At the interface of the affective, behavioral, and cognitive neurosciences: Decoding the emotional feelings of the brain

Abstract

This article summarizes recent conceptual and empirical advances in understanding basic affective processes of hte mammalian brain and how we might distinguish affective from cognitive processes. Six reasons are advanced for distinguishing the two types of consciousness, including (i) the presence of experienced valence, (ii) cortical sub-cortical locus of control, (iii) different developmental trajectories, (iv) informational vs organic considerations, (v) differences in bodily expressions, (vi) differences in cerebral laterality. The position is advanced that to make progress on understanding the neurobiological nature of affect, we need to utilize experimental strategies different from those that are common in cognitive science.

Introduction

It is generally accepted that emotional processes have many attributes including motor-expressive, sensory-perceptual, autonomic-hormonal, cognitive-attentional, and affective-feeling aspects. A general definition of emotion should include all these characteristics, phrased partly in neuroconceptual terms (Panksepp, 1982, Panksepp, 1992, Panksepp, 1993). If one were to ask non-scientists which of these attributes is most important, one would find the last two highest among most lists, with cognitive typically being first among those committed to intellectual views and affective being foremost among the more emotionally inclined (Panksepp, 1999a, Panksepp, 2000a).

The folk-psychological distinctions between these aspects of mind seem obvious to most, but it all begins to “flicker” when one tries to distinguish the two unambiguously in the laboratory; in most human experiences, they tend to go together. Hence, within the context of the cognitive revolution, many have begun to question the utility of the classic distinction. However, I would encourage us not to discard it, for this very distinction may help us unravel the neurobiological nature of the basic affective coloring of conscious existence, and thereby allow us the swiftest progress in elucidating the fundamental nature of those prepropositional experiences we share with many other animals. Thus, I will defend what has recently become a minority view among psychologists, but which, in my estimation, should remain accepted wisdom. At the same time, I must emphasize that “this view in no way seeks to deny their remarkable blending in our first person subjective experiences, nor the fact that cognitive abilities have co-evolved with affective processes in many higher regions of the brain” (Panksepp, 2000b, p. 29).

Regardless of one’s position on this contentious issue, in practical scientific terms our key concern should be whether the cognitive–affective distinction represents some real aspects of neuromental existence or whether it is a fictional parsing of neuropsychological space. I believe that focussed consideration of the basic affects, quite independently of the cognitive activities with which they always interact in the intact brain–mind, may promote a deeper and more substantive understanding of the feeling aspects of emotional processes than if we perpetually conflate the two. Let me give a paradigmatic example—many drugs can reduce feeding in humans and other animals, but only a few of them do so by simulating the good feeling of normal satiety. For practical clinical control of human weight and appetite problems, we must winnow those specific factors from the greater number that make animals sick, dysphoric or otherwise indisposed to eat. The pleasure of sensation strikes me as something that is fundamentally non-cognitive (also see Berridge’s contribution in this issue), even though we must, of course, talk about such entities in cognitive terms or we regress to communicating in various grunts, groans, and sighs. Still, we should not deny that hungers, thirst, and many other internal urges and feelings emerge from the ancient regions of all mammalian brains.

I would encourage us to make the evolutionary working-assumption that internally experienced affects are universal capacities of brains in all mammals, and that those seemingly intangible neurodynamic processes can, in fact, be elucidated by triangulating among (i) sensitive behavioral measures, and (ii) our understanding of their brain substrates in animals, combined judiciously with (iii) the study of related affective experiences and psychophysiological changes in humans (Panksepp, 1992, Panksepp, 1998a). It should be emphasized that without the third component, this type of approach could easily be dismissed as meaningless (as it commonly was during the behaviorist era). This strategy, even though it generates abundant new predictions for human research with direct implications for quality of life issues, has yet to be widely employed since many would claim that the experience of affective states is confined to human beings. In this context, it is important to emphasize that independently of that contentious issue, the strategy should still work even if animals have only a smidgen of the neural architecture that is essential for generating such internal experiences.

