Support the news
You might think the defining feature of psychopaths is that they're heartless: willing and sometimes eager to inflict suffering because they lack empathy. But a new Harvard-led study out in the journal Neuron highlights a less obvious aspect of the typical psychopath: poor decision-making.
Psychopaths' brains seem to be wired so that they are poor at taking into account how bad they'll feel in the future about what makes them feel good in the present, the study finds. And it suggests that perhaps, at the heart of the psychopath problem, is a brain that's poor at generating simulations — whether of other people's feelings or of the future.
Does this let psychopaths off the hook for their anti-social actions? No, but see how you feel after you read my conversation (below, lightly edited) with the study's senior author, Harvard associate professor Joshua Buckholtz. His research team gathered their data by trundling a mobile MRI scanner to prisons in the Midwest and scanning inmates' brains.
How would you sum up what you found?
Traditionally, the research done on psychopathy has really focused on the idea that psychopaths behave the way they do because they don't feel the same things that we feel. And that has trickled down into media representations of psychopaths as cold-blooded, almost alien predators.
But what we also know about psychopaths is that they make really bad decisions. Regardless of what's going on with their emotions — whether or not they experience the same kinds of emotions as non-psychopaths — we know they make terrible decisions, decisions that harm other people, decisions that are destructive to their own lives.
So we were really interested in the decision-making component, because that's an area that has received less attention. In particular, we focused on a really classic psychological paradigm called an inter-temporal choice, or delayed discounting task. You'll recognize these from the studies of Walter Mischel on delayed gratification.
The famous marshmallow tests: Do you want one marshmallow now or two later?
That's right. We asked people to choose between a sooner but smaller and a larger but later reward. And from their behavior we can estimate how steeply the value of delayed reward decays as you go forward in the future.
We gave a group of incarcerated offenders this kind of inter-temporal choice task and we scanned them. What this allowed us to do was to understand how their brains represent rewards at different phases in time.
We found that a region of the brain called the nucleus accumbens — which is really important for representing the value of different actions when we're faced with a decision-making problem — that the activity in that region is much higher in psychopathic individuals when they're asked to make these choices. That suggests that they might have some kind of present-focused bias, that the part of their brain that's responsible for saying "pick this one, not this one," is dysregulated in some way.
Then, we used a technique called functional connectivity, which is useful because we know that the brain is all about networks — individual brain regions don't act in isolation — and the brain is full of exquisitely intricate regulatory mechanisms. We thought maybe one of those regulatory mechanisms might be broken.
And sure enough, when we looked at connectivity in these individuals, we found a weaker connection between the striatum — the part of the brain that we found was overreactive during that inter-temporal choice task — and a part of the brain called the ventromedial prefrontal cortex, which is really important for mental time travel, for looking at how we're going to feel about something in the future and making decisions based on that. It's a core component of what it means to make goal-directed, future-oriented decisions.
These two regions normally talk. They negotiate the choices that we make based on a combination of. "How is choosing this action going to make me feel now?" and "How is choosing this action going to make me feel in the future?" But the connections between those regions were weaker in psychopaths.
Finally, we found that the degree of dysfunction in this regulatory circuit could predict how many times these folks had been convicted of crimes.
Is this the first time that a circuit that's key in psychopathy has been pinpointed?
There has been really elegant work by my colleague Kent Kiehl, who's on this paper, and Mike Koenigs at Wisconsin, showing that some of the anatomical connections in psychopathy appear to be weakened. The incremental value of our work is that we showed a different circuit than had previously been shown was weak, and that this has implications for how psychopaths represent reward. And that there was a real-world correlate of the degree of that circuit breakdown.
And with the real-world correlate being criminal convictions, could this be a measure that is key to whether a psychopath ends up in prison or not?
I want to make a really important distinction between demonstrating a mechanism, and showing that it has a real-world correlate, and using that to predict the future.
You know the famous Yogi Berra quote: Prediction is hard, especially about the future. Some colleagues and I have written and submitted a paper talking about this very thing: the use of a brain measure to predict violence, and how it's actually much trickier, for a number of different reasons, than we think. So I wouldn't want to leave you with the impression that we could scan a bunch of people and, based on their measures of circuit function, predict whether or not they're going to commit crime.
It's also worth highlighting that the way that this circuit appears to be dysfunctional in these individuals is very similar to the way that the circuit is dysfunctional in substance abusers, and compulsive overeaters, and people with gambling addiction. To me, what this says is that psychopaths are not alien, they're not unknowable. They behave in ways that one could predict, given the kinds of breakdowns in the way that their brain is wired.
The sentence in your paper that most stood out for me was about how this brain anatomy seems to degrade the capacity for mental simulations. So you don't have empathy because you can't imagine how other people feel, and then you're impulsive because you're not really imagining the results of your actions.
Right. That region of the brain, the ventromedial prefrontal cortex, is really important for perspective-taking and it's thought to be important for empathy, because it has the general function of simulation. It simulates other people and their mental states, it simulates ourselves in the future.
So the fact that this brain region appears not to work quite as well in psychopaths, and that it has weaker connections with other brain regions — like the ventral striatum in our paper, or the amygdala in other papers — suggest that this knocking-out or disrupting of this key node for simulation might have broad effects on behavior that include decision-making and interpersonal relations.
So do these new insights suggest new ways to deal with psychopathy?
Yes and no. There are interventions that appear to be useful for reducing this kind of impulsivity, something called episodic prospective training, or episodic simulation training. That's training people to really have a focus on, and very strongly represent, the future consequences of different choice options in the moment. That might have some benefit. But it's also the case that today's most effective interventions in psychopathy — and there are only one or two — seem to work best when done in adolescence.
I'm always uncomfortable when people want to take one single study and design an intervention based on it. For me, the main value of the work is contributing to a really robust, empirical foundation that, in aggregate, will lead to new intervention targets.
Should this change how we view psychopaths?
One thing that I would like to come out of this research is a de-mythologizing of psychopaths. How do you make a psychopath? You break their ability to have empathy and perspective-taking. You reduce their capacity to represent the cost of bad choices now in the future. For anyone who has this collection of brain wiring deficits, they will always produce that kind of behavior.
Strange as it sounds to say, this humanizes psychopaths because it's almost like a "there but for the grace of God" thing. If I had this broken circuit, I would make bad choices all the time, too. If I had this broken circuit, and deficits in empathy, and I grew up in an environment where pro-social behavior wasn't reinforced, and I had a lot of anti-social behavior models, that's how I would turn out. So it de-other-izes them.
But personal responsibility? Choices?
That's not a scientific question. Personal responsibility and free will — it's a question for philosophers, not scientists. Its like asking a scientist for proof of the existence of God. Our tools are just not built for that.