Reflections on the neuroscientific legacy of Jaak Panksepp (1943–2017)

Nature has placed mankind under the government of two sovereign masters, pain and pleasure … they govern us in all we do, in all we say, in all we think: every effort we can make to throw off our subjection will serve but to demonstrate and confirm it.

Jeremy Bentham

One of the truly great voices in modern neuroscience was silenced recently, when Jaak Panksepp suddenly died on April 18th of this year, following a fairly routine surgical procedure that went badly awry, generating a serious ischemic-hypoxic insult, a tragedy from which not even Jaak Panksepp could recover. Jaak had battled long and hard with recurring bouts of several cancers, and as is too often the case, it was actually the cancer therapy process that eventually killed him and not the cancer itself, although that distinction offers little solace. Nevertheless, the terrible and final suddenness of all this, despite Jaak’s age and medical history, was shocking and deeply upsetting to all of us who were close to Jaak. None of us is truly prepared for these departures at the end of life, even though we all know that they are coming eventually, and inevitably.

Jaak’s neuroscientific work was, as many now appreciate, centered at least in part around separation distress as a prototype emotional state. He has had much to say, over the many years of his career, about the nature of the very grief that all who knew Jaak well have been feeling since his passing. That grief still lies heavy on many hearts, of family, friends and colleagues. The personal and professional losses for me are almost impossible to separate, as I lost a senior colleague, a highly valued friend, and a mentor in neuroscience, who was decisively the largest and most positive single influence in the second half of my professional career. But neuroscience as a field has also lost at least as much – an irreplaceable leader, a brilliant scholar and scientist, and a most passionate advocate for an affect-centered view of the brain and mind. I take solace in my sense of great good fortune for having had 20+ years of close colleagueship and friendship with Jaak, sharing half a dozen week-long Affective Neuroscience seminars, many book chapters, dozens of larger conferences, and many other fruitful and sundry projects. I believe that over time, his intellectual legacy and its many gifts to the multiple sub-fields of mind and brain science will prove increasingly resonant and prescient, guiding generations of young scientists and clinicians.

Progress in science is intrinsically slow by its very nature. But once in a while, a highly creative genius comes along, whose insight and integration of large amounts of previously fragmented material, creating new paradigms and concepts, moves the field forward in sudden and deceptive leaps. In the increasingly “twinned” fields of neuroscience and psychology, Jaak was one of those people. By any conventional standard of scientific productivity, he was a giant contributor to the neuroscience of emotion and was staggeringly prolific, with over 450 empirical and review publications, and literally many dozens of book chapters in various edited volumes and compilations. His publication vitae centrally featured his 1998 textbook, Affective Neuroscience, increasingly regarded as a landmark work and a classic publication in neuroscience. This magnum opus was updated in the less technical and more accessible treatment, The Archaeology of Mind, co-authored with Lucy Biven (Panksepp and Biven, 2012). He also solo edited the well-regarded two-volume Advances in Biological Psychiatry in 1995–1996 (Panksepp, 1995, 1996), the more recent Textbook of Biological Psychiatry in 2004, and the highly technical compilation Handbook of the Hypothalamus with Peter Morgane, in four volumes (1979, 1980a, 1980b, 1981), where one could first see the flowering of Jaak’s powerful intellect and his encyclopedic grasp of neuroscience. Very few have read this particular four volume compilation on the hypothalamus, but despite its age, it is still to this day the most comprehensive collection of work on that critically important neural system and its complex functional networks, which are unquestionably central to emotion as well as homeostasis (the importance of the hypothalamus for emotion likely deriving from how homeostasis provided the evolutionary foundations for emotion). The very last of his many edited volumes, The Psychology and Neurobiology of Empathy, was completed with me in 2016. His huge body of work has had an enduring and pervasive influence on the fields of psychology, psychiatry, and both clinical as well as experimental neuroscience, and has inspired at least two generations of young scientists eager to unravel the mysteries of the brain and mind and the pivotal role of emotion in its neurodynamics.

Where and what were Jaak Panksepp’s greatest contributions? This is not an easy question to answer, because his interests spanned many traditional disciplinary boundaries, including such disparate topics as a theory of music, the fine-grained anatomy and connectivity of the mesodiencephalon, and the nature of laughter as an affective signal. And yet the common denominator in all of these was Jaak’s keen and unwavering focus on the nature of affect, and the sometimes surprising implications of a deep understanding of affect for every other important psychological and neuroscientific territory. The following list of Jaak’s major areas of scientific contribution has to be considered simply my personal scientific opinion, and therefore hardly definitive or unbiased, but I might outline three main areas of scientific contribution. Others – indeed, many others – might be added, but space considerations demand a shorter list. So here’s mine.

On a fundamental, intrinsic relationship between consciousness and emotion

As perhaps his most critical contribution, Jaak continually emphasized that emotion and affect must figure centrally in any potential neuro-architectures creating a conscious mind, what I would call an “affect-centric” view of the embodied mind. Affects are evolutionarily conserved core executive routines guiding the living mind. This strong conviction that emotion provided something much more basic than an interesting “coloration” to the neural foundations for consciousness contrasted starkly with the cognition-centered and predominantly sensory view of consciousness at the time of his early seminal work and publications. Instead of this conventional sensory/cognitive perspective on conscious architectures, Jaak conceptualized affects as signaling functional integrations foundational to any version of a conscious architecture, implying, significantly, that all strong affects are intrinsically conscious. These and other shared scientific convictions originally brought us together, and created an immediate intellectual bond that was to energize our many joint ventures and projects. Thus, it was no true coincidence that we met for the first time in the second biannual Tucson Series on the Science of Consciousness in 1996. It would be an understatement to suggest that affective neuroscience perspectives were underrepresented at the Tucson Conference in 1996. Jaak and I shared an immediate sense of dismay, tempered perhaps by a certain amusement, that the fields of psychology and neuroscience seemed at times so blissfully unaware that a mind without the internal “value compass” provided by emotion and affect had no basis for organizing basic behavioral, or for that matter, higher attentional and cognitive priorities. “What are they smoking?” we wondered over drinks at that first conference! By the end of that first conversation in 1996, it was clear that we shared an enormous amount, in terms of our basic conceptions of how the mind and brain might work, although it was also clear at that point that his empirical knowledge of brain anatomy and neurochemistry, especially in the deep subcortical regions of mesodiencephalon and brainstem, far exceeded mine.

