D. Freeman's Primitive Felt Theory

Biomedical engineer Daniel Freeman's Primitive Felt Theory (PFT) holds that phenomenal consciousness is constructed from a finite repertoire of phenomenologically irreducible felt states instantiated in evolutionarily conserved subcortical circuitry, not in the cerebral cortex. Distributed cortical networks compute the content of mental states (sensory, conceptual, and inferential information) largely outside awareness; their processed outputs descend through layer V corticofugal projections to integrative structures near the midline of the diencephalon and upper brainstem, where they are expressed as configurations of primitives by a physical mechanism not yet identified.

The theory's distinctive conjecture is that the repertoire that yields thirst, pain, and fear also yields the experience of abstract thought, so that cognitive phenomenology—the standing objection to subcortical theories—becomes a compressed, low-dimensional readout of cortical computation instead of evidence that experience occurs in cortex (Freeman, 2026). Human cognitive phenomenology—the subjective character of thinking—would therefore elaborate an older experiential architecture through changing combinations of primitive feelings rather than require a fundamentally new source of consciousness. Moreover, phenomenal unity requires a spatially cohesive or physically continuous process (Freeman, personal communication).

Computation Versus Instantiation

The theory separates two things that cortical theories tend to treat as one: the neural computations that fix what a mental state is about, and the physical process that gives the state its phenomenal character. Distributed cortical networks perform the representational and inferential work, but on Freeman's account these operations are not themselves experienced. What enters awareness is a felt readout of their results. The distinction concerns constitution, not simply neuroanatomy: cortex can be indispensable for a particular sophisticated experience without being the structure in which that experience is felt. Limited working memory and the sequential character of deliberation motivate a compressed experiential representation of much richer processing. Nevertheless, behavioral information rates and report limitations do not directly measure the total richness of phenomenal consciousness.

Global workspace and higher-order theories explain cognitive phenomenology readily because they place experience where cognition is computed; Freeman accepts the cortical computation and relocates only its experiential instantiation. He concedes that the generating mechanism remains unidentified, and the paper developing the argument presents itself as answering one specific challenge to existing subcortical accounts, not as a fundamentally new theory of consciousness (Freeman, 2026).

Primitive Felt States

"Primitive" is defined phenomenologically, not phylogenetically or computationally: a primitive state has its own distinctive what-it-is-like quality and is not experienced as a blend of more elementary feelings. The repertoire includes homeostatic and affective states such as thirst, pain, fatigue, fear, and rage, and Freeman extends it to evaluative states—familiarity, confidence, uncertainty, coherence, anticipation, obligation, urgency. Whether any given evaluative state is truly irreducible, he treats as an open empirical question. Mapping studies of self-reported feelings, which find homeostatic, emotional, and cognitive experiences distributed across a continuous space, are taken as supportive evidence (Nummenmaa et al., 2018).

On this view, the experience of thinking consists of sensory imagery, such as inner speech, together with rapidly shifting configurations of evaluative primitives. In the thought "I should leave now," familiarity, obligation, and mounting urgency are said to unfold alongside the words. That people with aphantasia (i.e., the inability to form mental images of objects that are not present) reason and deliberate normally suggests to Freeman that imagery is an auxiliary channel and the primitives are the more basic ones (Zeman et al., 2015).

Compression and Combinatorial Richness

According to Freeman, two considerations make a small repertoire sufficient. First is the mismatch between massively parallel cortical processing and the slow stream of conscious thought, estimated at roughly ten bits per second (Zheng and Meister, 2025); experience looks like a low-dimensional summary, not a transcript. Second, if experiential space has N independently varying dimensions, each resolved at B bits, the number of distinguishable states is 2^(B×N); thirty dimensions at eight levels each yield more than 10²⁷ configurations. Freeman uses large language models as an analogy for the architecture—high-dimensional internal representations, a narrow output channel—and marks the analogy as illustrative, not as a claim about neural implementation.

He uses the same numerical estimates to reply to the objection that distinct propositions might evoke the same blend of feelings, leaving thought without semantic specificity (Chudnoff, 2015). Different thoughts, he argues, occupy different coordinates, and the precise time-varying trajectory through experiential space individuates a concept or proposition. The theory is offered as neutral in the debate (Bayne and Montague, 2011) over whether thought has a rich, proprietary phenomenology (Pitt, 2004) or a thin one: either way it is built from the same primitives.