Now that the cognitive revolution is gradually giving way to an emotion revolution, investigators are gradually exhibiting a new taste for the pursuit of what was once deemed scientifically unpursuable—an understanding of what affective processes really are, even in non-human animals. The new brain imaging tools have been a major force in this transformation. Equally important has been the recognition that at a genetic and subcortical-brain organizational level, all mammals are strikingly similar. Thus, when we encounter homologies in organs, including mammalian brains, we may be able to reveal useful general principles for the whole class of animals by devoting a great deal of research effort to a single convenient species. Although this is not to deny the abundant species differences that exist in the details of all systems, but to recognize that those differences are bound to be much greater at cognitive (Hauser, 2000) than basic affective levels of brain organization (Panksepp, 1998a). Within such an intellectual framework, emotional systems of the brain can be studied in any representative species as long as they have not become vestigial—as separation–distress circuitry appear to have become in laboratory rats and fear in species who have no natural predators. There is now a growing recognition that while human neuroimaging can highlight many brain zones where we should focus our attentions, it will be through animal brain research that the underlying operating principles of the relevant neural systems must be revealed. Because of recent advances in neuroscience, few would be willing to decree that affective processes are meaningless concepts, as many neurobehaviorists still did just a few years ago.

Only through the integration of human and animal approaches can deep knowledge in this field be achieved. If we conflate emotions and cognitions too much in our thinking (even though, I must repeat, they may be quite impossible to separate unambiguously in most human inquiries), we may retard a fundamental understanding of affective processes and thereby, a real understanding of how cognitive activities are modified by emotional states. Many now agree that many cognitive processes are coded by mood congruent principles, and specific kinds of “affect logic” prevail in cognitive deliberations (Frank, 1988; Shand, 1920; Wimmer & Ciompi, 1995). I would submit that the brain substrates for affect are most easily decoded in animal models even though all conclusions must be provisional until validated in humans. At times I do fear that cognitive-imperialism, the prevailing view in mind sciences, will continue to suffocate the need for focused research on affective issues, and thereby, continue to delay a scientific analysis of such matters of foremost concern for understanding the existential inner qualities of human lives.

For instance, within the information-processing models of mind that the cognitive revolution continues to advance, there is an increasing, and in my estimation unjustified, tendency to assume that information-only strategies constitute an optimal methodology for understanding emotions. After all, it is tempting to focus on the finding that neurons do process “information” in remarkably similar ways regardless of their functions in the brain. Despite the fact that nothing resembling a Rosetta Stone has yet been found (Rieke, Warland, de Ruyter van Steveninck, & Bialek, 1997), abundant hope persists that digital neuronal codes will be found for psychological processes as instantiated in the temporal flow of action potentials. Since action potentials do constitute a universal neuronal “language,” many investigators still assume that the cognitive and affective processes of the brain are little more than variants on similar neurocomputational themes. According to such views, emerging conceptual frameworks such as “affective neuroscience” may seem redundant, unnecessary, and at times threatening to the hegemony of the increasingly popular cognitive-computational views of brain–mind functions (LeDoux, 1999, LeDoux, 2000).

Having coined the concept of “affective neuroscience” a decade ago (Panksepp, 1990a, Panksepp, 1991, Panksepp, 1992; also see Davidson, Jackson, & Kalin, 2000; Davidson & Sutton, 1995), followed by several theoretical syntheses (Panksepp, 1996, Panksepp, 1998a, Panksepp, 1998b, Panksepp, 2000a, Panksepp, 2000b), I would like to discuss why our willingness to distinguish affective and cognitive processes may advance incisive work in this long-neglected area of neuroscience. My main point is that affective feelings are, to a substantial degree, distinct neurobiological processes in terms of anatomical, neurochemical, and various functional criteria, including peripheral bodily interactions. Emotional and motivational feelings are unique experientially valenced “state spaces” that help organisms make cognitive choices—e.g., to find food when hungry, water when thirsty, warmth when cold, and companionship when lonely or lusty. If affective organic processes, ancient adaptive solutions that they are in brain evolution, are to a substantial degree distinct from those that mediate cognitive deliberation (even though they obviously co-evolved in recent brain evolution), then we must develop special strategies to understand them in neural terms. Biological solutions to such problems may promote the emergence of a solid foundation for the construction of a coherent mind science as well as providing a substantive grounding for psychiatric therapeutics, both pharmacological and psychological. It may also encourage investigators to better characterize how cognitive appraisals are modified by such noteworthy urges such as hunger, thirst, etc., as well as the more transient emotional storms.

