The somatic marker hypothesis and the possible functions of the prefrontal cortex

Study at a glance

• Authors: Antonio R. Damasio
• Year: 1996
• Journal: Philosophical Transactions of the Royal Society of London B: Biological Sciences, 351(1346), 1413-1420 — special issue, ‘Executive and Cognitive Functions of the Prefrontal Cortex’
• DOI: 10.1098/rstb.1996.0125
• N: Gambling experiment: 44 normal controls (21 women, 23 men; ages 20–79) vs. 6 ventromedial prefrontal patients (2 women, 4 men; ages 43–84); ~half high-school and half college educated in each group. SCR-to-emotional-images experiment: three groups (normal controls, non-frontal lesions, ventromedial lesions) — n’s not reported.
• Design: Single-author theoretical statement with summarized empirical support, published in a Royal Society discussion-meeting issue and followed by a transcribed discussion with B. J. Everitt and D. Bishop. No new data; the experiments are restatements of Damasio et al. (1990, 1991), Bechara et al. (1994, 1996), and Anderson et al. (1996). Supported by NIH NINDS P01 NS19632. JSTOR stable URL: https://www.jstor.org/stable/3069187 — no DOI printed on the scan.

Key findings

• The wiki’s earliest primary statement of the Somatic marker hypothesis, nine years before Bechara & Damasio (2005). Same core: marker signals arising in bioregulatory processes bias response selection at overt (conscious) and covert (non-conscious) levels. ‘Somatic’ is already chosen for the same stated reason — the markers relate to body-state structure and regulation even when they do not arise in the body proper.
• The as-if body loop is the stated default, not an escape hatch. Damasio holds the body loop is the original mechanism from both evolutionary and ontogenetic perspectives, but has been superseded by the as-if loop and is used less frequently than it. This is the reverse of the emphasis the wiki inherited from the 2005 paper, where the body loop governs decisions under ambiguity. See Body loop and as-if body loop — this is the ingest’s most consequential finding.
• Asked directly whether manipulating the periphery would change behaviour, Damasio says it should not much. Everitt’s discussion question puts it point-blank; the reply is that in the more frequently operational (as-if) mode, changes in the periphery ought to have relatively little impact on central functions related to emotion/feeling and reasoning/decision making. The qualification offered is developmental — body states are actively engaged during development and in non-average situations — plus a concession that cord-damage and peripheral-neuropathy evidence suggests some peripheral effect.
• No primary/secondary inducer distinction anywhere in the paper. This confirms from the primary source what the 2005 study page asserts on internal evidence: the two-trigger architecture is a later addition (Damasio 1995; fullest in Bechara, Damasio & Damasio 2003), not part of the 1996 framework.
• The amygdala’s role is the opposite of its 2005 role. Here it is a somatic effector — one of the central autonomic effectors that ventromedial cortices activate downstream, alongside hypothalamus and brainstem nuclei. In 2005 it is the trigger structure for primary inducers, upstream and parallel to VM. Same structure, inverted position in the causal chain, across nine years and no acknowledgement. Recorded on Amygdala.
• Specifies the network in three parts: (1) ventromedial frontal convergence zones recording links between dispositions for categorized situations and dispositions for their associated somatic states; (2) central autonomic effectors (amygdala, and structures acting on viscera, vascular bed, endocrine system, non-specific neurotransmitter systems); (3) somatosensory cortices — insula, SII, SI — and their interlocking projections, especially in the non-dominant hemisphere. Basal ganglia possibly mediate via somatomotor structures (Tranel & Damasio 1993).
• Grounds the whole architecture in dispositional knowledge and the convergence-zone framework (Damasio 1989a, b; Damasio & Damasio 1994): knowledge is stored coded, implicit, and non-topographically, and made explicit either as motor responses (some combinations of which constitute emotions) or as images. Four background assumptions and a four-way knowledge taxonomy (A: bioregulatory/body states, B: entities and facts, C: linkages between B and A, D: categorizations of A/B/C) do the framing work that the 2005 paper drops entirely.
