Must depression be irrational?

Abstract

The received view about depression in the philosophical literature is that it is defined, in part, by epistemic irrationality. This status is undeserved. The received view does not fully reflect current clinical thinking and is motivated by an overly simplistic, if not false, account of depression’s phenomenal character. Equally attractive, if not more so, is a view that says depression can be instantiated either rationally or irrationally. This rival view faces challenges of its own: it appears to entail that there are situations when not being depressed is rationally sub-optimal and that resilience to, and healthy coping strategies for avoiding, depression can be rationally remiss. I criticise an existing reply to these challenges before motivating a better one from the perspective of epistemic consequentialism.

1 Introduction

Harry has struggled to get out of bed for the past two weeks. Today, he clothes and feeds his two children and drives them to school. But he does so irritably, with great impatience. When the children inevitably drag their heels, quarrel, etc. he shouts at them, loudly and disparagingly. Back at home, he collapses in bed, exhausted, and feels enormous guilt for how he has acted. None of this is unusual. Recently, Harry has found himself feeling empty most days. Today, he is meant to go to the store to buy food for the week. He knows that this won’t happen. Also, the kids’ clothes need washing and the kitchen needs cleaning too. Then there are those bills. It is incomprehensible to Harry that he can see to these things.

Jan has found it difficult to get out of bed for the past two weeks too. She too feels empty and fatigued. She usually spends her commute busying away on her phone, catching up with friends and family on social media. Today, Jan can’t bring herself to look at the screen. The notifications mount up and it suddenly feels an overwhelming burden. In a meeting after lunch, a colleague makes a mildly dismissive remark about an idea of Jan’s and she nearly bursts into tears. Jan doesn’t think that she will go into work tomorrow and starts formulating excuses that she will later tell her boss. As with Harry, none of these are isolated incidents.

Harry and Jan are depressed. In this, they are not alone. In the United States, 2-3% of men and 5-9% of women suffer from depression, with the probability of developing the condition 5-12% and 10-25% for each group, respectively (Beck & Alford, 2009, p.4). Globally, more than 280 million people are estimated by the World Health Organisation to suffer from depression (WHO, 2023).

The issue I address here is the extent to which it is essential to depression that it involves an epistemically irrational response to life events.

Why might one think such a thing? Suppose that Harry is diagnosed with Major Depressive Disorder with a Seasonal Pattern and that Jan is diagnosed with Perinatal Depression. A change in season has caused Harry to be depressed. But that doesn’t give him a reason to be depressed. And Jan, we may suppose, has desperately wanted to become pregnant and start a family with her loving husband. Becoming pregnant has caused Jan to be depressed. But given her life plans, Jan instead has reason to feel joy, with slight trepidation perhaps. Like Harry’s depression, Jan’s is also ‘out of sync’ with the facts as she apprehends them. Moreover, depression aside, psychological disorders are often conceptualised, from philosophical and lay perspectives, as entailing failures of epistemic rationality (see Bortolotti, 2013 for an overview). As Arpaly (2005, p.284) puts it, “being told that one has a mental disorder is first and foremost being told that one has unwarranted mental states.”

So reflection on people like Harry and Jan, and psychological disorder more generally, may motivate the following view:

The Irrationality View of Depression:

Necessarily, depression is an irrational response to the (apparent) events in one’s life.

The Irrationality View says it is a conceptual truth that depression is without justification in adverse life events. Here are some statements of support in the philosophical literature:

[D]epression crucially involves states that are unwarranted by or disproportionate to events in the subject’s life… [N]ecessarily, if S has depression, then the states and events comprising S’s depressive symptoms are not an appropriate or proportionate response to antecedent events in her environment. (Davies, 2016, p.293)

[D]isproportion to circumstances and thus lack of [rational] explanation in terms of circumstances is an essential aspect of depressive disorder. (Wakefield & Demazeux, 2016, p.4)

[S]ymptoms that have some sufficient justification – for example, the loss of one’s job – and, crucially, that are a proportional response… are not indicative of [depression]. Conversely, those symptoms which do not have sufficient justification (e.g., those which arise ‘out of the blue’)… are indicative of [depression]. (Tully, 2019, p.116)

The Irrationality View might be sloganized as the claim that causes of depression never justify depression. Its defenders sometimes draw a contrast here between depression, qua psychological disorder, and grief, qua normal (or healthy) psychological episode. For although feelings of grief may spiral into illness and become pathological, they appear a reasonable response to the irreversible loss of something or someone of value (Cholbi, 2017).¹

Yet the Irrationality View is not without drawbacks. For one, the view entails that severe sexual or physical assault, loss of home and livelihood and refugee experience never justify being depressed. For some of us, this seems plainly incorrect. As Bortolotti (2013, p.483) suggests, “Some forms of anxiety and depression may seem reasonable responses to life events.” Graham (1990, p.408) agrees, claiming, in contrast to the Irrationality View, that “people sometimes suffer in circumstances so bad or terrible that it is beyond belief that they are not justified for depression.” Recently, Kate Abramson has offered gaslighting as circumstances that may reasonably cause one to be depressed. Abramson (2024, p.152) claims it is “outrageous” to say of someone who suffers depression as a result of gaslighting that they have ipso facto failed rationally. Rather, being depressed in such circumstances is, she claims, a “fitting evaluative response” (p.154).

So granted, it might be true of some depressed individuals, like Harry and Jan, that the cause of their depression does not also rationalise their depression. But that this is true of all instances of depression, and true of all depressed individuals, is not so obvious. If that is right, then a possibility worthy of serious consideration is that the Irrationality View is false. Instead, depression is justified on some occasions and other times not. For instance, in line with the claims of Bortolotti, Graham and Abramson, it might be that depression caused by, say, refugee experience, homelessness, gaslighting and other serious trauma is justified, whereas depression caused by change of season is not. And perhaps depression as a result of, say, pregnancy can be justified or unjustified depending on one’s life circumstances, how one has become pregnant, one’s prior desire to (not) have a child, one’s wider support network, and so on.²

Moreover, just as the situations in which one can be depressed are varied, so too is depression itself. Harry’s and Jan’s depressions are individuated by their causes: Harry has Major Depressive Disorder with a Seasonal Pattern and Jan has Perinatal Depression. By contrast, Psychotic Depression is depression that is individuated by its concomitants, e.g., delusions and hallucinations, rather than its cause. Different yet again is Dysthymia, a form of depression that is individuated both by its mild nature and the prolonged length of its episodes. Where depression is so heterogenous, it would be surprising if its relation to epistemic rationality were entirely homogenous (see also Arpaly, 2005, pp.287–288).

So where can one turn to find unequivocal opposition to the Irrationality View? Pockets of resistance exist, as evidenced above, but one must look hard to find them.³ I aim to level the playing field, showing a number of flaws in the Irrationality View and defending the following view instead:

The Mixed View of Depression:

Depression can be a rational or irrational response to the (apparent) events in one’s life.