A long-lasting theme in Jaak’s work was his ongoing attempt to provide antidotes to the reigning and popular “cognitive imperialisms” that we experienced at Tuscon, and which Jaak and I had both struggled against separately, long before we ever met. We shared a sense of astonishment that affects were not more universally appreciated as the true “reinforcers” referenced by behaviorism, and wondered how anyone in the many disparate fields of psychology and neuroscience could somehow miss this central “keystone” fact of our psychology. As consciousness had been, for the four previous decades, a mostly discredited subject under the still powerful aegis of behaviorism, we were joined in this early conviction in the mid-1990s by precious few in neuroscience, but with the notable exception of Antonio Damasio. Damasio, Joe LeDoux, and several other luminaries in emotion theory and consciousness studies all participated as commentators in one of my first ventures with Jaak, an ASSC (Association for the Scientific Study of Consciousness) web-seminar on Emotion and Consciousness, given in the summer of 1998, which generated lively debates still resonating to this day about exactly how to conceptualize the role of emotion in consciousness. Can consciousness and emotion be simply conceived of as largely orthogonal processes, as Joe LeDoux argued, with just a single intersection, when emotion created a conscious affective state (which LeDoux also believed was totally dependent on working memory)? We argued against any dependence of conscious affective experience on working memory, and also that this notable intersection was just the tip of the iceberg, in terms of the rich interweaving of affective signals into every aspect of conscious life – from what we find interesting, to what we avoid, what we crave and seek out, and from what we learn forever to what we forget next week – indeed, virtually every aspect of the mind. The crux of our early shared conviction was simply that affects (both homeostatic and emotional) and their underlying brain systems, had to be foundational in some poorly understood fashion for how neural activity gives rise to an organized agency – a “selfhood” in the brain. Furthermore, such a self would be an “agentic self” and not any version of a passive observer of sensory fields, with behavioral initiatives and basic motivation intrinsically organized around and by those emotional and homeostatic affects. In other words, it made sense to assume that all behavior was forever hinged to the pursuit of the many and varied positives, and in the service of minimizing the many varied negatives, of life, echoing Bentham’s early brilliant intuition, an insight essentially recapitulated by Freud 100 years later. Without those “value signals”, there is simply no other basis for understanding basic motivation or the fine and deep selectivity of behavior.

This set of early assumptions – which ran largely counter to the sensory and cognition-centric views in early consciousness studies – is being slowly but progressively validated in a large number of lines of research and theoretical work. In the early days of the resurgence of interest in consciousness from a neuroscience standpoint, around the time of his publication of his seminal work Affective Neuroscience, we found an unlikely early ally in the global workspace perspectives of cognitive theorist Bernie Baars and colleagues such as Jim Newman. This central neurologic dependency of conscious architectures upon affective signals has been revealed in the functional dependence of cortical systems on the integrations achieved in the mesodiencephalon to do any form of meaningful and organized cognitive work (see Merker, 2007 for the most definitive and exhaustive summary to date and Schiff, 2008). This affect-centric perspective on the Mind/Brain has been critically validated in the much-underappreciated syndrome of akinetic mutism, where collapse of emotion and motivation leaves a virtually rudderless and engine-less thalamocortical system and a disorder of consciousness (see Watt, 2012; Watt & Pincus, 2004 for reviews). And while one can say that neuroscience still lacks its “keystone in the arch” – a definitive and well-validated neural theory of mind as one of the core emergent properties and mysteries in Nature – increasing evidence argues that conscious architectures are highly distributed and non-local, existing as transient adaptive networks, flitting about in an “enchanted loom” in Sherrington’s haunting metaphor, with new transient and highly distributed alliances being formed on a moment-to-moment basis, but bridging the entire vertical axis of the brain (Schiff, 2008). As further support for Jaak’s affect-centric perspectives, he and I would both argue that it is now beyond meaningful dispute that the mesodiencephalon plays a critical role in such distributed networks. The structures in the upper brainstem and medial diencephalic territories are rich in ancient homeostatic and affective systems, and an extensive lesion of these mesodiencephalic territories will collapse consciousness totally and permanently (see again Schiff, 2008), even without a trace of damage to any aspect of the more telencephalic thalamocortical systems that are virtually the exclusive anatomical focus of cognitive neuroscience.

In other words, at this point in the evolution of a neuroscientific theory of mind, we think the evidence is strongly in favor of the assumption that affect is foundational, that it is largely based in brainstem and subcortical circuits, and constitutes a critical prerequisite for the development and creation of cognition, largely mediated by neocortical circuits. Indeed, just a full lesion of the tiny territory of midbrain PAG – at the ventral tip of the large group of brain affective systems running from paleocortex, through limbic regions, to hypothalamus and down into the brainstem, leaves the person “gone” in a comprehensive sense, with a permanent and severe akinetic mutism (see Watt & Pincus, 2004 for detailed case review). This may hint at PAG’s role in laying the neuroevolutionary foundations for those more evolved affective systems, with PAG sitting at the very bottom of the highly distributed affective networks, extending upward from reticular to hypothalamic and basal forebrain regions, and finally into more telencephalic limbic and paralimbic systems.

The neurodynamics of this profound vertical “system integration,” particularly in terms of a developmental trajectory that bootstraps the conscious mind from core homeostatic and affective systems, clearly remain to be more fully elucidated, but the ontogenetic development of consciousness for each of us appears to clearly recapitulate our phylogeny, as the brain builds networks from the inside (ventral and medial regions) out to the more telencephalic lateral and dorsal regions. Although the cortex still dominates cognitive neuroscience – and with good reason as it forms the neural substrate for virtually all cognitive content – I believe that future generations will strongly credit those handful of neuroscientific pioneers who kept pointing at neurodynamic integrations, still poorly understood, but achieved first in more ancient mesodiencephalic, tectal and tegmental systems, as foundational for all the cognitive activity at the top of the system. Jaak will likely be highly regarded and credited as one of those prescient pioneers.

On understanding emotion in terms of interactive but partially discrete prototype states, energized by an overarching, precedent and integrative SEEKING system

Jaak Panksepp also made pivotal contributions to unraveling one of the central mysteries of emotion, namely, that it functions as a central clearinghouse in which all emotional activators must compete and interact. This assertion is based on evidence for global and highly distributed neural system architectures for emotion (an underappreciated direct correlate of the global network ideas just reviewed). In other words, emotion cannot be allocated to any kind of simple “module” or any version of a local network, but has to be understood as a distributed integrative system. This realization made him rather fond of placing the affective primes in all capitals (“FEAR,” “RAGE,” “GRIEF,” “PLAY,” “SEEKING,” “LUST,” “CARE”), as his way of denoting this distributed or “system” property. Although this might have seemed semantically “quirky” to some (including even to this author at times ), it was energized by Jaak’s realization that more global system concepts were mandated, and where perhaps the various affective primes might function as “nodes” within an integrated system, creating thus a “neural clearinghouse” in which all the potential affective states could compete and interact in a variety of adaptively critical antagonistic and agonistic ways. Jaak would truly cringe when someone interpreted any of these capitalized primes as some kind of “module” in the brain. Instead, in his view, the negatives and positives would have to be understood as fundamentally interactive in terms of a basic antagonism (systems of like valance facilitating each other while those of opposing valance would be mutually inhibitory), but there was also evidence for more subtle versions of promotion and inhibition between systems of similar valence. For example, while playfulness might “set the table” so to speak for the activation of sexual desires, at some point excessive playfulness might actually inhibit activation of the LUST system. Such an image of complex agonistic and antagonistic systemic interaction between the primes fits beautifully with the phenomenology of emotion and emotional behaviors, as all kinds of emotional pushes and pulls clearly do agonize and antagonize one another in various ways – part of what gives our (and general mammalian) behavior its delightful and frustrating unpredictability.

