Many creative people have related that after discovering an unresolved problem, they are unable to immediately to solve this problem. After multiple attempts, they might become frustrated, give up trying to solve the problem, and then move on to other problems. Sometimes, after a few days, weeks, or months, suddenly the solution of this problem might come to them. As I mentioned, this experience has been termed "illumination," or the "aha" experience, by Wallas (1926). The ability to suddenly understand the solution of a problem suggests that the brain has been actively manipulating stored knowledge. Wallas called this process of subconscious knowledge manipulation "incubation." In his book Origins of Genius, Simonton (1999) quoted the famous French mathematician Poincare. As I mentioned in the relaxation section, Poincare wrote how he could not solve a mathematical problem and got his mind off this problem and then came up with the solution (i.e., "I went to spend a few days at the seaside and thought of something else. One morning walking on the bluff, the idea came to me."). On reflecting on this aha experience, Poincare posited that his sudden inspiration was "a manifest sign of long unconscious work."
During the latter part of the 19th century Freud wrote much about how unconscious activities influence our thoughts and actions. In the mid-20th century many academic psychology departments were immersed in Skinnerian behaviorism, and during this time, the construct that unconscious mental activity had an influence on human behavior was disregarded. In the past 20 or 30 years, however, neurop-sychologists have provided convincing evidence that the brain can mediate cognitive activity while the person who is performing this activity might not be consciously aware that this activity is taking place. For example, Bauer (1984) studied one of my patients who had prosopagnosia (failure to recognize the faces of people who were previously known to this man) from bilateral traumatic hematomas (blood clots) of the ventral (bottom) temporal and occipital lobes. These hematomas impaired the visual "what" system I previously described. In his hand, Bauer recorded the electrical resistance which changes when the hand sweats and the hand sweats when someone becomes alerted or aroused. Bauer (1984) showed pictures of some famous people to the patient and asked him either to name this person or to describe what made this person famous. This man was unable to recognize any of these faces, but when he was asked relevant questions about these faces, as opposed to irrelevant questions, this patient had an electrodermal response suggesting that he had unconscious or covert knowledge of these faces. For example, when shown a face of Richard Nixon and asked if this man was a famous athlete, the patient said that he did not recognize the face, and no electrodermal response was recorded. In contrast, when asked if this picture was former President Nixon, the patient again said that he did not recognize the face, but this time he did show an electrodermal response, suggesting that somewhere in his brain he had maintained a representation or image of Nixon's face. Although his visual system was able to access this representation, this stored face knowledge did not reach his conscious awareness.
Another example of unawareness in the presence of knowledge is a study performed by Marshall and Halligan (1988). There is a disorder called hemispatial neglect that causes patients to be unaware of stimuli presented in the half of space that is contralateral to a hemispheric injury (e.g., right hemisphere) (Heilman, Watson, & Valenstein, 2003). These investigators showed the participants who had left-sided unawareness a sheet of paper that contained pictures of two houses, one drawn above the other. The houses were identical except that the left side of one house was on fire. When participants with hemispatial neglect were asked in which house they would rather live, they said there were no differences between the houses, but when the examiner asked them still to point to the house in which they would rather live, they usually pointed to the house without flames. These results suggest that although these patients were not consciously aware of the flames coming from the left side of the house, some part of their brain did see the flames and reasoned that a house with flames would not be an ideal place to live.
Neuroscientists do not entirely know why we are aware of some things and unaware of others. However, recently, Meador, Ray, Echauz, Loring, and Vachtsevanos (2002) studied patients who were undergoing neurosurgery for epilepsy and thus had their cerebral cortex exposed and accessible. It had been posited that high-frequency (e.g., gamma of 30 to 50 Hz, as measured on the EEG) coherent neural activity integrates processing across distributed neuronal networks to achieve a unified conscious experience. Thus, to test this hypothesis, digital intracranial electrocorticographic recordings from implanted electrodes were obtained in six patients who were undergoing surgery for medically intractable epilepsy. These patients were stimulated in the hand opposite to these recording electrodes with simple near-threshold somatosensory (touch) stimuli. The patients were aware of some of these low-intensity stimuli and unaware of others. For those stimuli that were consciously perceived, there was gamma coherence in the primary somatosensory cortex that occurred approximately 150 to 300 milliseconds after these perceived contralateral hand stimuli were applied. For the stimuli that were not perceived stimuli, this gamma activity was not observed. These results suggest that conscious perception is dependent on coherent rapid cortical electrical activity.