Whereas the Irrationality View says that instantiations of depression are necessarily unjustified and hence irrational, the Mixed View says that depression can be instantiated either unjustifiably or justifiably. The Mixed View makes depression rationally continuous with various other mental states, e.g., beliefs, emotions, etc. insofar as these are also sometimes epistemically rational or otherwise well-supported, and other times not. In making a case that depression is sometimes rational, I join the ranks of several other authors who have recently argued that some psychological disorders are not intrinsically irrational and that their episodes are sometimes epistemically befitting one’s situation. This includes conditions as diverse as post-traumatic stress disorder (McSweeney, 2023), obsessive compulsive disorder (Haerle, 2023) and various eating disorders (Gadsby, 2023).

Here is the plan. In the next section, I clarify the nature of the disagreement between the Irrationality View and the Mixed View. In Sects. 3 and 4, I go on the offensive, arguing that the Irrationality View receives mixed support at best from current psychiatric thinking while also assuming an overly simplistic account of depression’s phenomenal character. In Sects. 5 and 6, I go on the defensive, supporting the Mixed View against two potential objections, and which might motivate a reductio of the view; I’ll assess, and find wanting, an existing defence of the Mixed View against these, due to Graham (1990), before motivating an alternative one via a broadly consequentialist approach to epistemic rationality.

Before beginning I want to be clear about my aims. Rather than proving the Mixed View true and the Irrationality View false, I will be satisfied to have pointed out some drawbacks of the Mixed View and to have defended the Irrationality View against objections. Thus, my aim is to even the odds between the views.

2 The irrationality view and the mixed view: clarifying the issues

To say that depression is epistemically irrational is to say that it fails to be reason-responsive. But psychological states can fail to be reason-responsive in at least two ways. First, a psychological state might fail to be reason-responsive in its aetiology. That is, a state may be epistemically irrational at its point of instantiation, having been formed in such a way that didn’t take into account of, or perhaps even went against, relevant evidence. Second, a psychological state may fail to be reason-responsive insofar as, once instantiated, it is impervious to reasoning. That is, once in place, the state can’t be modified by countervailing considerations. That is, a psychological state may be epistemically irrational in terms of its being maintained in the face of psychological defeaters. I shall call these different forms of irrationality aetiological irrationality (irrationality in causal history) and maintenance irrationality (irrationality in how an already instantiated state is sustained).⁴

The Irrationality View says that, necessarily, depression is epistemically irrational in that it is disproportionate to one’s (apparent) life events.⁵ Strictly speaking, this claim is ambiguous between the claim that depression involves aetiological irrationality and the claim that it involves maintenance irrationality (or both). However, examining the above remarks by defenders of the view suggests it is aetiological irrationality that characterises the Irrationality View. In the above quotes, the view’s defenders talk about depression being irrational at its point of instantiation, i.e. its necessarily being an inappropriate response to “antecedent events” in the person’s life (Davies, 2016, p.293), arising “out of the blue” (Tully, 2019, p.116) and lacking in any form of circumstantial “explanation” (Wakefield & Demazeux, 2016, p.4). These are claims about how episodes of depression are triggered, never mind how they continue to persist (or go out of existence). This is not to say that there are no interesting issues to examine in terms of depression’s relation to maintenance rationality. There certainly are. Insofar as depression characteristically involves not just low mood, but a loss of mood reactivity, the condition may, ipso facto, be pictured as involving maintenance irrationality. But as there are these two different dimensions to epistemic rationality, a psychological state may be epistemically rational along one such dimension but epistemically irrational along the other. For instance, a belief may be formed on good evidence, and so exhibit aetiological rationality, but then come to be sustained in the face of opposing evidence, and so be irrational in how it is maintained.

To be clear, then: the lens of epistemic rationality through which I am examining the debate between the Irrationality View and the Mixed View is aetiological irrationality, where this is a matter of a psychological state’s being either well- or ill-supported when first instantiated, independent of whether or not it continues to be well- or ill-supported by consequent events in a person’s life.

These two senses of epistemic rationality are often run together. But they should be distinguished. Here is a putative example that underscores how they might come apart in the case of depression: Harry’s depression might be said to be aetiologically irrational, since it is caused by mere change of season, and yet his depression might eventually become rational, depending upon the subsequent events in his life, say, he loses his job, is rendered homeless or suffers other forms of trauma. Converse cases, where aetiologically rational depression becomes maintained in an irrational way are also possible. For instance, depression as a result of severe assault may be aetiologically rational and yet end up being maintained in an irrational way due to the person’s developing paranoid, persecutory delusions.⁶ Of course, the extent to which one finds these examples plausible depends on one’s being antecedently sympathetic to the Mixed View.

The question of whether depression must be aetiologically irrational is orthogonal to the question of whether depression is an adaptation. The question of a mental disorder’s adaptiveness can be disambiguated in several ways. Plutynski (2023) has recently identified four: namely, a condition’s (i) having served some function in the evolutionary past and continuing to perform that function; (ii) having served some function in the evolutionary past, but no longer performing that function; (iii) having served some function at one stage of a person’s life, while no longer serving that function now and possibly even being maladaptive in the long run; and (iv) serving some function currently, but overall being less than optimal.

In the case of depression, psychiatrist Nesse (2000) has given perhaps the most well-known defence of the claim that the condition is an adaptation. Nesse claims that depression developed in circumstances where it was beneficial for our ancestors to cease futile or dangerous activities, but where it likewise benefitted them to temporarily withdraw from all undertakings rather than immediately commence new ones. On Nesse’s view, depression is a mechanism we developed to prevent us from jumping out of the frying pan and into the fire.

Nesse’s account says depression an adaptation, but, as Plutynski suggests, only in sense (ii). That is because accounts like Nesse’s are concerned with how there came to be the type of state that we call ‘depression’; they don’t speak to the question of whether and when tokens of such a state instantiated now might be epistemically rational in their aetiology. Strictly speaking, such accounts leave open whether all instantiations of depression are irrational. For instance, in the face of a theory like Nesse’s, about how depression arose in our ancestral past, defenders of the Mixed View may still motivate their position by pointing to the intuitive contrast between depression as a result of change of season (putatively ill-supported and hence irrational in aetiology) and depression as a result of serious trauma (putatively well-supported and hence rational in aetiology). Defenders of the Mixed View may rightly argue that this contrast goes unexplained by adaptive theories such as Nesse’s.

Another issue is whether the disagreement between the Irrationality View and the Mixed View is merely verbal. Consider Jan, and suppose that she does not have Perinatal Depression, but was instead the victim of a prolonged and aggravated assault. On the assumption that low mood and depleted motivation are now (sadly) an appropriate and proportionate response to Jan’s life-events, the Irrationality View is likely to require that we rescind the claim that Jan is depressed. (Recall the quotes above.) By contrast, defenders of the Mixed View will likely be open to the possibility that Jan is depressed. Yet both views agree on the substantive psychological facts; namely, that:

• a. Jan is distressed (is experiencing low mood, depleted motivation, a desire to socially withdraw, etc.);

• b. Jan’s being distressed is aetiologically rational, given the assault; and.

• c. The nature of that distress (low mood, depleted motivation, a desire to socially withdraw, etc.) is common to both Jan’s current condition and depression proper.

What the Irrationality View and the Mixed View disagree on is whether the following is true:

• d. Jan is depressed.