In this sense, I believe many emotion theorists, particularly in the social constructionist school, did not fully appreciate how the SEEKING system was conceptualized as a “special class of one” in Panksepp’s system of the seven emotional primes, and that it had to be the evolutionarily precedent or “master” system, a supposition recently gaining impressive validation by empirical work on dopamine in crayfish (Huber, Panksepp, Nathaniel, Alcaro, & Panksepp, 2011). In social constructional theories of emotion, the notion of “core affect” (i.e. Russell & Barrett, 1999), with an emphasis on dimensional aspects of approach/avoidance, arousal, and valence, is not actually operationally dissimilar to how Panksepp conceived of the integrative and superordinate role of the SEEKING system, although there are still important differences between Panksepp and social construction theories that cannot be glossed over. In particular, Panksepp parted company with constructional views if they went so far as to argue that the categorical or “emotional prime” systems (observing for example a principled distinction between PLAY and RAGE as neurobiologically meaningful) were simply a specious cultural convention or “meme” carried over from a pre-scientific “folk” nomenclature, but without any real neurologic and biological basis. (For more detail on opposing points of view, see Barrett, 2006, who challenges the idea that the emotional primes or “affective prototypes” reflect any version of true “natural kinds.”)

Panksepp’s theory of the emotional “primes” provides real ways of immediately and directly linking the SEEKING system to all the other primes. For example, FEAR is the seeking of safety, PLAY the seeking of rough-and-tumble joyful engagement, RAGE the seeking of an end (sometimes “with prejudice”) to any toxic agent or circumstance that might be the source of one’s frustration or injury, maternal CARE the seeking of safety, restored homeostasis, or good feelings in a dependent and vulnerable other (and thus indexing a proto-empathy system, as classically outlined in Panksepp, 1998 and more recently in Watt & Panksepp, 2016). In this critical sense, I sometimes felt that students as well as some critics of Panksepp did not always clearly appreciate that all of these classic affective “primes” or prototypes had to be, at least in some sense, specialized “resonances” or evolved “nodes” interacting within a more ancient and precedent generalized motivational arousal or SEEKING system. Perhaps this key insight was expressed too implicitly at times, although this precedent role of the SEEKING system is stated at several points in his magnum opus (Panksepp, 1998), and then made more explicitly and quite repeatedly in Archeology of Mind, such that it couldn't really be missed.

One underappreciated benefit – indeed one great beauty – of this set of concepts, from the standpoint of scientific theory, is that it provided an integrated neural substrate for the competitive interaction of the emotional primes, where the negatives and positives could easily inhibit each other, a clear functional fact of affective life that any good model had to explain. Additionally, the SEEKING system concept provided relatively navigable passageway from homeostasis to emotion, as the SEEKING system network contains, as one of its core nuclei, the lateral hypothalamus which receives abundant dopaminergic enervation from the midbrain, providing an immediate neural bridge for integrating homeostatic imbalances, learned behavior (via its extensive basal ganglia and mesolimbic connections), and sources of homeostatic relief. Thus, in its most ancient forms, as Jaak famously offered, “the SEEKING system gets thirsty animals to water, hungry animals to food, cold animals to warmer environments,” etc. The SEEKING system creates neural links between the experience of rewards and punishments, on the one hand, and predictors or cues, on the other, and binds both of those classes to basic behavioral routines, such that a hungry animal can activate motor routines to approach a source of food and then initiate consummatory routines, under the guiding influence of this master motivational system. Clinicians need only think of severe clinical depression (particularly the apathetic-retarded variety), in which almost all areas of motivation are compromised when SEEKING massively is inhibited, or mania, in which almost everything is amplified when that system is upregulated, to get a sense of the central role of the SEEKING system. It is hard to imagine a more elegant set of integrative predictive tools about emotion and motivation than what this set of basic ideas affords modern neuroscience, and where its potentially largest payoffs – in relationship to clinical issues – are just barely starting to be “cashed out” (see last sections).

Another underappreciated benefit of Jaak’s emotional system concepts was that they provided a highly plausible theoretical bridge for how homeostasis proper (likely the antecedent evolutionary process) might have given rise to the emotional primes; the evidence suggests that the primary “emotion command” systems emerged from basic evolutionary-predictive extensions of core homeostatic mandates. This was no coincidence, as Jaak cut his teeth on energy balance research. In other words, we could look at each “emotion-command” system as operating fundamentally in the service of a forward-looking extension of homeostasis and organism preservation. For example, fear reflects an anticipatory forward-looking extension of pain and the basic aversion to tissue damage. In fear, we are not yet damaged by an approaching predator or more powerful rival, but we might be, if we do not seek safety. We must therefore either freeze to avoid detection (a metabolically less costly solution to the problem), or flee, if this solution is not possible. Similarly, in separation distress, a young infant mammal finding itself suddenly separated from its caretakers is not yet metabolically compromised, but this separation from both its source of nourishment (metabolic supply) as well as protection from predation signals the probability of danger in the very near future. Therefore, we could speculate that separation distress emerged as an anticipatory forward-looking extension of these basic homeostatic routines to ensure the safety of vulnerable offspring. That the separation distress system relies on lowered opioidergic and oxytocinergic but upregulated CRF signals is consistent with these assumptions.

While this way of linking ancient homeostatic processes to emerging classic emotional primes (suggesting unequivocally that emotion is an evolutionary extension of homeostasis) remains still largely untested, and in need of further investigation, there is no meaningful evidence (that I am aware of) against it. It is also deeply consistent with how evolution functions to extend the brain as an increasingly sophisticated prediction engine, stacking new predictions on top of older successful behavioral routines. Evolution appears to operate from the highly conservative principle of having no need to “re-invent the wheel” in every new context, just tweak and enhance the existing operations a bit here and there to manage the new adaptive context. In fact, we could look at encephalization (the development of brains across species and through our extended phylogeny) as nothing more than the creation of more and more sophisticated prediction engines extending into increasingly complex sensorimotor directions (consistent, for example, with evidence that Wernicke’s area is specialized for the prediction of phonemic sequences (Hagoort & Indefrey, 2014).

Jaak’s early work on the mysteries of feeding and energy balance exposed him at the beginning of his career to the challenge of finding integration across highly varied and different classes of motivational “vectors,” including evolutionarily more ancient and more recent classes of motivations, and his SEEKING system concepts were the brilliant solution. This basic set of theoretical notions finds strong validation in the fact that opioids modulate pain and reward satiety (two very ancient homeostatic operations, presumably preceding the classic emotional primes in evolution), while in more complex and extended mammalian brains, opioids also modulate social bonding, play and separation distress (all critical emotional operations of a highly social brain). Both play and quieter forms of social comfort are high opioidergic states, while separation distress is organized in good part by the inverse condition of low opioidergic signals (validated in humans by Zubieta et al., 2003), coupled perhaps with high CRF signals. In this sense, separation distress might reflect an evolutionary extension of pain, while social comfort in a secure attachment perhaps is an evolutionary extension of a basic homeostatic wellness and satiety.

