Processing Stages Underlying Word Recognition in the Anteroventral Temporal Lobe

Processing Stages Underlying Word Recognition in the Anteroventral Temporal Lobe Halgren Wang Schomer Knake Marinkovic Wu Ulbert

Tag layer of the DHSS Meta-Tagging Project. The article text is reproduced from the staged copy named in sourceDesc and is not altered; the tags are held separately in standOff so that overlapping and nested spans are preserved.

Processing Stages Underlying Word Recognition in the Anteroventral Temporal Lobe 2006 NeuroImage 10.1016/j.neuroimage.2005.10.053 Halgren2006-processing-stages-underlying-word-recognitio Halgren2006-processing-stages-underlying-word-recognitio.txt, the project's staged copy 00c2997bbc32d0792cec5f4a707f374c80d309ee not-redistributable neuroscience

the paper's own words

the tag layer; a sentence written by the tagger about the paper, always anchored to a quotation but not itself a quotation

0 spans located by searching the stored text for the quote's exact words, allowing any whitespace between them the sequence of processing stages a word passes through in the anteroventral temporal lobe, read from which cortical layer is active at each moment Quotation withheld because this record is not cleared for redistribution. whether the sustained activity seen with scalp and macroelectrode methods is one extended process or a sequence of distinct stages too close together to resolve Quotation withheld because this record is not cleared for redistribution. a laminar sequence - a middle-layer sink from about 120 ms, alternating middle and superficial foci until about 800 ms - read as feedforward input giving way to feedback and associative input, with repetition suppressing the later stages laterally and enhancing them medially Quotation withheld because this record is not cleared for redistribution. mechanism Quotation withheld because this record is not cleared for redistribution. word recognition and repetition effects in the anteroventral temporal lobe Quotation withheld because this record is not cleared for redistribution. electrophysiology electrophysiology Quotation withheld because this record is not cleared for redistribution. mechanisms successive laminar sinks track a shift in which afferents dominate: first feedforward input from lower visual areas into middle layers, then local and top-down associative input onto superficial layers Quotation withheld because this record is not cleared for redistribution. mechanisms the first feedforward volley is too brief to unblock NMDA receptors, so it cannot support plasticity and therefore cannot show repetition effects Quotation withheld because this record is not cleared for redistribution. mechanisms repetition effects emerge from sustained polysynaptic interaction across an extended cortical network rather than from local plasticity at one synapse Quotation withheld because this record is not cleared for redistribution. mechanisms the late component of repetition suppression is produced by withdrawal of excitatory drive onto deep cortical layers, seen as a simultaneous collapse of the deep sink and the deep-layer firing Quotation withheld because this record is not cleared for redistribution. mechanisms the lateral-to-medial reversal of the repetition sign marks a division of labor, with lateral avTL supporting facilitated (primed) processing and entorhinal cortex supporting active recollective retrieval Quotation withheld because this record is not cleared for redistribution. an initial sink in the middle layers is read as feedforward input from lower visual areas, because those layers are where such input terminates Quotation withheld because this record is not cleared for redistribution. and the concurrent increase in firing is what makes it excitatory rather than inhibitory, since a sink alone does not decide that Quotation withheld because this record is not cleared for redistribution. the later superficial focus is read as local and feedback associative input, which is where those afferents terminate Quotation withheld because this record is not cleared for redistribution. and the medial structure shows two or three sinks at once, which may follow from its superficial and deep layers being anatomically distinct and physiologically somewhat isolated Quotation withheld because this record is not cleared for redistribution. Only one laminar probe per subject was used, so the entorhinal repetition enhancement might be idiosyncratic to that patient. Quotation withheld because this record is not cleared for redistribution. The entorhinal repetition-enhancement finding requires further replication. Quotation withheld because this record is not cleared for redistribution. The CSD analysis is probably blind to synaptic activity on layer IV stellate cells. Quotation