The work of Contreras and Llinas (2001) together with the observations of Meador and coworkers (2002) suggest that although conscious perception of threshold stimuli might be related to high-frequency neural activity, during this high-frequency activity neural networks that are not directly involved in the detection of these stimuli are actively inhibited. Thus, high-frequency activity appears to focus neuronal processing, thereby enhancing the signal-to-noise ratio and facilitating conscious awareness. As I mentioned, finding a creative solution to a problem often depends on the simultaneous activation of widely distributed modular networks that store diverse sets of representations, and high-frequency activity of the cerebral cortex might restrict the distribution of activated networks.
The electrical activity of the cerebral cortex can be influenced by neurotransmitters and neuromodulators. Norepinephrine has been proposed to influence the signal-to-noise ratio in the cerebral cortex. Studies have revealed that norepinephrine decreases spontaneous firing of neurons and augments neuronal activity evoked by an external stimulus (Waterhouse & Woodward, 1980). Such relative enhancement of responses to strong inputs relative to low-level or basal activity has been found in the cerebral cortex and is consistent with recent neural modeling work that posits norepinephrine acts to enhance signal-to-noise ratios in target systems (Servan-Schreiber, Printz, & Cohen, 1990). Hasselmo, Linster, Patil, Ma, and Cekic (1997) also demonstrated that norepinephrine changes the signal-to-noise ratio by suppressing intrinsic excitatory synaptic potentials relative to the potentials elicited by direct afferent input. The bias toward external input that occurs with high norepinephrine states may prevent asking "what if" questions of the networks that store cognitive representations. In addition, connectionist modeling has also suggested that a moderate amount of noise allows networks to settle into optimal solutions. Suppressing intrinsic excitatory potentials may prevent many association neurons that do not receive direct afferent input from achieving a firing threshold. The reduced activity of association neurons may lead to activation of relatively sparse, constricted, and nonoverlapping associative networks. As I mentioned, the activation of highly distributed representations might allow one to perform inference and generalization, processes that are critical to creativity.
Further support for this norepinephrine postulate comes from a study of the role of the locus coeruleus and norepinephrine system in the regulation of cognitive functions. Locus coeruleus-norepinephrine neurons in the brain stem give rise to an extensive set of projections to the cerebral cortex, limbic system, and thalamus. The locus coeruleus is a small brain stem nucleus, but probably enervates a greater variety of brain areas than any other single nucleus. Many of the strongest projections of the locus coeruleus-norepinephrine system are to the areas of the brain that are most important in attentional processing, such as the inferior parietal lobes (Morrison & Foote, 1986). Investigators have concluded that high levels of tonic locus coeruleus activity that induces increased levels of norepinephrine in the cortex favored "bottom-up" processing, which is important for sampling new stimuli and increasing behavioral responsiveness to unexpected or novel stimuli (Aston-Jones, Chiang, & Alexinsky, 1991). Although low levels of locus coeruleus activity have not been tested, we suggest they might be associated with "top-down" processing, which is critical for the innovation stage of creativity.
Aston-Jones and coworkers (1991) studied monkeys performing a visual discrimination task that requires focused attention. Their results suggest that locus coeruleus-norepinephrine neurons exhibit phasic or tonic modes of activity that closely correspond to good or poor performance on this task, respectively. On the basis of their studies, they developed a model that predicts alterations in electronic coupling among locus coeruleus cells that may produce the different modes of activity. According to this model there are two modes of locus coer-uleus firing: The phasic mode of locus coeruleus activity may promote focused or selective attention, and the tonic mode may produce a state of high behavioral flexibility or scanning attentiveness. According to our hypotheses it would be this latter mode that would be important in creative innovation.