Thus, the SEEKING system concept provides a ready and plausible evolutionary and functional bridge linking homeostasis and emotion, that is not equaled by any other putative neural system, to my knowledge. This illuminates Panksepp’s successful interpretation of the confusing results of the EBS experiments on the “self-stimulation system” (as originally labeled by Milner and Olds). In these experiments, electrical brain stimulation of SEEKING system trajectories (such as VTA, nucleus accumbens lateral hypothalamus or medial forebrain bundle) could generate disparate behavioral results, including just increased lever-pressing, but also gnawing, eating, drinking, and even copulation, or hunting behaviors (!). Based on this data, Jaak was not seduced by the simplistic behaviorist concept that this was “the reward system,” or that animals self-stimulated because it simply generated some kind of sensory pleasure. He even more strongly rejected the notion that this was the source of behavioral “reinforcement” but without any positive, pleasureable experience (as references to consciousness and subjective experience in behaviorism were deemed unscientific and strictly taboo). Instead, Jaak realized from many experiments that this system had to be conceptualized differently from these appealing oversimplifications, because the data suggested that this SEEKING system supported both the seeking of rewards but also the avoidance of punishments, and indeed virtually every type of motivated behavior.

Understanding of the putative phenomenology associated with SEEKING system’s activation has been a consistent stumbling block in neuroscience, and its designation as “the reward system” hasn't helped matters in this regard, with a clear tendency to assume that this system therefore must be the neural correlate for all pleasure(s). Jaak argued long and hard against this very idea, but this “meme” appears rather deeply entrenched, perhaps because of our deep penchant for and attraction to simplistic localization of function. Instead of such an appealing but ultimately misleading localization of function (perhaps epitomized in the widely quoted neo-phrenology that the amygdala is the neural “center” for fear in the brain), neuroscience consistently challenges us instead to understand how functions emerge from recursive and complex relationships between multiple neural systems, nuclei or regions, bound together into a concerted functional system by multiple connectivities, and by using signals from multiple modulators. There appear to be no exceptions to this principle of distributed networks (there is indeed no single neural system that is the “center” for anything in the brain), whether you are talking about neocortical systems that might help us to appreciate E = MC², or whether you are talking about systems that might control micturition. (The corollary to this principle that multiple neural “players” operate together in a concerted fashion to create a functional network is that all regions participate in multiple functional networks, supporting never one but several functionally valuable operations.)

Although obviously we can't ask the animals why they self-stimulate this system or how they feel when doing this (except to consistently observe that they clearly like it and that it is therefore unquestionably positively valanced), activation of this same system in humans creates the experience of “positive expectancy”, a kind of generalized and rather non-specific high-energy enthusiasm (but from there typically latching onto the ready pursuit of other more specific rewards). This of course led Jaak to initially call this an “expectancy” or “anticipatory” system. This suggests that while unambiguous phenomenological correlates emerging from the SEEKING system’s activity are somewhat elusive, they might be best described in terms of just “having an abundance of motivational and psychological energy for life” … and its many opportunities and challenges. This simple notion of course is totally consistent with the typical phenomenology associated with ingesting a moderate dose of any number of amphetamines, which act via their dopamine and norepinephrine reuptake inhibition, thereby amplifying the SEEKING system level of tonic activation. Put differently, Nature presumably could not allow the activation of the motivational core of the vertebrate and mammalian brain to be simply completely affectively neutral and “un-valenced”, and instead, a positive state was clearly selected, but one also clearly different from many and varied sensory rewards and pleasures. It would indeed be a deep contradiction in terms if basic motivational arousal – SEEKING and positive expectancy – did not have any rewarding quality to it at all, and instead were without valence. Indeed such a contradiction would likely guarantee creatures incapable of acting in a concerted fashion to match needs with opportunities, which likely would have led to extinction of that particular genetic line. These considerations explain, I believe, how it is that the activation of the SEEKING system is intrinsically rewarding without it being “a reward center” so to speak, or any version of “the reward system,” and how chemical or electrical stimulation of this system doesn't have a simple behavioral phenotype. In turn, this notion of a “SEEKING system” helps shed deep light on the problem of addiction, which reflects a progressive hijacking of that core motivational system by a single activity or drug, in which addiction becomes a deepening dysfunction within the brain’s motivational neurodynamics (see Zellner, Watt, Solms, & Panksepp, 2011).

In any case, these basic SEEKING system concepts were a brilliant solution by Jaak Panksepp to a host of thorny conceptual challenges and apparent empirical contradictions, with these ideas emerging as early as one of his chapters in Volume 3 (Behavioral Studies) of the Handbook of Hypothalamus in 1981. This SEEKING system concept allowed Jaak to achieve a theoretical integration and coherence where there had been previously only confusion and apparent contradiction. Within this assumption of the SEEKING system’s precedent and overarching role, Jaak realized that the behavioral and affective trajectories created by EBS of this system should be variable. The behavioral effects of EBS would of necessity depend on many factors, including species and animal personality phenotypes, adaptive context, and even the current “mind set” or mood of the animal, which would all influence which affective “prime” might become emergent from EBS of the mesolimbic mesocortical DA system. Thus, predatory animals like cats might be much more likely to go into a stalking mode, forager species like rats might start sniffing around and exploring the environment, and restrained animals might actually just try to escape more vigorously. Such phenotypic variability and behavior must have disappointed many a strict localizationist craving (dare one say SEEKING?) a plausible neo-phrenology. Unfortunately, this behavioral phenotypic variability resulting from EBS “self-stimulation” has been misconstrued by constructional theorists of emotion as hard evidence against the “categorical” or “affective prime” view. But it’s not – it is simply more evidence that not everyone has clearly understood the full implications of Jaak’s complex “affective prime” theory and the precedent, superordinate role of a “SEEKING” system, as a core motivational system underpinning all behavior.

Then there were the many discrete contributions (too many to specifically enumerate here) that Jaak made to understanding each affective system, aside from the overarching role of a SEEKING system. Most particularly, he made significant empirical contributions in relationship to separation distress/GRIEF, maternal CARE, the organism “defense states” of FEAR and RAGE, and perhaps his favorite system, PLAY. Table 1, below, first published in the ASSC Web-seminar on emotion and consciousness (from Watt, 1998) and featured in many of Jaak’s reviews, summarizes many of his basic empirical findings (table extracted largely from Panksepp, 1998).

Table 1. Neural correlates for the seven prototype emotion systems.

Affective prototype	Distributed neural networks and major structures	Neuromodulators
Generalized motivational arousal – SEEKING	Ventral Tegmental Area (VTA) to lateral hypothalamic to periaqueductal gray (PAG), with diffuse mesolimbic and mesocortical “extensions.” Nucleus accumbens as crucial basal ganglia processor for emotional “habit” systems and affective learning.	DA (+), glutamate (+), many neuropeptides including opioids, neurotensin, CCK, and many other facilitators
RAGE (affective attack)	medial amygdala to bed nucleus of stria terminalis (BNST) to anterior and ventromedial and perifornical hypothalamic to more dorsal PAG	Substance P (+) (? Ach, glutamate (+) as nonspecific modulators?)
FEAR	central & lateral amygdala to medial hypothalamic to dorsal PAG to nucleus reticularis pontine caudalis	Glutamate (+) and neuropeptides (DBI, CRF, CCK, alpha MSH, NPY)
LUST (sexuality)	BNST and corticomedial amygdala to preoptic and ventromedial hypothalamus to lateral ventral PAG	Sex steroids (+) (T/E), vasopressin, oxytocin
Nurturance/ maternal CARE	Anterior cingulate to bed nucleus of stria terminalis (BNST) to preoptic hypothalamic to VTA to more ventral PAG	Oxytocin (+), prolactin (+), dopamine, opioids
Separation distress/PANIC (social bonding)	Anterior cingulate/anterior thalamus to BNST/ventral septum to midline & dorsomedial thalamus to dorsal preoptic hypothalamic to more dorsal PAG (close to circuits for physical pain)	Opiods (−/+), oxytocin (−/+), prolactin (−/+) CRF (+) for separation distress, ACh (−)
PLAY/ (social joy & affection)	Parafascicular/centromedian thalamus, dorsomedial thalamus, posterior thalamus, projecting to ventral PAG (septum inhibitory re: play)	Opioids (+ in mod. amounts, – in large amounts), ACh (+), cannabinoids (+)