But with so much agreement between the views, especially on (c), one might think that disagreement over (d) is merely verbal and that nothing can be gained from pitting the two views against one another.

This is very much mistaken. As Davies (2016) points out in his defence of the Irrationality View, the latter is an instance of what he calls ‘psychiatric externalism.’ Psychiatric externalism says that mental disorder does not supervene on psychological features of a person alone, but is partially a matter of what relations they bear to the environment: in the case of depression, whether the person’s low mood, motivational depletion, desire to socially withdraw, etc. is a disproportionate response to antecedent life events. Crucially, the Irrationality View entails that two persons can be psychological duplicates, both with low mood, lack of motivation, desire to socially withdraw, etc. yet one be depressed and the other not, depending on whether these psychological states have insufficient grounding in life events or not. Davies writes:

Given a suitably dystopian context, devoid of any reason to be happy, even the severest of depressed mood might not qualify as pathological. The facts that determine whether or not an individual is depressed therefore must include facts about her environment. (p.293)

The Irrationality View is thus a substantial metaphysical thesis about the conditions under which depression is instantiated, just as traditional externalism in philosophy of mind is a substantial metaphysical thesis concerning the conditions under which psychological contents are instantiated (Putnam, 1975; Burge, 1979). So then the Mixed View, which denies the Irrationality View, must be a substantial metaphysical thesis too, and debate between them must likewise be considered at least somewhat weighty and not merely verbal. Indeed, one way to mark the difference between the Irrationality View and the Mixed View is to note that the latter is more internalist in nature. By claiming that depression can be both aetiologically rational or irrational, the Mixed View holds that factors about how one’s state(s) of mind happen to be caused are unimportant for whether one is depressed or not. It is worth emphasising: this claim is consistent with various depressions themselves being individuated by causal history. What makes someone depressed per se may be their psychological states alone (never mind the causes) but what makes the depression one kind rather than another (e.g., perinatal vs. delusional) can be a matter of relations to the environment. Compare: what makes something count as water is its intrinsic chemical composition—the presence of H₂O. Yet what makes something count as spring water, and not tap water, is determined by recent causal history. So the Mixed View can accommodate the idea that some depressions are individuated by their causes while depression per se is determined acausally, i.e. by intrinsic psychology alone.

But what is depression? The 5th and most recent edition of The American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders (2013, pp.160–161) says that someone is depressed if they present with either (i) low mood or (ii) diminished interest in, or pleasure derived from activities (anhedonia), in addition to having four or more of the following symptoms over a two-week period:

• iii. Weight loss or decrease/increase in appetite;

• iv. Insomnia or hypersomnia;

• v. Psychomotor agitation or retardation;

• vi. Fatigue;

• vii. Feelings of worthlessness or excessive or inappropriate guilt;

• viii. Diminished concentration; and.

• ix. Thoughts of death/suicide.

Some of these, particularly (iii) – (vi) do not seem legitimate targets for rational appraisal. Does this stop the Mixed View in its tracks? If it does, then it stops the Irrationality View from getting off the ground too, since criteria like (iii) – (vi) are not necessarily irrational. But arguably the DSM does not define mental health conditions. In the main, the DSM is a diagnostic tool. It outlines for clinicians criteria for attributing mental health conditions to patients. Some of its criteria can therefore be considered epistemic in nature; that is, as drawing the clinician’s attention to tell-tale signs of a condition, rather than specifying criteria metaphysically constitutive of disorder (see also Bueter, 2019, p.1067).

What, then, are the options for theories of depression? Graham (1990, pp.401–403) claims that depression is a combination of certain beliefs, emotions and desires. Ratcliffe (2014) identifies depression with a single state, a ‘existential feeling’ that, roughly, disrupts one’s sense of the world as allowing for possible actions. Tully (2017) claims that severe depression involves a complete lack of desires. Whiteley (forthcoming) claims that depression is an ‘altered’ or ‘global’ state of consciousness akin to dreaming. This is a paper about the rationality of depression and not its fundamental nature. That is a matter on which I aim to remain as neutral as possible throughout. Indeed, I am tempted to think that the fundamental nature of depression can only be settled once we agree upon answers to questions like ‘can depression ever be rational?’ We can all agree: depression typically involves low mood, lack of motivation, a desire to socially withdraw, negative self-directed beliefs, feelings of guilt and hopelessness, etc.⁷ But is that what depression is, in and of itself? Or are these depression’s effects? Could they be mere concomitants? Might they instead be depression’s causes? Knowing whether instantiations of depression are always irrational, or whether they can sometimes be rational, may help us decide between options here, both in terms of answering these particular questions and deciding between theories of depression’s nature more generally.

Granted, insofar as the Mixed View is at odds with externalism about depression it does have some metaphysical commitments, namely about the circumstances under which states of depression become instantiated. But this still leaves it open what particular state, or collection of states, depression is to be identified with. This is what I will not offer any positive remarks on.⁸ Moreover, the Mixed View has a significant commitment insofar as it entails that whatever kind of state (or collection of states) depression is to be identified with, that state (or collection of states) has a rationally appraisable aetiology. But in this respect, the Mixed View agrees with the Irrationality View. It is just that the two views disagree on matters of valence.

Some might prefer a view on which depression’s aetiology is, like that of perceptual experiences or pains, fully arational; that is, not the kind of thing one can rationally appraise in the first place. For instance, one may prefer a view like the following:

The Arationality View of Depression:

Depression is an arational response to the (apparent) events in one’s life.

On this view, it doesn’t make sense to enquire into the rationality of instantiations of depression.⁹ That would be to disagree with both the Irrationality View and the Mixed View. Since my focus here is on which of the latter two views to prefer, I leave substantive discussion of arational accounts of depression for another time.¹⁰ But it is worth mentioning an issue that might lead one to reject the debate between the Irrational and Mixed views as misguided and motivate supporting the Arationality View instead; namely, the fact that people don’t become depressed at will.

The fact that depression is not under one’s direct control or instantiated voluntarily should not be taken as decisive against the Mixed View (nor the Irrationality View). Arguably, beliefs cannot be formed at will, yet clearly beliefs can be justified or unjustified. Indeed, as Graham notes, in his discussion of the rationality of depression, there are a great many things that occur in the mind that we do not have direct control over, but which we take to be rationally appraisable:

Exam mistakes, offensive moral views are all nonvoluntary, but we don’t hesitate criticizing people for them, or judging them as unreasonable mistakes, unjustified views. The mere fact that depression is not directly voluntary does not mean it cannot be justified. (1990, p.410)

I offer two final remarks on my approach and the dialectic between the Irrationality View and the Mixed View.

First, I remain neutral on the precise nature of aetiological rationality, observing that this is, in some further sense to be explained, a matter concerning a mental state’s causal history (or counterfactual dependence on prior facts) serving as normative support for that mental state (see Sylvan, 2016; Korcz, 2021 for options). This is a matter for defenders of both the Irrationality View and Mixed View to say more on, but I leave that for another time.