withheld because this record is not cleared for redistribution. behavioral data for the Size Judgment task, the main implicit priming task, were lost. Quotation withheld because this record is not cleared for redistribution. the electrode targets and the duration of implantation were decided on clinical grounds with no reference to this experiment Quotation withheld because this record is not cleared for redistribution. the earlier non-invasive results remain consistent with a different account, because fields from separate sources superimpose in those recordings Quotation withheld because this record is not cleared for redistribution. the relation of unit responses in macaques to field potentials in humans is unknown, for three separate reasons Quotation withheld because this record is not cleared for redistribution. and the relationship between synaptic currents, action potentials and haemodynamics may vary between tasks and locations Quotation withheld because this record is not cleared for redistribution. the finding that feedforward and feedback information flow tends to involve different cortical layers, which is what lets the laminar data lead to hypotheses about the sequence of network interactions Quotation withheld because this record is not cleared for redistribution. the forward model that turns a laminar current pattern into a predicted surface potential Quotation withheld because this record is not cleared for redistribution. the electrodes themselves, placed for a purpose unrelated to the question Quotation withheld because this record is not cleared for redistribution. the sequential-then-simultaneous account of ventral stream activation, and the dispute about it Quotation withheld because this record is not cleared for redistribution. the learning-versus-retrieval comparison re-run on NEW words only - new words during learning against new words during recognition testing eliminates: repetition as the explanation of the larger retrieval response - if only new words are compared, nothing in either condition has been seen before · outcome (stated by the tagger, no quotation): held Quotation withheld because this record is not cleared for redistribution. a test with the formal properties of the memory tasks - visual presentation, key-press response - but with no words and no word repetition eliminates: the stimulus-response furniture of the testing situation: that entorhinal activity reflects looking at something and pressing a key rather than remembering · outcome (stated by the tagger, no quotation): held Quotation withheld because this record is not cleared for redistribution. inferotemporal inferotemporal Quotation withheld because this record is not cleared for redistribution. firing firing Quotation withheld because this record is not cleared for redistribution. repetition repetition Quotation withheld because this record is not cleared for redistribution. word word Quotation withheld because this record is not cleared for redistribution. THE SAMPLE WAS CHOSEN BY SOMEONE ELSE'S PURPOSE: the decision to implant, the electrode targets and the duration of implantation were made ENTIRELY ON CLINICAL GROUNDS WITHOUT REFERENCE TO THIS EXPERIMENT Quotation withheld because this record is not cleared for redistribution. THREE AMBIGUITIES IN THE STANDARD MEASURE, ENUMERATED: it cannot tell sustained widespread activity from separate areas too close to resolve, it cannot tell inhibition from excitation so simultaneous activity could be either, and the presynaptic cell of origin is difficult to infer Quotation withheld because this record is not cleared for redistribution. THE CORTICAL LAYER IS THE INFERENCE: since feedforward and feedback information flow tends to involve different layers, the laminar data lead to hypotheses about which is active Quotation withheld because this record is not cleared for redistribution. and two measures are combined to remove an ambiguity neither removes alone - the sink says where, the simultaneous firing says that it is excitatory Quotation withheld because this record is not cleared for redistribution. A FORWARD MODEL THAT JOINS THREE INSTRUMENTS AT THREE SCALES: the laminar pattern is used to PREDICT the surface potential, giving a positive peak near 200 ms, a variable one near 400 and a negative one near 600 Quotation withheld because this record is not cleared for redistribution. and the prediction matches what macroelectrodes actually record, so non-invasive magnetic recording, intracranial macroelectrodes and laminar arrays give a CONSISTENT MULTI-RESOLUTION VIEW of the same component Quotation withheld because this record is not cleared for redistribution. THE SIGN OF A PUBLISHED COMPONENT DEPENDS ON WHICH SIDE OF THE TISSUE THE ELECTRODE IS: the same potential is negative above the ventral surface and