back to his avowed reasons, he will change his view once those reasons are corrected.

But, in this world, the CEO is self-deceived. In taking his avowed reasons as his genuine reasons, we are therefore prevented from grasping what is actually behind the CEO’s views and the corporation’s actions. The driving reason behind his beliefs and actions is the desire to maximize profit regardless of human and environmental toll, not the considerations he invokes. In refusing to psychologize, we are therefore in the disastrous position of never identifying the root cause of the injustices the town is suffering. And without a grasp of the underlying reasons and causes, we will struggle to address this ongoing injustice or take the necessary steps to prevent the same injustices in future. For example, it will make less sense to pursue policy that counters or limits corporate wealth or power or a policy that takes the entity in question out of the private sector altogether and makes it public (if it provides a public service) because we will not identify the for-profit structure as the decisive factor in generating this state of affairs.

The harms of failing to psychologize the CEO in Toxic Waste are far weightier than the harms to the CEO’s dignity that result from prematurely psychologizing. First, the harms of failing to psychologize and, as a result, failing to identify the root cause of the injustice implicate many more people: all of the people whose lives are destroyed by the corporation’s actions, rather than harms to a single individual. Second, the harms do not have the abstract, difficult-to-quantify character of dignitary harms; they are physically devastating, potentially lethal impacts to scores of innocent people. Third, the dignitary harms are, at best, a wash: the corporation’s actions threaten the dignity of the town’s residents as well: these individuals are viewed and treated as disposable in the pursuit of profit, rather than possessing a dignity that demands respect.

And insofar as Toxic Waste is a familiar case of corporate malfeasance, and we are in the same position of lacking therapeutic knowledge of the relevant actors in most if not all cases of this kind, then Flowerree’s account will demand that we not psychologize in these other cases either. The same harms will therefore arise. As long as the corporate actors produce similar reasons to those in Toxic Waste and as long as we cannot psychologize them, then it will be far more difficult to address the root cause of this corporate wrongdoing and prevent its future occurrence, i.e., the profit motive that governs these institutions and that must be reckoned with to prevent these harms. The scale of the harms are therefore no longer confined to a single case; they increase exponentially. Given these considerations, it seems clear that the harms of psychologizing Flowerree cites are outweighed by the harms of not psychologizing.

A proponent of Flowerree’s account might nonetheless respond in two ways. First, they might argue that while these potential harms are clearly weighty, they need to be appropriately contextualized. I am assuming in the above scenario that the CEO is self-deceived. But this is not guaranteed. Rachel and others are not in the therapeutic position to know that the CEO is self-deceived, which means the above harms are speculative. If, however, the CEO is not self-deceived, then those harms do not follow. Furthermore, if the CEO’s avowed reasons are in fact their genuine reasons, then correcting those reasons is the appropriate and likely most effective response. Because Rachel and others are not in a position to know one way or the other, then they should err on the side of not psychologizing given the potential harms of doing so, just as we should err on the side of not psychologizing our friend in Xenophobia.

This reply, however, is unconvincing. While the harms I cite are speculative, so are Flowerree’s. We both accept that Rachel and others are not in a position of possessing therapeutic knowledge of the CEO. It is possible that the CEO is self-deceived, and it is possible that he isn’t. We therefore have to weigh the potential wrongs and harms of both possibilities: the harms of psychologizing when he is not self-deceived versus the harms of not psychologizing when he is. Only then can we decide which stance to adopt. And it seems clear that the potential harms of not psychologizing are far more substantial when the CEO is self-deceived than those of psychologizing when the CEO’s reasons are his actual reasons.

A second response might be that I have unfairly stacked the deck with my example: the target of Rachel’s psychologizing is unsympathetic, and the scale and nature of the harms of his actions are overwhelming. But if we make the target of psychologizing less cartoonish and the scale and nature of the harms less pronounced, my conclusion looks less tempting.

To properly answer this second reply, I need to show that the harms in Toxic Waste are not just a product of its idiosyncratic details. This is in fact the case: Toxic Waste contains an asymmetry that non-contingently generates less harm for the target agent of the psychologizing and greater harm to those affected by this agent. The asymmetry is that the target agent (i.e., here, the CEO) occupies a formal position of authority, and he is being psychologized by people who do not occupy such positions (i.e., the townspeople). While we can generate other cases where the target agent may be more sympathetic and their action’s consequences less shocking, any iteration of this asymmetry in formal authority will very likely involve the same disproportionate distribution of harms.