In my estimation, affective/emotional processes provide intrinsic values—organic “pressures” and “drives”—for the guidance of behavior. I believe such “energy” metaphors were prematurely discarded in psychology with the advent of digital computers and the information-processing revolution. To the best of our knowledge, the ancient analog processes that constitute the core of our emotional and motivational processes emerged largely from evolutionarily prepared “instinctual” action-generating systems as well as from homeostatic, visceral-type interoreceptors, situated largely subcortically (Panksepp, 1998a). These robust but slowly firing systems help generate “intentions in action,” to use Searle’s (1983) felicitous phrase, and in so doing may generate raw affective experiences without any need to interact with higher cognitive mechanisms. Cognitive processes, on the other hand, are linked closely to rapidly firing exteroceptive sensory systems, comparatively free of any intrinsic affects, which allow organisms to navigate effectively in space, time, and among the object of the world (often toward affective goals), yielding, in some species, “intentions to act” (Heyes & Dickinson, 1990).

In line with long-standing traditions, the way we perceive the external world and our resulting propositional thoughts about those perceptions are what constitute our cognitive terrain. Those functions emerge largely from higher, more recently evolved, neocortical regions of the brain. They are linked closely to what has traditionally been called the “somatic nervous system” concentrated in the thalamic–neocortical axis—the rapidly firing neuronal systems that interface organisms with the outside world. Its highest manifestation in our species is the capacity to use symbols and metaphor (Lakoff & Johnson, 1999) and to create “the prison-house of language” that is not well designed for cogent scientific discourse about the basic neuropsychological substrates for affect.

Affects reflect our internal feelings of goodness and badness, in the many varieties that those evolutionarily honed neurodynamic arise, typically through organismic interactions with the outside world. There are reasons to believe that affective feelings emerge largely from specific subcortical circuits where slowly firing neural systems abound, rich in various function-specific neuropeptides that are also abundant in the enteric nervous system of the viscera (Panksepp, 1993). Emotional responses, including their intrinsic affective attributes, probably emerge from “limbic” regions that are more evolutionarily conserved in vertebrates than those that mediate cognitive capacities (MacLean, 1990). In my estimation, the increasingly prevalent limbic-system bashing among emotion researchers reflects a misreading of the history of the field and the role of general concepts in promoting research and communication (e.g., LeDoux, 2000).

Of course, there is now extensive blending of affective and cognitive processes in many brain areas, and if conceptual distinction cannot be cashed out scientifically, they are bound to be counterproductive. I believe it has already been cashed out in the abundant number of new predictions concerning human affects that have arisen from animal brain research (Panksepp, 1998a). Still, there is a massive reluctance among most neuroscientists to utilize such information since many prefer to deny that the animals they study experience anything. That, I believe, is a hangover of Cartesian dualism along with the prevailing assumption that subjective brain–mind issues, since they cannot be directly measured, should not be deemed a topic of disciplined scientific discourse or inquiry. That is a rather counterproductive and narrow-minded view if neuroaffective processes do exist in animal brains and have causal efficacy in the long-term regulation of their behaviors. In fact, these subtle issues can now be empirically approached through cross-species theories that make explicit behavioral and brain predictions (e.g., Damasio, 1999; Panksepp, 1998a).