Second, defenders of the Mixed View may or may not agree with defenders of the Irrationality View that depression is necessarily pathological, a matter for medical intervention, or something that needs to be ‘treated’ another way. The view leaves all this open. The Mixed View may give the appearance of saying that depression, when rational, is a condition it is imprudent to seek relief from. But that is incorrect. Compare: people are not imprudent to seek relief from, say, justified anger, if it harms them (e.g., raises their blood pressure). And just as many different forms of treatment are available for high-blood pressure, some physician-prescribed and monitored and some not (e.g., yoga and exercise), the same is true of depression. The Mixed View leaves open how people with depression should be treated, but it certainly does not proscribe against treatment.

In the next two Sects. (3 and 4) I attempt to undercut motivations for the Irrationality View before moving onto a defence of the Mixed View (5 and 6).

3 Against the irrationality view I: the current clinical zeitgeist

In his defence of the Irrationality View, Davies (2016, p.293) speculates that it reflects “the current clinical zeitgeist.” In this section, I shall raise various doubts about that claim. In doing so, my aim is to undermine a potentially key motivation for the Irrationality View; namely, that it is supported by current clinical thinking about depression. In response, I will not claim that current psychiatry gives no support for the Irrationality View. But it does seem to me that current psychiatry gives ambivalent support for that view, at best.

3.1 The DSM

In Sect. 2, I outlined the DSM-5 criteria for depression. Put aside the question of whether that criteria is wholly epistemic in nature or whether some of it is metaphysically constitutive of depression. What I wish to draw attention do is this: the DSM gives no indication of it mattering to someone’s being depressed how symptoms meeting criteria (i) – (ix) are caused. By contrast, the driving thought behind the Irrationality View is that depression should not be judged present “without considering the nature of [the] trigger” (Wakefield et al., 2007, p.434). But insofar as the DSM-5 omits to mention the need for such an investigation when diagnosing depression, it is in conflict with the Irrationality View.

Naturally, this is a matter of dissatisfaction to defenders of the Irrationality View. For instance Jerome Wakefield, a chief defender of that view, has advocated that the DSM should include an overall ‘disproportionality to circumstance’ criterion that symptoms have to meet to be considered indicative of depression. Wakefield has repeatedly cautioned that, without such a criterion, psychiatry risks misdiagnosing ‘normal’ sadness, non-pathological grief in particular, as psychological disorder (Horwitz & Wakefield, 2007; Wakefield et al., 2007; Wakefield & Horwitz, 2016). I will have some things to say in the following section about the idea that extrinsic, aetiological facts alone can distinguish depression from grief; namely, that it is false. The point I am developing at present is that the Irrationality View is not as firmly reflected in current mainstream clinical thinking about depression as some of its defenders imply.

Granted, previous editions of the DSM were more consistent with the Irrationality View. They incorporated a bereavement exclusion: the DSM-III and DSM-IV stated that someone does not have depression if, despite having criteria-meeting symptoms, these can be explained by the recent death of a loved one. This did make extrinsic factors relevant to depression’s diagnosis. So was the Irrationality View, from the perspective of the DSM, at least previously a part of the ‘clinical zeitgeist?’ Not obviously. For there was no wider exclusion, covering other traumatic events like sexual or physical assault, homelessness, gaslighting, refugee experience, and so on. Despite appearances, these previous editions of the DSM were only partially consonant with the Irrationality View.

Indeed, as some clinicians rightly observed, the presence in earlier DSMs of a bereavement exclusion, but not an exclusion for other traumatic events was “not logically defensible” (Moran, 2011, n.p.). When updating depression’s criteria in the DSM-5 and resolving this tension, psychiatry could have taken one of two paths. It could have dropped the bereavement exclusion, abandoning any extrinsic, ‘disproportionality’ criteria as relevant for depression. Alternatively, the bereavement exclusion could have been extended to all adverse and traumatic life events; then, symptoms alone would never be sufficient for diagnosing depression. Instead, how those symptoms are caused—whether or not they are disproportionate responses—would have to be investigated before determining whether depression is present in a patient. If the history of psychiatry had gone down this second path, then the Irrationality View would reflect the current clinical zeitgeist. But as the above outline of the DSM-5’s depression criteria illustrates, The American Psychiatric Association did no such thing. They went with the first option: the DSM-5’s depression criteria entail that clinicians can determine depression’s instantiation on the basis of intrinsic, symptom-based features alone and ignore aetiological facts. Moreover, The American Psychiatric Association were motivated to go down this route on the grounds that the alternative, i.e. broadening the proportionality-based exclusion criteria, would be against the clinical status quo. A statement by a lead member of the DSM-5’s mood-disorder work group was emphatic about this. It claimed that additional proportionality exclusions would “represent a major shift, unsupported by a range of scientific evidence, in the nature of our concept of depression, as epidemiologic studies show that the majority of individuals develop major depression in the setting of psychosocial adversity” (reported in Moran, 2011, n.p.).

Granted, the DSM-II of 1968 stated that depression involves an “excessive” reaction to life events. This claim is very much in line with the Irrationality View. But if the best that a defender of the Irrationality View who wants to establish the psychiatric pedigree of their view can do is show that their theory is reflected in one of five editions of the DSM (and from over 50 years ago) then this can hardly be considered a victory. To be clear: none of this is to say that the DSM is beyond reproach. The point is simply that insofar as the DSM may be thought to represent ‘the current clinical zeitgeist’, the characterisation of depression in the DSM-II, and which seems most friendly to the Irrationality View, is a historical outlier.

3.2 Situational depression

This brings us to a second respect in which current psychiatry’s resistance to the Irrationality View is evident. The kind of depression identified above by the DSM-5’s mood-disorder work group as potentially overlooked, should it recommend an Irrationality View-friendly extension of the bereavement exclusion to other adverse life events, is so-called ‘situational’ or ‘reactive’ depression. This form of depression contrasts with depression caused solely by genetic or biological factors, also known as ‘nonsituational’ or ‘nonreactive’ depression.

Nonsituational forms of depression arise, from the patient’s point of view, out of the blue, as Harry’s and Jan’s did, bearing no obvious epistemic relation to life events; hence their seemingly irrational nature. It is this kind of depression that the Irrationality View seems to unwittingly consider paradigmatic, if not exhaustive of depression. But situational forms of depression are highlight prevalent. Consider the following remarks by psychiatrist William Coryell and his colleagues (1994):

Studies of bereavement, of response to natural disasters and economic depression, and of associations between medical diseases and depressive illness, all point to the existence of situational and reactive depressions. (pp.203-4)

Most clinicians are familiar with the depressed patient whose symptoms appear to be chronically intertwined with situational difficulties and adverse life events. (p.209)

This certainly sounds like a denial of the Irrationality View’s claim that, necessarily, depression must be without rationalisation and so must be aetiologically irrational in nature.

In reply, I anticipate that defenders of the Irrationality View will claim that they do, in fact, allow for situational forms of depression; such depression may be pictured, on their view, as a matter of one’s initial reaction to adverse life events spiralling out of control, with the symptoms taking on “a life of their own” (Horwitz & Wakefield, 2007, p.18). As Davies writes in his defence of the Irrationality View:

Depression can be (and often is) precipitated by events that would reasonably be classed as saddening or tragic. The point is that the subject’s response may become pathological if its severity or duration exceeds some contextually determined threshold for appropriateness or proportionality to these events. (2016, p.293; see also Tully, 2019, p.116)

But when Coryell and their colleagues characterise adverse life events as “chronically intertwined” with depression, they do not seem to be thinking about situational forms of depression in this way. That is, they do not seem to consider adverse life events as distal triggers of such depression—as if challenging life events first cause a psychological state that is proportionate and which only later, through some dysfunction in the patient, turns excessive and transforms into depression. Instead, by saying that depression can be “chronically intertwined” with challenging life events, Coryell and colleagues seem to be saying that situational depression can be fully explained by such events and that the latter aren’t mere distal causes of the former. (They are also saying that most clinicians are familiar with cases like this.)

