Based on work by Edmund Rolls, Leonard Kosiol, Deborah Budding, Jan Lauwereyns, David Zald & Scott Rauch
This synthesis of a series of books by researchers into emotion attempts to summarise how emotion functions in the brain. This is an area that was near to being taboo for much of the twentieth century, leaving much room for catching up. Attention is focused on how the orbitofrontal cortex assigns reward values to representations by other cortices, and how the basal ganglia integrate these values with its own inputs from the other parts of the cortex and the limbic system. Emotion, anticipation of rewards and enjoyment of the same are all based on subjective experience, and the key importance of these factors for behaviour suggests that subjective emotion is a common neural currency underlying the determination of behaviour. It is hard to distinguish a purely algorithmic basis for this processing, since the weighing of two subjective experiences seems to require the injection of initially arbitrary weights suggesting a non-computable or non-algorithmic element. This research points to the inadequacy of the Libet-based anti-freewill orthodoxy. As far as consciousness goes, it may tell us more about the function of consciousness, and how to distinguish between conscious and non-conscious entities, than it does about the hard problem of how consciousness actually arises. A curiosity of the books followed here is the lack of reference to the gamma synchrony, despite the growing weight of evidence from recent studies that this is correlated to conscious interactions between different parts of the brain. The evidence here is consistent with the concept that experience/qualia arises at a fundamental level and acts on behaviour through brain regions such as the orbitofrontal.
Modern descriptions of emotional processing in the brain revolve round a framework of ‘rewards’ and ‘punishers’, together referred to as ‘reinforcers’, with subjects working to gain rewards and to avoid punishers. Reinforcers are divided into primary reinforcers such as pain, and secondary reinforcers, where initially neutral stimuli come to be associated with pleasant or unpleasant experiences. The orbitofrontal, the amygdala, the cingulate cortex and the basal ganglia are all brain areas important in emotional processing.
Adaptive advantage: The adaptive advantage of emotions is that responses to situations do not have to be pre-specified by the genes, but can be learned from experience. If evolution had attempted to specify fixed responses for every possible stimuli there would have been an unmanageable explosion of programmes. The reinforcer defines a particular goal, but does not specify any particular action. This can be contrasted with existing robotic systems that do not have a goal, and are blind to their purpose.
Two-stage brain: The brain is envisaged as functioning in two stages. In the first stage, it produces a representation of external objects in the inferior temporal cortex, and similar representations in the auditory, somatosensory and other cortices. These representations are, however, neutral in terms of reward value. Thus visual representations in the inferior temporal or touch representations in the somatosensory cortex are shown to be neutral in terms of reward value until they have been projected to the amygdala and the orbitofrontal.
The orbitofrontal: The orbitofrontal cortex receives input from the visual, auditory, somatosensory and other association cortex, allowing it to sample the entire sensory range, and to integrate this into an assessment of reward values. In the orbitofrontal some neurons are specialised in dealing with primary reinforcers such as pain, while others are specialised in dealing with secondary reinforcers. Orbitofrontal neurons can reflect relative preferences for different stimuli. The subjective experience of one signal can be altered by another from a different modality. The impact of words can influence the subjective impression of an odour, and colours can also influence the perception of odour. Some inputs to the orbitofrontal arrive via the amygdala. There is seen to be a triangular system involving association cortex, amygdala and orbitofrontal.
Changes in reward value/prediction errors/memory: The orbitofrontal is particularly important where the reward value of a stimuli changes. The orbitofrontal is quicker to change its reward assessment than the amygdala. Patients who have suffered damage to the orbitofrontal have difficulty in establishing new and more appropriate preferences, and in daily life they tend to manifest socially inappropriate behaviour. The orbitofrontal is seen as at least one of the brain areas dealing with prediction error, that is comparing expected reward with actual reward, and adjusting future responses if there is a discrepancy. The orbitofrontal also signals the emotional significance of events to the hippocampal region, and this is important in establishing new memories. Another triangle is formed by the orbitofrontal, the amygdala and the thalamus is involved in laying down memories.
Correlation to subjective experience: The orbitofrontal is thought to encode the relative value of rewards. It responds more strongly to sensory inputs related to reward potential than to neutral stimuli. Thus it responds more to the pressure of velvet than the pressure of wood. Studies show that the level of orbitofrontal activity correlates to the subjective pleasure of the sensation rather than the strength of the signal being received. Activation in response to taste is seen to be in proportion to the subjective pleasantness of the taste, and in responding to faces, activity increases in line with the subjective attractiveness of the face.
