Peter Schiller once made an insightful comment on the operant training of behaving monkeys to investigate the visual system: “Perhaps we are just studying a monkey’s thirst for apple juice, and all this visual stuff is an epiphenomenon.” Indeed, early investigators such as Olds and Milner (1954) described a massive subcortical network that mediates reward that spans from the brain stem to the nucleus accumbens, and innervates vast parts of the neocortex and cerebellum (Fig. 1). Within this web of fibre bundles, reward circuits have been described in vertebrates, and homologous circuits exist in invertebrates (Breiter, Kahneman et al. 2001; Gallistel et al. 1981; Kofman and Yeomans 1989; Kubikova et al. 2010; Montague at el. 1995, 1996; Mogenson 1977; Olds 1958; Olds and Milner 1954; Pallikaras and Shizgal 2022; Pissadaki and Bolam 2013; Schultz, Dayan et al. 1997; Schulz et al. 1993; Solomon et al. 2017; Yeomans et al. 1988).
We have argued that the dopaminergic pathway is central to slowing down or speeding up transitions in consciousness (Tehovnik, Hasanbegović, Chen 2024). Subjects who are severely depressed to the point of being Parkinsonian have an interrupted stream of consciousness and movement (Sacks 1976, 2012; Varela 1999ab) and subjects who are schizophrenic have an accelerated stream of consciousness and movement (Varela 1999ab) that causes a detachment between a subject’s internally-generated reality and the outside world: they have great difficulty dissociating between self-generated and externally-generated movements (see Footnote 1; Blakemore et al. 2000; Frith, Wolpert et al. 2000; Lemaitre et al. 2016), i.e., they have a deficit in efference-copy programming, which is under the control of the cerebellum (Bell et al. 1997; De Zeeuw 2021; Fukutomi and Carlson 2020; Loyola et al. 2019; Shadmehr 2020; Tehovnik et al. 2021; Wang et al. 2023 ). Normally, the cerebellum is well integrated into the dopaminergic reward network (see Fig. 2) so that the stream of consciousness and movement are uniformly regulated by both the neocortex and cerebellum. Thus, declarative and motor processing are synchronized, a synchronization that breaks down in schizophrenic subjects (Parker and Andreason 2014)—who must also laugh at all their own jokes as though having originated from someone else.
Footnote 1: Subjects diagnosed as schizophrenic are more able to tickle themselves (as compared to controls) suggesting that their efference-copy mechanism is dysfunctional (Blakemore et al. 2000; Frith, Wolpert et al. 2000; Lemaitre et al. 2016).
Figure 1: Top panel shows the ‘reward’ afferents to the medial forebrain bundle from neocortical and brainstem and the bottom panel shows the ‘reward’ efferents from the medial forebrain bundle to the hippocampus, neocortex, and brain stem. From figure 5 of Pallikaras and Shizgal (2022) based on the rodent. (auto_065.jpg)
Figure 2: Efferents of the ventral tegmental area (VTA), whose nuclei occupy the most caudal part of the medial forebrain bundle, innervate the pontine nuclei and inferior olive (IO), and afferents of the cerebellar nuclei (dentate and fastigial) project to the ventral tegmental area, thereby integrating the cerebellum with the reward signal. The ventral tegmental area (or its homologue) sends dopaminergic projections to the forebrain in all vertebrates. From figure 3A of Parker and Andreason (2014) based on the rodent. (auto_204.jpg)