• Markers act on process as well as content. Emotion-related somatosensory patterns are proposed to boost attention and working memory, and to assist response inhibition — so a somatic marker failure could destabilize the representation of future outcomes rather than merely fail to tag it. Damasio raises this as a live alternative account of the gambling deficit, invoking dorsolateral prefrontal networks (Goldman-Rakic 1987).
• Somatic markers constrain the decision space rather than making the decision. Their function is to render the space manageable for logic-based cost-benefit analysis by rapidly rejecting or endorsing option-outcome pairs. Absent the marker, options and outcomes become virtually equalized and choice falls back entirely on logic over many pairs — necessarily slower, possibly ignoring prior experience. Slow-and-error-prone and random-and-impulsive deciding are both predicted.
• Predicts a dissociation the wiki should note as a LeDoux link. Somatic markers for unconditioned, basic stimuli (a startling noise, a flash of light) use a different and simpler network that bypasses cerebral cortex altogether and activates autonomic centres directly from thalamus — cited to Clugnet et al. (1988) and Farb, Ruggiero & LeDoux (1988), i.e. LeDoux’s own ‘low road’. Damasio states the formulation predicts a dissociation between responses to complex stimuli requiring cortical processing and basic stimuli not requiring it.
• SCR to emotionally charged images (Damasio et al. 1990, 1991; Tranel 1994; Tranel et al. 1995). Three groups — normal controls, non-frontal lesions, ventromedial lesions. All three had normal SCRs to startle stimuli and deep breath, establishing intact autonomic responsivity. Only the ventromedial group failed to react to emotionally charged images (social catastrophe, body mutilation) while responding to neutral ones (landscapes, abstract patterns) at floor like everyone else. The deficit is in emotional evocation, not autonomic capacity.
• The gambling task, with the n’s the 2005 review omits. 44 controls vs. 6 VM patients. Normals and non-frontal patients sample all decks, shift to the good decks about halfway through, and never abandon the strategy; VM patients keep playing the bad decks despite repeated losses until they lose the loan and must borrow. Deck structure given in full (A/B = $100percard,C/D=$50; penalties higher in A/B; A and B differ in punishment frequency vs. magnitude, as do C and D).
• Anticipatory SCR (Bechara et al. 1996). In the ~4 s window preceding a selection, normals generate high-amplitude SCRs before bad decks only, rising in amplitude as the task unfolds; VM patients generate none, while retaining normal SCRs to actual loss. Damasio reports that normals begin producing anticipatory SCRs to bad decks long before they can testify to any notion of the decks’ nature — the pre-hunch claim, stated here without the four-period knowledge probe that the 2005 review rests it on (that is Bechara et al. 1997, still a year away).
• Three candidate accounts of the VM deficit are named and two are ruled out. Hypersensitivity to reward, insensitivity to punishment, and general insensitivity to future consequences. Anderson et al. (1996) reversed the task design — punishment up front, unpredictable reward schedules — and the patients behaved the same, which cuts against the first two. Damasio’s preferred account is absence of covert and overt markers for future scenarios, failing to supply ‘positive’ signals that would counteract ‘negative’ ones.
• Damasio pre-empts the Maia & McClelland objection and rejects it on evolutionary grounds. He names the alternative — that normals first reason out which decks are bad and then generate an SCR indexed to that cognition — and finds it less plausible because of his view of the adaptive value of an automated marker device, plus the pre-hunch timing. Notably, the evolutionary argument is doing real work here; the empirical rebuttal is asserted from unpublished lab observations.
• Discussion (Bishop): the conditioning dissociation. Asked whether the gambling deficit is a failure to inhibit an earlier association rather than a failure to form associations, Damasio concedes the inhibition reading is conceivable but reports that most patients who fail the gambling task acquire classical conditioning normally — implying dissociable aspects of conditioning by task type. Relevant to Pavlovian defense conditioning.
• Explicitly disclaims scope: the hypothesis concerns the ventromedial sector, does not necessarily apply to prefrontal cortex as a whole, and should not be read as unifying frontal lobe function under a single mechanism. Patients with non-ventromedial frontal lesions may or may not be accountable by it.

