ABSTRACT
A number of very important challenges to the separation distress hypothesis of depression were discussed, clarified, and linked to various strands of existing literature. Foremost among those is the critical challenge to define a more comprehensive set of clinical precipitants, and the possibility that depression may have two prototype elicitors – social defeat and social loss, with hugely overlapping neurobiology, and currently inadequate differentiation in terms of biomarkers and fundamental neurobiology. The possibility of synergism versus exclusionary arguments around about various evolutionary and traditional psychiatric perspectives was advocated. Additional and relevant considerations around mammalian PLAY as an opponent process to depression, and the challenges posed by social constructivist views of emotion were also explored, as these were both neglected in the target article. Social constructivist views and prototype emotion theory are not incompatible; how cognition extensions of prototype ‘elicitors,’ as well as cognized versions of classic prototype affective behaviors might operate is also discussed. There is no way to defend social constructivist views of emotion in animals, and the enormous continuity – from subcortical architectures, neuromodulators and behavioral homologues – argues loudly against any assumption of a fundamental discontinuity between mammalian and human affective endowment.
Introduction
Let me just say first and foremost that I am immensely grateful to all the commentators. Although complements about the target article and validation of its argument are always appreciated, I am more grateful for the respectful challenges, criticisms, and scientific debate. These define how all good scientific discourse must be conducted – and they help immeasurably in the sharpening of arguments, the identification and discovery, and then potential mitigation of holes in arguments and data sets, and in the defining and refining of future predictions and empirical probes that can further modify and augment the original core group of hypotheses. Indeed, without incorporation of and then response to critical challenges to core hypotheses, it is hard to see how science would advance much at all! While I believe we are well past a situation in which any current data set on the table so to speak in either animal or clinical models truly falsifies the separation distress hypothesis of depression in a major way, I believe that the commentaries helped define a number of significant challenges, poorly defined aspects of the primary argument, and ways in which the argument might be tested, amplified, and strengthened. These might include (as a partial list, and in order in relationship to the responses to comment below but not necessarily ordered in importance):
The refining of the “prototypical stressor” question: the possibility that depression may indeed have two prototypical drivers – social defeat and separation distress, with very substantial overlap in their neurobiology but inadequate differentiation of biomarkers to discriminate them. This is an important question on the table (how can one distinguish neurobiologically between acute social defeat versus acute social loss – see details of my response to commentaries by Badcock and Friston (Citation2023);
That evolutionary and traditional “pathology-focused” psychiatric perspectives have been talking past each other but can come together in the target article argument that it is actually the disinhibition of the conserved mechanism that leads to depressive illness, overcoming the problem of depressive illness not being adaptive while not yielding the explanatory gaps found in traditional psychiatric perspectives that fail to account for depression’s ubiquity;
That many if not most evolutionary arguments, in an unfortunate lockstep with traditional psychiatric perspectives, have also shown the most unfortunate conflation of sadness with depression, thus (con)fusing and obfuscating the question of their potential adaptive bases;
As a general concern, cutting across numerous domains of biology and neuroscience, the problem of “knowledge silos,” resulting in artificial and often “tribally-maintained” distinctions between disciplines, perspectives and theories continues to exert a subtly and for many invisibly distorting influence on science. I have argued in other domains of neuroscience (specifically in relationship to our systematic failure around neurodegenerative disorders – Watt, Citation2014a) that such knowledge silos have a real and measurable negative effect on scientific progress. This is seen particularly in terms of the widespread and often unchallenged caricaturing of psychoanalysis, without appreciation for its massive contributions in a number of both psychological and more general health biology areas. This includes the role that early observational studies by psychoanalytic clinicians had in the elimination of warehousing of young infants (creating the syndrome of hospitalism), a psychotherapeutic method focused on early attachment and relationships, their internalization and their recapitulation, concepts now widely disseminated as virtual cultural memes, and the first truly systematic affect-centric theory of mind, even if that early articulation was muddled around the modeling of drives and instincts. Knowledge silos and artificial but “tribally” maintained opposition between psychological theories that must be eventually integrated and reconciled include the ongoing debate between prototype emotion theory and constructivism, as outlined in the response to the commentary by Davis and Montag (Citation2023). An increasing challenge for all of science but particularly for those working within neurobiological and psychological science is the sheer scale and massive size of the research and relevant knowledge bases – these make bridging and interdisciplinary work harder simply because many authors are unaware of all the many relevant findings, as the task of staying current even in a small area of science is formidable, while staying current in multiple disciplines is virtually impossible;
The critical role that various forms of play (both rough-and-tumble as well as cognitive humor) might have as an opponent process offered by the PLAY system to depression – neglected for reasons of space in the target article but fleshed out in the response to Siviy’s valuable comments;
How constructivism might be reconciled with prototype emotion theory, without harsh consequences to affective neuroscience concepts and without denying the informative and even transformative role of cognition;
How classical attachment theory might interdigitate with Panksepp's categories for prototype affective activation and work by Davis and colleagues on personality types to potentially inform more nuanced nosology and classification of attachment disorders beyond the classical concepts (as outlined in the response to commentary by Howard Steele); and
As an overarching principle – a poignant reminder of the raison d'être of this Journal – we should have a default hypothesis that the best kind of science of the mind/brain, one that bridges levels, practices a healthy and reverent versus sledgehammer reductionism, and respects complexity and emergence, will come from our success in bridging many hidden “knowledge silos”. As perhaps the latest potent example, a recent scientific meta-analytic study (Packheiser et al., Citation2024) in front page news (https://www.nytimes.com/2024/04/08/health/physical-touch-benefits-premature-babies.html) demonstrates that consensual touch is a powerful health variable, including its protective effect on risk for depression. This fine paper and its discussion section is without any mention on the part of the authors that there is anything like a “separation distress system,” with its potent mechanistic interdigitation with the stress and immune axes of the mind/brain, as a possible to probable nexus for the effects of prosocial touch. In any case, the focus for study in the meta-analysis itself – that touch and therefore social connection is a critical health variable – reflects the real softening of traditional knowledge silos between psychology, particularly the psychology of social bonds and social attachment, and biology. This softening of boundaries is similarly reflected in the paradigm-shifting emphasis, referenced in the target article, and documented in Holt-Lunstad and Golant (Citation2023), that secure social connection and social support is a primary health variable. This is all to the good, and as I indicated in my closing argument in the target article, an encouraging sign of very substantive progress towards affirming a much more social view of human existence, and of human health and illness.
Response to Paul Badcock and Karl Friston
The separation distress hypothesis vs. other social theories of depression? – Not a zero-Sum game – let's agree to not miss the forest for all the trees
It is a privilege to have the commentary and thoughts of Drs. Paul Badcock and Karl Friston (Citation2023) – and I appreciate both their gracious comments about the paper’s strengths, and their constructive challenges about its potential weaknesses. The challenges are especially valuable. Many of their expressed concerns and misgivings require significant unpacking and detailed responses, so please indulge the length of this response. I've created separate sections to frame and respond to their major concerns. Despite some of their concerns, I believe we are not that far apart – hopefully these clarifications will allay many of their concerns, confirm our areas of strong agreement, and point to ways that any remaining differences might be empirically adjudicated.
Can depression be explained adequately solely in terms of either separation distress or psychodynamics?
Let's start with their first challenge (second paragraph of their commentary) that “depressive presentations cannot be adequately explained solely in terms of either ‘separation distress’ or psychodynamic constructs.” Let me take on the first point down below in more detail, but I have to confess I am not quite sure how to respond to their second point, as I never claimed that depression was a defense, and I don't believe that any of the essential points of the separation distress hypothesis rely mechanistically on any psychodynamic concepts at all. So perhaps I missed something on this?
What I did note, and what is critically important to grant for the scientific record, is that psychoanalytically-oriented clinicians and theorists were the first in modern psychology to outline an intimate relationship between loss, especially traumatic loss, and depression. Indeed, classic work by Spitz (Citation1946) and Pinneau (Citation1955) on “hospitalism” (a term describing a refractory and devastating type of depressive withdrawal in infants and young children that had grave prognostic implications not merely for psychological development but for survival) goes strongly against the notion, presented in their commentary, that depression is mainly, or even primarily, relevant in adulthood. This still hugely under-referenced syndrome of hospitalism becomes some of the strongest prima fascia evidence that social bonding is a sine qua non in infancy and early childhood, as later argued by Bowlby (Citation1973, Citation1980), and the so-called “object relations school” in British psychoanalysis. Social bonds, in these early descriptive studies of profound neglect, were shown not simply important for psychological health, but likely essential for life and for stable life-regulation. Among other vitally important accomplishments, articulation and description of this devastating syndrome of hospitalism helped to finally end the grotesque warehousing of orphaned children and infants, long assumed benign, if infants and very young children were just physically cared for in terms of nutritional and other bodily needs. And although hospitalism can never be studied ethically, its associated lethality points to some form of fundamental homeostatic collapse (presumably hypothalamic, from the common feature of energy balance/appetitive collapse?) in the context of a devastating disruption and failure of any viable attachment process. I believe that this serious risk for early mortality was the underappreciated developmental antecedent, in a more vulnerable creature, for the major health effects, both for wellness and illness, in adults, for those securely connected versus vs. those who are not (discussed in pp. 146–148 of the target article and powerfully summarized in Holt-Lunstad & Golant, Citation2023). Additional important data points bearing on their skepticism about the powerful relevance of separation distress concepts for depression in childhood might include evidence for increases in depression in childhood (Lebrun-Harris et al., Citation2022), which are surely multifactorial, but where COVID-related isolation is a large factor, an association obviously supporting our central hypothesis, and also where the sibling relationship outlined in the target article between sickness behavior and depression may be critical, particularly where sickness behavior, temporally conjoined with loss of multiple rewards and social isolation, has become an all too frequent pathway into depression. Also not consistent with their challenge (that depression concepts are mainly relevant to the struggle of moving out into adulthood) is additional data showing that childhood and adolescent depression incidence, even pre COVID, isn’t far from many rolling estimates of adult penetration, with the same strong caveat that these, like the adult numbers, are likely serious underestimates (Bitsko et al., Citation2022; CDC, Citation2024) (with COVID clearly blowing these numbers up both from the stress of isolation and the intrinsic linkage of sickness behavior and depression).
The second and related point of skepticism is “the risk of infantilizing (italics added) depressive reactions by extrapolating a response to attachment dynamics – observed in infancy – to explain depressive outcomes later in life.” This challenge requires two clarifications of the impressive evidence base against any notion of a fundamental mechanistic discontinuity for depression in adults versus infants and children: (1) first of all, the environmental risk factors for depressive outcomes in adult life as outlined and summarized in the target article (pp. 126–130) have much to do with the prototype stressors of social loss/separation distress, as documented in previously quoted work by Slavich and colleagues in Citation2009 as well as Kendler et al. in Citation2003; and (2) even more importantly, the critical sensitization process linking early trauma and later vulnerability to depression is likely intimately related to and indeed driven by forced maternal separation, as shown in the animal models of Hennessy et al. (Citation2019), where they show that in rats and guinea pigs but also in primate animal models, early separation stress sensitizes and primes the innate immune system towards a disinhibited pro-inflammatory state, and also disorders the HPA axis regulation in the direction of glucocorticoid resistance (perhaps first outlined in Hennessy et al., Citation2001, then in Roque et al., Citation2014 showing a critical window and Sanchez et al., Citation2006 in primate models; see Hennessy et al., Citation2019 for extensive updated review and summary across several models). These animal model findings strongly align with the biomarkers identified for the effects of early abuse and neglect, especially disinhibition of the innate immune system but also GC resistance/HPA dysregulation (Nemeroff, Citation2016). Hennessy et al. (Citation2019) summarize this work by stating (p. 5) “HPA axis responses (to multiple forms of stress) may thus prime inflammatory responses to subsequent challenge” while separation trauma early on appears to prime sensitization and vulnerability to depressive stressors also though potentiation of pro-inflammatory signals in reaction to stressors.
Perhaps the authors believe that just or only in the human clinical situation, distinct from any animal model, that the adverse childhood experiences (ACE) priming of depression in later life occurs independently of any induction of potent or penetrant forms of separation distress. This possibility, which seems frankly unlikely, was also addressed in the review – both abuse and neglect but especially neglect appear to create penetrant forms of aloneness, likely activating separation distress circuitry, although one has to confess that there is a complex and not synonymous relationship between being alone and feeling alone. But still, given all this data aligning animal and human clinical biomarkers of abuse with known effects of separation distress, the authors' reluctance to connect the dots (separation trauma functioning as a critical early primer of vulnerability to later depressogenic stressors that may also involve the critical loss of social connection and comfort) is not easy for me to follow. (See later section on ACE for further elaboration on this question.)
Additionally, although they state at the end of their commentary that “separation distress is one of a suite of strategies (italics are mine) that have been favored by selection to return the individual to the state of social homeostasis,” I would be hard-pressed to describe another evolutionarily conserved system other than proto-empathy/maternal CARE (indeed the direct complement system to separation distress) that has equal importance in the restoration of social homeostasis for separated individuals, especially in our isolating-atomistic technological societies – perhaps an important part of the explanation for why depression has been such a difficult problem for Western healthcare systems (Hidaka, Citation2012; Smith, Citation2014).
The separation distress theory vs. other evolutionary perspectives … not zero sum?
The authors question whether they should “privilege” my account over other evolutionary views. I would say most definitely not! We have for far too long been seduced by the notion that not only evolutionary but other mechanistic and molecular explanations for depression are “either/or,” in some version of a zero-sum game, instead of pieces of an often times too-complex to fully grasp truth (and as argued in the target article, factors operate in strongly linked recursion rather than as competing isolated theoretical mechanisms). This is fortunately changing, as even mainstream views (Nemeroff, Citation2020) acknowledge this multifactorial nature. Additionally, in our original target article, Panksepp and I explicitly outlined how a separation-distress-shutdown benefit of depression could cascade forward in adolescent and young adult creatures into a losing-dominance-withdrawal benefit, while we explicitly eschewed simple primary factor explanations in understanding the molecular neurobiology of depression.
