The emotional feeling as a combination of two qualia: A neurophilosophical-based emotion theory

Abstract

It is argued that the emotional feeling comprises the following two emotional qualia. (1) A nucleus feeling or primary emotional quale, which is the phenomenological counterpart of the end product of appraisal by the central nervous system. (2) The experience of being urged to emotion-related reflection or secondary emotional quale, which is the phenomenological counterpart of the brain's decision to inhibit pre-programmed emotional behaviour, and to initiate emotion-related reflections. Different brain modules regulate these two qualia, and thus each can be experienced independently of the other. The primary emotional quale is related to activation of the amygdala, it is emotion specific, and neutral with respect to affect. The secondary emotional quale is related to activation of the orbito-prefrontal cortex (O-PFC), and includes affective aspects.

It is argued that emotional behaviour is regulated by the following three neural mechanisms, two of which two are directly related to the two qualia. (1) An evolutionary ancient system (amygdala-system), which comprises the amygdalae and subcortical nuclei, and which activates pre-programmed emotional behaviour. (2) An evolutionary recent system (PFC-system), comprising the prefrontal cortex, which inhibits pre-programmed emotional behaviour, activates emotional reflection, generates and evaluates behavioural alternatives. In contrast to the pre-programmed behaviour, the behavioural alternatives are more likely to serve long-term goals. (3) A default mechanism, which gives rise to default (i.e., “just do something”) behaviour. The first two systems are mutually competitive, while the third mechanism takes over if either the competition between the first two mechanisms, or the decision process of the PFC-system, takes too long. This default mechanism involves the function of the medial-prefrontal cortex (M-PFC).

Acknowledgements

The author is indebted to Conner Dolan, Ton Smolenaars, and Rob de Vries for their careful reading of the manuscript and critical remarks, and especially to Conner Dolan for correcting my English.

Notes

¹Experience refers to consciousness. Emotional experiences refer, therefore, to all contents of consciousness that are somehow related to emotions. Emotional feelings or emotional qualia are a subpopulation of these experiences; namely those that cannot be reduced to other things.

²The question as to whether some emotional feelings are blends of basic emotion feelings (Izard, 1972, Plutchik, 1980, as cited in Frijda 1986, p. 182) is irrelevant to the issue at hand. Even if we assume that some emotional feelings are blends then the composing elements can still be regarded as a combination of irreducible experiences, as feelings “an sich”, i.e., as unanalysable qualia.

³Damasio seemed to have changed his view in his latest book. Regarding emotional feelings, he wrote of “perception of a certain mode of thinking and of thoughts with certain themes” (Damasio, 2003, p. 86). This indicates a change from inducing thoughts to perceiving emotion-related thoughts.

⁴Marañon (1924) and Schachter and Singer (1962) merely demonstrated that people, who are already in an emotional state, and whose humeral feedback is experimentally increased, believe that that their emotional feeling is more intense. But so does false-bio-feedback, which leads people to believe that there is increased feedback in the absence of a real increase in physical responses (Barefoot & Straub, 1974; Valins, 1966).

⁵For those who have problems with analogy, I like to quote physicist Robert Oppenheimer (1956), as cited in Plutchik, 1994, p. 79): “Whether or not we talk of discovery or invention, analogy is inevitable in human thought, because we come to new things in science with what equipment we have, which is how we have learned to think, and above all how we have learned to think about the relatedness of things”.

⁶This sentence could be interpreted as indicating a dualistic point of view, qualia (mental contents) inducing something. It would therefore be more correct to rephrase it as: The brain-state corresponding with the neural decision that the specific stimuli are out there, could lead to other brain-states, which corresponds to the neural decision to group greens with greens and reds with reds. However, such sentences are hard to read, and I have therefore chosen a more comprehensible and shorthand formulation.

⁷It is like when we are looking at a green door. The colour quale is just “green” and nothing more. However, in humans with an intact brain this quale is accompanied by the cognition (which is not part of the quale) that the “green” is an attribute of the door.