Corticofugal Convergence and Phenomenal Unity

Freeman’s PFT builds on Merker’s efference cascade, in which descending axons of cortical layer-V pyramidal neuron, the principal output of cortical computation, convey cortical results to interconnected subcortical structures: striatal, thalamic, hypothalamic, and midbrain targets (Merker, 2013). He emphasizes the higher-order thalamic nuclei—the mediodorsal nucleus receiving frontal association cortex, the pulvinar receiving posterior sensory cortex—operating in tandem with the superior colliculus. Candidate substrates are therefore the upper brainstem, hypothalamus, and higher-order or intralaminar thalamus.

Convergence onto a compact region is also his approach to binding. Theories whose substrates are separated by centimeters of white matter must let functional connectivity alone carry phenomenal unity; Freeman proposes instead that unified experience requires a physical process with some spatial continuity. Near the midline, bilateral neural populations lie close enough for spatially extended processes, including extracellular ionic volume currents, to overlap. He presents volume currents as one candidate mechanism, not as an established or necessary component.

Evolutionary Continuity

Freeman contrasts two histories. In one, which he associates with higher-order views, experience appears only after cognition is sufficiently complex. In the other, which he favors, early vertebrates already possessed a limited set of felt states, and cortical expansion enriched content without adding a new experience-generating process. The adaptive role proposed for primitives is a common evaluative system: competing needs are translated into a single organism-level state that guides action selection, and later cortical representations project into that same valuation system. Primitive feeling and abstract thought are thus treated as different dynamical regimes of one process.

Relations and Predictions

The theory belongs with those of Merker, Panksepp, Damasio, and Solms. It differs by claiming that the experience of abstract thought, not only affect and basic perception, is generated subcortically through finite combinatorics, and by requiring spatial cohesion of the substrate. Metacognitive states that higher-order theories treat as conferring awareness are reinterpreted as primitives themselves. The central prediction is a dissociation: perturbing candidate subcortical nuclei, for instance with transcranial focused ultrasound (Freeman et al., 2026), should alter confidence, coherence, familiarity, or felt rightness while sparing the underlying cortical computation—a claim that may be empirically testable.

Predictions and Explanatory Limits

Primitive Felt Theory addresses the gap most often cited against subcortical accounts, and does so with a quantitative sufficiency argument and a stated test. Combinatorial capacity shows that a low-dimensional space could individuate thoughts, not that it does; the primitives are not enumerated and their irreducibility rests on introspective report; the mechanism that turns subcortical activity into experience is unspecified; and the volume-current proposal is conjectural.

Freeman predicts that perturbing candidate subcortical systems could change the felt qualities of thinking—confidence, coherence, familiarity, or uncertainty—while substantially sparing semantic processing and task accuracy. Transcranial focused ultrasound offers a possible experimental approach to deep targets (Freeman et al., 2026). Such experiments would need to distinguish changes in phenomenology from altered arousal, response criteria, metacognitive reporting, or cortical activity downstream of stimulation. A confidence–accuracy dissociation alone would not uniquely support PFT or establish where experience is constituted. The theory's standing will depend on whether the predicted dissociation between felt evaluation and cortical computation can be demonstrated.

PFT is therefore a provisional account of the organization and possible substrate of phenomenal consciousness. It explicitly addresses a difficulty within existing subcortical theories rather than claiming a fundamentally new theory. Together, the documents propose a finite phenomenal repertoire and a physical-unity hypothesis, but neither explains why the proposed physical processes should feel like anything at all.

References

Bayne, T., and Montague, M. (Eds.) (2011). Cognitive Phenomenology. Oxford: Oxford University Press.

Chudnoff, E. (2015). Cognitive Phenomenology. London: Routledge.

Freeman, D. K. (2026). Accounting for cognitive phenomenology in subcortical theories of consciousness. Preprint. [venue/DOI to be supplied]

Freeman, D. K., Odegaard, B., Yoo, S.-S., and Michel, M. (2026). Transcranial focused ultrasound for identifying the neural substrate of conscious perception. Neuroscience & Biobehavioral Reviews, 180, 106485.

Nummenmaa, L., Hari, R., Hietanen, J. K., and Glerean, E. (2018). Maps of subjective feelings. Proceedings of the National Academy of Sciences, 115(37), 9198–9203.

Pitt, D. (2004). The phenomenology of cognition, or what is it like to think that P? Philosophy and Phenomenological Research, 69(1), 1–36.

Zeman, A., Dewar, M., and Della Sala, S. (2015). Lives without imagery—Congenital aphantasia. Cortex, 73, 378–380.

Zheng, J., and Meister, M. (2025). The unbearable slowness of being: Why do we live at 10 bits/s? Neuron, 113(2), 192–204.