On the likely centrality of mammalian play in the construction of a social brain

Jaak Panksepp’s third area of critical contribution might be around placing the PLAY system more clearly front and center as a critical pro-social system in the mammalian and human brain. There are several indices suggesting that we are perhaps the most socially dependent creature on the planet, including centrally that profound social neglect (infants cared for in terms of basic physiologic needs but otherwise terribly neglected) appears consistently lethal to infants (Spitz, 1945). This syndrome of “hospitalism” has been validated in the context of other similar kinds of orphanages and infant “warehousing” (Ainsworth, 1962). This hidden vulnerability to neglect may be an underappreciated but unavoidable corollary to the developmental costs of having a huge cortex, which creates a long period of attendant and intrinsic helplessness in human infancy. Jaak argued strongly for the importance of play and its critical role in a variety of prosocial operations of the human mind and brain, which is still widely underappreciated. For all the contributions of various branches of psychoanalysis to an affect-centric view of the mind, even psychoanalysis (with some notable exceptions) has neglected this prototype emotional system, sometimes treating play as simply a proxy for aggression and/or dominance under the aegis of its original “dual instinct” theory. There is now so much accumulating evidence that this system is critical in a profound way for humans: it promotes attachment, signals trustworthiness (we trust playful people far more readily and much more immediately than dour, irritable or guarded people), and interacts powerfully in a supportive and facilitating way with LUST and CARE (proto-empathy) systems. Its cognitive extensions in humor ripple through much of our social discourse. In fact, we just can't seem to get enough of it! Despite all the endless mass media, pop psychology chatter, and confusion around what might create the basis for a good relationship, both men and women when surveyed about this question want first and foremost someone who is playful and has a good sense of humor (Chick, Yarnal, & Purrington, 2012). If the PLAY system in the brain is responsible for all forms of laughter, and if spontaneous smiling (as opposed to “social smiling”) is part of the PLAY affect continuum, and “on the way” to laughter so to speak (and where an amplification or strengthening of an initial “pro-smiling” stimulus often leads to outright laughter), this might well be the most underappreciated pro-social neural system in the human brain. As a thought experiment, imagine how deeply life would be “grayed out” and emptied, if all forms of laughter, play and humor were completely lost to us. If spontaneous smiling and laughter are essential social signals which indicate that “things are indeed okay,” if playful interactions are a powerful augmenter of attachment, a predictor of who we are going to like and want to be with, and also potently inhibit the stress axis of the brain, Jaak’s advocacy for the critical importance of this highly prosocial system, at a time when almost no one else in neuroscience was paying much systematic attention to play, playfulness, or humor, looks more and more prescient. On a more personal note, many have noted how delightfully playful and funny Jaak could be – clearly part of his charm. It would be indeed incongruous for a relatively less playful person to have become famous as “the rat tickler”– or to consider the possibility that playfulness in rats could be indexed by basic affective vocalizations which we could not hear without special technology, or to intuit that human play could be illuminated from looking at vocalizations from tickled rats!

An affective neuroscience foundation for clinical psychiatry, psychology and the mental health sciences

I’ve saved the best for last. This area of treatment-oriented implications for Jaak’s affective neuroscience indeed might prove to be, over the long run, the most historically important part of Jaak Panksepp’s scientific legacy. Affective neuroscience as a foundation for the mental health sciences has already been at various points clearly implied in much of the prior discussion and at some points even made explicit, but it’s worth fleshing out much more explicitly Jaak’s thoughts and hopes in this regard – efforts in which I was also quite invested, after 40 years in the mental health side of the medical-industrial complex, struggling with its massive blindspots and biases. It was again testimony to Jaak’s capacity to see patterns in the otherwise apparent “noise” of the scientific landscapes that he intuited that perhaps many if not most psychiatric disorders might reflect dysfunction in a prototype emotional system or systems. Table 2 below summarizes in basic terms this critical possibility which still has yet to be extensively and empirically “cashed out” (as he might say), but where the early returns, particularly in relationship to depression, have been very promising.

Table 2. Postulated relationships between prototype emotional systems, common emotional processes, and major psychiatric disorders (extracted and adapted from Panksepp, 2006).

Prototype affective system	Emergent emotions	Related emotional disorders
SEEKING (+ and −)	Interest	Obsessive compulsive disorder (+)
	Frustration	Paranoid schizophrenia
	Craving	Addictive personalities
	Apathy (−)	Retarded depression (−)
RAGE (− and +)	Anger	Aggression/explosive personality disorder
	Irritability	Depression (with ↑ PANIC)
	Contempt	Personality disorders, NOS (PD)
	Hatred	Psychopathic and other PD (see CARE)
FEAR (−)	Simple anxiety	Generalized anxiety disorders
	Worry	Multiple phobias
	Psychic trauma	PTSD and its variants
PANIC (−)	Separation distress	Panic attacks
	Sadness	Pathological grief/depression
	Guilt/shame	Lowered self esteem, depression
	Shyness	Agoraphobia
	Embarrassment	Social phobias, autism
PLAY (+/−)	Joy and glee	Mania (+)
	Happy playfulness	ADHD
	Irritability	Depression (−)
	Resilience/	Depression, PTSD (−)
	Affective Regulation	Increased stress diathesis (−)
LUST (+ and −)	Erotic feelings	Fetishes
	Jealousy (with PANIC)	Sexual addictions
CARE (+/−)	Nurturance	Dependency disorders (+)
	Love	Autistic aloofness (−)
	Attraction	Attachment disorders
	Empathy	Sociopathy (−)

Rather than any simple correspondence in which a complex psychiatric disorder (clearly a complex emergent phenotype) maps neatly onto an excess or deficit in any one prototype emotional system (which are “endophenotypes” but with currently very poorly mapped individual polymorphic genetic variations), it makes much more sense to view the relationships as multivariate, and where several systems, either pathologically inhibited or “boosted”, might contribute to the emergence of a formal psychiatric condition. For example, depression might emerge from a pathological over activation of separation distress, hypoactivity in several pro-social systems, and a resulting down regulation of SEEKING (discussed in much more detail below).