Furthermore, it has been argued in the psychiatric literature that the situational/nonsituational distinction might be scrapped entirely (Hirschfeld, 1981; Garvey et al., 1984; Kendler et al., 2010). For instance Kenneth Kendler and his colleagues (2010, p.778) are unequivocal in claiming that depression is “best diagnosed in a context-independent manner.” This, again, is at odds with the Irrationality View, though quite consonant with the DSM-5 and the view of depression encapsulated by the Mixed View.

A defender of the Irrationality View might support their view as follows: because psychiatrists enquire into a patient’s life events prior to, or around the time of, symptom onset, aetiological facts are important for the condition’s diagnosis. But defenders of the Irrationality View pursuing this reply must rule out that the clinician’s interest in depression’s aetiology is not its relevance for other matters, such as determining treatment choice. For instance, the type of therapy most effective for a patient is known to vary depending upon the context in which their symptoms arose (Maj, 2012).

In sum, the clinician’s deck seems at best a mixed bag, relative to the Irrationality View. One cannot assume that Irrationality View is particularly well supported by current psychiatric thinking. I now intensify the case against the Irrationality View further; namely, by arguing that the master argument for the view rests on an overly simplistic account of depression’s phenomenal character.

4 Against the irrationality view II: is depression just Irrational Sadness?

Defenders of the Irrationality View claim that one must appeal to extrinsic factors to individuate depression, where this means facts about how its symptoms are caused. More precisely, the Irrationality View’s defenders appeal to the (putative) aetiological irrationality of the patient’s symptoms as the marker of depression. In doing so, they imply that depression and ‘normal’ or ‘healthy’ distressing states are intrinsically or psychologically alike, and so cannot be told apart independent of aetiology. For instance, consider the title of Horwitz and Wakefield’s (2007) polemic against the DSM: The Loss of Sadness: How Psychiatry Transformed Normal Sorrow into Depressive Disorder. Davies (2016, p.293) similarly claims that a key reason for defending the Irrationality View is that divorcing depressive symptoms from their aetiology would leave one unable to “distinguish pathological depression from normal sadness.” Tully (2019, p.115) makes the same claim in their defence of the Irrationality View. Often, grief is proposed as the relevant contrast. Indeed, this is something of a ‘master argument’ for the Irrationality View: aetiological facts are necessary to distinguish depression (which is an abnormal and harmful response to life’s events) from similar distressing states like grief (which is, by contrast, normal and healthy).¹¹

Notice how the view’s defenders imply with such remarks that depression has no distinguishing intentional-cum-phenomenal features. For if it did, then appealing to aetiological facts would be unnecessary to individuate the condition; we would be able to distinguish depression from normal sadness, like grief, by first-person features alone.

However, depression and grief can be distinguished on first-person grounds alone. This makes appeal to aetiological factors unmotivated and so undercuts this master argument for the Irrationality View. Once again, the psychiatric literature seems to go against the Irrationality View here. We will focus on depression first.

Consider the following account of the first-person features of depression from Desmond Curran and Erich Guttmann’s (1945) Psychological Medicine:

The typical affect is that of sadness, but has been described as being qualitatively… different to that normally experienced… [T]he patient feels changes—strange, lifeless, detached, automatic. Sometimes, instead of feeling that they themselves have changed, the patients lay emphasis on a change in the outer world which seems dead or macabre. (p.157; cited in Kendler, 2016)

This idea of the external world feeling distant or dead is a leitmotif in much of the psychiatric literature on depression. When clinicians ask depressed patients to describe their symptoms, what is most frequently identified is lethargy and inability to act; but after that is a sense of detachment from the world (Healy, 1993; Rottenberg, 2005). Ratcliffe (2014) characterises the first-person experience of this ‘detachment’ as follows:

The practical significance of things is somehow diminished; they no longer offer up the usual possibilities for activity. Associated with this, there may be a sense of impossibility; possibilities appear as ‘there but impossible to actualize’. There can also be a sense of estrangement, as possibilities that are inaccessible to the self appear as ‘accessible to others with little effort’. Other people might continue to offer possibilities for communion, but these possibilities appear at the same time as ‘impossible for me to take up’. Together, these alterations in the possibility space constitute a feeling of isolation, which is experienced as irrevocable because depression does not include a sense of its own contingency. The resultant estrangement from the world amounts to a change in the sense of reality and belonging—things no longer appear available; they are strangely distant, not quite ‘there’ anymore. Certain kinds of possibility may also be heightened. A world that no longer offers up invitations to act can at the same time take the form of an all-enveloping threat, before which one is passive, helpless and alone. (2014 p.71)

What the above two quotes highlight is that depression involves an element of derealisation, of the world’s feeling distant, strange and no longer fully there for one (perhaps co-present with a painful awareness that it once was effortlessly there, and currently is for others). Can this distinguish depression from grief? If so, then first-person features suffice here and appeal to causes of these conditions is unnecessary.

Those who defend the Irrationality View might object: derealisation suffices to distinguish depression from much normal sadness, but not intense, debilitating sadness like grief. For in grief one also feels detached, helpless and motivationally depleted, perhaps to the same degree as in depression. And yet, they may argue, depression is mental illness and grief is normal and natural. Hence, extrinsic, aetiological factors, proportionality to circumstance in particular, are necessary to distinguish the two.

In reply, consider a further phenomenal feature of depression: the felt inescapability of one’s derealised state. In grief, one may feel detached from the world in the exact same way as if one were depressed. But in non-pathological grief at least, one often has a sense that the painfulness shall pass. Depression, however, feels inescapable insofar as one feels that it shall never lift. This is what Ratcliffe implies when saying that depression is unique in failing to include a sense of its own contingency. And he is not alone in characterising depression in this manner. The clinical psychiatrist Kay R. Jamison does the same. Crucially, she does so when contrasting depression with grief. In her depression memoir, she writes:

Even during the worst of my grief I had some sense that this would happen, that the weather would clear. I did not have this faith during the merciless months of depression. (2009, p.172)

It is here, in what we can call the felt inescapability of derealisation, that daylight can be found between first-personal features of grief and depression. The idea that grief and depression can only be distinguished by aetiological factors, as defenders of the Irrationality View claim, assumes an overly simplistic, if not false, picture of depression’s phenomenal character.

