The “thalamus as relay” framing held for most of the twentieth century and remains accurate for the primary sensory thalamic nuclei, where ascending information is faithfully reproduced (with modulation) before being passed to cortex. The framing has been substantially complicated by the recognition that most thalamic nuclei, by neuron count, are higher-order rather than primary, and that the higher-order thalamus is best understood as a partner to cortex in distributed cognition rather than a passive switchboard.

Sherman and Guillery’s distinction between “first-order” thalamic nuclei (driven by ascending input) and “higher-order” nuclei (driven by cortical layer V) has become the standard framework. Higher-order nuclei participate in cortico-thalamo-cortical loops that operate in parallel with direct cortico-cortical connections, potentially permitting cortex to coordinate its own activity through thalamic relays.

The thalamic reticular nucleus, a thin shell of GABAergic neurons that surrounds the dorsal thalamus laterally, gates ascending sensory information through inhibitory projections back onto the thalamic nuclei it surrounds. It is the only thalamic nucleus that does not project to cortex, and its role in attentional gating has made it a target of intense interest for work on selective attention.