Thus some aspects of orbitofrontal processing could be argued to only make sense in terms of qualia rather than any form of one-to-one relationship with physical objects. Responses are not directly linked to a physical sensation, but are proportionate to subjective preferences for sensations. If there is a choice of carrots or apples, carrots might be preferred and the top preference signal in the brain would correlate to carrots. However, if the range of choice was subsequently expanded to include bananas, the top preference signal could switch to bananas. This reaction looks to require some form of preferred qualia, referring to a previous tasting of bananas.
Common neural currency: The orbitofrontal can seen as the area that creates a common neural currency that can weigh up differing rewards such carrots, apples and bananas, or even rewards that have no common characteristics at all such as food, money and sex. It becomes hard to see how this could be achieved without the subjective nature of qualia providing a scenario in which the alternatives can be envisaged. P Visceral responses and emotions: The orbitofrontal and amygdala act on the autonomic and the endocrine systems when stimuli appear to have significance in terms of emotion or danger. Visceral responses as a result of this signalling are fed back to the brain. Studies suggest that visceral responses are integrated into goal-directed behaviour via the ventromedial prefrontal cortex (VMPFC). The insula and the orbitofrontal are also thought likely to map visceral responses, with feedback from the viscera influencing reward assessment via levels of comfort or discomfort.
There is considerable support for the idea that the body is the basis of all emotion. However, this looks difficult to square with the actual structure and nature of brain processing. While the bodily responses can certainly be seen to play a role, it is hard to see why all visual, auditory inputs, and the results of cognitive processing should have to wait on the laborious responses of the viscera, especially as it is the reward assessment areas of the brain that signal the viscera in the first place. If bodily emotion were the whole story, the orbitofrontal and amygdala would seem to be in a state of suspended activity between sending a signal to the autonomic system and getting signals back from the viscera. In the specific case of rapid phobic reactions in the amygdala, the idea seems to fail completely. A further objection to this theory is that bodily arousal does not provide a sufficient range to match the range of emotional responses. Emotional research, which often means animal research has tend to focus on the easy target of fear, which produces very definite bodily responses, whereas cognitive processing or visual and auditory sensations not related to immediate danger may produce a much less marked bodily response.
The more plausible view is that visceral responses are one aspect of many responses that are integrated in the orbitofrontal. Further to this, evolution seems to have altered the response system to visceral inputs when it came to primates. The visceral inputs no longer go via the pons structure in the brain stem, and this is argued to suggest a less automatic response to visceral inputs in primates including humans. It seems more likely that in line with most brain processes there is a complex feed forward and feedback between all parts of the system including the viscera and the orbitofrontal. The body-only theory seems to depend on a simple feed forward mechanism, which is alien to how brain processing appears to work.
Basal ganglia: The dorsolateral, orbitofrontal, amygdala, anterior cingulate, hippocampus and most areas of the cortex all project to the basal ganglia. Three groups of neurons in the orbitofrontal are involved in changing reward responses, and these also project to the basal ganglia. The orbitofrontal and anterior cingulate both project to the nucleus accumbens, which is the brain’s reward centre, and forms part of the basal ganglia. The suggestion appears to be that the assessment of reward values arises in the orbitofrontal, while the basal ganglia have to integrate these reward values with inputs from most parts of the cortex, the amygdala and the anterior cingulate before releasing or inhibiting behaviour.
The ventral striatum, which includes the nucleus accumbens (reward centre), is the largest part of the basal ganglia. This area is seen as having the function of integrating sensory input with motivational goals. Some regions of the cortex and particularly the orbitofrontal project to parts of the striatum known as striosomes. The amygdala also projects to the striosomes, and this is seen as constituting a limbic-basal ganglia circuit within the brain. The basal ganglia appear to integrate the reward assessment of the orbitofrontal, amygdala and other limbic areas with its reading of the environment, and on that basis it releases or inhibits the subjects behaviour. The basal ganglia are not thought to take part in cognitive computation as such, but instead act as a sort of mixer tap for the wide spread of inputs from the cortex and limbic system, and as such select or gate for material processed by the cortex, including the orbitofrontal. The basal ganglia have set responses to certain established stimuli, but with more novel stimuli or combinations of them, there has to be interaction between the cortex and the basal ganglia. The activity of the basal ganglia is seen as being related to, learning and working memory, because it reinforces successful responses, and prevents stimuli from distracting from the limited number of things that can be held in working memory at any one time.