However, this question indicates to me that I may not have clarified my evolutionary argument or its continuity and overlap with other evolutionary views adequately. They raise the possibility that my position might be non-trivially contrasted with: (1) the social navigation hypothesis of Watson and Andrews (Citation2002), as well as the social risk hypothesis of Allen and Badcock (Citation2003) and the social safety theory of Slavich and Irwin (Citation2014). Once again, careful review of these sources suggests synergism rather than competition. A global reduction of social risk (as hypothesized by Allen and Badcock) is not far away from Nesse's dominance hypothesis, and of course both of these are consistent with the assumption that a depressive mechanism might shut down a variety of problematic negative signals/states, either current or “prospective” (around the corner so to speak and a likely contingency which the depressive shutdown might mitigate). The reference to social safety theory as an alternative to my hypothesis is frankly a bit puzzling, given my extensive referencing of and acknowledged intellectual debt to George Slavich's work, and, most importantly, that I see almost no distance between his view and my own. Indeed, I regard Slavich's work very highly, and would argue that he has taken Bowlby's attachment theory and put a very substantial amount of neuro-immunological and neuro-endocrinological meat on the basic conceptual/psychological bones from Bowlby (that secure attachment is a psychological necessity from cradle-to-grave). So, I’m not sure that any real or substantive conceptual distance the authors appear to see, at least around my account vs. some of the other social brain evolutionary accounts, frankly exists.
Part of the problem is seduction by differential terminology, which compounds and interdigitates with the larger problem of siloed knowledge, especially between psychological and psychoanalytic vs. neurobiological perspectives – I regard the seemingly disparate notions of “social safety” (Slavich's term), “social comfort” (Jaak Panksepp's term) and “secure attachment” (Bowlby's term) as virtual synonyms, but offered by scientists and clinicians in venues that mostly don't read work in the other venues. For example, in George Slavich's (Citation2020) excellent articulation of social safety theory, neither Bowlby nor Panksepp gain a single reference. Does this mean we should downgrade our estimate of Slavich's work? I don’t believe so, but it is clear and powerful testimony to how siloed knowledge really poses challenges for progress in science – particularly if we simply assume different semantic formulations for the same process are truly oppositional.
Some real challenges around evolutionary perspectives – not all rosy agreement!
Despite this potentially broad consilience, it is not entirely clear how at least some aspects of these various evolutionary theories might integrate. For example, the notion from Watson and Andrews (Citation2002) that depression might show a “social rumination benefit,” as depressive cognition oftentimes contains massive psychological distortions of the social space that help ensure deepening isolation and psychological peril (see discussion of implications of the separation distress hypothesis for psychotherapy in target article). So, it is frankly hard, without any empirical testing of this, to know for sure that the “rumination” in depression actually yields anything other than people spinning their wheels. It's an open question how much people “ruminate meaningfully” on what led up, for example, to their defeat or loss (which may create true social learning opportunities), as opposed to focusing on various depressive cognitions (that they are simply “failures” or unlovable) – but also see discussion of this in last section of this response. (See also the invited commentary from Dr. Blinder who also raises the issue of rumination.)
Additionally, many evolutionary theories of depression (including work by both Watson and Andrew (Citation2002) and Hagen’s (Citation2011) overview of evolutionary theories of depression) continue to conflate sadness with depression – the same serial error made by psychiatry at large over generations of DSM. This means a conflation of the adaptive signaling into the social space of the saddened but perhaps not yet depressed person’s need for support, which has obvious massive benefit actually in the direction of the prevention of depression (!). I would argue that although depressed individuals may garner some support (but this is hugely dependent on environmental contingencies and hardly universal), openly sad individuals garner significantly more, and many depressed individuals are isolated or isolative and get little if any support. This conflation of the benefits of depression with those of sadness is quite regrettable, and can only contribute to further muddying of what are already turgid waters. It indicates that memes born of years of DSM conceptual mistakes and conflations (already extensively unpacked in the target article) continue to befuddle scholars of depression in many related disciplines, who may lack critical distance from these problems. This conflation is extensively discussed in the target article – in early stages of depression particularly following classic precipitating losses, individuals are sad and also becoming depressed. As they become more depressed, however, the sadness often times dries up and is replaced by simply feeling empty and hopeless. Sometimes individuals remain both sad and depressed – particularly in individuals where severe shutdown of the SEEKING system (indexed by degree of apathy and abulia) is not a prominent feature – as seen more commonly in anxious phenotype depressions.
Additionally, many evolutionary theories fail completely to distinguish between various putative adaptive benefits of a brief depressive shutdown from a lethal depressive illness. In the former, we “slink back to our cave” so to speak, but then we can and might retool, rethink and then reengage more effectively (often times with the benefit of a base of reestablished social support with greater appreciation for the dynamics of our loss or defeat). But the latter case (a potentially lethal illness), as I have argued in the target article, there are many possible and underappreciated positive feedback loops, not the least of which is a serious behavioral/affective withdrawal promoting a deepening emptiness and hopelessness, as loss and/or defeat cascades forward into the increasing failure to secure any rewards, once again particularly in those with severe global motivational deficit. This is a major blindspot in many evolutionary theories of depression – they fail to define or account for this critical functional gap between an adaptive mechanism, activated in a delimited form in a normative environment where the evolutionarily normative or predicted environment helps promote opponent processes, versus the trajectory when that same mechanism is removed from its normal terminators, due to a nonstandard, non-evolutionary environment and/or from massive “doses” of its prototype activators, which then goes on to have negative/morbid functional consequences.
Once again, the syndrome of sepsis is highly informative here – how could it be that a heavily conserved and selected set of mechanisms (immune defense and pro-inflammatory signaling) is quite capable of killing us? This of course is absolutely parallel and not unrelated mechanistically to the challenge of how a depressive mechanism that might be adaptive at times also yields a lethal illness. As with sepsis, this happens in the context of: (1) the intrinsic, profound (and often times underappreciated) functional trade-offs between various genetically encoded behavioral/immunological vectors (sickness behavior vs. normal default SEEKING modes – see extended discussion of this on page 125, second to last paragraph of the target article); (2) massive doses of classic inflammasome activators, yielding disinhibition of cytokine response, sufficient to derail even autonomic systems, which is increasingly implicated in lethality (“septic shock” – see Carrara et al., Citation2021); (3) the failure of opponent processes (including the adaptive immune system), which is likely multifactorial but where #2 is again important. The most important part of this puzzle box to grasp might be the baked-in functional trade-offs between selected mechanisms, such that they are in some important ways mutually antagonistic (discussed in detail in the target article on page 125 paragraph 3). Indeed, this distinction is critical – between the evolutionarily selected and putative “adaptive payoff” of the mechanism if activated in a delimited form vs. the deadly consequences of its over-recruitment with a protracted and sustained loss of functional balance. Many proponents of evolutionary theory sometimes forget that evolution crafted many counterposed and indeed antagonistic mechanisms – and that the maintenance of life is dependent on a context-sensitive activation of the “right” mechanism at the “right” time. Life appears to rest in a staggering collection of counterposed and counterbalancing “opponent processes” – a giant collection of rubber bands, if you will, keeping the system in equilibrium. This principle of “counterposition” has never been made more clear than in the paradoxes of addiction, where high doses of addicting drugs recruit a bunch of opponent processes rendering the addict completely miserable from a drug that once initially brought euphoria and affective relief. This group of opponent processes extends beyond the classically appreciated down-regulation of receptor sensitivity, as this is only one of many serious perturbations in substance abuse. My point is simply that functional trade-offs between a multitude of counterposed mechanisms are absolutely intrinsic and not incidental, a lesson that anyone who just had a too large meal and then tries to go out to exercise finds out immediately. Or as someone once put it, “you can't optimize for everything simultaneously.” This common explanatory gap in evolutionary theories of depression is part of why many in mainline psychiatry may summarily dismiss evolutionary arguments out of hand as simply naïve – pointing to the obvious and overwhelming evidence that depression often is a destructive, damaging, and in some cases lethal illness (Nemeroff, Citation2020). These distinctions go to the heart of why evolutionary and traditional clinical psychiatric perspectives on depression often talk past each other. They each have an important piece of the truth missed by the other perspective.
I would reference particularly my comments in the target article (page 144, last paragraph) about a highly relevant and underappreciated scotoma in our neurobiological understanding of what happens to animals in the “despair phase” when they are reunited, typically thawing out rather quickly. That this particular transition has not been studied may explain more about our treatment failure in depression than any other single hole in our understanding. And it absolutely goes to the heart of this distinction between an adaptive benefit resting in a delimited activation of a mechanism and a lethal consequence from an unrestrained and unopposed disinhibition of that same mechanism – the critical distinction between depression (as “brief depressive reaction”) as a conserved and protective mammalian shutdown mechanism with kindred relations to hibernation and sickness behavior, vs. depression as psychiatric illness with huge morbidity and mortality – and where this latter scenario must index the hidden but critical failure of opponent processes. A highly relevant question is, “how many folks with serious and refractory depressive illnesses are fundamentally socially isolated, understimulated, and virtually without any emotional support, and also suffering from classic pro-inflammatory lifestyle variables such as sedentary lifestyle, dysbiosis/pro-inflammatory diet patterns, poor sleep, etc.?” While Tom Insel's recent monograph (Citation2022) on the failure of our mental health medical industrial complex in relationship to chronic psychiatric illness acknowledges some of this, he fails to outline that this is not merely a failure of nonspecific support or even support systems in the society, it is mechanistically relevant to the syndromes we are trying to combat, part of the evidence that perhaps all chronic illness are at least in part about an “evolutionary discordance” (Cordain et al., Citation2005; Eaton & Eaton, Citation2002; Pontzer et al., Citation2018; Watt, Citation2014a). That we don’t have much if any systematic data on such obvious issues (what percentage of folks with major depression and particularly with refractory depression suffer from chronic social isolation and pro-inflammatory lifestyle variables) speaks to the neglect of the core relationships that the target article sought to expose. Put differently, how many folks in the early stages of a depressive shutdown have anything like a comfortable nest full of compensatory parental attention, (much licking and grooming!), the happy chirping of siblings at the reunion, and olfactory signals of safety to come home to, as they are starting to shut down and give up? While this might seem like a strange thought experiment and a weird mixed metaphor, it outlines a huge scotoma in our research approaches to depression. This question of what might constitute the naturalistic “depression terminators” and opponent processes has not even been asked in a systematic fashion within our mental health industrial complex but this again is a complementary blindspot to seeing depression as only an illness. But could this thought experiment help explain why we have had so little success and progress in the treatment of depression? (For further thoughts on a central opponent process to depression that is not widely appreciated as such, namely mammalian rough-and-tumble play, see commentary by Dr. Siviy and my response.)
Last but not least, I find it both surprising and regrettable that many evolutionary theories, despite the striking conservation of mammalian social behavior, fail to consider the homeostatic nature of social attachment, and the related probability that modern environments create an “evolutionary discordance” – a significant gap between our basic need for social connection and what is readily/culturally available (Hidaka, Citation2012; Slavich, Citation2020; Watt, Citation2014b). Indeed Jaak Panksepp's entire career and work showed us just how profoundly a small group of core social-emotional systems are conserved across all mammals – and one would suspect, if anything, systems supporting social bonding received upregulation in primates lines and then further upregulation in hominids (evolutionary “extensions”), with tribal cohesion perhaps one of the most critical of those evolutionary extensions and appearing essential to both primate and hominid success (Samson, Citation2023).
Social comfort is the ground state, to which the system seeks to return, with the intensity of motivation linked to the severity of departure from the ground state – this is the essence of a homeostatic mechanism. Of course, this begs the question as to what constitutes social comfort beyond simply physical safety and even more importantly and mysteriously, how the brain transduces and detects social comfort versus its absence – and Jaak Panksepp and I talked about this several times as a subject for much-needed research and review. The default hypothesis might be that in adult humans, social comfort emerges from a complex amalgam of empathy and social support relieving/modulating negative affects, while play, affection, and other positive affects are activated regularly (creating both a “safe” and a “good” space). In infant mammals, we must suspect and hypothesize a significantly more primitive version of social comfort emerging from a confluence of familiar olfactory signals, positive affect vocalizations (such as the well-studied 50 kHz vocalizations in rats), thermoregulatory signals, and perhaps especially the poorly appreciated C-fiber transduction of maternal licking and grooming, which has massive neurodynamic impacts in upper brainstem, limbic and basal forebrain systems (Panksepp, Citation1998; see also Sakamoto et al., Citation2021 for some of the flip side of this – how low (licking and grooming) mothering disinhibits nociception). That individuals will risk their lives to save the lives of and therefore their connections to their loved ones argues that this is a motive force without equal. Indeed, in view of the authors' preeminent role in the application of concepts such as Markov blankets and related prediction error concepts to neural networks, it is surprising to me that they do not endorse this kind of basic homeostatic concept of attachment themselves, a construct from which the primacy of separation distress concepts must logically emerge. If social comfort is a homeostat, separation distress is not one of many factors, as the authors appear to believe (“one among many equals”) promoting social homeostasis, it is the primary factor operating within the individual (and directly hinged to and complemented by empathy in a potential caretaker – and it is not a coincidence that separation distress and maternal CARE/proto-empathy are two critical and highly conserved prosocial affective/executive systems in the mammalian brain). Indeed, these systems are far more complementary than generally believed, such that maternal animals probably have their separation distress circuitry upregulated in relationship to vulnerable offspring who are always wandering into trouble so to speak, although this has not been well researched (Panksepp, Citation1998).
Is separation distress at the root of all depressions? Two cardinal depressogens?
As for some version of separation distress being assumed in my argument uniformly at the root of all depressions, I don’t believe that conserved mechanisms exist to serve a very narrow class of stressors but rather broad classes with several “prototypes” that presumably have some kind of “stressor continuity” or overlap (paradigmatically, there is no single foreign protein or pathogen signal to which the Toll-Like Receptor responds in its activation of innate immune response, with at least 8 known versions of the receptor that are differentially sensitive to pathogen molecular patterns). Separation distress in all its many forms is still a clear prototype for eliciting depression neurobiology. But I would readily grant, and did so in the paper, that there is more to depression induction than loss and separation trauma. I believe that at various points I clarified that social defeat and dominance failures (perhaps more critical for depression in males than in females?) – which are still fundamentally related to social comfort/security and therefore not unrelated to separation distress – were also potent depressogens. Indeed, I mentioned that chronic poverty is a form of “chronic social defeat” – helping to explain why it also is potently depressogenic. In this vein, I was a little bit surprised that the authors thought I was still shorting evolutionary perspectives on depression related to social defeat and dominance (as the term “social defeat” appears 6 times in the target article, while the term “dominance” appears 10 times).