What do I mean by “formal position of authority”? I mean a position where one has explicitly articulated duties, abilities, and entitlements as an officeholder within an institution and may be able to deploy resources and/or establish rules for that institution and any domains it governs.Footnote ⁸ Paradigmatic examples of formal positions of authority will include politicians, roles within economic structures such as corporations or treasury departments (e.g., CEOs, treasury secretary, etc.), judges, and military, intelligence, and police officials. I specifically have in mind formal positions of authority within institutions that have among the most substantial economic and political power in the society in question. The nature of formal positions of authority is such that agents who occupy these positions are able to cause comparatively more harm than those who do not. In virtue of their formal positions, agents have various powers and access to resources, as well as the entitlement to utilize these powers and resources, that non-office holders do not have. With this unique authority comes a comparatively far greater ability to cause harm. Agents without formal authority may contingently amass a great deal of power and resources, but they will not have done so in virtue of being non-formally authoritative.

At one point in his (2022), Smyth implicitly relies on this distinction. He distinguishes between a problematic case – reflexive psychologizing of student protestors as merely anger-fueled “snowflakes” – with the appropriate psychologizing of “the actions or utterances of a putatively racist judge” where we examine their conviction record:

I agree with Smyth, but we need to make explicit the relevant distinction at play. The judgment that the moral hazards of psychologizing the judge are easily outweighed by the potential harm of failing to psychologize follows directly from the formal authority the judge possesses. Conversely, the harms of this psychologizing are not nearly as easily outweighed when we target the student protestors precisely because they lack any such authority. To draw out the importance of this authority asymmetry, consider a further case:

Again, in this case, the interviewer and the members of the political community lack therapeutic knowledge of the politician’s motives. Perhaps they are wrong, and the politician is not self-deceived. Perhaps he really did change his mind for the reasons he gives and not because of the donations. If this turns out to be the case, then they will have harmed him in the ways Flowerree identifies (disrespect, silencing his reasons, and vulnerability to criticism).

But, again, this is not the only relevant possibility. We also have to consider what would happen if we fail to psychologize the politician when he is in fact self-deceived. In this world, where we take the politician’s reasons at face value, we never identify either the corruption or its source. And without identifying the actual reasons behind the change in the politician’s platform, it will be far more difficult to address this corruption in this or any future case. For example, it will make less sense to pursue more stringent campaign finance regulation to thwart the influence of these donations or pursue other measures such as the public funding of elections because we will take ourselves to have less evidence of corruption.

The results are devastating: the democratic process will have been undercut because the politician is not answerable to the will of the people they represent. And these harms will once again generalize to any similar cases, i.e., ones where the politician is self-deceived about their corruption. Given that we cannot identify this corruption when we fail to psychologize in such cases, we will also struggle to develop a meaningful solution for stopping it in future. Furthermore, if one agrees that the original policies that challenged corporate power were good and the revised policies that fail to do so are bad, then there are the harms that follow from the actual shift in policy as well – harms to those who are denied essential, life-saving medical care because they cannot afford it, harms to those whose livelihoods are destroyed by unchecked financial speculation, existential harms to humanity and the environment, etc. Of course, the latter harms depend on your first-order political views. You may well disagree with all of these claims. In that case, you can simply reverse Donations to have the politician shift from supporting policies you endorse to those you do not and similarly stipulate that this shift occurs because of corrupting influences concerning which the politician is self-deceived. Wide-scale, concrete harm aligning with your political commitments will follow in turn.

This distribution of harms directly mirrors those in Toxic Waste. This is again because of the politician’s powerful formal position of authority: their capacity for harm implicates more people and involves concrete, devastating impacts on individuals that are a direct result of this authority (i.e. passing policy that will directly affect the material conditions of people’s lives).Footnote ⁹ The world in which the politician is self-deceived and nonetheless psychologized involves harms only to a single individual and of a dignitary variety that are much more difficult to assess (and are, as before, a wash at best). Which of the two options minimizes harm overall seems clear.Footnote ¹⁰

