Two hemispheres but one brain

of the original article: A review of research with chicks, songbirds, rodents, and nonhuman primates indicates that the brain is lateralized for a number of behavioral functions. These findings can be understood in terms of three hypothetical brain processes derived from a brain model based on general systems theory: hemispheric activation, interhemispheric inhibition, and intcrhemisphcric coupling. Left-hemisphere activation occurs in songbirds and nonhuman primates in response to salient auditory or visual input, or when a communicative output is required. The right hemisphere is activated in rats when spatial performance is required, and in chicks when they are placed in an emotion-provoking situation. In rats and chicks interhemispheric activation and inhibition occur when there is an affective component in the environment (novelty, aversive conditioning) or when an emotional response is emitted (copulation, attack, killing). An interhemispheric coupling (correlation) found in rats and rabbits implies that the hemispheres are two major components in a control system with a negative feedback loop. Early-experience variables in rats can induce laterality in a symmetric brain or facilitate its development in an already biased brain. It is predicted that functional lateralization, when present, will be similar across species: the left hemisphere will tend to be involved in communicative functions while the right hemisphere will respond to spatial and affective information; both hemispheres will often interact via activation-inhibition mechanisms when affective or emotional processes are involved. Homologous brain areas and their connecting callosal fibers must be intact at birth and must remain intact throughout development for lateralization to reach its maximum level. Injury to any portion of this unit will result in hemispheric redundancy rather than specialization. One major function of early experience is to provide stimulation during development, which acts to enhance the growth and development of the corpus callosum, thereby giving rise to a more specialized brain. Two hemispheres but one brain G. Berlucchi Islituto di Fisiologia dell University di Pisa e Istituto di Neurofisiologia del CNR, Pisa, Italy Dencnberg's (1981a) theory about hemisphereic interaction suffers from a general lack of clear anatomical and physiological definitions. For example, the term "hemisphere" is loosely used as synonymous with "cortex," while according to the standard anatomical nomenclature the basal ganglia, the hippocampal formation, and the amygdaloid complex are also parts of each cerebral hemisphere of the mammalian brain. Expressions such as "hemispheric activation" and "interhemispheric inhibition" have little use unless one refers to specific anatomical entities and physiological processes. There are several recent studies on regional blood flow in the intact human brain to show that if "hemispheric activation" means increase in neuronal activity in certain or all parts of one hemisphere, with no change, or even a decrease, in neuronal activity in the other hemisphere, then it is indeed a rare occurrence, if it exists at all. Even during the execution of mental or behavioral tasks that are supposed to be strongly lateralized, changes in regional blood flow in corresponding areas of the two sides, which are strictly parallel to changes in neuronal activity, tend to be roughly equal (see, e.g., Larsen, Skinh0j & Lassen 1975; Prohovnik, Hakansson & Risberg 1980; Risbcrg, Halsey, Wills & Wilson 1975; Roland, Larsen, Lassen & Skinh0j 1980; Roland, Skinh0j, Lassen & Larsen 1980). If by "interhemispheric inhibition" one means that neuronal activity in one hemisphere reduces or suppresses activity in the other hemisphere, then the evidence that this takes place in the intact brain is again, at best, meager. Prohovnik et al. (1980) have noticed a conspicuous lack of negative correlations between the activities of homologous regions of the conscious human brain. Whether cross-inhibition occurs at subcortical levels is another problem. Risberg ctal. (1975), after testing the effects of a verbal task and a spatial task on regional cerebral blood flow (rCBF) in normal humans, concluded that "rCBF increases during 'unilateral' activation take place in both hemispheres, indicating the importance of integration of brain activity mediated by the cerebral commissures, especially the corpus callosum." This brings me to Denenberg's general misrepresentation of commissurai functions. In principle, his theory does not require major participation of the cerebral commissures in the putative mechanisms of interhemispheric inhibition and in the control of the negative feedback loop between the coupled hemispheres since the outputs.of the two hemispheres might compete at the motoneuronal level. Yet Denenberg makes it clear that he believes that the corpus callosum is the most likely candidate for fulfilling these