⁸This indicates that although the PFC patients cannot experience the urge to emotional reflection anymore, they are, somehow, still able to notice that this experience is missing. One could question whether this is possible. Sacks (1995) described a patient with cerebral achromatopsia (i.e., the loss of the capability to see colours due to a brain lesion). This had great consequences, since the patient was an artist, and after the accident he painted only in various shades of grey. Although his disability overwhelmed him in the period shortly after his accident, it became progressively more difficult for him to think in colours. Later his colour associations started to fade. However, the knowledge that someth ing was missing remained, and he started experimenting by adding single colours (which he could not see) to his black and white paintings. Likewise Ross and Rush (1981) described a patient with a very large lesion covering almost completely the right hemisphere. This resulted in a full flattening of affect. Although this patient showed emotional behaviour, her descriptions of her emotions lacked emotional content. However, emotional content reappeared when she described older emotional experiences (i.e., those predating the lesion). Finally Fricchione and Howanitz (1985) also presented a case study of a patient with a large lesion in the right hemisphere. Although she was incapable of describing her own emotional reactions, she was perfectly able to describe those of others.

⁹I have used the term “prefrontal patients” and various descriptions of subdivisions of the prefrontal cortex as if they were all the same, suggesting that there is a unitary prefrontal syndrome. This is inaccurate (Damasio & Anderson, 1993; Malloy & Duffy, 1994). Nonetheless, at various times, I choose to refer to prefrontal cortex (PFC) in general. This allows me to include the older literature concerning frontal lobotomy, which is so beautifully analysed by Trigg (1970). Since the introduction of the transorbital method by Freeman in 1946 by means of which mainly the orbitofrontal parts of the PFC were isolated from the rest of the brain, this older literature is, thus, mainly about the orbitofrontal part of the PFC (Freeman, 1971; Livingston, 1969). Most authors divide the prefrontal cortex into orbital, medial and dorsolateral subdivisions, suggesting a more or less clear localisation of functions within these PFC subdivisions. However, notwithstanding localisation of specific combinations of functions within the PFC, a strict localisational view of the prefrontal cortex is clearly inadequate (Malloy & Duffy, 1994). This is further complicated by the fact that various authors define the orbital, medial, and dorsolateral subdivisions differently. For instance Kolb and Whishaw (2003) locate the orbital-PFC in Brodmann's areas 11, 12, 13 and 14, medial PFC in areas 25 and 32, and the dorsolateral-PFC in areas 9 and 47. Whereas Fuster (1997) locates these subdivisions in areas 11 and 13; 8, 9, 10, 12, 24, and 32; and 8, 9, 10, and 46, respectively. Furthermore, comparative analysis of differences between pairs of uniovular and biovular twins has shown that there is only a low level of genetic determination in the functions of the frontal lobes (Luria, 1980). Finally, brain lesions are seldom restricted to one particular area. It is therefore difficult to pin down a specific prefrontal function to a specific part of the prefrontal cortex. However, there remain ample reasons to conclude that the orbito-prefrontal cortex is the prerequisite for the secondary quale. First, this idea is in line with the literature about the transorbital frontal lobotomy. Frontal lobotomy was especially successful in patients with strong emotional complaints (Kucharski, 1984). Unsurprisingly, the intervention was meant from the very beginning to induce reductions in emotions (Kolb & Whishaw, 1990). Second, the O-PFC is the region of the frontal lobe, which is associated with more dramatic changes in personality (Kolb & Whishaw, 2003). Lesions in this FPC division result, inter alia, in deficits in empathy (Bechara, 2004). Furthermore, the orbitofrontal cortex has been implicated in the decoding of negative as well as positive reinforcers (Rolls, 2000), and specific O-PFC activations in reaction to emotional stimuli have also been demonstrated (Taylor et al., 2002). Third, the prime function of the O-PFC for emotions is indicated by the fact that it is the only PFC part having direct reciprocal connections with the amygdalae. Fourth, the PFC is defined as that part of the frontal cortex receiving projections from the dorsomedial nucleus of the thalamus. This thalamic structure can be divided into the pars magnocellular and the parvocellular portion. The function for emotions of the pars magnocellularis is suggested by its afferents. In primates, it receives, inter alia, projections from the mesencephalic reticular formation, the amygdala, the peperiform cortex and the inferior temporal cortex (Fuster, 1997). This pars magnocellularis projects to the O-PFC. In contradistinction the parvocellular portion appears relatively free of afferents from other than the PFC itself. Furthermore, the parvocellular portion projects mainly to the cortex of the dorsolateral prefrontal convexity, which it is assumed is involved in cognitions, like planning, short-term memory, or executive functions for short (Fuster, 1997). Finally, one of the main orbito-PFC functions is behavioural inhibition (Fuster, 1997), which is necessary to create time for reflection. (The secondary emotional quale is the phenomenological representation of the inhibition of pre-programmed emotional behaviour the start of emotion-related reflections.) This does not mean that the reflection itself is regulated by the O-PFC. There are ample reasons to assume that this function is fulfilled by other PFC structures, as will be argued later. However, in case of spontaneous emotional reflection, activation of the O-PFC seems necessary in order to activate the other PFC structures.

