The default-mode, ego-functions and free-energy: a neurobiological account of Freudian ideas

This paper is a theoretical synthesis and narrative review rather than an empirical study. The investigators curated and integrated findings from multiple domains—resting-state and task fMRI, diffusion imaging, intracranial and scalp electrophysiology, psychopharmacology (notably serotonergic hallucinogens), clinical neuropsychiatry (schizophrenia, depression), epilepsy literature (temporal lobe aura), and developmental studies—to evaluate parallels between Freud's constructs and contemporary neuroscience. Their analytic strategy was conceptual mapping: (i) associating the Freudian secondary process and ego-functions with high-level intrinsic networks (chiefly the DMN) and their top-down suppressive role in a hierarchical inferential brain; (ii) associating the primary process with limbic/paralimbic activity and particular neurophysiological signatures observed in non-ordinary states; and (iii) using the free-energy principle and predictive coding as a formal bridge linking the two frameworks. The authors drew on empirical reports of BOLD resting-state networks, measures of functional and structural connectivity, intracranial EEG findings (theta and gamma activity), and clinical correlations (e.g. altered DMN suppression in psychiatric disorders) to support their reasoning. Where relevant, developmental and clinical findings (e.g. ontogenetic emergence of DMN connectivity, altered DMN dynamics in schizophrenia and depression) were incorporated to assess parallels with Freud's ontogenetic claims about the ego. The paper also discusses proposed empirical tests—such as using functional/effective connectivity measures to examine limbic influence on higher-order cortex during phasic events—but does not present new experimental data or formal meta-analytic procedures.

The synthesis yields a set of convergent observations and empirical claims supporting the proposed mapping between Freudian constructs and neurobiology. At a systems level, regions commonly identified as the DMN (medial prefrontal cortex, posterior cingulate cortex, inferior parietal lobule, lateral/inferior temporal cortex and medial temporal lobes) show high tonic metabolism at rest and deactivate during goal-directed tasks. Functional and structural connectivity analyses indicate that major DMN nodes are strongly interconnected and that this connectivity matures through ontogeny, paralleling Freud's account of ego development. The DMN is also reciprocally related (anti-correlated) to networks engaged by externally-directed attention; the investigators align this reciprocity with Freud's descriptions of ego-libido versus object-libido and the ‘‘give-and-take’’ between internally- and externally-oriented processes. Physiologically, recent simultaneous fMRI and intracranial recordings indicate that spontaneous BOLD fluctuations largely reflect neuronal activity—gamma-range local field potential power and multiunit firing—suggesting that large-scale DMN oscillations index ongoing cortical processes. The authors review data showing that cortical gamma, recorded in higher-level areas, can suppress lower-frequency activity in subordinate areas and that theta–gamma coupling (theta phase modulating gamma amplitude) is evident in limbic–cortical interactions. These dynamics are invoked to support a Helmholtzian account in which backward (top-down) connections convey predictions that suppress prediction-error (free-energy) encoded by forward connections. Turning to primary-process states, the investigators marshal phenomenological parallels among REM sleep, early/acute psychosis, the temporal-lobe aura and hallucinogenic drug states: altered time perception, fear, perceptual distortion/hallucinations, déjà vu and disturbances of self. Neurophysiological data cited include intracranial recordings showing bursts of high-amplitude low-frequency (often theta) activity in medial temporal and septal regions during recollective hallucinations, acute psychosis and under LSD/mescaline; similar hippocampal theta bursts appear in REM sleep. The authors propose that these limbic theta discharges index increased free-energy at subordinate levels and that, in normal waking cognition, cortical gamma activity entrains to and suppresses these limbic discharges, thereby implementing the secondary process. In non-ordinary states this hierarchical containment is compromised, permitting limbic activity to exert atypical influence on association and sensory cortices. Clinically relevant observations reviewed include reduced task-induced suppression of DMN activity in schizophrenia and depression, correlations between such reductions and symptom severity (e.g. hopelessness in depression), and developmental/clinical alterations in DMN connectivity (weaker in infants and elderly, altered in attention-deficit and impulse-control disorders). The investigators highlight these empirical patterns as consistent with Freud's descriptions of withdrawal, impaired reality-testing and ego-disturbance.

Carhart-Harris and Friston interpret their synthesis as evidence that Freudian constructs—particularly the secondary process and ego-functions—have plausible neurobiological substrates in the organisation and dynamics of large-scale intrinsic networks, with the DMN playing a central role. They argue that the free-energy principle and hierarchical Bayesian inference formalise Freud's older notion of the mind striving to minimise ‘‘sums of excitation’’, by showing how backward (top-down) predictions from higher cortical areas can suppress prediction-error in subordinate systems. In this view, the ego corresponds to high-level predictive systems that tonically contain endogenous limbic excitation and orient the organism appropriately to external stimuli via reciprocal interactions with attention and salience systems. The investigators position the primary process as an expression of increased limbic free-energy, manifesting phenomenologically in REM sleep, psychosis, temporal-lobe auras and hallucinogenic drug states and physiologically in bursts of limbic theta and altered theta–gamma coupling. They suggest that failure of top-down containment can account for symptoms such as hallucinations, delirium, altered time perception and disturbances of self, and that compensatory phenomena (for example delusions or withdrawal) might be understood as attempts to reduce overwhelming prediction-error. Carhart-Harris and Friston acknowledge that readers will differ in which aspects of the synthesis they find compelling: some may value the mapping between inferential coding and intrinsic networks, others the formal similarity between Freud's account and free-energy minimisation. The investigators advocate for grounding psychoanalytic constructs in measurable neurobiology, not to preserve theory for its own sake, but to provide heuristics for investigating non-ordinary states and psychopathology. They note limitations explicitly: the paper does not assess the clinical efficacy of psychoanalysis, some Freudian premises may ultimately be superseded as empirical knowledge develops, and their synthesis is intended as a framework to stimulate testable hypotheses rather than definitive proof. Empirical tests recommended include using functional and effective connectivity measures to detect increased limbic-to-cortical drive during phasic non-ordinary events and developing neurobiological signatures of ‘‘ego-disturbance’’ that might illuminate schizophrenia prodromes and depressive withdrawal.

The investigators conclude that multiple lines of evidence—developmental, functional-anatomical, electrophysiological and clinical—are consistent with mapping Freud's secondary process and ego-functions onto the default-mode network and its hierarchical suppressive role, while primary-process thinking aligns with limbic-driven, high-amplitude low-frequency activity characteristic of REM sleep, psychosis, temporal-lobe aura and hallucinogenic states. They emphasise that the free-energy principle provides a formal bridge between Freudian descriptions of ‘‘bound’’ versus ‘‘free’’ psychic energy and contemporary models of hierarchical inference in the brain. Finally, Carhart-Harris and Friston propose that this empirically-led synthesis can guide future research by specifying neurobiological measures (e.g. connectivity, oscillatory dynamics) to test hypotheses about ego-disturbance and the neurodynamics of non-ordinary states.