But I believe their criticism does point to how I did not adequately specify how these two classic “depressogens” of dominance failure/social defeat on the one hand, and social loss/separation distress on the other hand might be mechanistically connected: (1) significant reductions in opioidergic and oxytocinergic tone, which are characteristic of both submissive or defeated animals, as well as those undergoing separation (Panksepp, Citation1998, who emphasized that these classic prosocial peptides don't just terminate separation distress vocalizations, they also build social confidence in the direction of dominance, a fact which explains some of the otherwise surprising ways that oxytocin has been shown to be an anti-empathy molecule); (2) upregulated pro-inflammatory cytokines (Weber et al., Citation2017); (3) stress axis alterations, with typical but perhaps not universal hypercortisolemia/impaired GC feedback (Keeney et al., Citation2006). This very broad confluence across multiple conserved neurobiological systems of course raises interesting questions about what neurobiological markers might actually differentiate these two large classes of depressogen (defeat versus loss). I don't think we know. I do think that the evidence for such large overlapping effects from both social defeat and social loss as depressogens speaks to how we need either caretakers and close conspecifics/mates in some developmental stages while needing allies and subordinates (and therefore possibly tribes?) in others – in potentiating a depressive trajectory out of the negation of these two sets of needs, evolution may be commenting that, absent these critical resources, it might be better on average to shut down. These considerations, and the large neurobiological mechanistic overlap show that the connections between social loss and social defeat in the target article as depressogens are not, as they state, just “extrapolative” (page 161, last paragraph of their commentary), but are both substantive, and biologically mechanistic. An interesting unanswered question is what might actually differentiate social defeat and social loss, given such large mechanistic overlap in their biological footprint in depressogenesis (shared impact on HPA axis, upregulated innate immunity, lateral habenula dynamics, and down-regulated prosocial peptides). One candidate might be the upregulation of cholecystokinin (CCK) in social defeat (Becker et al., Citation2008), but this remains an important neurobiological question.
Of course, granting social defeat and social loss as approximately equal vectors, depending on the development stage, is still a fundamentally even profoundly social concept of depression, as both dominance failure and loss of a primary attachment are about our social and tribal embedding and status – security and social safety broadly defined. Additionally, I took pains to extend precipitating stressors to include anything activating guilt and shame (shame being a prototype response to dominance failure). I argued that both guilt and shame were cognized extensions of separation distress that require a relevant and functional self-mage and higher cognitive processes not available to infants, birds, and most other mammals. Like Badcock and Friston, and as I have clarified above, I'm also reluctant to embrace any notion that a single class of stressors fully explains depression – and I spent quite a bit of time in the target article clarifying the frequent trajectory into depression provided by its sibling shutdown mechanism, sickness behavior (again see Hennessy et al., Citation2019 for a more up-to-date summary of evidence linking pro-inflammatory priming of the immune system in early social stress and subsequent vulnerability to depression). All this becomes further compelling evidence of the evolutionary “sisterhood” of the two cardinal behavioral shutdown mechanisms of sickness behavior and depression.
However, I would readily grant that my heavy and perhaps for some too monocular focus on separation distress may be correctly interpreted as a form of “affirmative action,” for which “I beg the indulgence of the court.” It has been neglected as an epicenter in depressogenesis for far too long, and the literature reviewed in the target article makes it clear that the immune, endocrine, and neuropeptide correlates of depression are tied directly to the impact of separation distress on those systems. Indeed, the neuroscience in those areas is well past any skepticism that immune, HPA/endocrine, and social neuropeptide systems form the most critical triad of biological systems registering the state of the social world, and that these systems and their complex resonances with other conserved systems form the foundations for how social comfort vs separation distress and social defeat are transduced in the CNS. Prototype social stressors that are potently depressogenic centrally feature personally-targeted rejections as the single most potent depressogen identified empirically, as shown by Slavich et al. (Citation2009), and as discussed in detail in the essay, pp. 126–129. This does not remotely contradict the separation distress argument, but it does point out how events that create both lowered self-esteem/narcissistic injury conjoined with object loss may be the most powerful inducers of depression. When we come to believe that either our “badness” (guilt) or our inadequacy or defective nature (shame) ensures and guarantees that others do not and even cannot love or respect us – this may be a most depressing and untenable position, seeming to lock in separation and aloneness without hope of mitigation. Indeed, might these be examples of the most depressogenic possible predictions/ “dysfunctional priors” favored by the authors (see last section of this response)?
On the interpretation of how and why child abuse is such a potent predisposing factor
In any case, where the authors state that the “vulnerability to depression in adulthood might be related to childhood abuse as a vulnerability factor, without needing recourse to separation distress per se” (page 161 paragraph 3 of their commentary), they would need to: (1) produce solid evidence that the precipitants for depression in adulthood commonly or at least significantly separate out from the various forms of social defeat, social loss, and other classic factors as described above; and (2) refute the now substantive evidence that child abuse disinhibits the innate immune system, dysregulates the HPA axis, and downregulates prosocial peptides (see Nemeroff, Citation2016), exactly as protracted separation does in animal models. Their comment suggests the older and now largely refuted notion that abuse is some kind of “nonspecific vulnerability factor” for depression, unrelated to its induction of traumatic aloneness (separation distress), and unrelated to the causal neurobiological alternations in critical relevant systems outlined earlier (see Hennessy et al., Citation2019 for summary). I see no reason to believe that there is such a “nonspecific” effect of abuse – this appears beholden to the older and now discredited idea that stress or trauma simply causes psychological problems, or is a “software” or learning issue. It is now very clear that this is simply mistaken.
From both intermediate and long-term psychotherapeutic work with many hundreds of severe trauma victims and as described in the trauma literature (van der Kolk, Citation2015; see also work by Carolyn Spring), I believe that abuse generates a terrifying core of both fear but also unbearable aloneness – in a fashion that suggests deep linkages between these two types of affects. One is not only facing a life-threatening situation but also without any protective allies and without resources. Indeed, the problem exerting so much “shearing” force on the psyche is that the persons we might wish or expect to be protective are often doing the abuse, a shearing force of intolerable ambivalence, which extends from the abuse forward into all other relationships as a hidden systemic poison of basic trust. Instead of secure attachment at home, replicated in new relationships as that individual is moving out into the world, abused individuals struggle to detach from an abuser to escape the abuse but then are often swamped by aloneness (separation distress), thus oscillating between withdrawal and return to the abuse(r), or they find a new abuser with whom to recapitulate the childhood abuse and the chaotic cycles of withdrawal and reengagement. An enormous volume of literature validates that the PTSD spectrum predisposes to significant aloneness/loneliness, especially for folks with serial trauma and so-called complex PTSD, with subsequent heavy comorbidity with depression (Barbano et al., Citation2019; Dagan & Yager, Citation2019; Fox et al., Citation2022). Thus, I do not see either substantive clinical or biological evidence that severe early ACE (abuse and neglect) including severe early abuse operates as some kind of “nonspecific” factor for depression vulnerability, completely independent of intense separation distress induction and aloneness, especially given the evidence that neglect is a potent component of ACE, or any substantive mechanistic independence from the recursive biological factors outlined in the target article. These various neurobiological factors (alterations in immune, HPA axis, social peptide and monoamine systems) have been confirmed and replicated in the animal models referenced earlier that are the mechanistic perturbations from that separation distress. There may be still some important differences between complex abuse and simple neglect/abandonment, around additional activation of the fear system in the former versus the latter – I'm not suggesting that these are all neatly equivalent, but I am suggesting that the profound and untenable “aloneness problem” that is intrinsic to abuse has not received the attention it deserves. Recent reviews have suggested actually that a mono-focus on the fear system is not productive in the pharmacology of PTSD, and that ketamine is more effective than SSRIs, again suggesting large mechanistic overlap with depression (Abdallah et al., Citation2019), and what has been called a “dysconnectivity” phenotype. These considerations around shared induction stressors of course are also quite consistent with the large comorbidity and overlap between PTSD and depression (Radell et al., Citation2020), an overlap that may also account for an uncertain percentage of depressions labeled as refractory, a suspicion supported in the superficial, or even frankly absent, exploration of possible ACE/ abuse and trauma in both many depression studies and by many clinicians examining depressed patients (Nemeroff, Citation2020, Citation2016). All of these considerations powerfully argue that penetrant forms of separation distress are an epicenter at the heart of how trauma is transduced in the CNS – not a peripheral issue in other words, and that furthermore, abuse is not “nonspecific” in its mechanistic transduction and how those neurobiological factors impact a developmental trajectory. These considerations of course validate strongly the idea that the establishment of social comfort, and the accrued ability to better detect this kind of environment, is a critical treatment goal for trauma victims (van der Kolk, Citation2015). This may intrinsically collide with how trauma victims seek a counterdependent and isolated state, as a defense against revictimization, but of course, this is unsustainable because of the burden of unacknowledged and often times disowned separation distress and fundamental attachment and social need. In other words, the full destructive legacy of abuse cannot be reversed without the difficult achievement of secure attachment/social comfort, and not merely the achievement of physical safety. Many patients suffering from the long-term sequelae of abuse achieve the latter (physical safety and removal of themselves from abuse scenarios) long before the former (social comfort and secure attachments). All of these considerations argue that a prepotent and often times unacknowledged sustained activation of the separation distress system is a central challenge in the repair of trauma and PTSD, and further, that such activation of the separation distress system may indeed help to explain the enormous comorbidity of PTSD and depression.
How might depression actually shutdown separation distress? The SEEKING system revisited
Then there is the critical question that they are correct in stating was not explicated adequately in the target article. How exactly does depression shut down separation distress? It's astonishing that there has been so little detailed fine-grained empirical focus on this question (part of the general scientific neglect of separation distress both in human and in animal models), but the available information suggests down-regulation of the SEEKING system, interdigitating with the effects of cytokines and perhaps combined with other aspects of depression biology, terminates separation distress vocalizations, issuing in “passivity” or “despair” depending on the particular model and its description. This may also be an area where the unfortunate conventional terminology for the SEEKING system (the “reward system”) contributes to obfuscation of its foundation for all motivated behavior. Cytokines have to be considered important (Hennessy et al., Citation2001, Citation2009, Citation2017), as sickness behavior also shows this general blunting of motivation and emotion (but not pain and malaise!) quite well, perhaps especially in those with a high inflammation subtype of depression (see Drevets et al., Citation2022 for forward-looking treatment implications). In simple language, we reach a point where “we can't take it anymore,” and we simply give up protesting and crying out. Hennessy et al.’s (Citation2001, Citation2009, Citation2017) work on animal models clarifies that pro-inflammatory cytokines – with a complex neuromodulatory and neurodynamic impact – appear fully capable of terminating separation distress vocalizations. Given that upregulation of pro-inflammatory signaling directly impacts both aminergic reuptake and production in the direction of negative regulation, any hard distinction between a pro-inflammatory versus an aminergic arousal explanation for behavioral shutdown seems specious at the current state-of-the-art, but a hypo-dopaminergic basis seems at least part of the answer, consistent with the notion of a relative (and likely variable) shutdown of the SEEKING system. One of the deep nuggets of insight that is sometimes a little bit hidden in Jaak Panksepp's work and obscured by any quasi-neophrenological interpretations of his seven prototype emotional systems (that these are all neatly separate systems, or any version of “modules”) was the realization that a system for generalized emotional arousal (SEEKING), of necessity acted both as a general “volume control” and also an evolutionary foundation for all the other more differentiated emotional systems. This is not generally appreciated by many students of Panksepp's work.
It's worth remembering that the SEEKING system notion was originally articulated by Panksepp in relationship to the most ancient types of affects, namely homeostatic imbalances (classically hunger and thirst but also pain and thermoregulation, what Panksepp might have called “sensory affects”). It is worth reminding everyone that his early work on energy balance helped Jaak Panksepp to conceptualize a core distributed system, deep in the brain, generating something like “basic motivational arousal” – a difficult and slippery concept at best – what Jaak was fond of calling “a goad without a goal.” Goals are of necessity supplied by systems abutting or outside the SEEKING system proper, but in communication with it. Evidence for this precedent/foundational role of the SEEKING system is abundant, and at the same time often neglected, but it is (evolutionarily) baked into the operational logic of a generalized motivational arousal system. This system functions through the neurodynamic and neural network linkages between the registration of homeostatic imbalances/needs, reward cues (predictors), and linkages between both of those and executive-behavioral routines that are charged with the life-critical task of rectifying the homeostatic imbalance. Not only is the system integrative but it is also “graded” – capable of ramping up motivation in direct proportion to the homeostatic threat (Panksepp, Citation1998). Only if these critical classes of information (felt need, sensory/world mapping of satisfying opportunity/resource, and executive routines to secure the resource) are in some form of activating inter-neural register can we model how a conscious organism can do something as simple (and life-essential) as pursue food and then eat when hungry. Such integrations therefore should be appreciated as forming literally the neural backbone of the conscious mind.
This complex functional integration explains why the SEEKING system is a distributed network, including multiple brainstem areas, most especially PAG, several if not many hypothalamic areas, the VTA, and limbic basal ganglia regions such as nucleus accumbens and olfactory tubercle and all their connections. Jaak Panksepp also took pains to clarify that the notion of a prototype emotional system of necessity meant a distributed network. Extensive lesions of the VTA, along with extensive lesions of the medial forebrain bundle (the white matter “trunk line” for the SEEKING system connecting its ventral and dorsal components), PAG, and the nucleus accumbens all generate akinetic mutism, representing a complete collapse of motivation and emotion. The syndrome of akinetic mutism was sometimes conceptualized by Jaak Panksepp as perhaps the harshest and severest form of depression but is also a primary disorder of consciousness (Watt & Pincus, Citation2004). Additionally, relatively harsh chemical blockade of the dopamine system, as in moderate+ neuroleptic dosing, is fully capable of shutting down affective display/vocalization, again cutting across the classes of prototype emotion.
