Various lines of evidence suggest that the development and evolution of the mammalian isocortex cannot be easily explained without an understanding of correlative changes in surrounding areas of the telencephalic pallium and subpallium. These are close neighbours in a common morphogenetic field and are postulated as sources of some cortical neuron types (and even of whole cortical areas). There is equal need to explain relevant developmental evolutionary changes in the dorsal thalamus, the major source of afferent inputs to the telencephalon (to both the pallium and subpallium). The mammalian isocortex evolved within an initially small dorsal part of the pallium of vertebrates, surrounded by other pallial parts, including some with a non–cortical, nuclear structure. Nuclear pallial elements are markedly voluminous in reptiles and birds, where they build the dorsal ventricular ridge, or hypopallium, which has been recently divided molecularly and structurally into a lateral pallium and a ventral pallium. Afferent pallial connections are often simplified as consisting of thalamic fibres that project either to focal cell aggregates in the ventral pallium (predominant in reptiles and birds) or to corticoid areas in the dorsal pallium (predominant in mammals). Karten's hypothesis, put forward in 1969, on the formation of some isocortical areas postulates an embryonic translocation into the nascent isocortex of the ventropallial thalamorecipient foci and respective downstream ventropallial target populations, as specific layer IV, layers II– III, or layers V–VI neuron populations. This view is considered critically in the light of various recent data, contrasting with the alternative possibility of a parallel, separate evolution of the different pallial parts. The new scenario reveals as well a separately evolving tiered structure of the dorsal thalamus, some of whose parts receive input from midbrain sensory centres (collothalamic nuclei), whereas other parts receive oligosynaptic ‘lemniscal’ connections bypassing the midbrain (lemnothalamic nuclei). An ampler look into known hodological patterns from this viewpoint suggests that ancient collothalamic pathways, which target ventropallial foci, are largely conserved in mammals, while some emergent cortical connections can be established by means of new collaterals in some of these pathways. The lemnothalamic pathways, which typically target ancestrally the dorsopallial isocortex, show parallel increments of relative size and structural diversification of both the thalamic cell populations and the cortical recipient areas. The evolving lemnothalamic pathways may interact developmentally with collothalamic corticopetal collaterals in the modality–specific invasion of the emergent new areas of isocortex.