The preparatory activity that precedes movement execution can be thought of as a thinking process that anticipates the characteristics of an up-and-coming movement (Darlington and Lisberger 2020; also see James 1890). For both classical and operant conditioning, ‘preparatory loops’ depend on the neocortex and cerebellum, as demonstrated for both eyeblink conditioning and response discrimination in mammals. In the case of eye-blink conditioning, the critical circuit for mediating this response when a delay is imposed between the offset of the conditioned stimulus and the onset of the unconditioned stimulus (called trace conditioning) involves the following loop in rabbits: the medial prefrontal cortex, the pontine nuclei, the cerebellum and its nuclei, and finally the mediodorsal thalamus (Kalmbach et al. 2009; Siegel et al. 2013). For this circuit, it has been demonstrated that the neocortex is necessary to generate the preparatory activity between the offset of the conditioned stimulus and the onset of the unconditioned stimulus at the level of the cerebellar mossy fibres (which carry the conditioned-stimulus signal) and the climbing fibres (which carry the unconditioned-stimulus signal). The activity of the mossy fibres must be extended beyond the offset of the unconditioned stimulus for there to be an association between the conditioned stimulus and unconditioned stimulus, and the greater the duration between conditioned-stimulus offset and unconditioned-stimulus onset beyond 300 ms, the more vital the preparatory activity for making an association (Kalmbach et al. 2009).
In the case of response discrimination, mice were trained to detect a probe making contact with either rostral or caudal vibrissae (of the right side of the body) such that a rostral touch signaled an up-and-coming leftward licking response, and a caudal touch signaled an up-and-coming rightward licking response (Gao et al. 1918; Hasanbegović 2024; Zhu et al. 2023). Before licking, however, a mouse was required to delay the response for 1 second or so (an imposed memory period) such that thereafter the presentation of an auditory cue triggered the response. The critical circuit for the response included the anterolateral motor cortex, the pontine nuclei, the cerebellum including the cerebellar nuclei, and the ventromedial thalamus. At all the stations within this loop, preparatory activity was exhibited by the neurons during the delay period. As well, interference during the delay period at any loop location interrupted response discrimination, which suggests that the preparatory activity across the loop locations is shared to generate a common discrimination response.
Much data suggest that the learning of classical (i.e., trace conditioning, but see Footnote 1 for delay conditioning) and operant tasks depend on the neocortex (and hippocampus) and the cerebellum (Corkin 2002; Kassardjian et al. 2005; Kalmbach et al. 2009; Kim et al. 1995; Kimura 1993; Logothetis et al. 2012; Marr 1969, 1971; Ojemann 1991; Pavlides and Winson 1989; Pavlov 1927; Penfield 1975; Penfield and Roberts 1966; Sendhilnathan et al. 2020b; Squire et al. 2001; Takahara et al. 2003; Vanderwolf 2007; Wilson and McNaughton 1994). This conclusion does not depart from the notions advanced by Hebb (1949, 1961, 1968), who was mainly focused on the neocortex. And during his tenure as a researcher, the details of the cerebellum were just starting to be worked out, but many of the observations made by Hebb such as his interest in illusions and eye movements with respect to visual images (including prism adaptation) can now be explained by our understanding of loops passing through the cerebellum (for details see: Tehovnik, Hasanbegović, and Chen 2024; also: Tehovnik, Patel, Tolias et al. 2021).
Footnote 1: Delay conditioning (but not trace conditioning) can occur in the absence of neocortex (Gallistel et al. 2022; Kalmbach et al. 2009; Mauk and Thompson 1987). However, it was apparent to Pavlov (1927) that many types of classical conditioning (even if of the delay type) are permanently abolished in the absence of neocortex. The reason for this is that without neocortex there are many sensory associations that can never be made. For example, following neocortical damage in mammals, pattern and depth perception is abolished, and following such damage in humans any association dependent on language is an impossibility (Kimura 1993; Ojemann 1991; Penfield and Roberts 1966; Tehovnik, Patel, Tolias et al. 2021). Furthermore, if just V1 is damaged, animals (including rodents and primates) experience blindsight: they can detect (unconsciously) only visual stimuli of contrast greater than 95% (Tehovnik, Patel, Tolias et al. 2021). In short, in the absence of neocortex the only associations that are made are those requiring high threshold sensory activation, such as when evoking sensations of pain (see Baron and Devor 2022).
- Badges
- Science topic