This property of both foundation and modulation by the SEEKING system appears to hold whether these prototype emotional systems are defensive (classically of course FEAR and RAGE) but also if they are prosocial – the cardinal systems in this domain are PLAY, SEPARATION DISTRESS, and proto-empathy/ maternal CARE, perhaps many if not all of these being some kind of evolutionary extension of the original prosocial system, LUST. This analysis about the SEEKING system becomes more immediately relevant to depression and the question of how shutdown of separation distress might be affected if we consider that separation distress isn't simply a generalized sense of anxiety but contains at its core a panicked/agitated search for reunion. We speculated in our original essay (Watt & Panksepp, Citation2009) that the evolutionary selection pressure around this might show a very narrow time window in which this could be adaptive, functioning as a mixed signal, to alert both parents and predators. Indeed, if parents were not in the immediate vicinity, protracted vocalizations could have significant morbidity attached to them, explaining the evolutionary benefit of shutting down separation distress after a relatively brief interval of vocalization. In a motorically incompetent and highly immature and vulnerable creature, such a longed-for reunion of course could only be achieved with separation calls, which is why Panksepp argued that these vocalizations were the gold standard for indexing separation distress in animal models.
Application of active inference concepts, prediction error, and related notions to depression – overlap with classic psychoanalytic concepts?
In terms of a modeling of separation distress with homeostatic concepts (and perhaps also therefore potentially modeled by Markov blankets of the type investigated and favored by Badcock and Friston), social comfort is the desired homeostatic ground state to which the system seeks to return. Separation distress is the variable and graded sense of loss of, and departure from, that ideal ground state of social comfort. One might say that the evolutionary selection of attachment builds in a preconscious prediction that social bonding/parental proximity is a far safer place to be than in isolation. It's easy to see how this might have emerged in evolution from the simpler place preference demonstrated by all vertebrates. In autistic spectrum disorders, one sees the reappearance of the evolutionary antecedent – an exquisite sensitivity to place preference and its disruption, as the higher and superordinate process (variably described as secure attachment, or “object constancy” or the like) is not there to stabilize or provide a basic coordinate system. This is graphically represented in the modified Venn diagram taken from Affective Neuroscience (Panksepp, Citation1998) on page 115 of the target article, where thermoregulatory, nociceptive, and place preference mechanisms may have provided critical evolutionary antecedents to social bonding/separation distress.
In terms of their discussion of the role of prediction error and active inference, I believe that towards the end of their commentary, they articulate similar notions (page 163 paragraph 2) – identifying an obvious prediction error embedded in depressive shutdown and also recurrently referenced in cognitive theories of and therapy approaches to depression – “there is no point in trying, the outcome is already determined, and I will only experience more grief and my effort will be wasted.” Intimately related to this of course might be the prediction that “there is no point in seeking reconnection – I would only be rejected, abandoned, hurt, etc., just as I have been in the past.” Further examples of this that are very common in depression might be “people just don’t like me, don’t care about me, and there is nothing that I can do to change these things.” “Maybe it’s because I’m just bad” (expressing guilt) or “I just don’t have it, whatever it is” (expressing shame). “Why bother, things are hopeless for me?” While everyone recognizes these as classic depressive cognitions, they embed predictions that are globally destructive of or inhibitory of hope for rewards and thus motivational arousal/SEEKING, undercut or fragment any longing for or effort towards re-connection or social support, and help maintain the positive feedback relationships that energize and maintain depressive illness. These notions are sometimes conscious, but often pre-conscious, and because they are quite painful to experience unless one is very apathetic, they are often times shoved from awareness and out of active reflection, forming key items for psychotherapeutic work. I would argue that these depressive ideas are the leading candidates for what the authors term “aberrant priors” creating and maintaining a dysfunctional state (page 162, paragraph 3 of their commentary).
What the authors might call “dysfunctional or aberrant priors” (typified in the above depressive predictions) reside in what Panksepp called the “tertiary level of emotional processing,” where complex cognitive–affective amalgams involving affects dynamically linked to complex images of self and other in relation become powerful determinants of behavior (with these images in depression unconscious, or peri-conscious, and often linked to childhood trauma). This basic idea was also articulated within object relations theory (Kernberg, Citation1975) and associated child development work in psychoanalysis, that recurrent pathogenic affects were linked with internalized self and object images, an amalgam that functioned as “condensates” of trauma. These self and object images and attendant affects were reactivated in and by current social pressures that created situational analogs to the original trauma, a relationship that may help explain the primacy of right hemisphere processing in emotion. This was the very set of relationships, between perspectives thought totally orthogonal to each other, that actually drew me into this field some 40 years ago (Watt, Citation1986). The examination of that reactivation of traumatically-grounded self and object images linked to pathogenic affects, also giving rise to pathogenic distortions and defenses, all of that manifest in the therapy relationship itself, is at the core of psychoanalysis. Perhaps from the authors’ perspective and terminology, psychotherapy becomes the systematic repair of hidden but powerful “dysfunctional priors.” Are we really saying different things? Or does the radically different vocabulary just make it look that way?
This is only a small but relevant example of how creative individuals, who are all struggling to do justice to the staggering complexity of Nature, in this case, human neurobiology and psychology, might arrive at the same basic idea, despite vastly different intellectual traditions, and deceptively different vocabularies. I would like to close with this more hopeful message about seemingly competitive perspectives – indeed that potential for an eventually deep consilience of knowledge (but requiring much hard work) is the entire point of this Journal! Let us not be too taken in by just those kinds of apparent and seductive differences in terminology and emphasis. Although for a long time, psychiatry did not have substantive knowledge linking these levels of organization and discourse, and may have responded to that deficit by becoming reductionist, and losing focus on how biology, psychology and the social world might interact and interdigitate, we now have substantive bridgework that is being added on a daily basis. Indeed, this failure to moderate an early reductionism could be argued to be the complement error to the one that Freud and psychoanalysis made when it shelved attempts to bridge to neuroscience due to the immaturity of the knowledge base in the other field, and then simply forgot about the effort.
I hope that this response is helpful in clarifying very substantial agreement with the authors, as well as areas where we may still have some creative differences. All these questions beg for far more research into separation distress as a potent attractor within psychic landscapes – in striking contrast with its general neglect in mainstream psychiatric and psychological science. I hope the authors agree.
Response to comments by Barton Blinder MD PhD
Rumination – are its two forms somehow related?
First let me express my appreciation for the gracious comments by Dr. Blinder (Citation2023) about the depression paper – what he accurately refers to as “a project.” Indeed it was!
When I first looked at his commentary, of course I had assumed that Blinder was referring to the more common association of depression with cognitive rumination – when in fact he's referring to the original rumination as it were: postprandial rumination.
But it made me wonder if this interesting and neglected (out of rarity) syndrome in infants, children, and adolescents might be related to cognitive rumination, as one of the symptoms of depression within the cognitive architecture. They do appear to share a common feature of attempts to expel internal contents, and then reprocess, and then re-internalize, perhaps making them less toxic/upsetting/ “digestible.” Blinder's commentary provides some impressive data supporting the supposition that postprandial rumination may be an unusual syndrome related to low opioidergic tone and separation distress, particularly telling evidence that it is significantly inhibited by opioid agonists, and that it is seen in infants of opioid-addicted mothers who have down-regulation of the system from synthetic opioid exposure (summarized in Blinder, Citation2023 but also referenced in Chatpoor, Citation1984). It is interesting that this may fall, as Blinder notes, on a potential continuum of self-stimulatory behaviors, and may also be related to other eating disorders, all possibly reflecting regressive efforts at achieving self-soothing, associated with loss of social comfort/secure attachment and again, concomitant with low opioidergic tone. His many case examples suggest that postprandial rumination may indeed be strongly associated with losses, and with classic anaclitic depressive crisis ensuing from such losses. Rectification of social disconnection in the treatment of mood appears to significantly reduce postprandial rumination, from his case examples.
Somewhat in contrast to what has been suggested in some sectors of evolutionary psychology, cognitive forms of rumination appear to actually worsen or at least maintain depression – at least the kind of rumination typically seen in major depression (Cooney et al., Citation2010). Rumination appears to be associated with significant activation in amygdala, rostral anterior cingulate/medial prefrontal cortex, dorsolateral prefrontal cortex, posterior cingulate, and para-hippocampus when compared with simply abstract conceptual thinking as a control task. These findings may suggest that ruminative self-focus shows enhanced recruitment of limbic and medial and dorsolateral prefrontal regions in depression, but they did no MR spectroscopy, so it is not clear whether this is associated with lower opioidergic tone in limbic and paralimbic regions, compared to depressions without rumination. So this becomes an interesting question meriting further study. Another interesting question is, when we ruminate in depressive rumination, what determines its potential foci? I suspect a big attractor in the complex base is focusing on regret-driven review in terms of critical tipping points or other ways where we are drawn back to think about how the loss could have somehow been prevented. Other sources on rumination, stressing its connections to worrying and anxiety (perhaps to both fear and separation distress but this is unclear?) note its similarity with obsessional thought, and where people are recurrently focused on problems, symptoms, triggers, and stressors as opposed to solutions/adaptation (Nolen-Hoeksema et al., Citation2008; Papageorgiou & Wells, Citation2004). This may be part of the overlap between Generalized Anxiety Disorder (GAD) and depression, which may be complex: many folks diagnosed with GAD may in fact have anxious depressions of a mild sort that is being misdiagnosed (as discussed in Nemeroff, Citation2020); each disorder may predispose to the other, creating overlapping comorbidity.
I apologize to the author if this all seems off track, but I just wonder, in the wake of the author’s own comments, if postprandial rumination and depressive cognitive rumination share more than a semantic “tag.” Of course, this suggests a very interesting possibility that cognitive rumination, postprandial rumination, and other eating disorders may share a connection to separation distress dynamics, networks and neuromodulators. All of this becomes a fertile territory for more research that is both psychologically and biologically sophisticated, explicitly avoiding (our favored?) traditional siloed approaches. Hoping somebody reading this might be energized to track these questions down!
Response to Otto Kernberg commentary
Towards a future depth social neuroscience incorporating internalized object relations. Further thoughts on transference and the repetition compulsion
It is a major privilege to have the thoughts of Dr. Otto Kernberg (Citation2023) regarding the separation distress hypothesis of depression and its update review in the target article. I appreciate his many gracious comments. Indeed, his summation of my argument in the first two paragraphs of his commentary is easily more cogent than the paper's abstract! Kernberg's landmark work Borderline Conditions and Pathological Narcissism (Kernberg, Citation1975) formed a critical point of inspiration, inquiry and essential conceptual grounding for my doctoral dissertation (which explored both commonalities and differences across the nosology of primitive character disorders), so I feel gratified that I was perhaps able to repay some modest portion of my intellectual debt to him in this target article. In particular, I believe he articulates a critical issue when he states (Kernberg, Citation2023):
Separation distress originally would be motivated by actual abandonment, severe object loss, rejection, abuse and chronic attack and mistreatment in reality. But eventually, the intrapsychic psychological structures altered by realistic environmental trauma may become major pathogenic factors. I believe that the vicissitudes of the development of the world of internalized object relations, particularly the unconscious identification with sadistic object representations, may become major sources of internal abandonment signaled by guilt, shame, and projected self-accusations attributed to external objects. The effects of chronic traumatic experiences that trigger depressive reactions and strengthen self-critical and devastating self-devaluating characterological features affect the self-structure of the world of internalized object relations. (p. 169)
Different theoretical frames side, this dynamic, frequently seen leading up to depressive crisis, but often times not appreciated for what it might represent, of ruptured relationships driven by these pathogenic internalized object and self-images derived from actual experiences of traumatic loss and rejection, can and often does take more subtle forms, in which we invite or even provoke rejection and abandonment by virtue of our own toxic behaviors (often times demonstrating significant empathic failure when operating under the influence of chronic fears and the distorting effects of transferential processes). Such provoked ruptures in relationships are an example of underappreciated positive feedback between pathogenic internal structures and new cycles of social loss, trauma and dysfunction (what in classical psychoanalysis might be framed in terms of cycles of “projection and introjection”). Such positive feedback processes between pathogenic internalized object relationships, yielding harsh behavior and empathic failures that in turn promote social dysfunction/ruptures in attachment, is often times underappreciated by those providing social support to those suffering in the wake of these ruptures – until they are caught off guard by the brittle defensiveness, and misplaced retaliation of the person they are trying to support. Such reactions of course would not surprise (or at least not surprise as much!) a group of senior and well-seasoned psychotherapists, but friends and family of afflicted individuals with trauma and depression often times will comment about “how hard it is to take care of person X,” in good part because they just can't tolerate being in a position of real dependency. Or they can only tolerate such a position if they believe that they have near-total control over their caretakers, which of course is alienating, and can once again induce ruptures, and thus lead to further compounding of social loss and isolation. In any case, the ruptures serve to powerfully confirm and strengthen chronic fears of such rejection and abandonment – exactly of course what Freud was getting at in his articulation of a repetition compulsion, outlining our sometimes most uncanny “cookie-cutter”-like re-instantiation of original traumatic scenarios (Freud, Citation1914). Freud, in his reflections on this early in his theoretical journey, was genuinely impressed with the “uncanny” nature of this process, using the term several times in his discussions of this, as though these repetitions were even a bit “spooky.” Anyone who has done real psychotherapy and examined their own life in deep and psychodynamic detail (far and away the most valuable and essential training to be a psychotherapist) has to be impressed with the often uncanny nature of the “repetition compulsion.”
Although one can use this notion of a repetition compulsion too literally and concretely, it documents our powerful “gravitation to the familiar” (likely mediated by striatal procedural memory, perhaps most especially the ventral striatal complex consisting of the nucleus accumbens (NAc) and olfactory tubercle (OT) and their brainstem and paralimbic network connectivities, which may support much of “personality as affective predisposition”). One way to reconceptualize this classic concept of the repetition compulsion in the vocabulary of modern neuroscience is that it represents a kind of complex “place preference” mechanism (place preference is operative in all mammals), and in this specific instance, an unconscious “personality and social dynamics” preference mechanism. We're drawn back into these situations that are so deeply familiar, and some part of us seems “pre-consciously convinced,” without any real evidence, that this time through the mill, things will be better. Or as Shakespeare once put it (in Henry V), “once more into the breach!” Dynamic psychotherapy teaches us to be much more suspicious of our response to such “clarion calls” by calling attention to their patterning and frequency.