A useful contrast can be mapped between this more affective neuroscience view of the psychiatric and mental health landscapes (that most psychiatric disorders reflect dysfunction within one or more primary prototype emotional systems) with the current dominant “meme” in psychiatry that a steady “brute-force” stockpiling of molecular correlates, informed by a “mindless-brain” reductionism, would eventually provide a true and best mechanistic purchase on the primary clinical disorders for which psychiatry is struggling to offer effective treatments. This means especially in relationship to depression, but also bipolar disorder, schizophrenia, addiction, and PTSD – arguably the “Big Five” conditions breaking the bank in the increasingly overwhelmed mental health care systems in in the United States, with depression constituting far and away the biggest problem. Depression is now not only far and away the most common mental health/emotional/ psychiatric order in the world, it is now also the leading cause of disability, and the single most expensive disorder in Western societies, when one puts together the cost of treatment, with the cost of lost productivity and disability. The neurotransmitter hypothesis of the 1960s and early 70s – that all our commonplace psychiatric disorders might reflect simple neurotransmitter imbalances in a handful of monoamine systems (some version of basic synaptic excess or deficit) gave rise to the “Psychopharmacology Revolution” in mainstream Western psychiatry – which is, in my judgment anyway, a “revolution” from which psychiatry has not really recovered. Indeed, these concepts may have been more like a boat anchor that has carried psychiatry to a scientific bottom.

More nuanced and subtle psychiatric perspectives might point out – not unfairly – that I’m really just “knocking down a straw man” here, and that most senior psychiatric researchers no longer believe that any psychiatric condition maps neatly on to a handful of amine systems in any simple reductive fashion. That’s true, and it’s also true that every condition, including paradigmatically depression, is likely associated with an enormous number of interlinked and recursive neurochemical signaling shifts, only some of which we have mapped at all. Indeed depression appears associated with changes in virtually every neurochemical signaling system studied, including changes not simply in popular and conventionally appreciated monoamine systems, but in the amino acid transmitters systems of GABA and glutamate, in many, many neuropeptide systems (substance P, CCK, several opioid systems, and oxytocin just to name a few), in multiple hormones (particularly in CRF as the master peptide at the head of the stress axis and in glucocorticoids), in the signaling molecules of the immune system (cytokines), and last but certainly not least, in neurotrophins that appear to have mostly antagonistic relationships with pro-inflammatory signals. These signaling shifts are hugely recursive, meaning unequivocally that there can be no single “leader” or molecule orchestrating the entire cascade or recursion, and instead, the entire system of interlinked neural signaling controls appears to pivot around an axis we called “depression”. Depressions happen as a bunch of shifts appear to take place in parallel, but where increasing stress, increasing pro-inflammatory signals, decreasing mu opioidergic and oxytocinergic tone, increasing dynorphin signals, all leading to declining catecholamine particularly dopamine output, and increasing dysphoria, as these multiple shifts, all linked to social defeat and/or social loss, initiate, and then progressively drive, a deepening motivational shutdown, and eventually, a progressive and profound loss of virtually all rewards. Although the table below appears impressively comprehensive, it is likely a subset of the sum total of relationships organizing depression. Indeed, recent work on ketamine (Zanos et al., 2016) argues that the AMPA arm of the massive glutamate system (as the major ketamine metabolite has been shown to be an AMPA agonist) may be critically linked to both regulation of pro-inflammatory signals as well as regulation of the SEEKING system – relationships not covered in Table 3 from our 2009 position paper.

Table 3. Neurobiological factors forming an interactive depressive matrix?

Depressive factor	Driven by	Producing	Behavioral & symptomatic correlates
Increased CRF, hypercortisolemia, choleocystokinin & reduced BDNF	Multifactorial limbic influences on paraventricular nucleus promoting activation of HPA stress axis	Increased dynorphin, decreased 5-HT, reduced neuroplasticity/ HC atrophy. Intensification of Separation Distress. Disrupted ventral HC feedback on core affective regions?	Dysphoria, sleep & appetite loss. Reduced short-term memory, and other cognitive deficits?
Increased acetylcholine	Reduction of social and other rewards, reduced opioidergic tone, and any other social punishment.	Facilitation of separation distress circuitry and other negative emotions? Effects on other core variables?	Negative affect and excess attention to negativistic perceptions and thoughts?
Decreased mu opioids and oxytocin	Separation distress, other stressors, including physical illness and pain	Disinhibition/release of stress cascades; decreased 5-HT & DA; overdriven NE. Promotion of cytokine generation?	Anhedonia and sadness, reduced positive affect and reduced sense of connection? Suicidality?
Increased dynorphin in accumbens/VTA	Stress cascades	Down regulation of VTA and mesolimbic DA system.	Anhedonia, dysphoria, loss of and motivation
Increased cytokines	Acute but probably not chronic stress, acute reduction of opioidergic tone?	Promotion of stress cascades, decreased serotonergic and increased glutamatergic tone. Impairment of HPA axis negative feedback	Fatigue, malaise and appetitive losses. Increased cognitive disruption. Anhedonia?
Reduced serotonergic tone	Stress, increased corticosteroids, cytokines, decreased mu opioids.	Lowered dopaminergic and increased noradrenergic drive. Less functional segregation among brain systems.	Poor affective regulation? Impulsivity. Obsessive thought, and suicidality.
Diminished catecholaminergic (DA & NE) tone	Constitutional vulnerability, stress and poor reward availability	Reduced “signal-to-noise” processing in all sensory-perceptual and motor/executive systems.	Fatigue, diminished psychic “energy”: appetitive sluggishness, dysphoria. Impaired coordination of cognitive and emotional information processing

In terms of Jaak Panksepp’s prototype emotional systems, depression might thus emerge from an excess of panic PANIC/separation distress activation, conjoined to a notable shortage of PLAY and CARE/proto-empathy, both leading to a down-regulation of SEEKING and a generalized collapse of prosocial response and engagement. This set of issues was exhaustively summarized in our 2009 position paper in this Journal (Watt & Panksepp, 2009), and was partially codified in the “recursion” table above, in terms of a whole family of complex neuromodulatory interactions. But the basic point here is simply to emphasize a notion of neurodynamic causality in the mind/brain that moves us away from “soundbite” simplicity and the seduction of single or primary factor thinking as an adequate paradigm to meaningfully probe real causality in biological systems. Causality in biology, and therefore in the brain/mind, is never meaningfully explained in terms of single factors, but rather is grounded in the concerted operation of multiple factors in both positive and negative feedback – a principle that Jaak very much embraced. This type of more nuanced multifactorial thinking may not sell as well in popular media treatments, or in big Pharma advertising, but it’s much closer to the humbling truth of Animate Nature. When you consider that whatever factors we are able to map in this recursive fashion, they are likely to represent at best a partial subset of the genuine forest of interactive signals that are meaningfully operative in relationship to any conceivable biological phenomena of interest. Jaak was consistent in his warnings about premature certainty, premature closure, and the overarching importance of humility in the face of Nature, as a badly needed check on potential seduction by our most fashionable and current scientific illusions. In relationship to refractory problems like depression and Alzheimer’s disease, one frequently wonders if researchers are making the recurrent mistake of being seduced by a fundamental overconfidence in what we know. We know far less than we think we do, about so many subjects in neuroscience, starting with the confession that we simply don't truly understand how the mind emerges from neural activity, even though we have an impressive set of correlates for that.