The nucleus accumbens is part of the ventral striatum and constitutes the reward/pleasure centre of the brain. Dopamine-based activity in the nuclear accumbens is related to seeking reward and avoiding pain. Addictions are found to be related to a lack of natural activity in this area, with drugs of addiction working to enhance otherwise depressed activity.
The basal ganglia are particularly influenced by the neuromodulator, dopamine, and they receive excitatory inputs from most parts of the cortex, and feedback mainly inhibitory signals. Dopamine appears to play a part, both in the release/inhibition process, and the delivery of the subjective reward. The latter occurs in the nucleus accumbens. There is thus a system of loops between the basal ganglia and the cortex, including the orbitofrontal and the anterior cingulate.
The largest concentrations of dopamine in the brain are found in the basal ganglia, the amygdala and the prefrontal regions, particularly the orbitofrontal. The ventral striatum region of the basal ganglia, of which the nucleus accumbens is part, is highly active in anticipation of reward and also during reward. Dopamine acting on spiny neurons in the ventral striatum reduces inhibition, and releases the output of behaviour, while a reduced level of dopamine is inhibitory and reduces activity.
It has further been suggested that the use of neuromodulators by-passes the need to always rely on cognitive computation in the cortex. From the point of view of consciousness studies, it is apparent that these dopamine-rewards are registered in subjective consciousness, so one is effectively looking at a weighting of different subjective impulses. The dopamine producing neurons appeared to be influenced by the size and probability of rewards presumably based on information from areas such as the orbitofrontal and the amygdala.
Free won’t: An area of the basal ganglia known as the subthalamic nucleus (STN) is important from the point of view of the free will debate. Benjamin Libet, whose experiments indicated that some ‘voluntary’ movements were initiated before subjects were consciously aware of wishing to move, postulated that there could be a ‘free won’t‘ mechanism that blocked actions that began unconsciously, but were determined to be undesirable by the conscious mind. Libet did not have any ideas for a detailed mechanism, and the anti-freewill orthodoxy has not been keen to look for one. However, more recent studies show that the subthalamic nucleus has an inhibitory role in stopping behaviours whose execution has already begun. The dopamine/reward related nature of basal ganglia processing suggests that there is a subjective experience aspect to this ‘free won’t‘ mechanism.
Dorsolateral prefrontal: The orbitofrontal projects not only to the basal ganglia but also to the dorsolateral prefrontal, which is responsible for longer-term planning, and such decisions as deferring a short-term reward in favour of a larger longer-term reward. The connection with long-term planning is particularly important with respect to the free will debate. The modern orthodoxy as to the non-existence of free will relies heavily on Libet and similar experiments showing that the initiation of very trivial actions does not rely on conscious agency. However, here it is apparent that the more important longer-term planning is linked to subjectively experienced emotional processing in these brain areas.
The dorsolateral prefrontal is involved with attention, working memory, planning and executive functions. Where dorsolateral activity reflects preferences, it is found that the orbitofrontal has reflected them first, and these preferences have been projected from the orbitofrontal to the dorsolateral, where they can be utilised for planning or for deciding whether or not to defer short-term rewards. Discussions relative to free will are often overly simplistic, and do not even consider the situation where two stimuli conflict, a particular example being the conflict between short-term reward as against enjoyment deferred in order to receive a larger reward in the longer-term. In these instances the reward assessing functions of the orbitofrontal and the integrative role of the basal ganglia, both of which are linked to subjective experience, play a decisive role. It is argued that ethically based rewards for good or appropriate behaviour that are decided on by the dorsolateral have their basis in the processing of the orbitofrontal. Also relevant to the free will debate is the observation that increased activity in the dorsolateral prefrontal, which is related to planning and executive functions, correlates with attempts to overcome obsessive compulsive (OCD) activity rooted in the orbitofrontal and basal ganglia.
Tags: David Zald, Deborah Budding, Edmund Rools, emotional brain, Jan Lauwereyns, Leonard Kosiol, Scott Rauch Posted by