To directly test the hypothesis that norepinephrine modulates cognitive flexibility, we performed a study in our laboratories where we (Beversdorf, Hughes, Steinberg, Lewis, & Heilman, 1999) tested normal participants' ability to solve problems when treated with placebo, ephedrine, or propranolol. Ephedrine increases the levels of norepinephrine and propranolol a beta-noradrenergic blocker interferes with norepinephrine's influence on the brain. We used a test that relies heavily on cognitive flexibility: solving anagrams. In this task, normal participants are presented with a series of words in which the letter order has been scrambled and their task in each trial is to recognize the word that uses these letters. We found that this anagram task was performed better after participants took propranolol than after they took ephedrine. To learn if the increase in cognitive flexibility induced by propranolol was produced by central or peripheral nervous system blockade, Broome, Cheever, Hughes, and Beversdorf (2000) tested another group of normal participants' performance on an anagram task and compared the effects of propranolol, which enters the central nervous system, with nadolol, a purely peripheral adrenergic beta-blocker. Participants' solution times on the anagram test were more rapid with propranolol than with nadolol, suggesting that it is primarily the central nervous system beta-adrenergic blockade (i.e., blocking the ability of norepinephrine from influencing neuronal networks) that increases cognitive flexibility. These studies of the beta-blocker propranolol provide evidence that the process of finding innovative solutions (finding the thread that unites) is enhanced by reducing the effect of norepinephrine on the brain.
Foote, Berridge, Adams, and Pineda (1991) reviewed the evidence that the locus coeruleus, by way of its massively divergent efferent noradrenergic projections to the cerebral cortex, participates in arousal and can convert the EEG from nonalert to alert or aroused states. Further support for the adverse influence of catecholamine-induced arousal on creative innovation comes from EEG studies. Martindale and Hasenfus (1978) performed a series of experiments to investigate the relationship between arousal and creativity. To measure arousal these investigators used the electroencephalogram. Since the work of Berger, it has been known that with higher levels of arousal there is faster electrical brain activity as recorded by the EEG. In contrast, with relaxed wakefulness there is the slower, well-developed 8-to-12 cycles-per-second alpha activity. On the basis of participants' ability to write creative stories, Martindale and Hasenfus (1978) placed participants into either the creative group or the uncreative group. In the resting state there were no differences in the EEGs of these two groups, but during the time they were developing their stories (innovation stage) the creative participants demonstrated better-developed alpha activity than did the uncreative participants, who had more rapid activity, suggesting that the creative participants were operating at a lower level of arousal than were the less creative participants.
Unlike the many creative people who have mood disorders, people with autism have been shown to have a severe deficit of cognitive flexibility and impoverished creativity (Craig & Baron-Cohen, 1999). Studies of autism such as those that use "theory of mind" suggest that autism is associated with a deficit in top-down processing (Happe & Frith, 1996). For example, to test theory of mind, investigators might show relatively high-functioning people with autism a series of pictures or a video. In one such picture there are two women (a mother and her friend) walking, with the mother pushing a baby carriage that has an infant inside. This carriage has a hood so that a person has to look inside to see if the baby is in the carriage. They stop walking and the mother's friend goes inside her house to get something. While she is inside her house, the infant's grandmother comes out of another house, takes the infant out of the carriage, and brings the infant into her house. If asked, "When the mother's friend comes out of her house, will she think that the baby is still in the carriage?" normal people would say that without her looking into the carriage she will think the baby is still inside. People with autism, however, are likely to say, "No, the grandmother took the baby inside." Happe and Frith suspected that people with autism, when observing people's behavior, have trouble understanding these people's thought process, a deficit in top-down processing.
We wanted to learn whether people with autistic spectrum disorder have a loss of cognitive flexibility and impoverished creativity because they have a reduction in the breadth of their conceptual networks or decreased ability to activate large and highly distributed semantic-conceptual networks (conceptual constriction). To test this conceptual constriction hypothesis, we used a false-memory paradigm (Schacter,
Verfaellie, & Anes, 1997). When we test normal people's memory by having them recognize previously presented words that come from the same semantic class or overlapping categories (e.g., sweets, M&M's, caramel, Milky Way, Snickers, chocolate), many normal people will have false memories and think they were told to recall closely related words (e.g., candy) (Schacter et al., 1997). According to associative theory, the distributed representations of the concepts of the words to be remembered substantially overlap the distributed representations of concepts that represent the falsely recognized words. When we tested high-functioning autistic patients (Beversdorf et al., 2000) with this false-memory paradigm, these autistic participants recalled fewer false memories and discriminated true memories from false memories better than did the normal control participants. These results suggest that people with autism have constricted their semantic representations (i.e., relatively sparse and nonoverlapping). The reason for this constriction is unknown, but it might be related to a decrease in connectivity in these people's brains, and there is some evidence to support this connectivity hypothesis. For example, a Golgi analysis allows neuroscientists to examine the neuronal connectivity in brains. Upon examining the brains of people who were diagnosed with autism, using this Golgi analysis, Raymond, Bauman, and Kemper (1996) reported that there was a reduction of dendritic arborizations in the hippocampus. Defective dendritic arborizations could impair the formation of associative networks and can account for many of the signs of autism, such as diminished creativity, poor concept acquisition, impaired generalization, and a lack of cognitive flexibility, but as I mentioned earlier in this chapter, increased levels of norepinephrine can also restrict cognitive networks.