I suspect that the fault lies in defenders of the Irrationality View assuming that depression and grief are objectless moods sharing identical, intrinsic phenomenal qualities. On this assumption, the experience of depression and grief can only differ in the magnitude of those phenomenal qualities. On this account, neither depression nor grief has any intentional structure to speak of.¹² But this seems false to grief in particular. Grief has as its intentional object the absence in the external world of a particular individual, whether that is an object or person. One grieves for the loss of something. While depression might also be said to involve experiencing the absence of something—namely one’s agency or ability to act—this would still not suffice to make grief and depression phenomenally indistinguishable in ways that would motivate an appeal to extrinsic facts to distinguish the two. The phenomenal character of depression would remain primarily self-focussed in ways that grief is primarily other-focussed, though both would share a sense, perhaps secondarily or indirectly, of the external world as flat and colourless.

The suggestion here is that both depression and grief have something of a twofold phenomenology in which one thing, the world, feels dead to one through its absence of some other thing. However, in grief, the world feels dead through its being found to be absent of a loved one or thing. But in depression the world feels dead to one through its absence of oneself, in a manner of speaking, insofar as one’s ability to act in the world feels agonisingly absent.

My claim then is that the difference in phenomenal character between depression and grief goes beyond a difference in the felt inescapability of derealisation, as identified by Ratcliffe. It also includes a difference in what we may call self-/other-focussed intentionality. Crucially, the contrast here with grief is not a difference of mere degree, i.e. in intensity or duration of negative affect. The difference is one of intentional-cum-phenomenological structure. This is therefore a difference in kind.¹³

This is not an exhaustive account of the first-person, phenomenal differences (and similarities) between depression and grief.¹⁴ But thinking about the difference between depression and grief in terms of self-/other-focussed intentionality, in the above manner, is apt for a number of reasons. First, the appeal to a common twofold structure explains how one might mistakenly think there is no phenomenal difference between depression and grief, as defenders of the Irrationality View insinuate. Second, the twofold structure described above explains how, in the case of depression in particular, a patient’s experience may vacillate between the world’s feeling broken versus oneself feeling broken, a feature of the condition reflected in Curran’s and Guttman’s remarks, as well as Ratcliffe’s too. My suggestion is that this is akin to perceptual aspect-switching. For instance, when looking at a picture, one may alternate attention between what is depicted in the surface and the marks in which it is depicted (Gombrich, 1960; Budd, 1987) but one can also, arguably, experience both together, as two, united ‘folds’ of a single experience (Wollheim, 1980).

So individuating depression by reference to extrinsic facts, as the Irrationality View does, is unmotivated and redundant. Depression has first-person, phenomenal features of inescapability and self-focussed intentionality that already distinguish it from even the most intense forms of normal, healthy sadness. Appealing to extrinsic facts is simply not necessary here. Thus, the master argument for the Irrationality View relies on a naïve, if not false, view of the phenomenal character of depression.

5 How not to defend the mixed view

So far I have gone on the offensive against the Irrationality View. It’s now time to defend the Mixed View against objections.

5.1 Two problems

The Mixed View says that some events, traumatic ones in particular, are so unbearable that they rationalise depression. The view rejects the idea that it is a conceptual truth that depression is without justification in adverse life events. In doing so, it rejects the idea that depression is necessarily aetiologically irrational. It therefore also rejects the externalism of the Irrationality View.

What does the Mixed View say about individuals who undergo unbearable events, but do not become depressed? Suppose that Sam, having come face-to-face with some of the worst, most dispiriting and traumatic experiences that life can offer, does not, as a result, become depressed. Ordinarily, we would see this as fortunate. By being psychologically unscathed by trauma, Sam has narrowly avoided mental illness. Intuitively, there is nothing problematic in this at all—quite the opposite.

But defenders of the Mixed View seem forced to say otherwise. The Mixed View considers Sam, and those in similar situations, to be, as Graham says “justified for depression” (1990, p.400). So insofar as Sam is not, in fact, depressed, the Mixed View seems to entail that Sam is rationally sub-optimal. It would be aetiologically rational for Sam to be depressed, on the Mixed View, so insofar as they are not, they are missing out. After all, being merely justified for a mental state is not as good, from a rational point of view, as being justified for a mental state and then actually being in that state. Compare: it is financially good to be approved for a loan from the bank (in a situation in which you need funds quickly), but what really matters is that you are approved for the loan and the money then sits in your account. Merely having the loan approved, but not actually possessing the money, is financially sub-optimal. By a parity of reasoning, defenders of the Mixed View seem forced to say that Sam is likewise missing out, albeit rationally. But this seems absurd. It is good that Sam is not depressed and in no sense bad, rational or otherwise. Call this problem for the Mixed View ‘the problem of unfortunate near misses.’

There is a second problem in the vicinity when we consider what factors might prevent one for whom, on the Mixed View, it would be aetiologically rational to be depressed, from actually being depressed. First, it might be down to one’s character. Suppose that Priya is sufficiently resilient that intolerable events fail to cause her to be depressed. After a string of appalling experiences, she simply ‘toughens up’ further still. Mental resilience, psychological fortitude, etc. is normally thought to be a good thing. But the Mixed View seems forced to say that it is bad here, at least from a rational point of view. In this case, it has prevented Priya from reaching her full rational potential insofar as it bars her from being in a state that it is aetiologically rational for her to be in.

Someone in a similar position is Aaron. Having lived through traumatic events, he finds himself despondent to an extent he has never experienced before and worries he is sliding into depression. So he actively works to keep the condition at bay. Aaron talks through his experiences with a psychotherapist, takes increased exercise, has a meditation schedule, makes time for friends and family, etc. The Mixed View again seems forced to say that, from the point of view of rationality, Aaron is doing things wrong. His strategies for keeping depression at bay are, like Priya’s resilience, preventing him from being in a psychological state that it would be aetiologically rational for him to be in. The Mixed View seems forced to say that he would be more (aetiologically) rational if he simply yielded to his emerging depression and that his coping strategies are keeping him in a rationally sub-optimal situation.

In all this, the Mixed View seems to produce incorrect verdicts. It treats psychological resilience and coping strategies as rationally remiss insofar as they prevent persons for whom it would be rational to be depressed, like Priya and Aaron, from being, in fact, depressed. But the idea that mental resilience and healthy coping strategies are in any way bad is highly counterintuitive. Call this second problem ‘the problem of remiss resilience.’

Taken together, the problems of unfortunate near misses and remiss resilience constitute a dual-headed reductio of the Mixed View. And they have a common root: the Mixed View’s claim that depression is ever rationally instantiated in the first place. The Irrationality View says that it is a conceptual truth that depression is necessarily irrational. So, on that view, the problem of unfortunate near misses does not arise: on the Irrationality View, being depressed is never rational, so failing to be depressed is never rationally sub-optimal. Likewise, on the Irrationality View, psychological resilience to depression and healthy coping strategies are always rationally beneficial; they prevent people like Priya and Aaron from being in what is, on that view, an intrinsically irrational mental state. And for that reason, the problem of remiss resilience does not arise on the Irrationality View either. So, for all of the problems facing the Irrationality View, it appears to claim a substantial victory here over the Mixed View. How might a defender of the latter view reply?