But there is yet another relevant dimension for assessing cases with this asymmetry. So far, I have written as if, because we lack therapeutic knowledge of the target agent, there is roughly an even possibility that the target either is or is not self-deceived. But our background knowledge of many powerful formal institutions should in fact lead us to increase our confidence in self-deception verdicts. Consider Toxic Waste again. Publicly known examples of corporate greed resulting in atrocities are legion – to the point that such events are now standard Hollywood fare. There are also many well-known cases of corporations trumpeting their commitment to ethical conduct but nonetheless continuing to carry out these harms. In addition to specific examples, there is the straightforward structural explanation for this connection between corporations and wrongdoing. The goal of a corporation and the CEO is to maximize profits. If there is a conflict between the morally right decision and maximizing profit, profits are very likely to come first. This structural fact and these examples all give us excellent reasons to think that even if the CEO is sincere, they are likely deluded about their reasons. Such considerations require little specialized knowledge beyond a passing familiarity with the status quo or the history of how corporations behave. They are therefore in principle accessible to anyone deciding whether to psychologize a given agent. Similar considerations hold for politicians. It is not difficult to find examples of corrupt politicians who also vehemently deny this corruption. The structural conditions generating this corruption are straightforward: the politician’s role in generating and approving policy that directly affects the fortunes of special interests obviously gives those interests enormous incentive to influence politicians. In the case of psychologizing the formally authoritative, it is therefore not necessarily the case that this interpretation “is rarely accompanied by the kind of evidence that can ground justified belief in its causal hypotheses or its evaluative judgments,” as Smyth claims (Reference Smyth2022: 264).

Now, none of this is decisive in assessing either the CEO’s or politician’s psychology. Perhaps they really do believe that the company has not done anything wrong, that their changes in policy preferences are a result of responsiveness to the evidence, and that their reasons have nothing to do with their self-interest. It is certainly possible. But it is not likely. This is because it cuts against what we know about these institutions and how they function. While we are therefore rarely in a position to have therapeutic knowledge of such actors, we can nonetheless have more confidence in a judgment that they are self-deceived than the verdict that their avowed reasons are in fact their genuine reasons.Footnote ¹¹

At this point, Flowerree might argue that her account can in fact handle these possibilities because the most important consideration for being in a position to know that someone is self-deceived is whether the psychologizing explanation is reliable. Psychologizing is permissible for the therapist and the long-term romantic partner because their explanations stem from reliable mechanisms sufficient for knowledge: their honed, truth-tracking abilities accurately detect when the target agent is driven by something other than their avowed reasons. Flowerree might argue, then, that given that the psychologizing explanations in Toxic Waste and Donations are similarly reliable – for precisely the reasons I have given (i.e. the well-known corruption of these institutions) – then we are in a position to know that the CEO and politician (and any similar actors) are self-deceived. In fact, Flowerree herself seems to back away from the therapeutic model at one point and advert to reliability alone: “Other sources – such as an immediate awareness or a gut instinct – could put one in a position to know if they are sufficiently reliable. The exact epistemological story will depend on which indicators reliably reveal someone’s motivations” (978).

A more detailed account of this reliabilism may offer a path for the Flowerree-Smyth line, but I remain skeptical. There is a tension between invoking these reliable mechanisms and the paradigm cases of “being in a position to know” Flowerree centers. The latter involve extensive, self-conscious development of one’s truth-tracking abilities: the therapist and romantic partner are sufficiently reliable in large part because of their substantial exposure to and reflections on and refining of their understanding of the target agent’s behavior. It is this high epistemic bar for permissible psychologizing that makes Flowerree’s account so interesting: if we need to be in this sort of rarefied position, then a great deal of our psychologizing practices will be problematic – which is why Flowerree insists that being in this position is “rare.” But if proponents lean into a reliabilism where this long-term reflection is not necessary – if, for example, very general knowledge of formal institutional corruption or even attuned gut instinct can suffice – then it becomes much less clear why routine psychologizing practices are problematic.

For one, such knowledge does not require any contact with the psychologized agent, let alone the substantial interpersonal interaction that the paradigm cases evince. Flowerree might bite the bullet on this point and accept that the paradigm cases are misleading. But the problem goes deeper still. In this revised light, being in the relevant epistemic position is no longer an exceptional state since it is much easier to attain. For example, in Toxic Waste and Donation, if those carrying out the psychologizing count as utilizing sufficiently reliable mechanisms, then permissible psychologizing will hardly be a rare phenomenon; these types of cases are ubiquitous in our political and economic lives. And suppose the empirical literature on motivated cognition that I referenced at the start is right. It then seems plausible that many of our psychologizing practices may already be linked to reliable detection mechanisms insofar as self-deception would be rampant (and reliabilism would not require conscious access to this evidence on the psychologizer’s part, only an appropriate causal link to it).Footnote ¹² Of course, a great deal depends on how the reliabilism is developed and the empirical details and reality of these mechanisms. But whatever such an account looks like, further embracing reliability will both push us away from the paradigm cases of therapeutic knowledge and threaten the striking verdict that permissible psychologizing is rare.Footnote ¹³