Limitations

• Not an empirical paper, and more openly so than the 2005 one. It follows closely, in Damasio’s own framing, several prior reviews (Damasio 1994, 1995a), and the experiments are restatements. Two load-bearing citations are Society for Neuroscience abstracts or in-press (Anderson et al. 1996; Clugnet et al. 1988; Farb et al. 1988), and the claim that normals produce anticipatory SCRs before any conscious notion is sourced to ‘recent studies in our laboratory’ with no citation at all.
• The gambling groups are badly matched on age and size. 44 controls aged 20–79 against 6 patients aged 43–84. The control range extends two decades younger and the patient group is seven times smaller; no age-matched subgroup analysis is reported, and no statistics of any kind accompany the result. Since this wiki separately records that felt bodily emotion dampens with age, an age-confounded somatic-response comparison is not a small problem. The 2005 review inherits this evidence base and reports it as figures without n’s.
• The as-if default guts the paper’s own Jamesian credentials, and nothing here notices. If the periphery is bypassed in most circumstances and peripheral change ought to have little central impact, then the somatic marker hypothesis as stated in 1996 is a theory in which the body is a developmental scaffold that is subsequently dispensed with. Friedman (2010) files Damasio as Neo-Jamesian on the strength of bodily feedback preceding awareness and guiding behaviour. That filing is harder to sustain against this text than against the 2005 one. See Body loop and as-if body loop, William James.
• The covert-bias mechanism is a gesture. Covert somatic states are said to bias cognition through a non-specific neurotransmitter system ‘such as dopamine’ — offered twice, both times as an example rather than a claim, with no evidence and no alternative considered. The 2005 paper’s three-level architecture (striatum/dopamine covert, ACC/serotonin overt, lateral OFC-DLPFC conscious-pre-action) is nowhere here; what exists in 1996 is one unspecified neuromodulator standing in for the entire covert route.
• Interoception is assumed and never measured — the same gap as 2005, and the reason this framework has less purchase on this wiki than its citation count implies. The insula/SII/SI block is where somatic patterns are received and become feelings, so the architecture depends on interoceptive representation throughout, yet no interoceptive measure appears and individual differences are not considered. Compounding it: the whole apparatus rests on SCR, an efferent measure that records what the body did, not what the subject perceived.
• Autonomic evidence is skin conductance only, so the paper cannot address Autonomic specificity of emotion — but note it also does not need to in 1996, because the Background somatic states signal-to-noise model that requires valence discrimination does not yet exist. The 1996 framework is the version of the hypothesis whose evidence and whose claims are best matched to each other.
• The somatosensory laterality claim is dropped without trace. The 1996 network specifies insula/SII/SI projections ‘especially in the non-dominant hemisphere’ — a real anatomical commitment, stated without supporting citation, and absent from the 2005 paper, which instead lateralizes at the VM (left ↔ positive, right ↔ negative) on entirely different evidence. Two different laterality claims about two different structures across the same programme, neither reconciled with the other nor with Craig’s right-AIC asymmetry.
• The process/content claim is raised and abandoned. That somatic markers boost attention and working memory is stated as belief (‘I believe they may also assist with response inhibition’) and then invoked as an alternative explanation for the gambling deficit — meaning the paper contains two candidate mechanisms for its central result and does not choose between them or propose a test that would.
• The Discussion is more revealing than the article and is the reason to prefer this source over a later restatement. Everitt’s and Bishop’s questions are the two sharpest challenges to the framework — does the body matter, and is this response inhibition rather than marking — and both answers are concessive in ways the published reviews are not. This material exists only in the Royal Society format; it is not carried forward anywhere else in the wiki’s sources.

Connections

• Methods: Iowa Gambling Task, Skin conductance response (SCR), Pavlovian defense conditioning
• Concepts: Somatic marker hypothesis, Body loop and as-if body loop, Convergence zones and dispositional knowledge, Ventromedial prefrontal cortex (VM), Amygdala, Insular cortex, Primary and secondary inducers, Interoception, Autonomic specificity of emotion, Embodied selfhood, Survival circuits, Simulation map
• Debates: Does somatic feedback guide decisions?, Do basic emotions have distinct autonomic signatures?, parietal workspace?, inferential?
• Related studies: Bechara & Damasio (2005), Bechara et al. (2000), Dunn et al. (2006), Friedman (2010), Craig (2002), LeDoux (2012), Lindquist et al. (2012)

The Somatic marker hypothesis nine years before Bechara & Damasio (2005), single-authored, in a Royal Society discussion-meeting issue on prefrontal function — and followed by a transcribed exchange with B. J. Everitt and D. Bishop that is the most valuable eight paragraphs in the wiki’s somatic marker material.

The wiki did not need another statement of this framework. What it needed, and did not know it needed, was a timepoint. Having two primary sources nine years apart turns a set of claims into a trajectory, and the trajectory runs in a direction the wiki’s existing pages do not anticipate. See Antonio R. Damasio.