5.2 The hard-line reply

One reply on behalf of the Mixed View is the ‘hard-line’ reply. It digs in its heels and bites the bullet. Faced with the problem of unfortunate near misses, the hard-line reply says that if a person would be rationally better off depressed, by being in a state that it would be aetiologically rational for them to be in, then so be it. On this view, insofar as Sam, having lived through various traumatic events, fails to be depressed, they really are in a rationally sub-optimal situation. Moreover, insofar as Priya’s resilience and Aaron’s coping strategies prevent them from being depressed, resilience and coping strategies are, in fact, rationally problematic. On the hard-line reply, there really is, from a rational point of view, such a thing as unfortunate near-misses with depression and remiss resilience to the condition. On this view, a person’s failing to be depressed in sufficiently traumatic circumstances entails that something is wrong with them, just as the same might be said about a person’s failing to experience grief upon the death of a loved one.

How might these claims be substantiated? Consider the following remarks by Graham (1990):

[W]ere a person immune to depression in justifiably depressed circumstances, I think we should be inclined to think of him as psychologically deficient. Such an individual would either be self-deceived about his situation (‘It’s not a death camp but a training center’) or expressing some emotional confusion, or in some way other impaired. (p.419)

Seemingly, Graham defends the hard-line reply. He argues that someone for whom depression is aetiologically rational, but who is not in fact depressed, must thereby fail to comprehend the appallingness of their situation. Being justified for depression, life must be bad, terrible even, for persons like Sam, Priya and Aaron. But not actually being depressed, Graham thinks that they must have little idea of this fact. Accordingly, I take it that Graham’s view is that whatever has prevented such would-be depressives from being depressed is rationally problematic. Whether we are talking about sheer luck (in the case of Sam), resilience (in the case of Priya) or healthy coping strategies (in the case of Aaron), such factors preclude individuals for whom depression is justified from cognising their situation and wider world for what it is (p.419).

What should we make of the hard-line reply? Doubtless there are some for whom Graham’s comments apply. For instance, if sheer obliviousness explains why Sam fails to be depressed, then we should agree that they are psychologically deficient in some way. After all, if Sam is oblivious to great trauma and suffering in their life, then they would seem rationally sub-optimal to begin with. For instance, suppose that Sam is a frontline soldier and exhibits excessive cheerfulness or joyfulness in the face of horrific tragedy and inhumanity. Then, we might suspect that Sam’s broader evaluative outlook fails to accurately track the reality of the situation around them.

However, contrary to the hard-line reply, it is not true that the only way to make sense of Sam’s failure to be depressed in the face of trauma is by giving descriptions of them as rationally sub-par. Perhaps Sam is just built differently. They need not be psychologically resilient per se and they need not be unaware of the difficulties of their situation. For they need not be emotionally unaffected by the challenges they face; they simple need not be depressed as a result of their situation. Failing to be depressed in dire circumstances need not indicate that one is psychologically deficient. After all, Sam may still be upset, worried, sombre, etc.

So, as a general statement regarding those for whom it is aetiologically rational to be depressed, but who are not, Graham’s remarks do not ring true. Indeed, consider Priya and Aaron. As a psychologically resilient person, Priya need not be at all oblivious, self-deceived or emotionally stunted. After all, being psychologically resilient is not the same as being emotionally unintelligent. Moreover, as someone who takes themselves to be vulnerable to depression, it is unlikely that Aaron’s enacting various coping strategies entails, by itself, obliviousness about his situation. Quite the opposite is true: aaron exhibits by his actions a significant degree of self-awareness and attentiveness to his circumstances. Graham is certainly right that psychological resilience and coping strategies may prevent one from being attuned to the horrors of one’s situation. But they need not do so necessarily. A person can be fully aware that life is bad for them, but attempt to meet its challenges head on instead of shrinking away, all without being psychologically deficient.

One way that Graham attempts to support his hard-line reply is by reference to the thesis of ‘depressive realism’ (1990, p.416 though he doesn’t use the term; see also Martin, 1999 and Ratnayake, 2022). Depressive realism is the thesis that through involving a negative outlook, depression attenuates one’s overly positive biases. The result is more accurate and sober assessments of reality and one’s own abilities. According to depressive realism, depression is something of a rational upgrade, as the hard-line reply insinuates. As a result, factors that prevent justified, would-be depressives like Sam, Priya and Aaron from being depressed—whether that is luck, psychological resilience or coping strategies—must be considered rationally remiss. On the line of argument currently being considered, that is not because of their ‘backward looking’ property of preventing justified instantiations of depression. Rather, barriers to depression are now being criticised for their having the ‘forward looking’ property of preventing people like Sam, Priya and Aaron from going on to have a greater number of true beliefs (or more accurate degrees of belief) than they would have otherwise.

The evidence for depressive realism, and which Graham approvingly cites, is the performance of depressed subjects on tests measuring the ability to accurately judge one’s capacity to control certain events, e.g., a light’s turning on (Alloy & Abramson, 1979; see Moore & Fresco, 2012 for a review). Strikingly, depressed individuals have been found to perform better in such tests than non-depressed individuals. The latter subjects were instead found to suffer an ‘illusion of control’, overestimating their ability to influence, and so having false beliefs about their ability to control, the relevant happenings. Since the depressed individuals in these tests had fewer false beliefs, perhaps depression can, from the point of view of rationality, be a good thing.

We shouldn’t be so quick. Although some join Graham in affirming the thesis of depressive realism (Ratnayake, 2022), experiments on the effect have been criticised on multiple fronts, by philosophers (Bermudez 2001, pp.484-6; Ratcliffe, 2014, pp.272-3) and psychologists (Ackermann & DeRubeis, 1991; Allan et al., 2007) alike. Misgivings run the gamut, from concerns about a lack of objective measures in the experiments, to worries about the limitedness of their conclusions, to objectons about their inapplicability to real-life situations. More problematic for Graham is that the effect does not reliably replicate. Some psychologists found no illusion of control whatsoever, in depressed or non-depressed groups (Bryson et al., 1984). Others have reported the opposite effect; namely, that an illusion of control is present in both depressed and non-depressed groups (Kapçi & Cramer, 1999). These results undercut the thesis of depressive realism. They thereby undercut Graham’s support for the hard-line reply that depression is a rational upgrade.

So, Graham’s hard-line reply to the dual reductio objection outlined in 5.i is not only inapplicable to certain people who fail to be depressed, it is ill-supported more generally. Thus, it is worth exploring other solutions to these problems. At the very least, it is worth seeing whether the Mixed View has the flexibility to handle the problems of 5.i in ways that do not require bullet-biting on whether people who have lived through serious trauma but fail to be depressed are, ipso facto, rationally sub-optimal.

6 How to defend the mixed view

In this final section, I develop a different version of the Mixed View, one that offers a novel response to the problems of unfortunate near misses and remiss resilience. This version of the Mixed View departs from Graham’s hard-line version by attempting to vindicate the idea that it is never rationally fortunate to be depressed and that resilience to depression and coping strategies are never rationally remiss. These claims are also affirmed by the Irrationality View. So, if the Mixed View can be shown to make room for them too, then the playing field between the two views will be levelled somewhat.

First, consider that there are a multitude of empirically confirmed, rational harms of depression. As a recent meta-analytic review puts it, depression is “associated with significant moderate deficits across all tasks within the domains of executive function, memory and attention” (Rock et al., 2014, p.2034). A small sample of these cognitive impairments, and some of their implications for rationality, includes the following:

• (1) Depressed individuals perform worse, on average, than non-depressed individuals on tasks measuring the ability to discard false beliefs (Silberman et al., 1983).