Ultimately, the repetition compulsion was believed by Freud to operate in pursuit of mastery and repair, but it is very much a two-edged sword. It is also a mechanism that can drive depression and despair, by virtue of re-instantiation of serial trauma, and of course this can drive increasingly self-destructive behaviors and maladaptive defenses, including addiction as a stopgap affective regulator which obviously has huge morbidity and mortality all by itself, and thus even early death, either from addiction, suicide, or other disastrous contingencies. Unfortunately, both in some current as well as in classical accounts, treatments of the repetition compulsion often leave out the manner in which our own defensive structure, blind spots, acting out, and empathic failures contribute to the re-instantiation and repetition. In that critically important sense, we are not simply victims, we are proactive unconscious agents in our own undoing and defeat – the repetition compulsion in a sense unites the heroic and tragic themes in human life together in some kind of amalgam. From the perspective of biological science, it is an “uncanny” manifestation of positive feedback in complex systems between internalized object relations and the social/interpersonal milieu – I oftentimes think of it as analogous to pathogenic protein folding and prion-like templating in neurodegenerative disorders in which abnormally folded proteins induce similar confirmations in naïve monomers of that protein, explaining how pathogenic and pathologically folded proteins spread throughout the brain's networks. (Others may, or may not, regard that metaphor as useful!)
I still find it both unsettling but also amazing when people provide dismissive summaries of psychoanalysis (part of the general problem of the widespread caricature of psychoanalysis which is often uncritically accepted as accurate, furthering the ignorance and prejudice – see again my commentary response to Badcock and Friston). Many of these pejorative summations of psychoanalytic theory are top-heavy in their emphasis of classic structural model concepts, and in emphasizing the original and admittedly Procrustean instinct theory concepts (substantively updated by Panksepp, Citation1998), contributions from neurobiology to an updated metapsychological theory of instincts, recently very ably summarized by Kernberg himself (Citation2022) in a recent target article publication in Neurosychoanalysis. And yet, these caricatured views of psychoanalysis (which virtually never appreciate or reference the updated affect theory) often neglect or leave out completely the clinical core concepts of transference and the repetition compulsion, as the dynamos that they are in interpersonal and social dynamics. We carry our entire attachment history, for better or for worse, around inside of us as a kind of fundamental social meta-operating system.
Indeed, the residues of that very earliest component of our attachment history, preverbal experience as infants and very young children, has been termed, in the poignant expression from Alvin Frank (Citation1969), “the unrememberable and the unforgettable.” Although this has received almost no research attention relative to its likely importance, the “unrememberable and unforgettable” qualities of our early attachment are presumably operative through various gating (inhibiting and amplifying), epigenetic and perhaps several other unknown potentiating effects on basic negative and positive affective systems. This early history, if positive and demonstrating strong and joyous attachment to caretaking figures, can leave us with a deep fundamental resilience, optimism, and prosocial orientation that can carry us through some of the worst of life’s traumas and disappointments, or it can undercut profoundly our resilience, and leave us hampered with seriously lowered thresholds for negative affective states (fear, rage and separation distress), thereby constituting a lifelong burden on affective regulation, which we may struggle to mitigate in any substantive fashion. This can saddle those most unfortunate individuals with a deep and for some unshakable conviction that relationships are mostly about pain. Much work on depression of course argues that resilience emerges from interactions between many genes with generally weak penetrance with multidimensional environmental variables, and this adds another layer of complexity and unknowns in our clinical psychotherapeutic work with patients, where we virtually never know about the genetic components. Another intrinsic challenge in the science of affective development of course is to separate out how much these affective trends emerge from later social experiences, versus hidden and virtually unmappable preverbal ones, as these all immediately create a complex alloy in any living mind/brain, as do subsequent and later experiences in any social attachment. In any case, we repeat and recapitulate that history whether mostly successful, mostly unsuccessful, or mixed as is often the case, in our current social relations. We are often trying to win battles we lost a long time ago, in a most unconscious and sometimes tragically self-destructive fashion, by reinstating, inviting and even provoking the same toxic behavior that we experienced in childhood traumatic scenarios, with the readily available and subtly “entrained and molded” social proxies for our parents. In that sense, I agree with Kernberg (Citation2023) when he states:
I believe that psychotherapies that focus on understanding transference developments and, particularly, on resolving negative transference reactions are fundamental in the psychotherapeutic treatment of depressed patients. Many depressed patients are prone to feeling easily rejected, abandoned, or criticized by the therapist. The projection of the patient’s own self-devaluation onto the therapist, the conviction of depressed patients that they don’t have the right to be helped or even to live, and that, at times, so may the therapist think as well, offer great difficulties in many psychotherapeutic treatments. (p. 170)
We await the potentially fruitful clinical applications of these ideas, including additional probing of the separation distress hypothesis in terms of an empirical study on psychotherapy efficacy perhaps looking at how psychotherapists might pay special attention to the vicissitudes of loss and the activation of separation distress as a primary psychotherapeutic focus. Such studies can have a primary outcome variable measuring reduction of chronic or acute separation distress and relative achievement of social comfort/stable attachment, and its comparison with other theoretical approaches in terms of antidepressant efficacy, along with indexing protection against relapse, improved mood ratings, etc.
Response to commentary from Davis and Montag
Unraveling the roots of depression – it's complicated. Thoughts on slower progress than we might like in affective neuroscience
It is both a pleasure and a privilege to hear feedback about the target article from Davis and Montag (Citation2023), as long-time contributors, in various ventures over the years, to affective neuroscience research. The authors have made significant contributions, most notably in their fine work bridging affective neuroscience to personality theory (see especially Davis & Panksepp, Citation2018), with recent follow-up position papers outlining the status of the field of personality research using the Affective Neuroscience Personality Scales (ANPS) in Montag et al. (Citation2021), and Montag et al. (Citation2022). Bringing this rich base of experience and understanding, Davis and Montag point out a couple of important questions and difficulties around the separation distress hypothesis. These include empirical data sets that they reported in their commentary, validating the supposition that elevated levels of separation distress conjoined with global reductions of motivational tone (otherwise known as a low activation state of the SEEKING system) predicted depression in their cohort (Montag et al., Citation2017). A complication in all this was that earlier versions of the separation distress/GRIEF scale of the ANPS did not adequately discriminate between “protest” and “despair” phases, or between sadness proper and depression (as unpacked and discussed in Montag et al., Citation2022).
I regret not catching these references in my literature review, as they certainly are worthy of mention, particularly given the dearth of work on separation distress as a prototype emotional state and major ramp into depression and other psychiatric morbidities. Additionally, as they point out, the findings go to the heart of the argument that protracted separation distress activates systems that shut down motivation and the separation distress system with it (see extended discussion of this still inadequately understood question of how depression might actually terminate separation distress, in response to commentary of Badcock & Friston, Citation2023). In this important way, I believe increasing apathy, abulia and deepening emptiness cancels high activation of the separation distress/PANIC/GRIEF system. In other words, depressed patients stop protesting or crying, and particularly if showing a retarded/apathetic phenotype, they may observe that they are indeed not sad, and they even sometimes comment that they would wish to be sad again, so that they could feel anything other than the emptiness and blackness of the depression.
This is an underappreciated phenotypic distinction, mentioned in the paper, but worth repeating, and it contrasts meaningfully with patients who are both sad and depressed, part of the difficulty in depression subtyping which lacks both clear categorical (differential symptom clusters) as well as definitive biological (biomarker) knowledge and discrimination. A promising pathway in progress in subtyping may be for example differential inflammatory markers (Drevets et al., Citation2022) but there seems little chance that expensive technologies such as metabolic brain imaging, or invasive technologies such as CSF protein assays, will get much traction in the current healthcare environment where many healthcare and mental health systems are literally swamped with depressed patients and where the cost-effectiveness of this subtype phenotyping remains very much debatable. All of this of course goes to the problem of meaningful depression subtyping, a can of worms which we “kicked down the road” so to speak 15 years ago (to mix metaphors). Progress looks remarkably refractory here, despite a lot of scientific effort in the intervening decade and a half. One wonders if in fact subtyping might possibly reflect differential (+/−) contributions from several systems listed in the complex recursion outlined in the “Depressive Matrix” Table (table #2, on page 121 of the target article). Such differential contributions might be consistent with evidence that the HPA axis alterations, and immune changes are both rather variable, as were the aminergic system alterations, as indexed in the classical assays for amine metabolites.
I also appreciate the authors’ attention to the (underappreciated) complementary process to separation distress, namely maternal CARE, which appears to be based in a proto-empathy system, and where separation distress/attachment, and thus proximity sensitivity, may actually be upregulated in maternal mammals to explain their high degrees of adaptive protectiveness. This may be especially critical when dealing with young that are either motorically competent and therefore good at wandering into trouble, or conversely, not motorically competent, helpless and therefore easy stationary targets for predation.
I resonate with the authors’ disappointment that affective neuroscience concepts, particularly those pioneered by Jaak Panksepp, are not more mainstream. I recently had an interesting reality check on this, as I received a requested copy from one of the primary authors of the massive compendium review/position paper “The Human Affectome” (Schiller et al., Citation2022). This was a paper with roughly 180 listed authors, 750+ references, and 124 pages, with authors from just about every area of affective science. The principal authors group (15 primary authors) aimed to articulate bridging concepts that would integrate disparate fields, methods, assumptions, perspectives, and ontologies of affective science. I would encourage other readers of this position paper to read this paper and to draw their own conclusions, but I found the paper quite discouraging, at least in terms of how one might index scientific progress, with no meaningful consensus, conclusions, or integrative perspectives. The authors did manage to catalog and list, and in a constructive fashion, define many of the “buzzwords” so to speak in affective science – central organizing concepts (see pp. 35–44 – such as valence, arousal, embodied, dynamical systems, affectome, enactive, homeostasis, etc.). Those sections of their position paper might be potentially quite valuable to students starting out in the field. But the conclusions – speaking to a lack of real consensus – were remarkably “watery,” specifically, that there were such things as “affective concerns,” which might be related to “comfort zones,” and “affective features,” that referred to “monitoring of the affective process.” Needless to say, it is difficult to know how the notion of “affective concerns” would apply to the affects of animals or for that matter young children and infants – speaking to the not-so-subtle effects of what Jaak Panksepp called “cognitive imperialism” on our vocabulary and terminology. Although the abstract was disappointing and not particularly substantive in terms of offering any clear conceptual integration or overview, I assumed that by reading the entire paper I would be enlightened and a viable conceptual overview – or at least some kind of outline of a potential way forward – would emerge. This expectation was also quite disappointed.
My tentative conclusions are: (1) perhaps having 180 authors on an endlessly contentious subject such as human emotion may virtually guarantee that there will be no substantive agreement; (2) there was no principled evolutionary perspective in the overview/abstract provided by the principal authors, despite referencing evolutionary principles as foundational (that a primary job of the organism so to speak is to survive and maintain its integrity – although the promotion of procreation was notable by absence as an essential, irreplaceable dimension of any putative selection process); (3) related to possible evolutionary hierarchies missing in action, there was no meaningful outlining of the relationship as well as the separation between emotion and cognition, or how the apparent conflict between current basic emotions theory versus constructivist views (as the latest permutation of the attempt to make emotion just another cognitive domain?) might eventually be reconciled. This was something that Jaak Panksepp cared deeply about, and where he articulated a pathway towards reconciliation. This pathway towards reconciliation of course emphasized evolutionary hierarchies, and that constructivist insights applied to the last layer of the onion – the tertiary level of processing where affects are intrinsically linked to high-level cognition – but could not characterize or account for the conserved nature of prototype emotional systems for either the primary process levels, or the secondary process learning effects, which are driven by the valence of the primary process (behavioral learning falls apart when stimuli reach “valence neutrality”). Almost paradoxically, the authors mentioned significant indebtedness to Jaak Panksepp in their Acknowledgments preface, while seeming to incorporate only a single one of his most salient concepts (the notion of “comfort zones”), incorporated without adequate explication or much developed understanding, I might add.
Ultimately, I believe one has to simply accept certain sobering facts about science and scientific progress – which may index their own version of deep emotional truths about us humans. Under the best of circumstances, the rate of progress in science is generally glacial. That glaciality shows further slowing in established sciences, versus in more infant or adolescent sciences, where progress is sometimes more rapid. Of course, one would like to believe that rational arguments, buttressed by a cumulative weight of evidence, would be enough to convince people of the basic truth in Jaak Panksepp's primary emotions theory, particularly given the massive weight of evidence for fundamental mammalian conservation across the various prototypes of emotional state. However, the remarkable fragmentation in the many disparate fields dealing with emotion, rather than clearly yielding to integrative efforts over many decades, instead seems to at least as much metastasize, and defy those efforts. I do think that constructivist views – which show if anything increasing popularity and penetration in psychology – appeal to the same easily seduced human vulnerability that energizes cognitive imperialism. As someone once observed, many of the most powerful paradigm shifts in different sciences are simply a different version of the Copernican revolution – the systematic falsification and rolling back of human grandiosity, that we are not the center of the universe, that we are not that separate from the rest of the animal kingdom, and that we share far more, including an enormous overlap of genes and core biological mechanisms at the heart of our cellular biology, with organisms (such as yeast!) that we may consider “inferior,” or even beneath concern. As I stated in the target article about the widely held deep skepticism about the potential value of any animal models in depression:
It is worth reminding ourselves that in relationship to our similarities with and our differences from other mammals, human grandiosity has been sorely disappointed by the general trends within most scientific research into our mammalian brothers and sisters. This has been shown in relation to a host of putatively unique human abilities, including language, empathy, and social bonds. In other words, most research finds that the differences are more quantitative than qualitative, and that primitive precursors of even our higher cognitive abilities, where the evolutionary distance is arguably greater than in any other functional domain, reside widely in other mammalian genomes, and further, that many unique functions appear to emerge from the interaction of those individually modest differences. (p. 143)
These questions in emotion theory end up frankly being a bit of a projective test, just like politics and religion. I know – science should not work that way – but I guess that's my point. While science has a built-in course correction mechanism lacking in politics and religion (critical hypothesis testing and weight of evidence), it takes a while, sometimes a long while, to overcome highly popular and seductive memes. Ultimately, I believe that the overextension of constructivist views will be falsified, although at the tertiary level (see discussion of Badcock and Friston commentary), there is little question about how critical cognition is both in terms of activating complex cognized emotions as well as the cognizing of emotional outputs/behaviors. Even in the admittedly constructed tertiary levels, I still believe we are talking about derivative families of affects, where the prototype emotional system is still relevant, even if its inputs (triggering perceptions and cognitions) and outputs (solutions to adaptive problems) are both much more complex, subtle and differentiated relative to what we see in animal models of those same prototype systems. Even around inputs and outputs that are cognized, I believe that science will eventually validate, for example, that cognized forms of fear or rage emerge from events in the environment that reflect danger or loss of safety in a complex social and cognitive space, or that reflect attacks, threats, injuries, or other classic precipitators for rage respectively. In other words, I believe that neuroscience will eventually confirm that the systems for prototype emotion are sensitized to the various ways in which we can map cognitive extensions of the original triggers, creating cognized and obviously modulated versions of the prototype state. This of course was my fundamental argument around the cognitive extensions of separation distress in shame and guilt. This is the essence of Jaak Panksepp’s concept of the tertiary level of the prototype emotional system.