positive when recorded by subdural strips immediately below it, and the geometry explains both Quotation withheld because this record is not cleared for redistribution. THE RELATION BETWEEN TWO MEASURES IS NOT FIXED: here the sustained response carries the effects seen with haemodynamics, while in another area it is the phasic response that matches - so the relationship MAY VARY BETWEEN TASKS AND LOCATION Quotation withheld because this record is not cleared for redistribution. THREE REASONS THE CROSS-SPECIES COMPARISON CANNOT SETTLE THE DISPUTE: the physiological measures differ, macaques lack language, and the areas expanded substantially in evolution Quotation withheld because this record is not cleared for redistribution. AND THE EARLIER RESULTS SURVIVE A RIVAL READING: because fields from different sources superimpose in magnetic and macroelectrode recordings, the same data are consistent with the temporal components arising in different regions Quotation withheld because this record is not cleared for redistribution. THE SAME-COLUMN CLAIM IS LICENSED BY A SIMULATED DECAY CONSTANT: amplitude falls tenfold in about 250 micrometers, which is what lets multiple stages be called properties of one microdomain Quotation withheld because this record is not cleared for redistribution. AND THE SAME TRANSITION TAKES 500 MS HERE AND 20 MS IN INTERICTAL SPIKES - the same tissue and the same measure, twenty-five times apart depending on what drives it Quotation withheld because this record is not cleared for redistribution. OPPOSITE-SIGNED EFFECTS IN ADJACENT STRUCTURES: repetition suppresses the lateral areas and enhances the medial one, which is read as incidental priming against active retrieval Quotation withheld because this record is not cleared for redistribution. and the effect is absent from the first 100 ms, which is taken to mean it arises from sustained interaction across a network rather than focal plasticity in one antecedent cell group Quotation withheld because this record is not cleared for redistribution. with the model systems declared possibly unable to show it - the human interactions may be selective and information-specific in a way that electrical stimulation in anesthetized animals or in vitro cannot replicate Quotation withheld because this record is not cleared for redistribution. claim_type: mechanism a second reading reaching the same value. Recorded: mechanism — note: a second reading reaching the same value from a different sentence. Not applied - a person decides. Quotation withheld because this record is not cleared for redistribution. method: current source density and multiunit activity resolved across cortical layers, with the laminar pattern used to predict what the surface potential should look like a second reading DIFFERING from the recorded value. Recorded: electrophysiology — note: a second reading, verbatim-verified, disagreeing with the value on the record. Not applied - a person decides. Quotation withheld because this record is not cleared for redistribution. phenomenon: the multi-stage cortical response to a word, and its repetition effects a second reading DIFFERING from the recorded value. Recorded: word recognition and repetition effects in the anteroventral temporal lobe — note: a second reading, verbatim-verified, disagreeing with the value on the record. Not applied - a person decides. Quotation withheld because this record is not cleared for redistribution. replicates → Miller and Desimone, 1994; Xiang and Brown, 1998 (macaque repetition suppression) Quotation withheld because this record is not cleared for redistribution. builds-on → Halgren et al., 1994; Smith et al., 1986 (intracranial macro-electrode avTL recordings) Quotation withheld because this record is not cleared for redistribution. distinguishes-from → scalp and macroelectrode recordings, which cannot resolve whether sustained activity is one process or several differs in: THE AXIS IS WHAT THE MEASUREMENT CAN RESOLVE, NOT WHAT HAPPENED. 'Quotation withheld.' Two accounts that predict the same recording, separated by an instrument nobody has - the missing-experiment shape, stated as a resolution limit. Quotation withheld because this record is not cleared for redistribution. supports → the canonical cortical circuit, in which a middle-layer sink is followed by a superficial one across modalities and species Quotation withheld because this record is not cleared for redistribution. confirms → the repetition suppression of firing reported in macaque temporal cortex and of haemodynamic activity in humans Quotation withheld because this record is not cleared for redistribution. reconciles → the opposite polarities recorded above and below the ventral temporal surface, by deriving both from one laminar pattern Quotation withheld because this record is not cleared for redistribution. qualifies → the assumption that a fixed relationship holds between synaptic currents, spikes and haemodynamics differs in: THE AXIS IS THE TIME WINDOW. 