• (2) Depressed individuals are more likely, on average, to have ‘overly-general’, i.e. less fine-grained, memories than non-depressed individuals (Evans et al., 1992).

• (3) Depressed individuals perform worse, on average, than non-depressed individuals on tasks measuring the ability to generate mental imagery of the future (Williams et al., 1996).

• (4) Depressed individuals perform worse, on average, than non-depressed individuals on tasks measuring social, problem-solving abilities (Goddard et al., 1996).

Let us take these in turn. (1) suggests that depression puts one at risk of having a greater number of false beliefs than one might have otherwise. (2) suggests that depression risks attenuating the number of propositions about the past that one is justified in believing. (3) suggests that depression risks attenuating the number of futurity-related propositions that one is justified in believing, as well as hampering decision-making and planning, insofar as these may involve imagery (Nanay, 2016). (4) suggests that depression risks interfering with one’s ability to relate to and understand others; indeed, impaired psychosocial functioning is often considered a “core feature” of depression (Rock et al., 2014, p.2035).

Moreover, these cognitive impairments often persist after remission from depression (Hasselbalch et al., 2011). That is, they don’t simply go away once one is no longer depressed. So depression, however else one characterises it, entails significant and long-standing rational degradation. Depression is most definitely not a rational upgrade, as claimed by Graham and friends of depressive realism.

All this leaves defenders of the Mixed View who wish to refuse Graham’s hard-line variant in something of a bind: on the one hand, they claim it is intuitive to think that some people (those who have suffered great trauma) are aetiologically rational in being depressed. On the other hand, we’ve just seen that depression involves rational degradation. So is it rationally good to be depressed or isn’t it? I think that a defender of the Mixed View can see their way clear of this dilemma by adopting a broadly consequentialist view of epistemic rationality.

To begin, compare the rational merit of being in a state that it is aetiologically rational for one to be in, with the rational disvalue of the harms represented by (1)–(4). I think it is clear that the latter greatly outweighs the former. We should agree: it is rationally beneficial to be in a state which, given the circumstances, it is aetiologically to be in. However, this is not to say that depression is, from a rational point of view, a good thing en toto. If the rational costs of being depressed are severe, which they appear to be, going by (1)–(4), then it would be overall better, from a rational point of view, for people like Sam, Priya and Aaron not to be depressed. So although factors like luck, psychological resilience and coping strategies may, for those that have them, prevent aetiologically rational instantiations of depression, these factors are rationally beneficial (in the context of depression at least) since they prevent a dramatic increase in overall rational disvalue for a person. Analogously, it may be financially good to be approved for a loan from the bank and then have the money sitting in one’s account. But it is not financially good en toto if having that extra money causes one to lose money overall, e.g., if the interest and repayment fees are so exorbitant that they wipe out most of one’s savings. This is how I claim things stand with depression. Thus, the Mixed View can accommodate the common sense idea that, when it comes to depression, luck, psychological resilience and coping strategies are good things insofar as they prevent one from being depressed, given all the rational harms that come with it. Most important of all, the Irrationality View doesn’t claim a victory here over the Mixed View. Both views can agree: it is not rationally better to be depressed than not, and luck, resilience and coping strategies are not rationally remiss, but valuable. If you want to affirm these claims, you need not reject the Mixed View for the Irrationality View.

This solution to the dual reductio of unfortunate near misses and remiss resilience bears the hallmarks of epistemic consequentialism (see Percival, 2002; Dunn & Ahlstrom-Vij, 2017; Singer, 2018), a view that has provided a useful perspective on other forms of psychological disorder (Bortolotti & Miyazono, 2016). Accordingly, we can call it the ‘consequentialist reply.’ Here are some important consequentialist features.

First, in handling the problem of unfortunate near misses, the view endorses a trade-off. It says that some rational value ought to be given up, if doing so will secure greater rational gains. The reply doesn’t, or at least need not, reject the idea that it is sub-optimal for a person to fail to be in a state that, given the circumstances, it is aetiologically rational for them to be in. But the reply insists that the additional rational good of actually being in that state may be sacrificed, if actually being in that state means being in a situation that is overall rationally worse for the person.

Second, in handling the problem of remiss resilience, the reply says that barriers to depression like resilience and coping strategies are rationally valuable in a particular way; namely, teleologically, in a forward-looking, means-ends manner. In particular, resilience and coping strategies are pictured as rationally valuable insofar as they are tools for the creation of rationally good situations and the evasion of rationally bad ones. That is, the rational goodness of resilience and coping strategies is (in this context) a matter of them being means for avoiding the rational harms of (1)–(4) being actualised by virtue of them preventing depression from being instantiated.

Third, the reply is couched in terms of totalizing the relevant rational outcomes. When thinking about depression, the reply urges us to focus on total rational goodness and badness. The reply claims that the degree to which a person’s overall rationality would be compromised by depression, as evidenced by (1)–(4), eclipses the rational goodness of being in a state that it is aetiologically rational for one to be in. It is important to be clear on how this last claim does not go against the idea at the heart of the Mixed View: depression can remain an aetiologically rational response to the (apparent) events in one’s life; it is just that, it is overall rationally sub-optimal to be depressed, given (1)–(4). These claims are not contrary to one another since the latter concerns the total rational value of a situation in which someone is depressed, while the former concerns the rational value of one element of the situation: its aetiology.

In sum, then, contrary to appearances, the Mixed View is not forced to say, with Graham, that barriers to depression e.g., luck, resilience or coping strategies, are rationally problematic. It is epistemically rational for people who have experienced trauma to be depressed; yet at the same time, such individuals are not rationally sub-optimal if they fail to in fact be depressed. If one finds this combination of claims attractive, then the Mixed View, coupled with epistemic consequentialism, shows how they may both be affirmed. The Irrationality View need not be affirmed along the hard-line manner endorsed by Graham.¹⁵

7 Conclusion

Where the Irrationality View sees in depressed individuals only rational failings, the Mixed View seeks to acknowledge the dreadfulness of some people’s lives. Sometimes things can be so bad as to make depression a reasonable response from the point of view of epistemic rationality. Those who are depressed need not be responding incorrectly to the facts as they see them—they may be perfectly attuned to the horrors of their situation. I have argued that this view is a serious competitor to the Irrationality View in terms of how we ought to understand the relationship between depression and epistemic rationality. In particular, I have argued that the Irrationality View does not obviously receive greater support from the current clinical literature than the Mixed View. Moreover, I have argued that the Irrationality View assumes a simplistic account of depression’s phenomenal character. In terms of supporting the Mixed View, I have argued that epistemic consequentialism offers a novel response to two problems that beset one of the few existing defences of the theory, i.e. Graham’s. Granted, this still leaves a great deal up in the air, regarding aetiological rationality, depression itself, epistemic consequentialism, and so on. I leave these challenging questions for another time.

In sum, the Irrationality View does not obviously deserve to be the preferred account of the rationality of depression. The Mixed View is equally as attractive, if not more so in certain respects.