In separate, recent work, Flowerree has in fact noted the value of considerations relating to power and third-party harm. She considers a principle of charity that “enjoins us to refrain from endorsing the interpretation that attributes ill will” to others and argues the principle should be rejected because it can prove dangerous in non-ideal contexts of oppression (Reference Flowerree and Timmons2025, 101). She gives the example of the target of our interpretation being “a powerful person” where their “bad will has the propensity to cause harm”; in this case, “it is important to consider the harm that could be done to others by overlooking a bad will” (106). This seems exactly right. But note that Flowerree limits the relevance of these considerations to attributions of “ill will” and other disparaging agential interpretations. On this view, the potential for wronging a powerful individual by attributing to them ill intent is typically outweighed by harms others may endure if we fail to make this attribution. But the same considerations are not taken to apply to psychologizing interpretations. An interpretation that says this figure is unaware of the actual reasons driving their beliefs and actions would apparently wrong them in a way where these same considerations do not come into play – perhaps because an attribution of ill will at least treats an individual as an agent, rather than viewing them through a causal lens, and the latter is simply a bridge too far. But I see no reason to draw this kind of line. The harms to third parties implicated by our interpretation of others are a relevant consideration whether our interpretation attributes ill will or dismissively psychologizes.

Moreover, it is arguably more important to factor in these considerations when deciding whether to psychologize: few individuals, even among those in powerful formal positions of authority, have straightforward “ill will.” While some of course match this description, many assume that they are doing what is morally right and that they believe and act according to the reasons they give. They would reject the notion that their actual reasons for belief and action are a product of the larger material forces that shape their formal authority. Causal and psychologizing explanations are therefore especially important because they draw our attention to the larger structural forces responsible for the majority of harms, while attributions of ill will keep us focused on – and perhaps distracted by – individual bad actors. But, either way, whether we are deciding to attribute ill intent or to psychologize, considerations of power and harm to others should play a crucial role in our interpretations.

There remains, however, a further way to push back here. It might be argued that my cases are still simply too easy – that the real world is far messier than my toy examples imply, leaving little room for concrete applications of my principle. In the next section, I respond to these concerns by moving even farther from the realm of the hypothetical. I discuss a real-world case where agents seem to currently operate with a version of Flowerree’s principle and where doing so leads not only to the harms discussed so far, but particularly pernicious forms of propaganda.

4. Psychologizing formal authority in the real-world

The case concerns crime reporting in US media, and I turn here to the work of Alec Karaktasanis – civil rights attorney, activist, and media critic who has meticulously documented the media’s uncritical treatment of politicians and police officials’ stated reasons for the policies they advance. For example, it is regularly reported without further comment that political and police officials are implementing archetypal War-on-Drugs policy – increasing police, prosecutors, and incarceration – in order to protect public health and address the harms of prohibited drug use. Associated Press: “Lawmakers in multiple states across the country are looking to increase fentanyl penalties in an effort to combat what has been called the deadliest overdose crisis in U.S. history” (Chandler Reference Chandler2023 ); ABC News: “Senior government officials are concerned that Congress may not pass enough funding to secure the border, as federal authorities work under immense pressure to stop illegal narcotics smuggling” (Owen, Villareal, Scholz Reference Owen, Villareal and Scholz2023); The New York Times: “‘We need more tools – legislation to modernize our customs laws and enhance penalties – and resources such as more officers, more technology, and more money to fuel our fight,” Customs and Border Protection’s commissioner, Troy A. Miller, said in a statement’” (Kitroeff and Gebeloff Reference Kitroeff and Gebeloff2024).

But social scientific evidence points overwhelmingly to the failure of such policies to meaningfully address these issues (Wood et al. Reference Wood, Werb, Marshall, Montaner and Kerr2009; Provost Reference Provost2024; Mofokeng Reference Mofokeng2024). In fact, it is well documented that these policies make these problems worse: use of prohibited drugs has only increased, as has their potency. The harms of the forty-year War-on-Drugs are also well known: explosions in incarceration and attendant harms for the millions convicted and their families (loss of housing, employment, benefits, etc.), trillions in dollars spent, the slaughter of hundreds of thousands of individuals abroad, among many others (Karakatsanis Reference Karakatsanis2025).

The standard avowed reasons of politicians and police for pursuing these policies are therefore entirely at odds with the evidence. How, then, should we address this discrepancy? On Flowerree’s approach, because we lack therapeutic knowledge of these officials, it would be wrong to psychologize them, and we should therefore view and report these reasons as their genuine reasons. Flowerree might interject that her account does not preclude reporters from discussing the above facts. In fact, she might argue that this is a prime example of how we should take agents’ reasoning at face value precisely in order to rebut it. Refraining from psychologizing certainly does not require adopting an interlocutor’s reasons.

