It Wasn’t Me: How Coercion Affects Agency
The slippery slope of neuro-excuses (or blaming a “bad” brain)
By Sofia Deleniv
In the Nuremberg Trials of 1945 and 1946, some of the most high-ranking members of the military and political leadership of Nazi Germany were tried for war crimes committed during World War II. Shockingly, multiple defendants pleaded innocent by claiming that they were not responsible for the atrocities they committed since they were just following orders from their superiors. All of them were convicted.
Since then, courts have repeatedly dealt with war criminals pleading innocence on grounds of coercion, such as the famous case of former U.S. Army officer William Calley, who was charged with the cold-blooded murder of 109 Vietnamese civilians in the tiny village of My Lai in 1968. International law generally does not regard the so-called “superior orders” defense as valid justification for committing crimes, so things don’t usually end well for the defendants. Nonetheless, the popularity of this plea among perpetrators raises an important question: Do we have any scientific reason to believe that following orders alters people’s agency?
These findings point to the possibility that complying with coercive instructions diminishes our core sense of agency. It’s as if, on some level, we experience these actions as not belonging to ourselves.
It seems so. An experiment published in the journal Current Biology in March provides some evidence that being coerced into committing a hurtful action, such as pressing a button to deliver electric shock, produces a pattern of behavior and brain response that is surprisingly similar to what is observed during passive movements — that is, when someone literally holds an individual’s hand and pushes that button for them.
In the experiment, pairs of participants sat across from each other and took turns being the “agent” or “victim,” each with a starting capital of 20 British pounds. Some agents were asked to decide whether they wanted to earn five pence by administering a painful electric shock to the victim’s wrist (physical harm), while other agents were asked to decide whether they wanted to earn those five pence by removing them from the victim’s own funds (financial harm). The researchers then attempted to estimate how strongly individuals experienced themselves as the agents behind these decisions when making them freely, as opposed to when an experimenter stood right next to them and dictated which choices to make.
Of course, we can’t learn much from an experiment if we just ask individuals whether they felt that a choice made under pressure was their own. Particularly in court, the most plausible explanation for pleading innocent on grounds of following orders is that it’s more socially acceptable, and potentially more forgivable, than owning up to doing something atrocious.
When it comes to such socially sensitive issues, psychologists often rely on implicit measures that aim to capture behaviors that are untainted by personal motivations or fears of social acceptance. For instance, you might want to be brave and tell me that you’re not nervous about seeing a giant spider, but the increased sweat on your fingertips generated by your autonomic nervous system would tell me otherwise. So when it comes to agency — what can we measure that doesn’t involve asking?
In the past, researchers have observed that when people intentionally commit an action (e.g. press a light switch), they perceive the delay between the action and its outcome (e.g. light comes on) as shorter than when that action was made passively (e.g. someone held their hand to press the switch).
The researchers behind the recent Current Biology publication reasoned that this could be useful for testing whether coercion really does, on some subconscious level, reduce people’s sense of agency. They designed the experiment so that whenever participants pressed a button to make a free choice or to follow an order, a beeping noise would play after a brief period of time. Participants were then asked to estimate how long they felt they had to wait between pressing the button and hearing this beep (the rationale being that the greater this time estimate, the less the individual experienced “ownership” of the action when pressing the button).
Individuals consistently reported the delays that occurred between button presses and resulting beeps as longer when they were ordered which button to press, compared to when they made their own choices. Individuals reported similar experiences of prolonged delays between button presses and beeps when an experimenter literally held their fingers and pushed the button down for them. These findings point to the possibility that complying with coercive instructions (even though technically the choice to refuse is always there) diminishes our core sense of agency. It’s as if, on some level, we experience these actions as not belonging to ourselves.
The researchers went further, using electroencephalography (EEG) to examine whether the brains of free or coerced agents responded differently to the sound of the beep. When the researchers focused on the auditory N1 — a hallmark brain response to sounds — they found that it was substantially weaker when participants heard a tone produced by a button pressed under coercion. Interestingly, a similar reduction in the strength of the N1 signal was observed when participants’ fingers were literally held by an experimenter and used to press down a button.
What does it all mean? Do we fundamentally experience ourselves as puppets when coerced into making decisions and — more importantly — should it matter? The authors of the study indicated that their results “may have profound implications for social and legal responsibility.” Given how routinely brain scans are used to argue against the culpability of defendants in criminal court cases, I wouldn’t be surprised if evidence for disturbed agency under coercion also eventually makes it to the courtroom. I find this prospect quite worrying.
It’s tempting to justify actions with “neuro-excuses” — if the brain did it, it’s not really the person’s fault. This is fundamentally false, because everything we ever think or do is underpinned by brain phenomena. But rather than assuming that in all cases an abhorrent or kind action was produced by the workings of the brain, we appear to treat this organ as a companion that only occasionally takes over the steering wheel that we normally hold ourselves. It is essential that we’re critical about this if we’re going to assess someone’s culpability based on what we see happening in their brain.
Here’s a brief thought experiment. Consider psychopaths. In 2014, a group of researchers examining brain activity in individuals with psychopathic traits found that those who were most comfortable with saying fearful or immoral things to other people had the weakest activations in their amygdala — a collection of brain cells known to be critical for experiencing fear and anxiety. Such dysfunctional patterns of brain activity indicate that these individuals are biologically less capable of feeling fear, and thus empathizing with others’ suffering.
This study was one of countless others that have reported abnormalities in the brains of psychopaths, and it should not come as much of a surprise. Abnormal brain function is exactly what places some individuals at risk of deciding to commit abnormal actions. So now, if you were part of a jury presented with a scan detailing atypical activations in a psychopathic criminal’s brain, what would you think? Should neuroscientific evidence enter the courtroom? Or might it pave the way toward a slippery slope at the bottom of which we come to excuse away all bad actions as rooted in “bad” brains?
Sofia Deleniv is currently pursuing a PhD in Neuroscience at the University of Oxford in the UK, where she researches mouse brain physiology. Driven by a slight frustration with the superficial manner in which the media reports scientific findings, she decided to start The Neurosphere blog in 2015. Follow her on Twitter.
