The somatic marker hypothesis: emotion as the machinery of choosing, and the experiment that was challenged

Origin

Damasio, a neurologist then at the University of Iowa, published Descartes’ Error in 1994 [1]. The book’s starting point was a patient he called Elliot, who had lost part of his ventromedial prefrontal cortex to a tumour. Elliot’s intelligence, memory and reasoning tested as normal. His life fell apart. He could analyse a decision indefinitely and could not make one, and the decisions he did make were poor.

Damasio’s reading was that Elliot had lost the emotional signal that normally narrows the field. Without it, every option was weighted the same, and reasoning alone could not break the tie. The hypothesis generalised from there: the body produces emotional responses to anticipated outcomes, those responses are registered as markers on the options, and the markers guide choice below the level of deliberate thought.

The word somatic is doing work. The marker is bodily, a change in heart rate or skin conductance, not a thought.

The problem Damasio was solving was the Cartesian one in his title. Western thought had treated reason and emotion as opposed. His patients suggested they were parts of one system.

What the evidence shows

The lesion evidence is real. Patients with ventromedial prefrontal damage show the pattern Damasio described: intact performance on standard reasoning tests, impaired real-world decision making, and reduced emotional response to outcomes [1][4]. This part of the record is not in dispute.

The Iowa Gambling Task is the experimental core. Bechara, Damasio, Damasio and Anderson published the task in 1994 [4]. Participants draw cards from four decks. Two decks pay well per card and carry occasional large losses that make them bad overall. Two pay less per card and are good overall. Healthy participants learn to prefer the good decks. Patients with ventromedial damage keep drawing from the bad ones.

The 1997 Science paper made the strong claim. Bechara and colleagues reported that healthy participants began generating anticipatory skin conductance responses before drawing from the bad decks, and began choosing advantageously, before they could articulate why [5]. The title stated it: Deciding Advantageously Before Knowing the Advantageous Strategy. The body knew before the mind did. That finding is the one that travelled into marketing.

Marketing’s use of it rests on that 1997 finding. The argument runs: if emotional signals guide choice before conscious reasoning, then advertising that produces a feeling is making its case in the channel that decides. Binet and Field’s finding that emotionally driven campaigns outperform rational ones over the long term is often attached to Damasio as the mechanism [6]. That attachment is an inference across literatures, and this entry says so.

The challenge

The challenge here is serious and this entry would be dishonest without it.

Maia and McClelland, 2004. They re-ran the Iowa Gambling Task with more detailed questioning and found that participants had much more conscious knowledge of the deck structure, and much earlier, than Bechara’s 1997 paper had reported [2]. The participants were not deciding before they knew. They knew, and the original questionnaire had been too coarse to detect it. If that is right, the central claim of the 1997 paper, choice ahead of knowledge, does not stand as stated.

Dunn, Dalgleish and Lawrence, 2006. Their critical review of the somatic marker hypothesis went through the evidence line by line and found it equivocal [3]. The lesion findings were solid. The claim that bodily signals cause good decisions, rather than accompany them, was not demonstrated. Alternative accounts, in which the ventromedial cortex does other work that happens to affect both emotion and choice, fit the data as well.

Damasio and colleagues replied, and the exchange continues. The honest state is this. Emotion is involved in decision making, and patients who lose certain prefrontal function lose both emotional response and decision quality. That is established. The specific claim that bodily markers precede and drive choice, which is the claim marketing borrows, has had its best experiment challenged on method and has not been re-established to the same standard.

The literature is split and should be left split. A marketer citing Damasio should cite the challenge in the same breath.

Where I see this go wrong

Damasio is cited for a finding that belongs to someone else. “Emotion drives decisions” gets attributed to Descartes’ Error in decks and briefs. The evidence that emotional advertising works over the long term is Binet and Field’s, from campaign effectiveness data [6]. The evidence that emotion is involved in choice is broad and comes from many sources. Damasio is one theorist of mechanism, and his mechanism is the contested part. Cite the finding you are actually using.

The 1997 paper is quoted without the 2004 reply. Deciding before knowing is a wonderful phrase and it is in a thousand presentations. The paper that undercut it is in almost none. A reference work that promised documented studies has to carry both.

Neuroscience is used as a seal. A brain scan in the deck makes the brand strategy feel like science. The somatic marker hypothesis is one of the two or three ideas most used this way, and it is one of the least settled. The scan is not the evidence. The behaviour is, and the behavioural evidence for emotional advertising is good enough on its own without borrowing a disputed mechanism.

Having spent years in media before running an agency, I notice that the industry has always wanted a scientific reason for things it already knew worked. Emotional advertising worked before Damasio and will work if his hypothesis falls. The appetite for the mechanism is about status, not about the work.

What to do with it

1. When you want to argue that emotion matters in your category, cite Binet and Field and the campaign data, not Damasio. The first is a finding about advertising. The second is a disputed theory about the brain.
2. If you cite Damasio, cite Maia and McClelland and Dunn, Dalgleish and Lawrence beside him. Anyone who reads the first and not the others is quoting half a literature.
3. Do not put a brain scan in a brand deck unless it is your scan, from your study, with your sample size on it. A borrowed image of a lit-up cortex is decoration dressed as proof.
4. Test emotional response in your own category with behavioural measures: recall, preference, choice. Skin conductance is a research instrument and a poor marketing one.
5. Separate the two claims every time. That emotion is involved in choice: well supported. That bodily markers precede and cause choice: contested. Your strategy can rest on the first without needing the second.
6. Read the 1994 book, the 1997 paper and the 2004 reply, in order. Together they are a short course in how a mechanism becomes a slogan and what gets lost on the way.

If you run the company rather than the marketing

You will be told that buyers decide emotionally and rationalise afterwards, and it will come with a neuroscientist’s name attached.

The first half is reasonably well supported by advertising effectiveness data, and it is a good reason to take emotional work seriously in your category. The attached name is where to be careful. The specific brain mechanism usually cited has had its central experiment challenged, and the field has not re-established it. That does not make the advice wrong. It means the advice is standing on the campaign data, not on the brain scan, and anyone who presents the scan as the proof has not read the literature they are quoting.

The useful question is simple. Show me the behavioural evidence for emotional work in a category like ours. If it exists, fund the work. If the only evidence offered is a picture of a brain, ask for the study, the sample and the date, and watch how quickly the conversation returns to the campaign data it should have started with.