Tracing the Evolution of Consciousness Through Neuroscience
Second Lecture
The question “When did consciousness first appear on Earth?” is no longer dismissible as metaphysical. It has become a tractable, indeed urgent, empirical problem because the answer determines whether we are investigating a late, fragile, uniquely human miracle or a deep, robust, continuously graded capacity that reaches far back into the history of mobile multicellular life. The past twenty-five years of comparative, developmental, and palaeo-neurobiology have converged on a radically continuity-based picture: phenomenal consciousness is not an abrupt event that occurred with the first Homo sapiens, nor with the appearance of cortex, nor even with the origin of mammals. Its essential ingredients—affectively charged, unified sensory-motor representation—are ancient, traceable at least to the early Cambrian, and possibly to the Ediacaran–Cambrian boundary 540 million years ago.
Begin with the functional argument that forced this revision. The Cambrian explosion produced the first animals with image-forming eyes, distance chemoreceptors, rapid locomotion, and active predation. Suddenly, selection pressure was no longer simply for metabolic efficiency or reproductive speed but for flexible, adaptive behaviour in a three-dimensional world full of rapidly changing threats and opportunities. Björn Merker (2005, 2013) and others pointed out that such behaviour demands an internal evaluation system that can weigh competing motivational states (hunger vs. fear vs. mating) and select a single course of action. Reflex arcs and simple tropisms are insufficient; a centralised “decision marketplace” is required. Affect—the felt valence of good/bad, approach/avoid—becomes the common currency of that marketplace. Once organisms can feel, even dimly, that one outcome is better than another, a new adaptive landscape opens: learning accelerates, novelty-seeking becomes possible, and the stage is set for increasingly rich phenomenal worlds.
The neural hardware for this minimal affective consciousness appears with the first true vertebrates. The lamprey, a jawless fish that diverged 500 million years ago, possesses a pallium, a striatum, a tectum, and a functional analogue of the basal ganglia–thalamus–cortex loop. It has dopamine, serotonin, and opioid systems organised almost exactly as in mammals. Optogenetic and lesion studies in zebrafish larvae (a more experimentally tractable vertebrate) show that blocking the pretectum or the pallial amygdala homologue abolishes orienting to whole-field motion while leaving local reflexes intact—mirroring the dissociation seen in human hydranencephalic children who lack cortex yet smile, cry, track faces, and show affective preferences. The mesodiencephalic core—superior colliculus, pretectum, thalamus, hypothalamus, and basal forebrain—appears sufficient for a primitive phenomenal field: a unified sensory-motor space imbued with valence.
Feinberg and Mallatt (2016, 2019, 2023) have formalised this into a rigorous set of neurobiological criteria for basic consciousness:
(1) hierarchical sensory mapping with isomorphisms (retinotectal, somatotopic, tonotopic), (2) global workspace-like integration across modalities, (3) affective valence mediated by brainstem neuromodulators, (4) attention-like selection mechanisms, and (5) flexible memory-guided behaviour. These criteria are met by all vertebrates and by cephalopods, but not by protostomes (arthropods, nematodes) whose nervous systems, however sophisticated, lack the requisite recurrent mesodiencephalic circuitry. Recent work on octopuses (Crook, 2021; Godfrey-Smith & Caruso, 2023) complicates the picture: they show pain behaviour, long-term potentiation in the vertical lobe, and trade-off decisions that satisfy Cabanac’s criterion of affective consciousness. Their distributed nervous system (two-thirds of neurons in the arms) may support multiple semi-independent phenomenal perspectives rather than a single unified field, but the presence of some form of felt experience is now difficult to deny.
Reptiles, birds, and mammals inherited and elaborated this vertebrate core. Birds, despite lacking a laminated neocortex, possess a pallium with recurrent circuitry functionally equivalent to mammalian cortex (the dorsal ventricular ridge and hyperpallium). Single-unit recordings in crows and parrots show concept cells, mirror neurons, and theory-of-mind-like representations indistinguishable from primates. The discovery of REM sleep in lizards (Libourel et al., 2018) and of slow-wave/REM cycles in octopuses (Medina et al., 2023) further erodes the idea that dreaming—and thus offline phenomenal experience—is a mammalian privilege.
Mammals added two decisive innovations: the six-layered isocortex and the thalamocortical system. The dramatic expansion of cortex in primates, especially humans, is not the origin of consciousness but its most extreme elaboration. The default-mode network, metacognition, narrative self, and linguistic recursion are late arrivals, built on top of an affective core that was already conscious half a billion years ago.
Developmental neurobiology reveals the same continuity in slow motion. Human consciousness does not switch on at birth; it assembles gradually over months in utero and years postnatally. Fetal lambs at 0.7 gestation show only discontinuous trace alternant EEG; REM-like states appear after 0.85 gestation. In human preterm neonates, noxious stimulation before 25 weeks evokes only spinal and brainstem responses; specific cortical activation emerges between 28–34 weeks. Lagercrantz and Changeux’s thalamic switch hypothesis (2010, updated 2023) remains the leading model: around 24–26 weeks, massive thalamocortical afferents invade layer IV, transforming the cortex from a collection of isolated columns into a recurrent network capable of sustained ignition. Preterm infants born before this ingrowth show permanently altered functional connectivity and higher rates of neurodevelopmental disorder.
Resting-state fMRI in preterm infants reveals proto-default-mode and salience networks by 30 weeks, with gradual anti-correlation emerging across childhood. The slow maturation of long-range white-matter tracts (corpus callosum, superior longitudinal fasciculus) parallels the emergence of episodic memory, theory of mind, and reflective self-awareness. Autism spectrum conditions are increasingly understood as failures of long-range integration with preserved or enhanced local precision; schizophrenia as excessive salience-driven override of top-down prediction. Both can be traced to disruptions in the same developmental cascades that build the conscious brain.
Evolutionary game theory provides the functional rationale. Ginsburg and Jablonka’s Unlimited Associative Learning (UAL) criterion (2021) marks the transition from mere adaptive behaviour to experiential learning: the ability to form novel, compound, cross-modal traces with motivational valence. Once organisms can learn that a new combination of cues predicts reward or danger, a centralised evaluation system becomes adaptive. Phenomenal feel is the internal signal that enables apples-to-oranges comparison across motivational systems. Graziano’s attention-schema theory adds that in social species, modelling one’s own attentional state became necessary to predict conspecifics; the side-effect was a simplified internal model of awareness itself—the persistent illusion of a unitary self.
The most radical hypothesis is that consciousness is not for online decision-making (unconscious systems decide faster) but for learning about the consequences of decisions across vastly different timescales and modalities. Sensorimotor contingency theory (O’Regan) and predictive-processing accounts converge here: phenomenal experience is the brain’s way of tracking the structure of its own lawful interaction with the world. Damage those contingencies (deafferentation, vestibular loss) and qualia vanish or transform. The felt redness of red is the specific way visual exploration is perturbed by surfaces that reflect long wavelengths.
Ultimately, evolutionary and developmental neurobiology dissolve the pseudo-problem of emergence. There is no magical moment when consciousness appears ex nihilo. Instead, different grades and kinds of experience shade into one another across phylogeny and ontogeny. The bright, noisy, language-saturated consciousness of an adult human is the far end of a continuum that begins with the dim, affect-laden flicker in a zebrafish larva 500 million years in the making. The biological perspective thus returns us to the same conclusion reached by the Indian non-dual traditions from an entirely different route: consciousness is not a late addition to an otherwise unconscious universe; it is the primordial light in which the universe, as experienced, first becomes possible.
Tanmoy Bhattacharyya