But this approach – of accepting these reasons in order to rebut them – is itself a part of the problem. Karakatsanis points out that because the reasoning of politicians and police is taken at face value, treated in the very best case scenario as wrongheaded but still genuine, the actual primary causes of these disastrous policies are rarely, if ever considered:

This real-world failure to psychologize therefore leaves audiences with, at best, a fundamentally warped understanding of drug and crime policy, where the message relayed is that politicians and police pursue these policies because they have a noble aim of protecting public health and addressing the harms of prohibited drug use (despite all evidence to the contrary); their actions are not a product of the causal factors that much more plausibly drive their decisions.

In addition to this first-order misinformation, Karakatsanis notes three more general features of this reporting that have insidious propagandistic effects that are worth emphasizing. The first is that the media’s failure to dismissively psychologize leads to a dangerous mythologizing of these officials. On a non-psychologizing approach, officials are portrayed as well-intentioned figures, genuinely concerned with the public good who are therefore, at worst, simply misguided. Because they are not psychologized, their flawed reasoning – if it is described as such – is not treated as a product of corruption, except in rare cases. But in a world where special interests do play a decisive role in shaping our politics, this portrayal misleads readers about their political reality. A second effect is that readers are left with the false impression that policy decisions are the product of individual actors’ decision-making rather than larger structural conditions. From this reporting, readers have the impression that the decisive – even the only – factor in shaping policy is the reasoning given by the officials, which is not the case. Powerful interests implicated by policy play an enormous role in shaping why the policy exists, what it looks like, and how it will be enacted. This entire process, however, is obscured if we assume that the reasons of individual officials are the primary factor in determining policy. Even, then, if we refrain from psychologizing and nonetheless correctly judge in cases such as Toxic Waste that the company wrongfully harmed the community, we still do not identify the material interests and forces that cause these sorts of cases because we identify the culprit as individual actors’ bad reasoning, in turn hampering efforts to locate and address the actual root causes. A third propagandistic effect of non-psychologizing is that readers end up with a misguided notion of how the problem of mitigating the harms of drug use should be addressed. If officials’ reasoning is the problem, then the appropriate response is to correct that reasoning through argument and debate. If, however, the primary problem is officials’ accountability to corrupting interests and their avowed reasons are, if not simply a cynical ploy, a product of self-deception, correcting their reasoning will be largely pointless. Challenging those interests directly and removing their influence from our politics becomes the fitting response.

I said that reporters often operate with a Flowerree-esque principle. Media scholars have shown that mainstream media consistently relay the statements of government sources uncritically – that, as Michael Schudson puts it, “[t]he press as a whole during Watergate was, as before and since, primarily an establishment institution with few ambitions to rock establishment boats” (Reference Schudson1992: 123). Taking the reasons of officials at face value is not only a matter of practice, but often stems from a deeper commitment concerning the inappropriateness of questioning the aims of US state actors: “The media may feel obligated to carry extremely dubious stories and mute criticism in order not to offend their sources and disturb a close relationship” (Herman and Chomsky Reference Herman and Chomsky1988: 22). The economic and social proximity of reporters to the officials they cover contributes to this problem, as do the larger structural conditions of media ownership that facilitate this proximity (Herman and Chomsky Reference Herman and Chomsky1988; McChesney Reference McChesney2015; Pickard Reference Pickard2020).

The media’s failure to consistently psychologize the formally authoritative is morally and politically destructive, generating and spreading harmful propaganda on issues of enormous importance. It is worth noting that this type of propaganda is particularly insidious. It is not just an individual piece of propaganda – a political message or commercial advertisement, for example. Institutions and their actors that constitutively produce propaganda (such as government officials and the police) are treated by a default non-psychologizing principle as fundamentally animated by the reasons they give rather than ulterior motives. Propaganda is therefore not just misidentified here; the very producers of propaganda are not – and cannot – be identified as such.Footnote ¹⁴

