Researchers at Baylor College of Medicine have published findings in Nature that directly unsettle a foundational assumption of both clinical medicine and consciousness science: that general anesthesia suppresses complex cognition. Consciousness is a fundamental component of cognition, the paper begins, but the degree to which higher-order pattern recognition relies on it remains disputed; the study then demonstrates the persistence of oddball discrimination, semantic processing, and online prediction in individuals under general-anesthesia-induced loss of consciousness. The paper, titled "Plasticity and Language in the Anaesthetized Human Hippocampus" and published May 6 in Nature volume 654, was led by first author Dr. Kalman A. Katlowitz, a neurosurgery resident at Baylor, and senior author Dr. Sameer Sheth, professor and Cullen Foundation Endowed chair of neurosurgery there.
The experimental design was spare and carefully staged. Katlowitz and colleagues performed intraoperative recordings using high-density Neuropixels microelectrodes in seven patients with drug-resistant temporal lobe epilepsy undergoing anteromesial temporal lobectomy, recording from the anterior hippocampal body and isolating 651 neuronal units in total. Neuropixels microelectrodes captured very high-resolution data from individual neurons and had not been used on the hippocampus before this study. In the first experiment, the team played repeating tones occasionally interrupted by an "oddball" sound; hippocampal neurons and local oscillations retained some detection of oddball tones, and this effect size grew over the course of roughly ten minutes, demonstrating representational plasticity. The team then escalated the stimulus. Surgeons played episodes of the Moth Radio Hour, a storytelling podcast; neurons in the hippocampus responded to spoken language in a way that looked very similar to what the team would have expected from waking humans, firing in patterns that suggested calculations about what kind of word would come next, verb, noun, and so on. Those results extend prior work by demonstrating change over a timescale of several minutes normally associated with wakeful learning, linkage to a plausible biocomputational model that avoids executive control presumably diminished during anesthesia, and availability of language information beyond the level of auditory processing.
The implications reach well beyond the operating room. Global workspace theory proposes that conscious processing results from coherent neuronal activity between widely distributed brain regions, predicting that transitions between conscious and nonconscious states involve abrupt, nonlinear changes in the level of coherence across that distributed network. The Baylor data complicate that picture: the results suggest that awake-like semantic responses, and at least some contextualization, can occur in the absence of conscious awareness, meaning the hippocampus executes what had been considered signature conscious operations without any access to whatever global broadcast is supposed to make experience possible. It is further proof, as one analyst put it, that even fairly complex neural processes are not the same as consciousness. The authors themselves resist overclaiming: they note that consciousness may depend on broader coordination across brain regions rather than activity within a single structure like the hippocampus.
Skeptics have raised pointed methodological concerns that the authors largely acknowledge. The investigators emphasized that the study involved a small cohort of patients with epilepsy undergoing surgery, used propofol-based total intravenous anesthesia only, and may not generalize to other anesthetics or nonconscious states such as sleep or coma; they also noted that the observed processes may not be unique to the hippocampus and that the study could not determine whether changes in tone-identity encoding reflected compensatory plasticity or independent adaptive effects. Dr. Kirill Nourski, a professor of neurosurgery at the University of Iowa, told Time that the findings are surprising and that sophisticated processing of sound signals happening under anesthesia in the hippocampus is not something he, or others in the field, would have predicted. The practical frontier the team is already eyeing: the work may contribute to future communication technologies, including speech prosthetics, and Katlowitz specifically asks whether these signals could be used to run a speech prosthetic for parts of the brain damaged by stroke or injury.
In short: a brain that is, by every clinical measure, switched off can apparently still read the room, which means the clinical definition of "off" needs urgent revision.