¹⁰Pre-programmed either by evolution or by learning (conditioning) processes.

¹¹Sifneos (1973) coined the term alexithymia, in order to name a complex of features, referring to sever reductions in both the affective and the cognitive aspects of the emotional experience.

¹²Although the BVAQ is meant to measures alexithymia, given appropriate reversal scoring, it can serve as a measure of emotional experience.

¹³It is not my intention to claim that appraisal takes place only in amygdalae. For instance, Iversen et al. (2000) stress the fact that the “emotion identifying” modules for visual stimuli in the inferior temporal cortex can make the appropriate responses autonomously, and then send this information to the amygdalae. It is further likely that there are other neo-cortical “emotion identifying” modules, for other sensory modalities than visual. The point I want to make here is that the amygdala seems to be the key structure in the appraisal and in the subcortical regulation of emotions, and is therefore the prime candidate for the primary emotional quale. However, this primary emotional quale is just a nonverbal “gut-feeling”. Neocortical structures involved in “emotion identification” could be a source of confirmation of such gut-feelings, thus facilitating the verbalisation of these feelings. Furthermore, although there are good reasons to assume that the appraisal takes place in the amygdala, this does not prove that the primary emotional quale is created by the amygdala. It remains possible that this occurs in (all or some of) the neural structure, to which the amygdala projects, especially since these projections are specific for the various emotions. In other words, the amygdalae are necessary, but may not be sufficient for the production of the primary emotional quale. The same may be true of the orbito-prefrontal cortex in relation to the secondary emotional quale. However, the still unsolved mystery of how the qualia reach consciousness is of no importance to this contribution. The amygdalae, and orbito-prefrontal cortices are the key structures, since they seem to be the nuclei where the brain conclusions are created, which are, on the phenomenological level, represented by the two qualia.

¹⁴The fear-inducing elements of The Shining consist mainly of facial expressions by the actors, and, in my opinion this is also largely true of The Silence of the Lambs. Furthermore, it has been demonstrated that the amygdalae are involved in recognition of emotion facial expressions (Alphonse & Tranel, 2003; Alphonse et al., 1999; Hare et al., 2005), which could explain the reduced fear reaction.

¹⁵This is in line with William (2002) who argued and demonstrated experimentally that animals do not have an episodic memory, do not have a sense of time, are cognitively stuck in time, are unable to engage in “mental time-travelling”, and therefore unable to anticipate future events. The ability to mentally travel in time enables subjects to evaluate the consequences of various behavioural options in advance. Organisms lacking the ability to “mentally time-travel”, cannot evaluate the long-term consequences of various behavioural options, and thus, have no use for the inhibition of pre-programmed behaviour, and, therefore, no need for the secondary emotional qua le.

¹⁶For those who view the qualia as epiphenomena, as I do myself, it seems irrational to consider emotional qualia as products of evolutionary development, since there is no selection pressure thinkable for something that has no effect. However, there is good agreement that the emotional process in animals provides a positive fitness value. It is also clear that the capacity to postpone pre-programmed emotional behaviour in order to create time for reflection on a behavioural option that is more suitable to the long-term goals has an additional fitness value. Selection pressures for these capacities are therefore plausible. We could consider the two emotional qualia epiphenomenal by-products of two different emotional capacities, which have developed at different times in vertebrate evolution.

¹⁷By stating that the two systems are in competition with one another, I do not mean that there is only inhibition between these systems. They also exchange information (LeDoux, 1996). The competition determines which of these systems finally induces the emotional behaviour. The fMRI measurements represent average neural activities over relatively long time periods (i.e., relative to speed of brain processes). Specifically, the results of Phan et al. (2002) do not exclude mutual facilitation in the shorter term.