Otherwise, we are led to the completely contradictory statements that evolution saw no point in building a basic distinction (precognitively) between situations of play and situations of rage. The notion that those differences are all cognitively constructed makes no sense from the standpoint of evolution providing a kind of “prepackaged” if primitive solution – it would be fully analogous to the situation in which evolution gave us no capacity to discriminate between various homeostatic challenges such as hunger vs. thirst vs. pain vs. thermoregulation. We would only know that something was wrong; we could not possibly know the difference between hunger and thirst or between hunger and any other homeostatic affect. We would thus be completely blinded about what we needed to do to survive. Prototypes for emotion, if they really don’t exist, and it is all just verbally “made up” categories, would mean that there is no built-in discrimination of adaptive contexts, no prepackaged executive routine of any kind other than simple reflexes, no point in creating a “basal level predisposition” that it is better to fight, or to run, or to play, or have sex. No such thing as instinctual behavior in any substantive sense – and although psychoanalysis and other psychologies of emotion have struggled under a straitjacket of rigid instinct theory, there is a sweet spot where instinctual systems undergo learning, extension, and even profound modification from cognition – as Jaak Panksepp stressed in his primary, secondary and tertiary level emotion system concepts. These concepts from Panksepp are still the best approximation for how to resolve and integrate social and cognitive constructionism with the undeniable evidence (at least again in my estimation) that emotion must still come to us in some version of primary flavors, the same flavors we share with our closest relatives, other mammals, and where evidence for verbal construction in creating these categories of behavior and response in other mammals is pretty much zero.
We humans are such a tribal and fractious species – even in science – that once tribal kingdoms are established with litmus tests for loyalty to the tribe/school of thought, it may not be realistic to expect much easy or rapid progress. Progress may come over the very long term (many decades), and only from research that directly nibbles at the borders between the putatively different perspectives and schools, showing that they are indeed more reconcilable and more synthetic, or at least overlapping and complementary, than had been assumed, perhaps as in the classic metaphor describing different aspects of the elephant (see my response to commentary by Badcock and Friston). Fashionable memes are thus constructed and propagated forward under the potent incentives of tenure, publication, scientific status, and the increasing reification of scientific concepts, and eventually may create large territories of what I have called “siloed knowledge” – while the reconciliation and interdigitation of such areas of presumed silo is the raison d'être of this Journal. I still regularly encounter a version of this kind of scientific prejudice, when people in various biological sciences look askance at anything related to psychoanalysis, and understand that only in terms of a caricature.
Response to commentary by Steve Siviy
Play – the most important and also most neglected opponent process to depression?
(In this response, I will follow the Pankseppian convention of referring to the distributed and conserved neural mammalian system in all CAPS (“PLAY”), while the behavioral and affective process of play (a “play bout”) is placed in conventional lower case.)
I much appreciate the thoughtful comments by Dr. Steven Siviy (Citation2023). Hopefully, his commentary and this response will rectify the review paper’s relative neglect of mammalian play and its potentially intimate relevance for depression and mood regulation in general. In strong agreement with Siviy, I believe that the PLAY system has cardinal relevance for just about every aspect of the social mind/brain. In reflecting on the relative neglect of PLAY within the target article, although this emerged largely from my aversion to enlarging the paper any further (as it was already WAY too long), I was still surprised, given my great affection for and strong scientific interest in the subject. Additionally, I count the identification of mammalian playfulness as a conserved prototype emotional system (a “neuroendophenotype”) as one of Panksepp's greatest and most original contributions to the understanding of human emotion. It may also be the most neglected of all the prototype emotional systems within both behavioral neuroscience and in clinical practice areas as well, and thus the most globally underestimated in terms of its overall importance. Yet in many surveys of both men and women, the one thing men and women agree on as a most valued attribute (top three or four) in a partner is a sense of humor (although tellingly, and consistent with its neglect and residence “below the waterline,” some surveys don’t even index this behavioral/affective quality – additionally there are some fascinating gender differences, as discussed in Nicholson, Citation2012). The subject is a place of significant personal epiphany – after many discussions with Jaak about PLAY as a prototype in the mid to late 90’s, just before AN (Panksepp, Citation1998) was published, it dawned on me that my early choices for partners were less-than-optimally playful (no doubt part of my repetition compulsion!).
A prototype emotional system of rough-and-tumble play in mammals forms the foundation for its tertiary/cognized extension in humans of the wondrously varied spice of verbal humor and the endless ways in which such humor (like sleep?) “knits up the raveled sleeve of care,” and thus provides us with so many things we may take for granted. Play and humor reduce our stress, reaffirm a sense and place of safety (as play is terminated in any dangerous or fearful situation, and play and humor powerfully disarm mistrust), teach us reciprocity and social rules/turn taking, strengthen social bonds (we may miss most all those who make us laugh the hardest), and help build a sense of reciprocity and empathy (neither party can be too dominant, or play terminates). All of these may make a huge difference in our daily quality of life. It is the only prototype system whose primary effect is making life more fun! A very powerful thought experiment would be to imagine what life would be like with all forms of play and laughter stripped out. Not a life many of us would want to live. And it is not hard to see how this might impoverish a host of other critical social and prosocial processes, and yet despite all this, play, and any concept of a PLAY system in the brain, remain almost invisible as topics within most clinical psychiatric landscapes. Most of us familiar with Jaak Panksepp’s work and legacy regard this as a hidden scientific tragedy, as a deeper appreciation for PLAY's many fingered trajectories within the social brain could pay huge dividends (see partial list at the end).
So this much-appreciated commentary by Steve Siviy gives me an opportunity to mitigate my neglect of the subject – as target article’s length and existing overinclusion made PLAY one of many unfortunate casualties. I would agree with Siviy that play is probably the opponent process – the polar opposite in a complex affective space – to depression. PLAY system activation thus may be the original antidepressant and its regular activation may build in resilience to depression in ways that we simply do not yet understand. However, the discovery of its activation of neurotrophins (particularly IGF and BDNF) provides an immediate and powerful link to most other antidepressants (Burgdorf et al., Citation2010; Gordon et al., Citation2003), as upregulated neurotrophin signaling appears to be a mechanistic common denominator across many different types of antidepressant interventions. Neurotrophin upregulation is increasingly thought the primary antidepressant mechanism in so called mainline (SSRI/SNRI) antidepressant drugs. Play also appears to promote attachment and strength of social bonds (although this issue has also been under-researched), suggesting that creatures (and humans!) playing together are less likely to lose each other and that relationship ruptures are therefore perhaps prevented by successful and “regulated” play, where a lopsided dominance is avoided. There is also a primary protective role for empathy in this preservation of relationships over the long term (empathy which play again likely promotes, also an understudied issue; see Watt & Panksepp, 2016, especially chapter 1, for overviews).
It is not surprising in view of the relative neglect of PLAY as a prototype neuroemotional system that it is fairly easy to frame important questions that currently have very little in the way of empirical answers. What might be a short list of important questions about play, the PLAY system and its relevance for depression?
Could play function directly as an antidepressant? And is the antidepressant effect modulated by the severity of depression? In other words, in severe depressions, does play have a lesser impact, while in milder depressions might it have a stronger, more easily discernible impact? Also, in a properly randomized clinical trial, could you test both verbal humor as well as supervised rough-and-tumble play (both singly and then in combination?) in terms of their impact on a quantified depressed mood? Blinding of course would be an issue, but meaningful controls might include non-aerobic physical activity such as balance exercise, versus listening to educational materials instead of highly rated verbal humor or cartoon clips.
Although GYLX-13 (Moskal et al., Citation2014) failed to clear phase 3, can we use the drug discovery methods from its development to create and clinically test many new additional antidepressant drugs, using rough-and-tumble play gene expression, protein translation, etc.?
As Panksepp (Citation1998) mentions in most of his classic monographs on play, and going back to Plato, rough-and-tumble play promotes turn taking/reciprocity (and probably empathy) via the delimiting of excessive dominance (while paradoxically still promoting dominance SEEKING, skills and practices). Could supervised and optimized play prevent children at risk for psychopathy from realizing that phenotype even when they have high genotypic and other risk factors? This suggests that play could deflect a very destructive developmental trajectory, one with enormous and underappreciated social and economic costs and burden. However, this possibility of moderating psychopathy has also seen no formal study – we don't even know absolutely for sure that rough-and-tumble play promotes empathy although the anecdotal evidence is strongly in support of that, at least as a default hypothesis. Given the current desert of validated options to prevent callous unemotional children from transitioning to adult psychopathy, this question of a preventative effect of rough-and-tumble play constitutes an urgently needed probe of a potentially high yield hypothesis.
Is a hidden aspect of the depressogenic impact of ACE (adverse childhood experiences) the currently invisible effects of “play deserts”? Would concomitant healthy doses of rough-and-tumble play mitigate the effects of abuse and neglect, acting as a protective factor in ACE? Related directly to these questions, might PLAY activation and promotion of both humor and/or rough-and-tumble play have a neglected role in the treatment of PTSD, especially in children, but also in adults?
A prediction emerging from our first instantiation of the separation distress hypothesis (Watt & Panksepp, Citation2009) is that diminution of a baseline level of playfulness is a prodrome to depression, but once again, this has also never been tested. Could the self-rating/monitoring of playfulness and its decline alert an individual that they might be at risk for entering a depressive trajectory? Would playful re-engagement or extra doses of play act as preventative in those just starting to enter a depressive trajectory?
Is there such thing as a high play genotype, and how might that be characterized? Indeed, this is an amazing and large hole in the science. One would readily predict that it probably emerges (like almost all predisposing genotypes in the behavioral and affective landscape) from a lot of genes with weak penetrance but this is not really understood or well characterized at all. Our frank ignorance about this becomes again just part of the enormous evidence that this prototype emotional system has little mainline scientific visibility within much of the mental health community, relative to its huge importance for our social brain development.
What are the relationships between a high play genotype and high play phenotypes? One very obvious prediction, still not meaningfully probed or tested, is that a high play genotype in a playful environment yields a very playful person, which we would predict makes these individuals socially very successful … and very funny! Are there different subtypes of high play phenotype expression (for example, those in competitive athletics vs. those in entertainment)? A default (easy) prediction might be that mismatches between a suboptimal environment (a grumpy environment … or worse?) for someone with high play genotype would create significant problems, including more activation/disinhibition of the RAGE system. However, are there ways in which high play genotype individuals might unconsciously and almost “magnetically” seek out and find playful companions and environs to mitigate this mismatch?
Perhaps these questions might help energize a young clinician-researcher interested in play, who could then report back to this Journal or other peer-reviewed forums with some valuable empirical results!
Response to commentary by Howard Steele
Further perspectives on Bowlby's seminal contributions and the nature of interdisciplinary work
Terminology “translation” across disciplines and perspectives
I'd like to thank Prof. Steele for his thoughtful comments about the target article. It is clear that he is highly familiar with and steeped in Bowlby's framework and terminology, and it may “cost” some gear-shifting to bridge to the affective neuroscience terminology, but I think there are major cost–benefit advantages in a “full bridging” – as defined below. I believe that at least a basic translation of terms across the perspectives of psychoanalytic attachment theory and affective neuroscience still is reasonably uncomplicated and straightforward (see for example Solms, Citation2021 and Kernberg, Citation2022 for more detailed and well developed and neurobiologically grounded metapsychologies), with the most important concept simply being that when Bowlby talked about the “activation of the attachment system,” this would essentially mean the activation of separation distress. This equivalence emerged directly from Bowlby's notion that this “attachment instinct” was defined or operationalized simply by proximity sensing and proximity need. Bonded creatures, especially maternal-infant pairs, appeared not comfortable with significant or sustained loss of proximity – obviously any ethologist worth their salt could observe this easily in both mammalian mothers and infants. Indeed, this is clearly how they behave, and those that don’t behave in this fashion unfortunately show relatively rapid loss of the infant’s life, so it is not as though this equivalence didn't have a solid observational/ empirical foundation. I believe however that this formulation (that the “attachment instinct” is confined to/defined by proximity sensing and proximity maintenance) is ultimately too limiting, and omits major relevant process that social bonding and attachment necessarily bind together, particularly PLAY (see response to commentary by Siviy), but also empathy, which Panksepp conceptualized as residing at the core of maternal CARE (see first chapter in Watt & Panksepp, 2016). Panksepp's frame of reference as outlined in his seminal magnum opus, Affective Neuroscience (Citation1998), unpacked social bonds from the overly homogenized psychoanalytic drive vocabulary to something more differentiated. In the separation distress hypothesis target article, I mentioned that maternal care and a very protective stance could emerge in part from postpartum up-regulation of separation distress system – a hypothesis that has never been tested, but surely is consistent with ethological evidence of the ubiquity of highly protective maternal behavior in the mammalian kingdom, and which may indeed even be part of the puzzle box of postpartum depression.
