There's a scene from a disaster movie I watched as a small child that I'll never forget. As passengers attempted to flee a ship sinking and on fire, they reached a missing section of walkway above a perilous drop. It looked like they were all doomed, but one of the male passengers turned himself into a human bridge. The others walked across him to safety. He then dropped to his death, exhausted. I can't remember the name of the film, and I can't remember why the passengers couldn't just jump across the gap. But the self sacrifice of the man was seared in my memory – his sheer heroism, his willingness to risk his own life for others.
So it is that a new brain imaging study caught my attention. Marco Zanon and his colleagues scanned the brains of 43 young adults (30 women) while they took part in a virtual reality (VR) experience of a disaster. Wearing VR goggles and headphones, each participant began the study by meeting up with what they thought were three other volunteers in a virtual waiting room. In fact, these other avatars were computer controlled. After exploring the room for a while, the participants were surprised by the sound of a fire alarm. Having earlier been instructed to behave as they would in the real world, the participants raced to evacuate the building. Simulated smoke, flames and coughing and heart beat sound effects added to the drama.
Crucially, near the building exit, their "life energy" bar nearly depleted, the participants encountered one of the other people they'd met in the waiting room, finding them trapped injured under a fallen filing cabinet and surely doomed to die. Each participant faced the same choice – try to save the stricken individual (they'd earlier learned that objects could be moved by tapping a joystick key; saving the other human required 150 such button presses), or plough on to safety. Throughout this VR experience the researchers scanned the participants' brains. They used an approach known as independent component analysis, which is about looking for networks of correlated activity across the brain.
There were 16 heroes, including 11 women, who rescued the trapped man. Nineteen others, including 12 women, passed by without helping. The remaining 8 tried to help but gave up – they were omitted from the subsequent brain analysis because they were so few in number. Zanon and his colleagues identified three functional hubs in the brain that they said were differently activated in the heroes and the more selfish folk. The first was more active (throughout the VR experience) in the selfish participants, and took in the anterior insula and the anterior mid-cingulate cortex (areas buried deep in the cerebral cortex), but also included other regions such as the thalamus and the cerebellum. Zanon's team said this network has previously been associated with finding things salient, which is itself a state associated with anxiety. Activity in this network has also been linked with harm avoidance, the researchers said. In other words, greater activity in this functional hub may reflect the fact that the selfish participants felt more endangered (there was a trend for them to report feeling more anxious than the heroes, but this didn't reach statistical significance) and a greater motivation to protect themselves.
Two other networks were more active in the heroes, specifically while they encountered the trapped victim. The first included areas such as the medial orbitofrontal and anterior cingulate cortex, activity which the researchers said is involved in taking other people's perspective. The second included an area over the junction of the temporal and parietal lobes, and has previously been linked with thinking about other people, and distinguishing the self from others. The obvious interpretation is that the greater activity in these networks in the brains of heroes reflects their greater empathy for the trapped victim. However, the researchers admitted it's also possible this brain activity could reflect increased concern for one's reputation.
What to make of this study? I admired its ambition. So much research into altruism and so-called "pro-social" behaviours depends on financial games, in which people's generosity or trust is measured. Or researchers pretend to drop pens on the floor and they see if participants will bend down to help pick them up. These methods are obviously a far cry from real-world heroism. And so it is refreshing to see a study that involved a little existential drama, albeit in a virtual reality context. Unfortunately, in most other regards I found this research disappointing – it just seemed to involve so much speculation. Take the first brain network (the one involving the anterior mid cingulate and insula) – the researchers identified this as a salience network, and linked it with increased anxiety. But one could just as easily interpret this activity as being involved in empathy, given that other research has linked the anterior insular cortex with this function. In which case, how come the selfish people showed more empathy-related activity? Perhaps we should remember, as the Neuroskeptic blogger recently pointed out, that the link between activity intensity and function isn't straightforward – perhaps the selfish participants in this study showed extra activity in this network because they had to work harder to empathise with the victim whereas for the heroes this concern came more naturally. But now I'm doing it – speculating wildly about the meaning of recorded brain activity.
In the end what really have we learned from this research? It feels like we have preconceptions about heroes – that they have more feeling for other people, for example – and then the brain scan results are interpreted in line with those prior beliefs. This makes me think again (I've made this point in previous blog posts) that actually far more psychology research is needed to lay the groundwork, in this case on extreme acts of bravery, and perhaps then with more sophisticated psychological understanding we would be better placed to explore the neurophysiological correlates of heroism. Even then, if our aim is to understand heroism, is the brain really the place to look? I can't help feeling skeptical – I'd be interested to hear what you think. To be fair to the researchers, they do admit that they can't draw "definitive conclusions" from their findings, and they express the modest hope that their study "might inspire new hypotheses or experimental protocols".
Meanwhile, if anyone knows what that ship-based disaster movie was that I saw as a child (I watched it on TV in the 80s), please let me know!