In any case, these ideas about the multifactorial nature of biological causality, along with the universality of depression as a response to chronic stress in all mammals studied so far, support our idea that rather than any reductive image of depression as “chemical imbalance,” depression must reflect a evolutionarily conserved response mechanism in the prosocial brain, indeed perhaps one of its most basic cornerstone vulnerabilities, particularly when deprived of social support, subject to chronic social defeat, or chronic separation distress. This evolutionarily conserved response mechanism, disinhibited, and removed from its normal adaptive controls (and perhaps even epigenetically promoted as outlined in Otsuki, Uchida, Hobara, Yamagata, & Watanabe, 2012), forms the basis for our huge and costly societal problem of “depressive illness.” And of course the fact that Western societies are inundated with acute depression and even disabling levels of chronic depressive illness suggests that the social isolation which is a pandemic in technological societies is literally killing us (and also that pro-inflammatory Western lifestyles – particularly our sedentary, sleep deprived, and alien dietary habits – are likely also a factor in both the promotion and risk for depression – see Hidaka (2012)). But this is still a very different perspective on depression than simply labeling it as an “illness caused by a chemical imbalance” – an image that fosters the still fundamentally unproven but widely accepted illusion that psychopharmacology is nearly always your best and first option.

If these statements about depression being related to recursively-linked shifts in many, many neuromodulatory systems are fundamentally true, then what’s wrong with the “chemical imbalance” meme? Well, the problems associated with it are neither simple nor obvious. As we stated in our review of depression (Watt & Panksepp, 2009, p. 10),

the prevailing radical reductionism in mainstream psychiatry still envisions that one can go from molecules and brain details to psychiatric diagnostic categories, with no psychologically meaningful neuroscience of emotions in between, almost as if the psychological properties of the brain and its adaptive mandates are somehow irrelevant to the analysis.

Although the “chemical imbalance” meme has been exposed as a profound oversimplification that is probably more wrong than right, the relative scientific bankruptcy of the original reductive idea contrasts painfully with the staggering hundreds of billions of dollars worth of profits derived from it.

I would also argue that a principal negative consequence driven by this “chemical imbalance” meme has been a slow but progressively severe displacement of psychotherapy and other socially guided and informed treatment interventions for depression within the mental health treatment landscape in the United States (Leichsenring, Steinert, & Hoyer, 2016). We have never had more antidepressant drugs, both on and off patent, at any time in modern medical history, with antidepressant prescribing rising nearly 400% since 1988 (Lee, Paz-Filho, Mastronardi, Licinio, & Wong, 2016), and yet there is every evidence that depression is causing more and not less suffering and disability, and becoming an increasingly refractory problem, despite, or possibly because of, the widespread application of serotonergic, noradrenergic and mixed serotonergic/noradrenergic reuptake inhibition drugs, and the closely associated systematic inattention to social, lifestyle, and environmental variables (Hidaka, 2012). There is additionally evidence from now several meta-analyses (Kirsch et al., 2008) that antidepressants only separate from placebo in moderate to severe depressions, while failing to do so in mild to moderate depressions, not necessarily because they are more effective in more severe depressions but simply because the placebo mechanism with which they are competing (“positive expectancy” – indexing SEEKING activation) is itself falling apart as depression worsens (again, see Kirsch et al., 2008 but also Ioannidis, 2008).

This suggests that we have had a takeover of the mental health system in relationship to depression by a treatment meme (first-line antidepressant usage) that in fact does not have an empirical basis, where benefits have been inflated and risks (particularly promotion of obesity – Lee et al., 2016) minimized. This is more than troubling. These findings and other meta-analytic treatments of this question of antidepressant efficacy (see again Ioannidis, 2008) are in general mostly suppressed and not widely disseminated within mainline psychiatry. Adding to the concern is the unexplored possibility, which Jaak and I talked about many times, as a potentially explosive biomedical scandal yet to be clarified, that antidepressant drug usage may over time actually deteriorate stress resilience, leading to a greater vulnerability to relapse and recurrence of depression in those chronically exposed to serotonergic antidepressants. This is an exceptionally difficult problem to research, and there is still inadequate data – to say nothing of ambivalent motivation within mainline American psychiatry – to explore and answer this critical question. It remains a possibility, yet to be falsified or proven. Even if it turns out to be not true, there is already plenty of reason to be concerned about how completely co-opted the mental health system has been by the commercial incentives emanating from big Pharma.

The proof of any scientific hypothesis or idea as they say is “in the pudding” – one can only judge the scientific value of ideas by what useful and heuristic testable predictions they enable and what novel findings they generate and lead to, and, in relationship to any illness or healthcare challenge, to what interventions with “disease-modifying” impact are thereby potentially created. To my knowledge, few (or no) novel or significantly more effective treatments for depression have been discovered, where their discovery was guided by a simple “chemical imbalance” reductionist meme, in relationship to virtually any single modulatory system. The psychopharmacology landscape of psychiatry has been largely dominated by an initial serendipity followed by big Pharma’s drug design methodology, where commercial efforts are typically aimed at simply cloning previous generations of blockbuster drugs coming off patent (which have been almost entirely serotonergic and noradrenergic promoters of various types). Contrast this with the fact that affective neuroscience perspectives have yielded three novel antidepressant treatments in the relatively short space of 15 years – GLYX-13 (a peptide created from studies of PLAY gene expression), EBS of the medial forebrain bundle (a concept directly emerging from EBS in animals of SEEKING system components), and buprenorphine (guided principally by the separation distress and SEEKING system down regulation hypotheses as outlined in Watt & Panksepp, 2009 and first speculated about in Panksepp, 1998). The most interesting novel treatment for depression in the last 10 years – discovered truly by serendipity—has been ketamine. While its mechanism of action was initially assumed to be NMDA antagonism, this has now been well falsified. Instead, evidence favors the idea that AMPA agonism from a major ketamine metabolite, 2–6 norhydroxyketamine, is responsible for its antidepressant effects (Zanos et al., 2016), although there is also evidence that its central anti-inflammatory effects may also be contributory (Tan, Wang, Chen, Long, & Zou, 2017), and with rapid normalization of SEEKING circuitry perhaps emerging from this unique combination of properties – although this remains an appealing hypothesis still in need of empirical confirmation or falsification. Big Pharma is now eagerly trying to find oral compounds with similar AMPA agonist properties.

And last but not least, a true scientific theory of “instincts” (AKA prototype emotional systems) to update psychoanalytic metapsychology and further the integration of psychodynamic and psychoanalytic ideas with neuropsychiatry and neuropsychology)

This has already been ably covered in this journal by several authors in many ways, perhaps most ambitiously by Solms (2013) but also by Yovell (2008). For example, Yovell’s treatment of romantic love (Yovell, 2008) looked at the updating of drive metapsychology in light of findings from neuroscience. Jaak’s work illuminating families of prototype affective systems could first expand Freud’s original dual instinct theory into two families of “drive” systems – one family clearly concerned with the primary goal of organism defense (FEAR and RAGE systems), while a second family of prototype emotional systems has classically prosocial objectives (maternal CARE/proto-empathy, PANIC/separation distress, LUST, and PLAY/social joy). An unfinished project that Jaak and I talked about tackling at some point would have been to summarize what might be known about the neural and developmental mechanisms whereby the prototype elicitors of these classic affective systems could be extended through learning and cognition, such that FEAR for example could be extended to and activated by threats that were not simply representing physical dangers, in other words, where there might be the threat of damage to the “body of the psyche” (our self image) instead of to the body proper.