'The repetition-induced decrease in IT and PR reported here Quotation withheld.' The fixed relationship between synaptic currents and spiking is qualified for one epoch and left standing for the rest. Quotation withheld because this record is not cleared for redistribution. extends → the guinea pig laminar result that the initial sink is monosynaptic and the later one polysynaptic, to human word processing Quotation withheld because this record is not cleared for redistribution. sink —part-of→ layer sink layer strength of the relation: 1 · note: AND TWO MEASURES REMOVE AN AMBIGUITY NEITHER REMOVES ALONE: the sink says where, the SIMULTANEOUS FIRING says it is excitatory Audit 2026-09-28: DOUBTFUL - read against its span and kept, but the span localizes the initial sink to middle layers; part-of 'layer' is loose. Quotation withheld because this record is not cleared for redistribution. entorhinal —causes→ retrieval entorhinal retrieval strength of the relation: 1 Audit 2026-09-28: DOUBTFUL - read against its span and kept, but the span is cut before entorhinal activity is linked to retrieval. Quotation withheld because this record is not cleared for redistribution. EPSC —operationalised-as→ sink EPSC sink strength of the relation: 1 · note: a phasic middle-layer current first, then long-lasting ones first in the middle and later in the superficial layers Audit 2026-09-28: DOUBTFUL - read against its span and kept, but the span interprets the sink as reflecting EPSCs; operationalised-as is loose. Quotation withheld because this record is not cleared for redistribution. CSD —causes→ MEG CSD MEG strength of the relation: 1 · note: A FORWARD MODEL THAT JOINS THREE INSTRUMENTS: the laminar pattern is used to PREDICT the surface potential, positive near 200 ms, variable near 400, negative near 600 Quotation withheld because this record is not cleared for redistribution. column —operationalised-as→ CSD column CSD strength of the relation: 1 Audit 2026-09-28: DOUBTFUL - read against its span and kept, but the span says column ERPs can be estimated from laminar CSD; operationalised-as is loose. Quotation withheld because this record is not cleared for redistribution. distance —causes→ CSD distance CSD strength of the relation: 1 · note: THE SAME-COLUMN CLAIM IS LICENSED BY A SIMULATED DECAY CONSTANT: tenfold in about 250 micrometers Audit 2026-09-28: DOUBTFUL - read against its span and kept, but the span says CSD declines steeply with distance; the edge carries no direction. Quotation withheld because this record is not cleared for redistribution. repetition —causes→ firing repetition firing holds when: regardless of whether repetition was task-relevant; the effect was largest when repetition was implicit · direction on the measure: reduces · strength of the relation: 1 · note: and the effect is LARGEST when the repetition is implicit, presumably because encoding was deeper then Quotation withheld because this record is not cleared for redistribution. t-test t-test Quotation withheld because this record is not cleared for redistribution. one-sample t-test one-sample t-test Quotation withheld because this record is not cleared for redistribution. statistical-significance statistical-significance Quotation withheld because this record is not cleared for redistribution. standard deviation standard deviation Quotation withheld because this record is not cleared for redistribution. CSD CSD Quotation withheld because this record is not cleared for redistribution. simulations simulations Quotation withheld because this record is not cleared for redistribution. In all tasks, word-evoked activity began with a sharp current sink in putative layer IV arising abruptly from baseline. Quotation withheld because this record is not cleared for redistribution. Evoked activity was identical to new and old words for the first ~100 ms. Quotation withheld because this record is not cleared for redistribution. Repetition effects appeared only from ~220 ms onward, as a prolongation of the layer IV sink to new words. Quotation withheld because this record is not cleared for redistribution. Repetition suppressed the layer IV synaptic sink in IT and the deep-layer firing in PR. Quotation withheld because this record is not cleared for redistribution. Entorhinal cortex showed the opposite sign of effect: synaptic currents and firing were larger to repeated words. Quotation withheld because this record is not cleared for redistribution. Entorhinal current sinks were much smaller during learning than during retrieval. Quotation withheld because this record is not cleared for redistribution. Multiple