Some, but not all, studies have suggested that autism is associated with an increase of central nervous system monoamines, including norepinephrine (Gillberg & Svennerholm, 1987), and as I mentioned earlier, this increase can account for a constriction of associative networks and conceptual representations. People suffering with autism frequently manifest behaviors that suggest that these people might be hyperaroused (Fankhauser, Karumanchi, German, Yates, & Karumanchi, 1992). For example, they exhibit stereotyped body movements, self-stimulation, hypervigilance, and hyperactivity. Arousal is, in part, mediated by the neurotransmitter norepinephrine. Whereas direct examination of brain norepinephrine has not yet been performed in autism, treating autistic patients with either clonidine (an alpha 2 adrenergic receptor agonist), which reduces the excretion of norepinephrine (Fankhauser et al., 1992), or beta-blockers, which block the effect of norepinephrine on the brain, may improve the symptoms associated with autism (Rately et al., 1987).
EEG studies of autistic patients also suggest that some of these people might be in chronically high states of physiological arousal (Hutt, Hutt, Lee, & Ounsted, 1965). In contrast, studies of physiological arousal in depression, as determined by EEG power analysis, have revealed that depressed patients have reduced arousal that is altered with treatment (Knott, Mahoney, & Evans, 2000; Nieber & Schlegel, 1992). If, as I suggested previously, lower levels of physiological arousal allow people to increase the extent of their concept representations and increase their cognitive flexibility, then it would follow that people with depression might have a propensity to be creative and patients with autistic spectrum disorder would tend to have limited creativity.
As mentioned, increased locus coeruleus activity is associated with increased levels of cerebral norepinephrine and this increased norepi-nephrine induces physiological arousal as determined by changes in the EEG. Neurophysiological studies have revealed that stimulation of only certain parts of the cerebral cortex induces these arousal changes in the EEG. These areas include the frontal cortex and the anterior cingulate gyrus (Segundo, Naguet, & Buser, 1955), which are strongly interconnected.
Functional imaging studies of patients with depression can provide some insight into the mechanisms by which depression might provide the basis for creative inspiration. Many such studies have shown that depression is associated with reduced cerebral blood flow (i.e., reduced synaptic activity) in the dorsolateral prefrontal cortex and the anterior cingulate gyrus (Liotti & Mayberg, 2001). The observed reductions in dorsolateral prefrontal blood flow might be related to the relative failure of depressed patients to spontaneously engage the environment (attend to and develop thoughts or plans about environmental interactions) either because they are pathologically disengaged or because they are engaged in internally defined plans (e.g., introspection, rumination). The reduced activity in the dorsolateral prefrontal cortex and the anterior cingulate cortex that occurs in depressed patients might be important for creative innovation because, as I mentioned earlier, the frontal lobes and anterior cingulate gyrus are the primary cortical areas controlling the locus coeruleus. Reduced activity in these frontal and cingulate regions, by virtue of the reduced input to locus coeruleus, could provide the basis for reductions of cortical norepinephrine, associated reductions in signal-to-noise ratio, and the recruitment of widely distributed networks that contain a rich variety of representations.
Whereas depression might facilitate creative innovation, the verification and production portions of creative endeavors are often associated with high arousal. Thus, these stages of creativity often must await the resolution of depression. In addition to the frontal lobe's crucial role in modulating the activity of the locus coeruleus, the frontal lobes play an important role in many other aspects of creative endeavors. In the next chapter I discuss some of the important functions that the frontal lobes play in many other aspects of creative endeavors.
Was this article helpful?
Are You Depressed? Heard the horror stories about anti-depressants and how they can just make things worse? Are you sick of being over medicated, glazed over and too fat from taking too many happy pills? Do you hate the dry mouth, the mania and mood swings and sleep disturbances that can come with taking a prescribed mood elevator?