To round out this section, I want to include a final point of clarification on this first plank of my principle that we should default to psychologizing those in powerful formal positions of authority. Although the default I advocate for is the opposite of Flowerree’s, it too has defeating conditions. We should default to psychologizing these agents unless we have good, positive reasons for believing the institution they inhabit is trustworthy. What makes an institution trustworthy is, of course, controversial (Weinstock Reference Weinstock, Archard, Deveaux, Mason and Weinstock2013; Hawley Reference Hawley, Faulkner and Simpson2017). But specific criteria for navigating these judgments are not hard to come by. Relevant considerations might include: the institution takes effective steps to combat corruption and special interests from influencing its work or members; it follows expert consensus; it includes meaningful public input on matters of public concern; it subjects its own procedures and policies to routine review and reform; it has robust avenues for whistleblowing; it has a track record of success in its domain, etc. If an institution has proved itself on some or all of these scores over time, then we have good, positive reasons for believing the institution and its members are trustworthy, and we should therefore opt to take their reasons at face value, all else being equal. The risks of the harms of failing to psychologize are minimized in these cases: it is far less likely that the institution is suffering from corruption or that its members are systematically deceived. Making these judgments is of course difficult in many cases, and institutions may meet some of these criteria without meeting others. But that we can make such judgments of trustworthiness and that they can be better or worse should be granted. When these conditions obtain, it will make sense to eschew dismissive psychologizing of the formally authoritative. But, at the risk of understatement, it is far from clear how many formal institutions in our societies meet these criteria. As such, the tool of psychologizing remains not only permissible to deploy, but recommended as a crucial check against potential harms.Footnote ¹⁵

5. Psychologizing the non-formally authoritative

In this final section, I defend the second plank of my principle – that we should be comparatively more hesitant to psychologize those who lack powerful formal positions of authority than those who hold these positions. In one sense, a defense of this plank simply falls out of the arguments I have already given. Non-formally powerful figures do not have the same capacity for harm and are not occupationally bound to institutions with track records of corruption. The harms they experience from premature psychologizing are therefore not as obviously outweighed by the harms of failing to psychologize. But here I want to directly articulate the case for why psychologizing the non-formally powerful – that is, psychologizing ordinary individuals – does pose a risk of encouraging forms of political disenfranchisement, denying individuals full-fledged, equal political participation because they are taken to lack the necessary rational capacity for this participation.

First, though, some qualifications. Earlier, I glossed Flowerree’s account of the harm of disrespect that follows from psychologizing by claiming that it consists in dehumanizing one’s target: we do not recognize a role for the characteristically human capacity of giving and asking for normative reasons in guiding the agent’s behavior in the relevant context. Of course, as Flowerree herself acknowledges with the defeating conditions, this lack of recognition need not be nefarious. When my long-term partner is well aware that I become curmudgeonly when I have not slept well, it is permissible for them to point out that my current frustration has more to do with my lack of sleep than my avowed reasons.

But an important worry also arises at this point. There is an enormous psychological and social scientific literature pointing to rampant and systematic human epistemic irrationality and rationalization – a literature that often relies on the very psychologizing explanations Flowerree and Smyth raise concerns about.Footnote ¹⁶ It might be argued, then, that routine psychologizing is permissible simply because it accurately represents human behavior. Furthermore, by focusing on the morality of psychologizing explanations rather than its empirical credentials, we seem to be pursuing a non-sequitur: an armchair moral objection to what is a question of empirical truth.

It would be a disastrous move, in my view, to deny or attempt to overturn this empirical work. It is therefore important to clarify the precise nature of my claim before delving into the specifics of this second plank of my principle. First, to recognize that there is a certain moral risk to psychologizing is compatible with thinking that this is a risk that is very often worth taking or that it is a risk whose importance and severity vary substantially with context. Second, my second plank is weaker than Flowerree’s or Smyth’s. It says that there is a certain moral and political risk to premature, public dismissive psychologizing of non-formally authoritative individuals, but I remain agnostic on whether a default of not psychologizing is called for (as, for example, Flowerree wants). How much weight this moral and political risk should have in any particular case or for any particular project is not something I seek to dictate. With these clarifications in mind, I turn now to the distinct risk of political disenfranchisement that can follow from the routine psychologizing of the non-formally authoritative.

When we take a historical perspective on this practice, it is striking the degree to which denying a guiding role for an agent’s avowed reasons plays a role in rationalizing injustice and atrocity. Representations of groups as failing to possess a genuine capacity for rationality and instead being driven by “subrational” impulses that are claimed to be the product of scientific investigation have historically been used to rationalize slavery, the denial of women’s suffrage, and colonial and imperial projects:

• - From the pro-slavery volume Cotton is King: Abolitionists “seem to have overlooked the fact, that the Africans, of savage origin, could not be elevated at once to an equality with the American people, by the mere force of legal enactments. More than this was needed, for their elevation, as all are now, reluctantly, compelled to acknowledge. Emancipation, unaccompanied by the means of intellectual and moral culture, is of but little value. The savage, liberated from bondage, is a savage still” (Christy Reference Christy1856: 45).

• - Senator George Vest on women’s suffrage: “The great evil in this country to-day is in emotional suffrage. The great danger to-day is in excitable suffrage. If the voters of this country could think always coolly, and if they could deliberate, if they could go by judgment and not by passion, our institutions would survive forever…Women are essentially emotional. It is no disparagement to them they are so…What we want in this country is to avoid emotional suffrage, and what we need is to put more logic into public affairs and less feeling” (Blair et al. Reference Blair, Brown, Dolph, Vest and Hoar1887: 43–44).