processing stages over hundreds of ms are generated within the same IT, PR and EC microdomains, roughly the size of a cortical column. Quotation withheld because this record is not cleared for redistribution. Simultaneously increasing multi-unit firing showed that the recorded current sinks are excitatory synaptic currents rather than passive returns of an inhibitory current elsewhere. Quotation withheld because this record is not cleared for redistribution. The MUA response in the visuomotor task was a brief decrease from baseline, while during word recognition it was a prolonged increase. Quotation withheld because this record is not cleared for redistribution. Repetition effects in PR were similar whether repeated words were recognition targets or incidental to a size-judgment task. Quotation withheld because this record is not cleared for redistribution. different extrinsic and associative inputs target different cortical layers, which is what makes layer resolution into stage resolution Quotation withheld because this record is not cleared for redistribution. the initial sink begins about 120 ms and peaks about 170, and the focus alternates between middle and superficial layers until about 800 ms Quotation withheld because this record is not cleared for redistribution. AND TWO MEASURES REMOVE AN AMBIGUITY NEITHER REMOVES ALONE: the sink says where, the simultaneous firing says it is excitatory Quotation withheld because this record is not cleared for redistribution. repetition does NOT affect the initial middle-layer sink and DOES decrease the later activity Quotation withheld because this record is not cleared for redistribution. THREE AMBIGUITIES IN THE STANDARD MEASURE, ENUMERATED: one process or several too close to resolve, inhibition or excitation, and which presynaptic cell Quotation withheld because this record is not cleared for redistribution. THREE REASONS THE CROSS-SPECIES COMPARISON CANNOT SETTLE THE DISPUTE: different measures, no language in macaques, and evolutionary expansion of the areas Quotation withheld because this record is not cleared for redistribution. THE CORTICAL LAYER IS THE INFERENCE: since the two modes involve different layers, the laminar data lead to hypotheses about which is active Quotation withheld because this record is not cleared for redistribution. with repetition studied both overtly and implicitly, embedded inside a semantic task Quotation withheld because this record is not cleared for redistribution. THE SAMPLE WAS CHOSEN BY SOMEONE ELSE'S PURPOSE: targets and duration decided entirely on clinical grounds without reference to this experiment Quotation withheld because this record is not cleared for redistribution. OPPOSITE-SIGNED EFFECTS IN ADJACENT STRUCTURES: repetition suppresses the lateral areas and enhances the medial one Quotation withheld because this record is not cleared for redistribution. a phasic middle-layer current first, then long-lasting ones first in the middle and later in the superficial layers Quotation withheld because this record is not cleared for redistribution. the same pattern is seen for visual, somatosensory and auditory input in animals, which is what makes the circuit canonical Quotation withheld because this record is not cleared for redistribution. with the animal work supplying the reading - initial sink monosynaptic, later sink polysynaptic Quotation withheld because this record is not cleared for redistribution. AND THE DELAY FOR THE SINK TO RISE FROM MIDDLE TO UPPER CORTICAL LAYERS WAS AS LONG AS ~500 ms in the current study, whereas in human interictal spikes it is ~20ms Quotation withheld because this record is not cleared for redistribution. two or three sinks are present at once in the medial structure, whose superficial and deep layers are separated by a cell-free lamina Quotation withheld because this record is not cleared for redistribution. with the model systems declared possibly unable to reproduce it Quotation withheld because this record is not cleared for redistribution. A FORWARD MODEL THAT JOINS THREE INSTRUMENTS: the laminar pattern predicts a positive surface peak near 200 ms, a variable one near 400 and a negative one near 600 Quotation withheld because this record is not cleared for redistribution. THE SIGN OF A PUBLISHED COMPONENT DEPENDS ON WHICH SIDE OF THE TISSUE THE ELECTRODE IS, and one laminar pattern explains both Quotation withheld because this record is not cleared for redistribution. AND THE PREDICTION HOLDS: three methods at three scales give a consistent multi-resolution view of the same component Quotation withheld because this record is not cleared for redistribution. AND THE EARLIER RESULTS SURVIVE A RIVAL READING, because fields from different sources superimpose Quotation withheld because this record is not cleared for