A basic conception of the cognitive extension of prototype states and their activators is already visible in Jaak’s now widely appreciated concept of a tripartite level neural system (as seen in the above graphic reproduced various forms in many of his more recent review papers, including this version in Watt & Panksepp, 2016), with primary process affectivity at the bottom, secondary process learning (supported principally in the basal ganglia) that allows extensions of the primary systems simply in space and time, and tertiary process higher cognitive extensions, supported in limbic and heteromodal zones. The tertiary paleocortical and heteromodal cognitive systems also allow for complex blends, modulation, inhibition, and increasing top-down cognitive control over both the activation and regulation of emotion.

But there are many unanswered and important questions about all this. For example, how does a largely subcortical PLAY system, originally and prototypically activated by tickling and rough-and-tumble play, become the neural foundation for a highly cognized system of verbal humor? These are neither easy nor trivial questions, and indeed, they are rarely even asked. These questions go to the heart of the challenge to figure out how classic prototype affective systems might neurodevelopmentally percolate upwards and extend their influence into the furthest reaches of the higher cognitive mind. I have long suspected (originally argued in relationship to the central psychoanalytic problem of transference in Watt, 1986) that this process of “cognitive extension” took place via a modeling of “analogues”, possibly explaining why right hemisphere systems may take the lead in affective appraisal and activation – a basic correlation of hemispheric laterality yet to be adequately explained, but where the comparison of “whole templates” and large sensorimotor gestalts or “snapshots” would be an adaptive way to extend the prototype system. For example, the rapid flipping and inversion of expectancy, as comically instantiated in the unexpected inversion from a position of putative alpha male dominance to being a total klutz (the most universally highly rated humorous scenario in many studies of humor, as seen in classic “banana peel” scenarios – Veatch, 1998) would be a cognitive analogue to the flipping and rapid alternating dominance reversals of physical rough-and-tumble play, and is also consistent with how rough-and-tumble play is terminated when animals cannot take turns and share dominance – where one animal insists too much on being “top dog” (Panksepp, 1998). There is nothing all that funny about total dominance (and instead it typically activates mixtures of envy, efforts at identification, and/or FEAR, and occasionally, LUST), but when one goes from a putative high dominance to no dominance at all (e.g. on one’s back and with one’s feet in the air), we often times find that downright hilarious – as long as no one is truly hurt. Analogical cognition might be the best way to extend the activation of prototype states into more complex cognitive territories. But of course this remains just an appealing but unproven speculation.

In any case, this mostly ignored question of mechanisms creating neurocognitive “extensions” of these classic prototype systems is a rich vein of ore yet to be anything more than minimally “mined out” – and may eventually fully explain how the higher cognitive mind adaptively recruits these many prototype emotional systems to deal with complex analogues to the original prototype stressors/elicitors. Such analogical cognition, and where the right hemisphere clearly has some unique advantages in this kind of processing, may help us to quickly recruit for example FEAR and/or RAGE systems (hopefully modulated and restrained) when a powerful rival suddenly intrudes into our intellectual – but not physical – property). While such neurocognitive extensions of the prototypes and their classic triggers seems intuitively obvious and with good face validity when framed in this fashion, there has been, to date, little real study of or empirical interest in these issues. It’s a deceptive gap in our knowledge of the vast emotion–cognition border.

As another critically relevant example of how important these neurocognitive extensions of prototype states might prove to be, consider the frequently appreciated clinical evidence that intense shame and guilt are often significant triggers for depressive episodes, and, consistent with our separation distress hypothesis of depression, that they are also neurocognitive extensions of separation distress. In shame, we are exposed as defective and lacking in whatever qualities ensure love and respect, while in guilt we must stand outside the door and make penance for having harmed someone of value, diminishing our value. Both are complex cognized emotions “built” from linking the prototype state of separation distress to images of self and other in a dynamic relationship. One might also emphasize that these neurocognitive extensions of a prototype state are the psychodynamically fundamental and constitutive currency of “internalized object relationships,” to use the classic psychoanalytic terminology. These are core subject matters for psychotherapists and psychoanalysts, particularly in any psychoanalytically guided therapy process, where the reactivation of “internalized object relations”, recapitulating early attachment dynamics, is the central focus of treatment, and where their reactivation and then systematic elucidation gives the therapist and the patient critical opportunities for reworking the patient’s dysfunctional affective machinery. These basic ideas about critical processes sitting on the emotion–cognition border are of course not new – indeed, they have been around in the psychoanalytic literature in some fashion for many, many decades (see Fairbairn, 1952; Kernberg, 1976; Watt, 1986), but there is now an opportunity to more fully explore this convergence between these old psychoanalytic ideas and the emotion–cognition border as conceptualized by Jaak Panksepp.

Some closing and more personal thoughts

What is my overall conclusion from this enormous body of work by Jaak Panksepp? It’s very hard of course to judge any great thinker’s scientific legacy while that legacy is still unfolding, but collectively, these contributions have had a pervasive and enduring impact on many aspects of the science of the Mind/Brain, particularly those related to emotion and affect. Given the enormous number of subjects within the Mind/Brain sciences that Jaak addressed over the course of his career, many people might generate many different take-home messages from Panksepp’s work. But as I stated at the beginning, it is virtually impossible for me to separate the personal from the professional in relationship to Jaak and his work. The most enduring aspect of Jaak’s legacy, for me personally, is something rarely made explicit in his work, but which we talked about on quite a number of occasions as being more implicit: a deep reverence for an animate Nature, which may both energize the best science and may be an emotional endpoint of the best science as well. Spirit appears to be matter profoundly organized into truly dizzying levels of complexity – that such wondrous properties as minds with affects emerge from a physical Nature, over a long and still in many ways mysterious evolutionary course, deepens a sense of awe and wonder. As Jaak himself stated many times, affects are deep and profound evolutionary gifts, “voices of the genes” in his phrase, conserved value signals that provide our only real compasses keeping us on track in life, if we can understand, regulate and integrate them. As he might have emphasized, this family of negative and positive affective systems work best when the defensive systems of FEAR and RAGE are tightly constrained, mobilized appropriately in the context of real (and not merely imagined) threats to body and psyche, which are then competently and adaptively resolved. This adaptive problem-solving, energized by the negative/defensive systems, allows the positive pro-social affective systems to quickly reactivate and then flourish in the social field, where they ensure both reproductive success and many rich subjective rewards: one would hope for large helpings of PLAY/social joy, CARE/proto-empathy when we are in trouble or when others close to us are in trouble, and only enough separation distress to make it clear to us who really matters to us and how much. Although I feel a keen sense of loss and an enduring grief at his passing, I am also immensely grateful for having encountered Jaak Panksepp in my life’s travels, and for the great personal and intellectual treasures that have emerged from that relationship.

Disclosure statement

No potential conflict of interest was reported by the author.