What is already here, and what is not

Worth stating plainly before the analysis, because the wiki’s summary of this framework was built from the 2005 paper and it is easy to read the earlier one as merely thinner.

component	1996	2005
marker signals bias response selection, overtly and covertly	yes	yes
”somatic” chosen over “emotional,” same stated reason	yes	yes
body loop / as-if body loop	yes — as-if is the default	yes — loops indexed to decision type
VM as convergence zone coupling situations to somatic states	yes	yes
convergence-zone framework	yes, foundational	absent
amygdala’s role	downstream somatic effector	upstream trigger for primary inducers
Primary and secondary inducers	absent	central
posterior→anterior VM gradients (time/abstractness/probability/valence)	absent	central
Background somatic states and the signal-to-noise model of mood	absent	central
neuroeconomics (delay discounting, prospect theory)	absent	central
three-level biasing (striatum/ACC/lateral OFC)	absent — one unnamed neuromodulator	central
markers boost attention and working memory	yes	dropped to a subordinate remark
insula/SII/SI laterality (non-dominant hemisphere)	yes	dropped; laterality moves to VM

The bottom half of that table is why the framework’s citation count grew. The third row is why this ingest matters.

The finding: in 1996, the body is optional by design

The wiki’s Body loop and as-if body loop page treats the as-if loop as the hypothesis’s structural weak point — the move that lets any decision showing no peripheral change be attributed to intra-cerebral simulation — and credits the 2005 indexing claim (ambiguity → body loop, certainty → as-if loop) with rescuing falsifiability. That framing is correct about the 2005 paper. It is the wrong shape for this one.

Damasio’s 1996 position is not that the as-if loop is available when the body is quiet. It is that the as-if loop is what the body loop evolved and develops into: the body loop is original from both evolutionary and ontogenetic perspectives, has been superseded, and is used less frequently than its replacement. The peripheral route is the ancestral implementation. The simulation is the upgrade.

And then Everitt asks the question the wiki has been waiting for someone to ask. To what extent do peripheral somatic changes contribute to the risk-taking in the gambling test — if you manipulated them experimentally, would behaviour change?

The answer is no, mostly. In the more frequently operational mode, changes in the periphery ought to have relatively little impact on central functions related to emotion/feeling and reasoning/decision making. Damasio qualifies it three ways — the role varies between developmental and adult phases; body states are actively engaged during development and in non-average situations; cord-damage and peripheral-neuropathy evidence suggests there may be some peripheral effect despite the obligate incompleteness of such lesions — and grants that the final answer requires further study.

This is the founder of the field’s most cited bodily theory of decision-making saying, on the record, in response to a direct experimental challenge, that the body is largely bypassed. Nothing in this wiki says it as clearly, and none of the wiki’s secondhand sources report it.

Three consequences, kept separate because they are separately arguable:

1. The Neo-Jamesian filing gets harder. Friedman (2010) files Damasio as Neo-Jamesian because bodily feedback precedes awareness and guides behaviour. Against the 1996 text, the feedback is usually not bodily. What survives is a claim about somatosensory representation preceding awareness — which is a claim about brain states that were once caused by bodies. James’s peripheralism is not obviously in there. See William James.

2. The 2005 indexing is a later concession, not a clarification. Read forward from 1996, the certainty→as-if / ambiguity→body mapping looks less like a specification of a standing view and more like a reversal of one: the body reacquires a job it had been retired from. The wiki should stop describing the indexing as “the constraint the framework always had, stated late.” It is a change of position. Whether Damasio regarded it as one is unknowable from these two sources — neither cites the other on this point, and 2005 does not mention that anything was revised.

3. The developmental qualification is the interesting part, and it is the wiki’s best unexploited link. Body states are actively engaged during development, then the periphery is bypassed. That is precisely the structure of primary → secondary induction as stated in 2005 (you must have felt pain to imagine it; once acquired, secondary induction becomes independent of primary) — which means the 1996 discussion already contains the developmental dependency nine years before the architecture built on it. It is also, and more interestingly, the structure of the Simulation map: a body model built from bodily experience and then run offline. Farb et al. and Damasio are describing the same developmental arc from different literatures, and the 2005 paper — which has the inducer architecture but not this framing — is the one place it could have been said and is not.