• - Lord Cromer, England’s colonial administrator in Egypt, on “the Oriental mind”: “Want of accuracy, which easily degenerate into untruthfulness, is in fact the main characteristic of the Oriental mind. The European is a close reasoner; his statement of fact devoid of any ambiguity; he is a natural logician, albeit he may not have studied logic; he is by nature sceptical and requires proof before he can accept the truth of any proposition; his trained intelligence works like a piece of mechanism. The mind of the Oriental, on the other hand, like his picturesque streets is eminently wanting in symmetry. His reasoning is of the most slipshod description…[Arabs today] are singularly deficient in the logical faculty. They are often incapable of drawing the most obvious conclusions from any simple premises of which they may admit the truth. Endeavor to elicit a plain statement of facts from any ordinary Egyptian. His explanation will generally be lengthy, and wanting in lucidity. He will probably contradict himself half-a-dozen times before he has finished his story. He will often break down under the mildest process of cross-examination” (Cromer Reference Cromer1916: 146; cited in Said Reference Said1979: 38).

Recent work in political philosophy on colonialism echoes the role of this kind of dismissive psychologizing in facilitating injustice: “[T]he fundamental wrong of colonialism is treating normal adult human beings as inferior in the sense of assuming them to be unfit to manage their own affairs, treating them as if they were minors or mentally incompetent” (Agrawal and Buchanan Reference Agrawal and Buchanan2025: 2). This type of psychologizing is also not a relic of the distant past. Consider, for example, the ongoing US War on Terror, where the image of Arabs and Muslims as characteristically paranoid, violent, backward, and irrational (in contrast to a sober, peaceful, enlightened, and rational “West”) has proved instrumental in rationalizing the horrors of US foreign policy and marginalizing critiques of this policy (Aistorpe Reference Aistorpe2016). Deep-seated anger toward and skepticism of the US in the Middle East is therefore not a product of the US propping up authoritarian regimes, destroying democratic movements, or the theft of resources, but issues from a supposedly deep-seated irrationality that manifests itself, for example, in a unique susceptibility to a “culture of conspiracy and misinformation” that contrasts with a Western democratic culture of exposing “falsehoods, prejudices, and dishonest propaganda,” as the US 2006 National Security Strategy put it (Bush Reference Bush2006: 10–11). Thomas Friedman argues that this irrational paranoia means “bin Laden touches something deep in the Arab-Muslim soul, even among those who condemn his murders. They still root for him as the one man who was not intimidated by America’s overweening power” (Friedman Reference Friedman2002).

Of course, one might argue that the problem here is simply that these are cases of bad psychologizing: there was no evidence in favor of psychologizing these groups, and any illusion otherwise was a product of bigotry. These cases are therefore not counterexamples to the importance of routine psychologizing; they simply show that we should carry out good psychologizing, i.e., psychologizing when the evidence calls for it.

There is certainly something right about this reply. But it is also too easy. If it is the case that an individual’s or group’s avowed reasons should be regularly or systematically disregarded, then it becomes natural to ask in what sense they are entitled to self-government rather than needing to have their interests and lives determined by others. This is, after all, an extremely influential and long-standing thread in the history of political philosophy beginning with The Republic and continuing to the present day with the renewed fascination with epistocracy, or rule by experts. According to proponents of the latter, the results of “political psychology” tell us “that most citizens process political information in deeply biased, partisan, motivated ways rather than in dispassionate, rational ways” (Brennan Reference Brennan2016: 37). Democracy is therefore unjust because it allows epistemically irrational or non-reason/evidence-responsive ordinary individuals to have political power through voting and other democratic measures, subjecting others to incompetent government. We should instead embrace epistocracy, disenfranchising ordinary citizens through, for example, voter qualification exams, ensuring that only the knowledgeable rule. Brennan is undeterred by either the history of this practice as a tool of Jim Crow-era domination of Black Americans or by what its contemporary application would look like where “the people who pass the exam would be disproportionately white, upper-middle- to upper-class, educated, employed males” given that this demographic apparently scores highest on such tests (228).Footnote ¹⁷ For Brennan, voter exams are not intrinsically racist; broader social and political conditions are discriminatory, and these conditions should be changed – though he acknowledges that, if implemented now, these measures will produce the disproportionate disenfranchisement of Black Americans, the poor, and other minority groups.