If you consider Panksepp's argument for six conserved prototype emotional systems (two defensive plus four prosocial), plus a seventh master system (SEEKING), this suggests that social attachment emerges from multiple conserved processes or interactions between somewhat discrete affective process “bins” so to speak. This would include of course not just separation distress, but also the extremely important promoter of social bonding, namely mammalian PLAY, and the central process intervening when there is a problem, namely proto-empathy/maternal CARE, to which sexuality (LUST) is added later in development as creatures approach and then reach reproductive competence. This large group of diverse but interacting processes enables one to research attachment in a significantly more differentiated fashion. In this vein, one might worry that labeling separation distress as “activation of the attachment system” could actually confuse people about what social bonds and attachment might really reflect. Bonded creatures, for example, clearly are more likely to engage in a reciprocity of smiling/laughing interactions (PLAY), and such playful interactions, although this has been only modestly researched, clearly promote social preference and approach (seeking out of the more playful other), and thus acting as critical glue for social bonding. Trevarthyn, as early as Citation1979, and before much of Panksepp’s seminal work, outlined how these playful “proto-conversations” indexing a critical “intersubjectivity” (which he argued was heavily dependent on a mutuality of smiling), tickling and play responses were critical to successful and healthy maternal-infant attachment. Work by Ed Tronick (Citation1989) and colleagues have also extended these notions on the micro-structure of attachment, looking at the ultrafine grain of infant and maternal behavior in these proto-conversations and how they might reflect attunement, empathy, maternal responsiveness, or the lack thereof.
As a practical implication of these considerations, I would wonder specifically about how much the traditional attachment disorders nosology (ambivalent/anxious-preoccupied attachment, avoidant-dismissive attachment, and disorganized attachment) could be more deeply analyzed or unpacked in terms of relative activations of the various prototype emotional systems in both parties in a parental-child dyad, how those patterns might be recapitulated in later attachments, as well as various types of defensive response to separation distress, lack of empathy, distancing/ defensive activation of the distressed partner, etc. I have been both a bit surprised and perhaps also a bit disappointed that this classical nosology of attachment disorder from Bowlby and colleagues has never really evolved much beyond its original form(s). Work by Montag et al. (Citation2022) for example might provide some important models for such work.
One strongly suspects that some or most of these attachment style variables may have a significant genetic predisposition, particularly the trajectories for avoidant versus labile/ambivalent/disinhibited styles, reflecting genetic endowment impacting affective regulation/affective activation. Increasingly, there is strong evidence that pathologies of attachment may reflect differential innate immune activity, particularly in view of evidence for example that people with borderline personality disorder suffer from significant pro-inflammatory, neurotrophin and HPA axis abnormalities (see for example Forte et al., Citation2023; Petruso et al., Citation2023). How much this is cause versus effect in this relationship of course is an open question, especially given the accumulating in large body of evidence cited in the target article about the impact of adverse childhood experiences on social peptide, endocrine, BDNF, and immune regulation, but these distinctions may be somewhat specious, in view of the evidence for circular causality/recursion. This means that syndromes may emerge from the nontrivial amplification of positive feedback cycles between: (1) various neurobiological derailments, (2) gravitation to or entrapment in poor social environments that may recapitulate both early abuse and neglect, and (3) social (relationship) dysfunction within those environments creating more negative affects. Obviously, it is a resource-demanding scientific project to examine the evolution of attachment pathology over a longer development course looking at these three streams of causality, but it might offer significant return on the investment. To my knowledge, such a longer-term survey using these interdisciplinary affective neuroscience tools and concepts has never been done, despite the widespread meme that psychiatry is based on a “biopsychosocial model.” But it is an example of again of how, at the risk of repeating myself, that we have far too many hidden and protected “knowledge silos,” too many points of view that regard themselves as independent from their true neighboring (and fertilizing?) disciplines – this was a central theme of the target article. Indeed, I would count my target article successful even if people learned next to nothing about depression, but simply began to be less comfortable with their favored and mostly insular perspectives on depression or other syndromes.Footnote1 And only research using the tools of both disciplines can begin to more fully appreciate the interactions between biology and psychology (most especially as we saw in the depression mechanistic causality space, the HPA axis, the gut-brain axis, numerous social neuropeptide systems classically relevant in attachment such as opioids and oxytocin, the immune system, and the complex glutamatergic axis of the brain – see Depressive Matrix recursion table on page 121 in Watt, Citation2023). Indeed, work on these critical neurobiological variables in borderline conditions is already well underway, and these kinds of investigations can only deepen our understanding of the emergent processes – the vicissitudes of attachment and its disorders.
On core anxieties … or precedent motivations?
Much time is spent in psychology courses debating the questions raised by Prof. Steele about what might be “the core anxiety in human life” – whether it is “annihilation anxiety,” perhaps in modern affective neuroscience terms indexing activation of the fear system, or whether it is separation distress. It's hard to discuss or evaluate these questions in the abstract (without a specific adaptive context), but combat experience consistently demonstrates that individuals will risk and even sacrifice their lives to protect their mates, suggesting that preservation of social connection trumps even the strongest of classic homeostatic (pain) and self-preservation (fear) instincts. Or perhaps it simply reflects a kind of unconscious calculus that it is better to die a beloved and well-remembered hero than it is to survive as a shamed coward. In any case, the hedonic calculus in these scenarios suggests a relative primacy of social bonds and connection to others, even over other powerful prototype emotional and homeostatic “instinct” systems.
Another way to think about this course is that in our evolutionary trajectory, where social bonds had a huge influence on survival, this motivational system became precedent, and although it clearly interacts with other systems in a variety of antagonistic and agonistic ways, the mandate for social bonding and the preservation of those connections is – at least some of the time – able to trump other fears and concerns, to secure its primary agenda as a motivational sine qua non. This simply underscores what has been appreciated ever since Aristotle – that humans are social beings to their core, and social bonds, and their extension into tribes and communities and eventually into the creation of larger social aggregates (exemplified by modern nation-states) became a defining feature of human adaptation, and a/the central engine in our survival (Samson, Citation2023). Another way to think about this in evolutionary terms is that although originally mammals were prey species for many millions of years, we transitioned to apex predator in the context of the relative simultaneity of (1) group/tribal social bonds as extensions of original family and dyadic bonds, (2) the ability to make tools and weapons, (3) the capacity to communicate and plan, and (4) the emergence of coordinated hunting and predatory behavior underpinned by such bonded group cohesion in early hunter gatherers. This created an all-for-one and one-for-all kind of ethos, still plainly visible in many forms of human tribalism, against which no other larger, faster, stronger, and previously dominant apex predator really had much of a chance. Just 10 or so humans with primitive weapons, acting in a cohesive fashion to protect one another, were a lethal and eventually successful opposition for all the other previously dominant apex predators. And although we have no clear data on what happened to the other hominid species with which we initially shared the planet for 200,000+ years, one has to wonder if they were also but inevitably no match in a variety of ways for tribally-coordinated and armed human aggression. Whether that came from better weapons, better communication, better tribal cohesion, or some combination of those, remains a mystery. What seems hard to contest is the role that extended social bonds beyond the family, but still hinged to that familial core, played in this transition to the dominance of Homo sapiens. Aristotle, Bowlby, and Jaak Panksepp all were very much on the right, and same, track.
Conclusions
My main hope would be that the target article, the commentaries, and the responses stimulate people's curiosity and interest in the literally bottomless well of complexity in the mind/brain, with the puzzle box of depression as a great example of that complexity. Depression represents a primary challenge to long-term health in Western societies, where our current track record suggests we are not approaching this problem with either the respect it deserves, or with sufficient resources or adequate circumspection for the harsh limitations of traditional mainline antidepressants. We need a complete overhaul of our default treatment of depression away from simply placing people on SSRIs or SNRIs, a practice which is grounded in the obfuscation of depression's real neurobiology in separation distress/social defeat, and in the deep interdigitation of the social brain with the endocrine and immune systems. A secondary objective might be to energize attention to the benefits of lifestyle overhaul for folks suffering from depression, reducing a pervasive pro-inflammatory gene/environment mismatch, towards improved diet/microbiome, exercise, sleep and social support, as foundational protection against all the diseases of aging as well as depression. This would place primary prevention where it should be in our healthcare system – in the driver’s seat, instead of being still generally neglected. A tertiary objective would be to help healthcare systems reconceptualize and mitigate the challenges of poverty on mental health, promote the much-needed recognition that poverty presents a form of chronic social defeat, and thus must be seen as a primary depressogen.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
1 As perhaps one of the outstanding examples of this, with massive implications for our serial and compounding failure to develop disease-modifying interventions in relationship to all the neurodegenerative disorders, Alzheimer’s research has been for decades myopically focused on the amyloid hypothesis. And although this is changing under the sheer massive weight of contradictory evidence against amyloid deposition as a simple single factor explanation (while undeniably contributing to the recursion puzzle box of neurodegeneration), the science of aging, as the proper precedent science for any disease of aging, has been largely marginalized in many treatments of the major neurodegenerative disorders, again, until more recently.
References
- Abdallah, C. G., Averill, L. A., Akiki, T. J., Raza, M., Averill, C. L., Gomaa, H., Adikey, A., & Krystal, J. H. (2019). The neurobiology and pharmacotherapy of posttraumatic stress disorder. Annual Review of Pharmacology and Toxicology, 59(1), 171–189. https://doi.org/10.1146/annurev-pharmtox-010818-021701
- Allen, N. B., & Badcock, P. B. T. (2003). The social risk hypothesis of depressed mood: Evolutionary, psychosocial, and neurobiological perspectives. Psychological Bulletin, 129(6), 887–913. https://doi.org/10.1037/0033-2909.129.6.887
- Badcock, P. B., & Friston, K. J. (2023). The separation distress hypothesis: The ultimate theory of depression? Neuropsychoanalysis, 25(2), 161–164. https://doi.org/10.1080/15294145.2023.2262471
- Barbano, A. C., van der Mei, W. F., deRoon-Cassini, T. A., Grauer, E., Lowe, S. R., Matsuoka, Y. J., O'Donnell, M., Olff, M., Qi, W., Ratanatharathorn, A., Schnyder, U., Seedat, S., Kessler, R. C., Koenen, K. C., & Shalev, A. Y., & International Consortium to Prevent PTSD. (2019). Differentiating PTSD from anxiety and depression: Lessons from the ICD-11 PTSD diagnostic criteria. Depression and Anxiety, 36(6), 490–498. https://doi.org/10.1002/da.22881
- Becker, C., Zeau, B., Rivat, C., Blugeot, A., Hamon, M., & Benoliel, J. J. (2008). Repeated social defeat-induced depression-like behavioral and biological alterations in rats: Involvement of cholecystokinin. Molecular Psychiatry, 13(12), 1079–1092. https://doi.org/10.1038/sj.mp.4002097
- Bitsko, R. H., Claussen, A. H., Lichtstein, J., Black, L. J., Everett Jones, S., Danielson, M. D., Hoenig, J. M., Davis Jack, S. P., Brody, D. J., Gyawali, S., Maenner, M. M., Warner, M., Holland, K. M., Perou, R., Crosby, A. E., Blumberg, S. J., Avenevoli, S., Kaminski, J. W., & Ghandour, R. M. (2022). Mental health surveillance among children — United States, 2013–2019. MMWR Supplements, 71(2), 1–42. https://www.cdc.gov/mmwr/volumes/71/su/su7102a1.htm
- Blinder, B. J. (2023). Commentary on Douglas Watt’s “The separation distress hypothesis of depression - update and systematic review” with some additional thoughts on rumination and object loss. Neuropsychoanalysis, 25(2), 165–168. https://doi.org/10.1080/15294145.2023.2264302
- Bowlby, J. (1973). Attachment and loss, Vol. 2: Separation. Basic Books
- Bowlby, J. (1980). Attachment and loss, Vol. 3: Loss, sadness & depression. Basic Books.
- Burgdorf, J., Kroes, R. A., Beinfeld, M. C., Panksepp, J., & Moskal, J. R. (2010). Uncovering the molecular basis of positive affect using rough-and-tumble play in rats: A role for insulin-like growth factor I. Neuroscience, 168(3), 769–777. https://doi.org/10.1016/j.neuroscience.2010.03.045
- Carrara, M., Ferrario, M., Bollen Pinto, B., & Herpain, A. (2021, May 17). The autonomic nervous system in septic shock and its role as a future therapeutic target: A narrative review. Annals of Intensive Care, 11(1), 80. https://doi.org/10.1186/s13613-021-00869-7
- Chatoor, I., Dickson, L., & Einhorn, A. (1984). Rumination: etiology and treatment. Pediatric Annals, 13(12), 924–929.
- Cooney, R. E., Joormann, J., Eugène, F., Dennis, E. L., & Gotlib, I. H. (2010). Neural correlates of rumination in depression. Cognitive, Affective, & Behavioral Neuroscience, 10(4), 470–478. https://doi.org/10.3758/CABN.10.4.470
- Cordain, L., Eaton, S. B., Sebastian, A., Mann, N., Lindeberg, S., Watkins, B. A., O’Keefe, J. H., & Brand-Miller, J. (2005). Origins and evolution of the western diet: Health implications for the 21st century. The American Journal of Clinical Nutrition, 81(2), 341–354. https://doi.org/10.1093/ajcn.81.2.341
- Dagan, Y., & Yager, J. (2019). Addressing loneliness in complex PTSD. Journal of Nervous & Mental Disease, 207(6), 433–439. https://doi.org/10.1097/NMD.0000000000000992
- Davis, K., & Panksepp, J. (2018). The emotional foundations of personality: A neurobiological and evolutionary approach. WW Norton.
- Davis, K. L., & Montag, C. (2023). Unraveling the roots of depression – it’s complicated. Neuropsychoanalysis, 25(2), 171–174. https://doi.org/10.1080/15294145.2023.2261454
- Drevets, W. C., Wittenberg, G. M., Bullmore, E. T., & Manji, H. K. (2022). Immune targets for therapeutic development in depression: Towards precision medicine. Nature Reviews Drug Discovery, 21(3), 224–244. https://doi.org/10.1038/s41573-021-00368-1
- Eaton, S. B., & Eaton, S. B., III. (2002). Hunter gatherers and human health. In R. B. Lee & R. Daly (Eds.), The Cambridge encyclopedia of hunter gatherers (pp. 449–457). Cambridge University Press.