Convergence zones: the framework’s discarded foundation

The 1996 paper’s theoretical base is not somatic markers. It is dispositional knowledge and the convergence-zone framework (Damasio 1989a, b; Damasio & Damasio 1994), from which somatic markers are derived as a special case: the VM cortex holds records of temporal conjunctions of activity in other neural units, and a somatic marker is what you get when one of the conjoined units is a bioregulatory state rather than a sensory one.

This matters for two reasons.

It is the wiki’s answer to a question Ventromedial prefrontal cortex (VM) raises and cannot settle. That page notes the most promising unexplored link in this material — that triggering a somatic state from a thought and issuing a descending interoceptive prediction (Seth & Friston’s Visceromotor areas (VMAs)) may be one operation in two vocabularies. The convergence-zone framework is the missing middle term, and it makes the conjecture look better and worse at once. Better: a convergence zone is explicitly a re-activation device — it holds no content, only the potential to reactivate patterns elsewhere by acting on the appropriate cortical or subcortical structures, which is structurally a generative model. Worse: Damasio’s re-activation is retrodictive, reconstructing a state that belonged to a past conjunction, with no comparator, no prediction error, and no precision. See inferential?, where the somatic marker hypothesis still has no position — but now the reason is nameable rather than merely absent.

And it is gone by 2005. The convergence-zone scaffolding does not appear in the Games and Economic Behavior paper, which uses “convergence–divergence zone” as a phrase without the framework behind it. A theory that started as a corollary of a general architecture of memory ends as a free-standing account of decision-making. The wiki should record that the somatic marker hypothesis got less theoretically embedded as it got more famous.

The amygdala changes jobs

Flagged for Amygdala, where four incompatible readings of the structure are already recorded and this makes the fourth internally unstable.

In 1996 the amygdala appears in exactly one role: a central autonomic effector. Ventromedial cortices activate somatic effectors in amygdala, hypothalamus, and brainstem nuclei — the amygdala is downstream, an output stage that executes a somatic state on the viscera when VM tells it to.

In 2005 the amygdala is the trigger structure for primary inducers — upstream, parallel to VM, the fast obligatory route that lays down the patterns VM later re-activates. The whole double dissociation depends on it being there.

Same structure, inverted position in the causal chain, nine years, no acknowledgement in either text. Note the 1996 paper does already contain a fast subcortical route (thalamus → autonomic centres, bypassing cortex) for basic unconditioned stimuli, citing Clugnet et al. (1988) and Farb, Ruggiero & LeDoux (1988) — so the raw material for the 2005 role was in hand and was used for something else: a predicted dissociation between basic and complex stimuli, not a trigger architecture. The wiki should treat “the amygdala’s role in the somatic marker hypothesis” as a moving target rather than a fixed reading.

(The Farb of Farb, Ruggiero & LeDoux 1988 is Claudia Farb, LeDoux’s anatomist collaborator. No relation to Norman A. S. Farb. Noted because this wiki’s raw/ tree makes that collision likely to recur.)

What this does to the debate page

Does somatic feedback guide decisions? is organized around three questions that can come apart: is the signal causal, must it be felt, does perceiving it help. This paper moves the second one, and it moves it from an unexpected direction.

The debate page records the framework’s answer to “must it be felt” as no — the striatal level is covert, the anticipatory SCRs bias without awareness — and concludes that the hypothesis is a theory about somatic signalling with interoception left as an assumption. That stands. But 1996 adds a stronger and separate claim: not merely that the signal need not be felt, but that it need not be peripheral. Those are different retreats and the wiki had them merged.

A theory can be about unfelt bodily signals and still be a bodily theory — that is the version the debate page currently argues with, and SCR is an appropriate measure for it. A theory in which the periphery is bypassed in most circumstances is not a bodily theory at all; it is a theory about somatosensory cortex. The 1996 text is the second one. Recorded on the debate page as a fourth row in the escape-hatch table, and as the sharpest version of the concern that the as-if loop gives the game away — because here it is not an escape hatch being used, it is the mechanism being advertised.

Damasio also pre-empts the Maia & McClelland objection here, which the debate page notes the wiki holds only secondhand and cannot yet state responsibly. He names the alternative precisely — that normals reason out which decks are bad and generate the SCR from that cognition, so the marker indexes knowledge rather than preceding it — and rejects it on two grounds: his perspective on the evolutionary biology and adaptive value of an automated marker device, and unpublished lab observations that normals produce anticipatory SCRs before they can testify to any notion of the decks. The wiki still has not read Maia & McClelland. But it can now record that the objection is not one the framework failed to anticipate; it is one the framework answered in 1996 with an evolutionary argument and an uncited observation, eight years before the critique was published.