I do not have space here for anything like a proper discussion of epistocracy. I emphasize this link because for those of us who are not proponents of epistocracy, who are committed to democracy as the preferred political system of government, and repudiate epistocracy precisely because of its anti-democratic nature, then this also gives you the reason to take seriously the political and moral risks posed by the premature psychologizing of non-formally authoritative individuals. There is a non-contingent link between this kind of psychologizing and political disenfranchisement. It is non-contingent because the inference that disenfranchisement is warranted for individuals who are not capable of being guided by their avowed reasons has appeared and continues to appear plausible to many. If we are worried about this kind of disenfranchisement, then premature psychologizing implicates an important moral and political risk – of possible threats to an individual or group’s political participation and representation – that researchers should be aware of and consider in formulating and evaluating their research.

It might be objected, though, that Brennan is nonetheless right about the consensus in political psychology, and that the above worry – about dictating scientific practice from the armchair – rears its head here. But, again, my view is not that any of this evidence should be minimized or disregarded. My view is only that there is a risk to the premature psychologizing of ordinary individuals that all of us, including researchers, should be aware of and weigh in our analysis of human behavior. This is a risk that accounts such as Flowerree’s and Smyth’s help bring to light, even if I disagree with the scope of their views. To acknowledge such a risk is not some exceptional corrective; it is in fact part and parcel of scientific practice. In testing a hypothesis, our threshold for evidential significance is always partly a product of our judgments about the risk of being wrong: if we place the bar for evidential significance high, we risk generating an increased number of false negatives. Requiring a stronger signal to verify the target phenomenon means we miss instances where this signal is less apparent. Alternatively, if we place the bar for evidential significance low, we allow for more false positives: we err on the side of treating ambiguous cases as genuine instances of the phenomenon. Heather Douglas has influentially articulated this notion of inductive risk and its pervasiveness in scientific practice. She also argues that, given the value-laden nature of scientific practice, the “expert should make clear why they view the evidence as sufficient for any claims made, both in terms of the strength of the evidence and in terms of the risks they are concerned about if they make an inaccurate claim…[This depends] on the values they used” (Reference Douglas2016: 74–5). What type of risk we are willing to countenance and how to settle on our threshold depend on the sort of risk we think we should accept, i.e., our values. But such judgments are rarely made explicit. In the specific case of deciding when to conclude that a dismissive psychologizing verdict of a non-formally authoritative individual is called for, the considerations discussed in this section might prompt us to raise the evidential threshold for concluding that psychologizing is apt.

For an example of how an eschewal of this type of psychologizing in academic research can be both theoretically and practically productive, consider philosopher Maya Goldenberg’s work on vaccine skepticism. Goldenberg’s research is animated by a concern about the reflexive treatment of vaccine skepticism as part of what is taken to be a larger “War on Science” that a great deal of the public is construed as taking up. On Goldenberg’s view, these “metaphors frame the issue as us versus them, good versus evil. Such framing minimizes the need to understand the perspective of the other, or to find compromise. Vaccine hesitators and refusers are uncharitably represented in popular media, and sometimes in academic sources, as scientifically illiterate, irrational, and willfully antiscience and anti-expertise. All the while, the actual concerns of vaccine hesitators are dismissed or ignored, leaving little room for workable solutions” (Reference Goldenberg2020: 12). By taking the testimony and avowed reasons of vaccine hesitators and skeptics seriously, Goldenberg arrives at a number of surprising insights – that parents take themselves to be following the advice of “their pediatricians and other frontline health workers to actively engage in their children’s healthcare and to be experts on their own child,” that given this role and concern for their children, they often conclude that “the presence of rare but serious adverse events [represents] a safety priority rather than, as health officials may see it, a reasonable risk,” and finally that the driving source of vaccine skepticism is falling public trust in scientific institutions rather than brute irrationality (33 and 31). Premature, dismissive psychologizing of this demographic, on Goldenberg’s account, therefore leads to a general demonization that is both problematic in its own right and that impairs our ability to effectively tackle vaccine hesitancy and skepticism.

Other researchers, however, may not judge that the weight or probability of a risk of demonization or disenfranchisement is particularly salient for altering their evidential standards for their projects. And this may be just the right view given relevant considerations and their value commitments. Alternatively, they may judge that the risk is mitigated because, for example, their research consistently pursues the psychologizing of both the formally authoritative and non-formally authoritative, rather than singling out the latter. Again, I have no interest in dictating exactly how researchers should take this phenomenon of the moral risk of psychologizing into account or how it should affect any specific research projects; I only want to argue that it is worth taking into account, even if it will be judged in many cases that the risk is tolerable – even clearly tolerable – or outweighed by competing considerations.