- Forte, A. R. C. C., Lessa, P. H. C., Chaves Filho, A. J. M., Aquino, P. E. A., Brito, L. M., Pinheiro, L. C., Juruena, M. F., Lucena, D. F., de Rezende, P. H. F., & de Vasconcelos, S. M. M. (2023). Oxidative stress and inflammatory process in borderline personality disorder (BPD): A narrative review. Brazilian Journal of Medical and Biological Research, 56, e12484. https://doi.org/10.1590/1414-431X2023e12484
- Fox, R., Hyland, P., Coogan, A. N., Cloitre, M., & McHugh Power, J. (2022). Posttraumatic stress disorder, complex PTSD and subtypes of loneliness among older adults. Journal of Clinical Psychology, 78(2), 321–342. https://doi.org/10.1002/jclp.23225
- Frank, A. (1969). The unrememberable and the unforgettable. The Psychoanalytic Study of the Child, 24(1), 48–77. https://doi.org/10.1080/00797308.1969.11822686
- Freud, S. (1914g). Remembering, repeating and working-through. In The standard edition of the complete the collected works of Sigmund Freud (Vol. 12, pp. 147–156). Hogarth.
- Gordon, N. S., Burke, S., Akil, H., Watson, S. J., & Panksepp, J. (2003). Socially-induced brain “fertilization”: Play promotes BDNF transcription in the amygdala and dorsolateral frontal cortex in juvenile rats. Neuroscience Letters, 341(1), 17–20. https://doi.org/10.1016/s0304-3940(03)00158-7
- Hagen, E. H. (2011). Evolutionary theories of depression: A critical review. The Canadian Journal of Psychiatry, 56(12), 716–726. https://doi.org/10.1177/070674371105601203
- Hennessy, M. B., Chun, K., & Capitanio, J. P. (2017). Depressive like behavior, its sensitization, social buffering, and altered cytokine responses in rhesus macaques moved from outdoor social groups to indoor housing. Social Neuroscience, 12(1), 65–75. https://doi.org/10.1080/17470919.2016.1145595
- Hennessy, M. B., Deak, T., & Schiml-Webb, P. A. (2001). Stress-induced sickness behaviors: An alternative hypothesis for responses during maternal separation. Developmental Psychobiology, 39(2), 76–83. https://doi.org/10.1002/dev.1031
- Hennessy, M. B., Schiml, P. A., Berberich, K., Beasley, N. L., & Deak, T. (2019). Early attachment disruption, inflammation, and vulnerability for depression in rodent and primate models. Frontiers in Behavioral Neuroscience, 7(12), 314. https://doi.org/10.3389/fnbeh.2018.00314
- Hennessy, M. B., Schiml-Webb, P. A., & Deak, T. (2009). Separation, sickness, and depression: A new perspective on an old animal model. Current Directions in Psychological Science, 18(4), 227–231. https://doi.org/10.1111/j.1467-8721.2009.01641.x
- Hidaka, B. H. (2012). Depression as a disease of modernity: Explanations for increasing prevalence. Journal of Affective Disorders, 140(3), 205–214. https://doi.org/10.1016/j.jad.2011.12.036
- Holt-Lunstad, J., & Golant, S.2023). Our epidemic of loneliness and isolation: U.S. surgeon general’s advisory on the healing effects of social connection and community. US Gov Health and Human Services. https://www.hhs.gov/sites/ default/files/surgeon-general-social-connection-advisory.pdf
- Insel, T. (2022). Healing: Our path from mental illness to mental health. Penguin Press.
- Keeney, A., Jessop, D. S., Harbuz, M. S., Marsden, C. A., Hogg, S., & Blackburn-Munro, R. E. (2006). Differential effects of acute and chronic social defeat stress on hypothalamic-pituitary-adrenal axis function and hippocampal serotonin release in mice. Journal of Neuroendocrinology, 18(5), 330–338. https://doi.org/10.1111/j.1365-2826.2006.01422.x
- Kendler, K. S., Hettema, J. M., Butera, F., Gardner, C. O., & Prescott, C. A. (2003). Life event dimensions of loss, humiliation, entrapment, and danger in the prediction of onsets of major depression and generalized anxiety. Archives of General Psychiatry, 60(8), 789–796. https://doi.org/10.1001/archpsyc.60.8.789
- Kernberg, O. (1975). Borderline conditions and pathological narcissism. Jason Aronson.
- Kernberg, O. (2022). Some implications of new developments in neurobiology for psychoanalytic object relations theory. Neuropsychoanalysis, 24(1), 3–12. https://doi.org/10.1080/15294145.2021.1995609
- Kernberg, O. (2023). Commentary on “The separation distress hypothesis of depression – an update and systematic review”. Neuropsychoanalysis, 25(2), 169–170. https://doi.org/10.1080/15294145.2023.2261944
- Lebrun-Harris, L. A., Ghandour, R. M., Kogan, M. D., & Warren, M. D. (2022). Five-year trends in US children’s health and well-being, 2016–2020. JAMA Pediatrics, 176(7), e220056. https://doi.org/10.1001/jamapediatrics.2022.0056
- Montag, C., Elhai, J. D., & Davis, K. L. (2021). A comprehensive review of studies using the affective neuroscience personality scales in the psychological and psychiatric sciences. Neuroscience & Biobehavioral Reviews, 125, 160–167. https://doi.org/10.1016/j.neubiorev.2021.02.019
- Montag, C., Solms, M., Stelzel, C., & Davis, K. L. (2022). The future of the affective neuroscience personality scales: A reflection on seven pressing matters. Personality Neuroscience, 5, e10. https://doi.org/10.1017/pen.2022.2
- Montag, C., Widenhorn-Müller, K., Panksepp, J., & Kiefer, M. (2017). Individual differences in Affective Neuroscience Personality Scale (ANPS) primary emotional traits and depressive tendencies. Comprehensive Psychiatry, 73, 136–142. https://doi.org/10.1016/j.comppsych.2016.11.007
- Moskal, J. R., Burch, R., Burgdorf, J. S., Kroes, R. A., Stanton, P. K., Disterhoft, J. F., & Leander, J. D. (2014). GLYX-13, an NMDA receptor glycine site functional partial agonist enhances cognition and produces antidepressant effects without the psychotomimetic side effects of NMDA receptor antagonists. Expert Opinion on Investigational Drugs, 23(2), 243–254. https://doi.org/10.1517/13543784.2014.852536
- National Center for Health Statistics. U.S. Census Bureau, Household Pulse Survey, 2020–2024. Anxiety and Depression. Generated interactively: from https://www.cdc.gov/nchs/covid19/pulse/mental-health.htm
- Nemeroff, C. B. (2016). Paradise lost: The neurobiological and clinical consequences of child abuse and neglect. Neuron, 89(5), 892–909. https://doi.org/10.1016/j.neuron.2016.01.019
- Nemeroff, C. B. (2020). The state of our understanding of the pathophysiology and optimal treatment of depression: Glass half full or half empty? American Journal of Psychiatry, 177(8), 671–685. https://doi.org/10.1176/appi.ajp.2020.20060845
- Nicholson, C. (2012). The humor gap. Scientific American. https://www.scientificamerican.com/article/the-humor-gap-2012-10-23/
- Nolen-Hoeksema, S., Wisco, B. E., & Lyubomirsky, S. (2008). Rethinking rumination. Perspectives on Psychological Science, 3(5), 400–424. https://doi.org/10.1111/j.1745-6924.2008.00088
- Packheiser, J., Hartmann, H., Fredriksen, K., Gazzola, V., Keysers, C., & Michon, F. (2024). A systematic review and multivariate meta-analysis of the physical and mental health benefits of touch interventions. Nature Human Behaviour, https://doi.org/10.1038/s41562-024-01841-8
- Panksepp, J. (1998). Affective neuroscience: The foundations of human and animal emotion. Oxford University Press.
- Papageorgiou, C., & Wells, A. (2004). Nature, functions, and beliefs about depressive rumination. In C. Papageorgiou & A. Wells (Eds.), Depressive rumination: Nature, theory and treatment (pp. 1–20). John Wiley & Sons.
- Petruso, F., Giff, A. E., Milano, B. A., De Rossi, M. M., & Saccaro, L. F. (2023). Inflammation and emotion regulation: A narrative review of evidence and mechanisms in emotion dysregulation disorders. Neuronal Signaling, 7(4), NS20220077. https://doi.org/10.1042/NS20220077
- Pinneau, S. R. (1955). The infantile disorders of hospitalism and anaclitic depression. Psychological Bulletin, 52(5), 429–452. https://doi.org/10.1037/h0045987
- Pontzer, H., Wood, B. M., & Raichlen, D. A. (2018). Hunter-gatherers as models in public health. Obesity Reviews, 19(Suppl 1), 24–35. https://doi.org/10.1111/obr.12785
- Radell, M. L., Hamza, E. A., & Moustafa, A. A. (2020). Depression in post-traumatic stress disorder. Reviews in the Neurosciences, 31(7), 703–722. https://doi.org/10.1515/revneuro-2020-0006
- Roque, S., Mesquita, A. R., Palha, J. A., Sousa, N., & Correia Neves, M. (2014). The behavioral and immunological impact of maternal separation: A matter of timing. Frontiers in Behavioral Neuroscience, 8, 192. https://doi.org/10.3389/fnbeh.2014.00192
- Sakamoto, T., Ishio, Y., Ishida, Y., Mogi, K., & Kikusui, T. (2021). Low maternal licking/grooming stimulation increases pain sensitivity in male mouse offspring. Experimental Animals, 70(1), 13–21. https://doi.org/10.1538/expanim.20-0030
- Samson, D. R. (2023). Our tribal future: How to channel our foundational human instincts into a force for good. McMillan Publishers: St. Martin's Press.
- Sanchez, M. M. (2006). The impact of early adverse care on HPA axis development: Nonhuman primate models. Hormones and Behavior, 50(4), 623–631. https://doi.org/10.1016/j.yhbeh.2006.06.012
- Schiller, D., Yu, A. N. C., Alia-Klein, N., Becker, S., Cromwell, H. C., Dolcos, F., Eslinger, P. J., Frewen, P., Kemp, A. H., Pace-Schott, E. F., Raber, J., Silton, R. L., Stefanova, E., Williams, J. H. G., Abe, N., Aghajani, M., Albrecht, F., Alexander, R., Anders, S., … Lowe, L. (2022). The human affectome. Neuroscience and Biobehavioral Reviews, https://doi.org/10.31234/osf.io/9nu32
- Siviy, S. (2023). Adding a layer of playfulness? Neuropsychoanalysis, 25(2), 175–176. https://doi.org/10.1080/15294145.2023.2263002
- Slavich, G. M. (2020). Social safety theory: A biologically based evolutionary perspective on life stress, health, and behavior. Annual Review of Clinical Psychology, 16(1), 265–295. https://doi.org/10.1146/annurev-clinpsy-032816-045159
- Slavich, G. M., & Irwin, M. R. (2014). From stress to inflammation and major depressive disorder: A social signal transduction theory of depression. Psychological Bulletin, 140(3), 774–815. https://doi.org/10.1037/a0035302
- Slavich, G. M., Thornton, T., Torres, L. D., Monroe, S. M., & Gotlib, I. H. (2009). Targeted rejection predicts hastened onset of major depression. Journal of Social and Clinical Psychology, 28(2), 223–243. https://doi.org/10.1521/jscp.2009.28.2.223
- Smith, K. (2014). Mental health: A world of depression. Nature, 515(7526), 180–181. https://doi.org/10.1038/515180a
- Solms, M. (2021). Revision of Drive Theory. Journal of the American Psychoanalytic Association, 69(6), 1033-1091. https://doi.org/10.1177/00030651211057041
- Spitz, R. A. (1946). Hospitalism; A follow-up report on investigation described in volume I, 1945. The Psychoanalytic Study of the Child, 2(1), 113–117. https://doi.org/10.1080/00797308.1946.11823540.
- Trevarthen, C. B. (1979). Communication and cooperation in early infancy: A description of primary intersubjectivity. In M. Bullowa (Ed.), Before speech. Cambridge University Press.
- Tronick, E. Z.1989). Emotions and emotional communication in infants. American Psychologist, 44(2), 112–119. https://doi.org/10.1037/0003-066X.44.2.112
- van der Kolk, B. (Ed.). (2015). The body keeps the score: Brain, mind, and body in the healing of trauma. Penguin Books. ISBN: 978-0143127741.
- Watson, P. J., & Andrews, P. W. (2002). Toward a revised evolutionary adaptationist analysis of depression: The social navigation hypothesis. Journal of Affective Disorders, 72(1), 1–14. https://doi.org/10.1016/s0165-0327(01)00459-1
- Watt, D. F. (1986). Transference – a right hemisphere event?: The boundary between psychoanalytic metapsychology and neuropsychology. Psychoanalysis and Contemporary Thought, 9(1), 43–77. [No DOI available].
- Watt, D. F. (2014a). The biology of aging: Implications for understanding the diseases of aging and healthcare in the 21st century. Contributed chapter. In A. K. Nair & M. N. Sabbagh (Eds.), Geriatric neurology (pp. 3–37). Wiley-Blackwell.
- Watt, D. F. (2014b). Depression in the elderly: Interactions with aging, stress, chronic pain, inflammation, and neurodegenerative disorders. Contributed chapter. In A. K. Nair & M. N. Sabbagh (Eds.), Geriatric neurology (pp. 288–301). Wiley-Blackwell.
- Watt, D. F., & Panksepp, J. (2009). Depression: An evolutionarily conserved mechanism to terminate protracted separation distress. A review of aminergic, peptidergic and neural network perspectives. (Target article with invited commentaries). Neuropsychoanalysis, 11(1), 7–51. https://doi.org/10.1080/15294145.2009.10773593
- Watt, D. F. (2023). The separation distress hypothesis of depression – an update and systematic review. Neuropsychoanalysis, 25(2), 103–159. https://doi.org/10.1080/15294145.2023.2240340.
- Watt, D. F., & Pincus, D. (2004). Neural substrates of consciousness: Implications for clinical psychiatry. In J. Panksepp (Ed.), Textbook of Biological Psychiatry (75–110). Wiley Press.
- Weber, M. D., Godbout, J. P., & Sheridan, J. F. (2017). Repeated social defeat, neuroinflammation, and behavior: Monocytes carry the signal. Neuropsychopharmacology, 42(1), 46–61. https://doi.org/10.1038/npp.2016.102