redistribution. THE SAME-COLUMN CLAIM IS LICENSED BY A SIMULATED DECAY CONSTANT: tenfold in about 250 micrometers Quotation withheld because this record is not cleared for redistribution. and other structures are active throughout, so the column is one participant in an extended network Quotation withheld because this record is not cleared for redistribution. the repetition effect was largest when repetition was implicit, presumably because encoding was deeper Quotation withheld because this record is not cleared for redistribution. THE RELATION BETWEEN TWO MEASURES IS NOT FIXED: the sustained response carries the effects here, the phasic one in another area, so the relationship may vary between tasks and location Quotation withheld because this record is not cleared for redistribution. anteroventral temporal lobe (avTL) anteroventral temporal lobe (avTL) Quotation withheld because this record is not cleared for redistribution. CSD analysis CSD analysis Quotation withheld because this record is not cleared for redistribution. population neuronal firing population neuronal firing Quotation withheld because this record is not cleared for redistribution. current sink current sink Quotation withheld because this record is not cleared for redistribution. N400 N400 WHOSE RESULT THIS IS: studies typically find - THE SUBJECT OF THIS RESULT IS AN INDEFINITE STUDY, NOT THIS PAPER. 'Previous studies reported', 'studies in guinea pig showed', 'recent studies have reported' - somebody else's work under a common noun rather than a name, with a citation in the sentence and no first-person marker. The quote is verbatim and the claim is true; only the attribution is wrong. FOUND BY MEASURING MY OWN CUE'S RECALL: the first sweep required a NAMED author as the subject ('Sperling (1960) demonstrated') and found 17. Reading eleven citation-bearing findings on 'Plakke2016' by hand showed four the cue could not see - 'A study examining recognition found', 'studies of typical human subjects have demonstrated', 'A follow up study found' - so the named cue's 17 was a floor and not a count. A CUE'S FALSE POSITIVES TELL YOU NOTHING ABOUT ITS FALSE NEGATIVES, and the only way to find these was to read a record's whole layer. Quotation withheld because this record is not cleared for redistribution. ‘countercurrent’ pattern ‘countercurrent’ pattern Quotation withheld because this record is not cleared for redistribution. CSD CSD Quotation withheld because this record is not cleared for redistribution. sink sink Quotation withheld because this record is not cleared for redistribution. N400 N400 Quotation withheld because this record is not cleared for redistribution. column column Quotation withheld because this record is not cleared for redistribution. cortical layer cortical layer Quotation withheld because this record is not cleared for redistribution. CSD CSD Quotation withheld because this record is not cleared for redistribution. population neuronal firing population neuronal firing Quotation withheld because this record is not cleared for redistribution. sink sink Quotation withheld because this record is not cleared for redistribution. latency latency Quotation withheld because this record is not cleared for redistribution. polarity polarity Quotation withheld because this record is not cleared for redistribution. repetition repetition Quotation withheld because this record is not cleared for redistribution. 111 of 120 — repeating words correctly detected by patient 2 in the Word Recognition task, at a mean reaction time of 762 ms Quotation withheld because this record is not cleared for redistribution. ~120ms — latency at which the laminar response first differed significantly from baseline Quotation withheld because this record is not cleared for redistribution. ten fold in ~250μm — simulated fall-off of CSD amplitude with distance, the basis for claiming column-level localization Quotation withheld because this record is not cleared for redistribution. ~250μm — the distance over which the signal falls tenfold - the simulated constant that licenses calling multiple stages properties of one column Quotation withheld because this record is not cleared for redistribution. ~500 ms — the delay for the sink to rise from middle to upper cortical layers in the current study, whereas in human interictal spikes it is ~20ms Quotation withheld because this record is not cleared for redistribution. ~120ms — the onset of the initial middle-layer sink, peaking near 170 ms Quotation withheld because this record is not cleared for redistribution.

The full text of this article is not redistributable, so it is not included here. The evidence sentences are withheld with it; the tag layer and its grounding status remain. The article itself is at https://doi.org/10.1016/j.neuroimage.2005.10.053.