Hippocampal neurons carried word-type information in patients under general anesthesia

The improvement in the tone test is the easier result to interpret. A response that sharpens over a session is not a fixed reflex to sound onset, and the researchers read it as neural plasticity ^[8].

The story result needs more care. The researchers say the word-category patterns suggest the hippocampus was responding to meaningful features of language ^[11]. Sheth, a neurosurgeon at Baylor St. Luke's Medical Center ^[13], put the prediction finding this way: "The brain appears to anticipate what comes next in a story, even without conscious awareness," ^[15] said Sheth.

That sentence is a reading of a decoding result. Information that can be used to predict upcoming words ^[12] is something an analyst extracts from recorded neurons. The decoding does not show whether the brain used that information for anything.

The Baylor summary of the Nature paper ^[2] does not give the number of patients, how anesthetic depth was confirmed, or whether awake patients served as a comparison. Those details decide how strong the word-category result is, since hundreds of neurons ^[4] can come from few patients or many.

Sheth said: "Our findings show that the brain is far more active and capable during unconsciousness than previously thought." ^[14] We think the data as described support a narrower sentence. Under general anesthesia, hippocampal neurons carry information about word class ^[10] and about what comes next ^[12]. If anesthetic depth held and the same pattern appears outside the hippocampus, the assumption that language processing needs consciousness would be hard to keep. From one structure, it is a suggestive result.

The recording method may end up mattering more than the finding. The researchers say Neuropixels probes had not previously been used to investigate this region in this way ^[5].