Primary and secondary inducers
The division of labour that lets the somatic-marker-hypothesis explain a double dissociation rather than merely accommodate it. From Bechara & Damasio (2005); see antoine-bechara, antonio-damasio.
A later addition, now confirmed as one. The 2005 study page called this the paper’s main theoretical addition over Damasio (1994), on internal evidence. The Damasio (1996) ingest settles it against a primary source: the distinction appears nowhere in the 1996 statement of the framework, which has no inducer types at all — and in which the amygdala occupies the opposite position in the causal chain, a downstream somatic effector that VM activates rather than an upstream trigger that seeds VM’s material.
Two things follow, and they cut against the elegance claimed below. First, the double dissociation this architecture “predicts rather than accommodates” was already in hand when the architecture was built — Bechara et al. (1996) is cited in the 1996 paper, and the amygdala half (Bechara et al. 1999) preceded the 2005 statement by six years. The prediction is retrospective. That does not make the account wrong; a framework that reorganizes to fit a known dissociation is doing ordinary theoretical work. But the wiki should not credit it with foresight it did not exercise.
Second, the developmental dependency was not new in 2005 either. It is stated in the 1996 discussion in a different vocabulary — body states are actively engaged during development, then the periphery is bypassed — as a remark about the body loop rather than about inducers. The 2005 architecture reuses a claim Damasio had already made about a different pair of mechanisms. See body-loop-and-as-if-body-loop.
The distinction
Primary inducers are stimuli that, once present in the immediate environment, “automatically and obligatorily elicit a somatic response.” Two kinds are lumped together, which is worth noticing: innate or learned objects (a snake, or a stimulus predictive of one), and concepts or knowledge that have come through learning to trigger obligatorily — hearing you won a lottery, or that your savings are gone. The “aha” of solving a puzzle is also listed as a primary inducer, which stretches the category about as far as it goes.
Secondary inducers are “entities generated by the recall of a personal or hypothetical emotional event” — thoughts and memories of a primary inducer that, brought to working memory, elicit a somatic state. Remembering the snake; imagining losing the money; imagining being attacked by a bear or winning an award. Hypothetical events count, so this is not merely memory: it is simulation.
The trigger structures differ, and this is the whole payoff:
| primary inducers | secondary inducers | |
|---|---|---|
| trigger structure | amygdala | VM cortex |
| speed | fast, automatic, obligatory | deliberate, slow |
| duration | short-lived, habituates quickly | lasts a long time |
| state produced | full strength | fainter re-activation of the stored pattern |
| processing style | ”low-order,” relatively automatic | ”high-order,” driven by thought and reflection |
The developmental dependency
The asymmetry that makes the account more than a relabelling: normal secondary induction requires prior normal primary induction, but not the reverse.
The mechanism proposed: once a somatic state has been triggered by a primary inducer and experienced at least once, a pattern for that state forms in brainstem nuclei (e.g. the parabrachial nucleus) and in somatosensing cortices (insula/SII, SI, cingulate). A secondary inducer then works by re-activating that stored pattern. No pattern, nothing to re-activate.
Bechara & Damasio’s illustration: if burning your hand on a stove did not induce pain, you would not know how painful boiling water should feel. But if you once could feel pain and built the representation, you will still avoid painful situations after losing the capacity — because secondary induction, once acquired, “becomes less dependent on primary induction.”
This is what predicts the lesion pattern in the iowa-gambling-task: amygdala damage removes the primary trigger and therefore starves secondary induction of its raw material (amygdala patients generate no SCRs at all, not even to actual reward and punishment); VM damage leaves primary induction intact and removes only the secondary trigger (VM patients respond to reward and punishment but never anticipate). Amygdala damage means the patient “can no longer register how painful it feels when one loses money,” which in turn “misleads” the VM cortex about how painful a loss should feel.
The VM cortex as convergence–divergence zone
How a secondary inducer becomes a somatic state, per the paper: the VM cortex holds neuron ensembles that couple (a) a category of event held in high-order association cortices, to (b) the effector structures that execute the somatic state, and to (c) the neural patterns of the non-conscious (brainstem/PBN) or conscious (insula/SII, SI) feeling of it. In short, the VM cortex couples knowledge of events to “what it feels like” to be in a given situation.
Sometimes it couples only to covert effectors — basal forebrain or brainstem, no felt state. The anticipatory SCRs of the pre-hunch period are the example: pondering a deck (a secondary inducer) elicits a covert somatic response that biases the choice “without any awareness of why the choice was made.”
What to watch
The two systems are inseparable in an intact brain. Bechara & Damasio concede this directly: a photograph of a sick baby is a primary inducer and generates thoughts (picturing your own child) that act as secondary inducers, at the same time. The distinction is therefore only visible through lesions — which is a methodological strength for the Iowa programme and a limitation on how far the distinction can be exported to healthy-subject work, including most of the interoceptive literature this wiki otherwise runs on.
The primary category is heterogeneous. A snake, a lottery win, and the satisfaction of solving a puzzle are grouped by their obligatoriness, not by anything about their content, route, or learning history. Whether the amygdala triggers all three alike is not shown; it is the kind of claim LeDoux (2012) would reject on principle, since he denies “the amygdala” is an explanatory unit at all and would want the nucleus-level story for each. Nothing in this wiki tests it.
Relation to the simulation-map. Farb et al.’s layered “as-if” body representation and Damasio’s stored somatic pattern are attempting the same job — a body model that can be run without the body. They come from different literatures and are never connected in either source. Worth noting rather than merging: the simulation map is built for predictive-coding interoception, the somatic pattern for decision biasing, and neither author cites the other on it.