functions, although he gives little evidence in support of his belief. He certainly cannot use as evidence for callosal inhibition the finding of Glick (1975) that a section of the anterior corpus callosum in rats increased the number of rotations induced by amphetamine administration. According to the authors this phenomenon results from the interruption of a pathway between the right and left caudate nuclei that passes through the ventral corpus callosum, resulting in the accentuation or the establishment of neurochemical striatal asymmetries. Their suggestion is that this pathway serves to synchronize the activities of the two corpora striata, rather than, as Denenberg instead suggests, to mediate cross-inhibition between higher cortical centers. I am amazed by the complete lack of consideration on Denenberg's part of the enormous literature on unilateral cortical spreading depression (see e.g. BureS, BureSovd & Kfivdnek 1974; and the commentary of BureS, BureJovii & Kfivdnek 1981), which, in my view, gives very little support for the hypothesis of hemispheric dominance and interhemispheric inhibition on rats. To the extent that Denenberg's model is a neurodynamic one (i.e., to the extent that his putative mechanisms of hemispheric interaction are physiological influences rather than long-term biochemical changes), reversible inactivation of one hemisphere by spreading depression would provide an excellent means to test his ideas directly. THE. BEHAVIORAL AND BRAIN SCIENCES (1983) 1 171 Continuing Commentary To return to the physiology of the cortical commissures, I have recently reviewed a massive body of evidence indicating that the functional significance of these connections lies in the unification of sensory information coming from the two halves of the body or the visual field and, in general, the equalization of activity in corresponding regions of the two hemispheres (Berlucchi 1981a). This is fully in accord with Sperry's idea that the corpus callusom has a unifying role in conscious experience and cortical activity (Sperry 1974). As an example, visual receptive fields of neurons in the inferotemporal cortex of the monkey extend across the vertical midline of the visual field and have the same response properties throughout (see Gross, Bender & Mishkin 1979). Since the input to these neurons from the contralateral visual field is conveyed by thalamocortical projections or intrahemispheric corticocortical projections, and the input from the ipsilateral visual field is conveyed by the corpus callosum or anterior commissure, such continuous and homogeneous receptive fields provide a good indication of the integration between intrahemispheric and interhemispheric inputs, and no evidence for interhemispheric inhibition. I fail to see the analogy between the competition between the inputs from the two eyes in the lateral developing geniculate nucleus and the proposed action or inaction of the corpus callosum in development and maturation, as suggested by Denenberg (1981a on pp. 52 ff.). Whatever evidence is available on the development of callosal fibers indicates the callosal connections are modified according to the functional needs of their target areas, and exert no definite influence on the organization and connectivity of such areas during development and maturation. Thus, callosal connections uniting cortical points representing eccentric portions of the visual field are dropped in the postnatal period in the cat (see Innocenti 1980), probably because it would be meaningless to receive incongruent visual information from the far right and left fields at the same cortical point (see Berlucchi 1981a). The general rule appears to be that callosal and noncallosal inputs to the same cortical area are blended and integrated into a functional whole, and where there seems to be a segregation of callosal and noncallosal inputs, as in the somatic sensory cortex of the rat, there is no competition between the two sets of inputs during development (see Wise & Jones 1978). I like Denenberg's suggestion that homologous regions of the two hemispheres make up a functional unit because of their commissural connections, but my view of such a functional unit is very different from his (Berlucchi 1981b). I believe that corresponding cytoarchitectural areas of the two sides have exactly the same physiological organization and general function in behavior, and the commissural links serve to ensure a yoked synchronous and congruent activity in the two members of each pair of areas. Hemispheric asymmetry arises when the two areas constituting a bilateral functional unit are asymmetrical (a good example is the temporoparietal transitional area in the human brain; see Galaburda, Sanides & Geschwind 1978), so that the area on one side is larger or has a higher neuronal density than that on the other. Under these conditions, I submit, the callosal connections of such areas would also be asymmetrical, the projections arising from the larger area b