The evidence, at its actual size

Two experiments carry the paper, and this source gives numbers the 2005 review does not.

SCR to emotionally charged images. Three groups: normal controls, non-frontal lesions, ventromedial lesions. The control condition is the good part of the design — SCRs to startle stimuli (loud noises) and to deep breath were normal in all three groups, so the ventromedial group’s failure to respond to social-catastrophe and body-mutilation images is not an autonomic-capacity deficit. That is a clean dissociation and the paper does not oversell it. No n’s are reported.

The gambling task. 44 controls (21 women, 23 men, ages 20–79) against 6 ventromedial patients (2 women, 4 men, ages 43–84). Normals shift to the good decks about halfway through and never abandon the strategy; VM patients keep playing A and B until the loan is gone and they must borrow. Full deck structure is given (see Iowa Gambling Task).

Hold the sample in view. Six patients, seven times fewer than controls, with a control range extending two decades younger and no age-matched analysis and no statistics. This wiki records elsewhere that felt bodily emotion dampens steadily from 18 to 90 — so an age-confounded comparison of somatic responsivity is not a pedantic objection. It is the confound most likely to produce this exact result. That is not an argument that the finding is wrong; the effect is large, the patients’ real-life impairment is independently documented, and the Anderson et al. (1996) reversal makes the two obvious alternative accounts unlikely. It is an argument about what the finding is worth on its own, which is: less than fifteen years of citation implies, and roughly what a six-patient lesion study should be worth.

The Iowa Gambling Task page’s limitation that the 2005 review reports the dissociation “as figures without n’s” is now partly answered — the n’s exist, they are in the 1996 paper, and they are small.

A dissociation Bishop extracts and Damasio concedes

The second discussion exchange, and the one that most deserves a follow-up nobody in this wiki has run. Bishop’s question is sharp: subjects did not merely learn a stimulus-response mapping, they had to change one, since everyone starts on A and B because those decks pay more. So is the VM deficit a failure to inhibit an earlier association rather than a failure to form associations?

Damasio grants the inhibition reading is conceivable — and then makes a claim that goes further than the question: most patients who fail the gambling task acquire classical conditioning normally. If that holds, the gambling deficit is not a conditioning deficit, and “conditioning” is not one thing. Filed on Pavlovian defense conditioning, where it sits interestingly beside LeDoux’s position — the two authors agree that conditioning fractionates by task, and reach it from opposite ends (LeDoux from nucleus-level anatomy, Damasio from a lesion group’s dissociated performance).

The concession Damasio does not make is the more telling one. His reply to Bishop is that the intriguing thing is not whether inhibition failed but that such a failure would not be compensated by the patients’ realization that their strategy is losing — reasoning unaided by the marker does not prevail in guiding behaviour. That is the 1997 knowledge/performance dissociation stated as an intuition a year before Bechara et al. tested it, and it shows where the framework’s conviction came from: not from the data, which did not exist yet, but from the clinical observation that these patients say the right thing and do the wrong thing.

Not a contradiction — a trajectory

Nothing recorded in this wiki is overturned. The existing pages attribute the indexing claim, the inducer architecture, the VM gradients, and the background-state model to Bechara & Damasio (2005), and every one of those attributions is accurate about that paper. This source does not contest them; it predates them.

What it does is establish that four of the framework’s load-bearing components did not exist in 1996, that two components present in 1996 were silently dropped, and that the as-if loop’s status inverted between the two texts. The wiki’s Somatic marker hypothesis page presents the framework as a single architecture. It is better read as two, and the second is not obviously a refinement of the first — it is more articulated, less theoretically embedded, and it gives the body back a role its author had written out.

Recorded across Body loop and as-if body loop (the as-if default and the Everitt exchange), Amygdala (the role inversion), Convergence zones and dispositional knowledge (new page for the discarded foundation), Ventromedial prefrontal cortex (VM) (the 1996 network and the scope disclaimer), Iowa Gambling Task (the sample), Does somatic feedback guide decisions? (the peripheral-bypass row), and Antonio R. Damasio (the trajectory).