¹⁸Of course, one could assume that the PFC-emotion-system is just an extension of the amygdala-system, i.e., one emotion-regulating system, rather than two competing systems. However, this would not sit well with the results of Phan et al. (2002). More importantly, the two systems work according to different principles. In essence the amygdala system is a penny in the slot machine. This is clearly seen in lower vertebrates, e.g., in the effects “fear substance”, discussed above. Such emotional stimulus response behaviour is not exceptional. For instance, “Boar Mate”, i.e., artificial male pig pheromones, induces the receptive sow to stand immobile in a receptive posture, even if the boar is manifestly absent. Furthermore, sexual behaviour may be triggered in some animal species by the colour green, regardless of whether the green comes from lush vegetation or a green-painted environment (Crews, 2002). Furthermore, although the actual emotional behaviour in higher vertebrates may be regulated by more than one stimulus, it remains fixed to a high degree, as is indicated by behaviour-sequence studies. For instance, Bermond (1982) demonstrated that the type of aggressive behaviour displayed in rats is directly linked to the stimulus properties of, and the behaviour exhibited by, the opponent. In contradistinction, the PFC-system actively generates new behavioural alternatives and evaluates these alternatives.

¹⁹It is reasonable to assume that the default mechanism will be influenced by the ongoing emotional neural activity elsewhere in the brain, specifically the neural activity related to the stimulation of pre-programmed emotional behaviour, and the neural activity related to the generation of emotional behavioural alternatives. Thus, although the behaviour induced by this system is often irrational in the eyes of the observer, it could, for the one who is executing this behaviour, still be related to the ongoing emotion.

²⁰This term was coined by Gerrit François Makkink in the 1930s (in Dutch, the term is “Oversprong”, literally “jumping over” behaviour). In 1940, Nico Tinbergen introduced it into the field of ethology, where it is now an accepted concept.

²¹It should be realised that this line of thought was not spontaneous, but initiated by the person, who presented the patient with the two options. It therefore differs from the secondary-quale-related line of thoughts, which is missing in this patient.

²²The to the secondary-quale-related reflection is a continuous process till a solution is reached or the default mechanism takes over. It is, therefore, this spontaneous emotion-related reflection, which gives us the experience of a sustained emotional feeling. Further, since the secondary emotional quale is, by its very nature, directed at the future (the selection of behaviour that is more adapted to the long-term goals), it probably underlies the general “experience of endurance”. LeDoux (1996, p. 177) stated about PFC patients: “They are glued to the present and unable to project themselves into the future”. See also Footnote 15.

²³It surprises me that LeDoux added the arousal and bodily responses to the emotional feeling. For instance, on page 292, he stated: “At minimum, it takes a second or two for signals to travel from the brain to the viscera and then for the viscera to respond and for the signals created by these responses to return to the brain. For some systems the delay is longer. Visceral organs are made up of what is called ‘smooth muscle’, which responds much more slowly than striated muscles, and for some hormones (like adrenal steroids) the effects on the brain can require the synthesis of new proteins and can take hours to be achieved”. Cannon (1927) refuted feedback theories, on the grounds that this time delay was too long compared to the time needed for the emotional experience to emerge.

²⁴Panksepp's statement that it is almost impossible to evoke emotional feelings (including pain) by electrically stimulating the human neocortex is correct. In fact, human emotional experiences after electrostimulation of subcortical structures have been documented. For instance, Heath (1964) described a depressive patient in whom an aggressive mood was replaced by an elation mood within 15 seconds after stimulation of the septum, and Mook (1996) described a patient in which electrostimulation of the amygdala resulted in aggressive feelings. However, such data do not refute the idea that the secondary emotional quale is a product of neocortical activation. The structures mentioned are part of the limbic system, and it is known that this system is well connected to the PFC. Specifically, its orbitofrontal part is so well connected that nowadays the orbitofrontal PFC is seen as part of the limbic system itself (DeLong, 2000; Iversen et al., 2000). For these reasons it has to be assumed that electrostimulation of subcortical-limbic structures results also in an activation of the orbitofrontal PFC, which could explain the emotional experience.

²⁵Panksepp (1998, p. 190), like Bermond (2001), also stated that animals do not have emotional behavioural options. If this is true, there is no need for emotional behavioural inhibitions, and thus no need for the secondary emotional quale (see also Footnote 15).

²⁶Emotional qualia are a subpopulation of emotional experiences (see Footnote 1). Since Frijda refers to emotional experiences whereas I refer to emotional qualia it is inevitable that Frijda includes more elements than I do.

²⁷This is the figure on page 454 of Frijda's 1986 book.