I study the psychology of morality, including the cognitive mechanisms we use to make moral judgments, their developmental origins, and their evolutionary history.  On this page, I have grouped my research under a few broad categories:

1. (1)How do we use information about causation (“who done it?”) and mental states like intention (“did they mean to?”) to make moral judgments?
   

2. (2)How do moral intuitions - that is, gut feelings - give rise to explicit moral principles such as those used in philosophy and the law?
   

3. (3)How do moral judgments affect the way that we understand and describe actions and events?
   

4. (4)What does it all mean?  Implications for philosophy, law, and the rest of us.
   

Certain publications and areas of research don’t fit well under this schema, and so they aren’t discussed here.  Please visit my publications page for a complete list of publications.

You may also be interested in visiting these online research projects:

The Online Games Project

The Moral Sense Test

Causation and intention in moral judgment  ❧ 

Here’s my reading of Massachusetts law:  If you drink, drive, fall asleep at the wheel and  run into some bushes, you pay a $250 fine.  If you drink, drive, fall asleep at the wheel and run into a person, killing her, you face 2.5 to 15 years in prison.  From one perspective this is crazy: the random chance of what you happen to hit makes the difference between a small fine and years in prison.  But would it be OK to send the bush-hitter to prison for years?  Would it be OK to give the person-killer nothing more than a ticket?  We seem to be stuck in a dilemma: differential punishments seem unfair, but identical punishments seem unjust.  The philosophers Bernard Williams and Thomas Nagel played an important role in showing how cases like this lead to a dilemma (click for more).

Research I’ve conducted suggests that cases like this strike us as a dilemma because two distinct processes of moral judgment treat them very differently (Cushman 2008).  One process assesses causal responsibility for harm, and it plays an important role in shaping our decisions about punishment.  It focuses on the very different consequences -- a broken law and damaged bush, versus a dead person -- and assigns different punishments to them.  Another process assesses people’s mental states.  It focuses on the fact that the bush-hitter and tree-hitter engaged in identical behavior with the very same state of mind.  Judgments about moral wrongness and moral permissibility are strongly dominated by this mental state process.  This case strikes us as a dilemma because the causal responsibility process and the mental state process are in conflict.  Indeed, there are hints of neurological evidence for this conflict between systems (Young et al 2007), and ongoing work by Liane Young targets the neural underpinnings of mental state reasoning in moral judgment.

In current research I am testing the development of these two processes, studying them in the context of economic games, and attempting to understand them in terms of their evolutionary functions.  For instance, one study of actual punishing and rewarding behavior in the lab found that people tend to punish accidentally ‘stingy’ allocations of money caused by attempts at generosity, and reward accidental ‘generous’ allocations of money caused by caused by attempts to be stingy (Cushman et al 2009).

Another active area of research is the role of causal and intentional representations in generating certain patterns of moral intuition (Cushman et al 2006; Cushman et al in press).  For instance, people think that it is worse to harm somebody as a means to and end than as a side-effect.  And, people think that it is worse to harm somebody actively than to merely allow them to be harmed.  In current research I am testing the hypothesis that these moral judgments ultimately depend on the way that we assess causal responsibility for harm and intent to harm (Cushman & Greene, in press). 

I also recently joined Joshua Greene on a study showing a curious interaction between assessments of intention to harm and the use of personal force in harming (Greene et al 2009).  In short: it is judged particularly wrongful to harm somebody by directly applying your body, if your intent is to harm them.  Why might that be?  We hypothesize that these judgments might be a product of monitoring goal-oriented physical actions, and rejecting “action plans” that involve harm to another.  Thus, our moral judgments of other people’s actions might be produced by using a mechanism designed to regulate our own “action plans” (see also Cushman & Greene, in press).  This hypothesis is an active area of research.

Intuition vs. reasoning  ❧

A focal point of recent research in moral psychology is the so-called “Trolley Problem”, introduced by the philosopher Phillipa Foot.  It involves two contrasting situations.  In both situations, a runaway train is heading down a railroad track towards five people.  In one situation, the only way to save them is to send it down a side-track where it will kill one person.  In the other situation, the only way to save them is to push a person in front of the train, causing it to slow down in time to save the five.  Most people think flipping the switch is OK, but pushing the man is wrong (Hauser et al 2007).  The problem is: Why?

While a great majority of people have the intuition that flipping the switch is OK and pushing the man is wrong, most people can’t give you a reason why (Cushman et al 2006).  This is a case of what the psychologist Jonathan Haidt has termed “moral dumbfounding.”  It would be natural to assume that we make specific moral judgments by reasoning from general moral principles.  And, there is evidence consistent this view -- for instance, people are able to tell you exactly why it is worse to actively harm a person than to merely allow them to be harmed (Cushman et al 2006).  But moral dumbfounding suggests that the opposite may be true as well: we often construct general moral principles by consulting our intuitions regarding specific moral cases. 

This seems to be as true of philosophy as of ordinary people.  Philosopher Eric Schwitzgebel and I tested both moral reasoning and post-hoc rationalization among professional academic philosophers, comparing them with academics in other disciplines as well as non-philosophers (Schwitzgebel & Cushman, in press).  We found that philosophers were no less susceptible to order effects than non-philosophers in their moral judgments even for familiar cases like the trolley problem, suggesting that philosophical expertise can fail to guard against unwanted biases.  (Order effects, in this context, means that you judge two cases differently depending on what order you see them in).  Subsequently, philosophers were more likely to endorse abstract, general principles that comported with the order effects they exhibited during moral judgment.  In short, it looks like philosophers engage in post-hoc rationalization just as much as the rest of us (in fact, probably a little more).

Research by the philosopher and neuroscientist Joshua Greene suggests that some of our moral judgments may be supported by affective processes (i.e. emotions).  In a study of individuals with damage to brain regions implicated in affective processing, we found systematic differences in the way that they judged cases like the “push a man” version of the trolley problem (Koenigs et al 2007).  On the other hand, they gave very typical responses to cases like the “flip a switch” case.  This study and others like it provide evidence that we accomplish moral judgment by at least two different systems: an affective, intuitive system and a controlled reasoning system (reviewed in Cushman et al 2009).

The relationship between intuition and reasoning in moral judgment was the subject of my doctoral dissertation.  The core discussion is presented more concisely in a recent chapter that draws directly from my dissertation: Cushman & Greene, in press.

How morality changes our view of actions and events  ❧

When we make a moral judgment of somebody’s action, we usually care a lot about whether it was performed intentionally.  But the philosopher Joshua Knobe has demonstrated that just the opposite relationship also holds: when we make a judgment about whether an action was performed intentionally, it depends on our moral judgment!  For example, consider the following pair of cases.  In one case, the Vice-President of a company comes to the Chairman and proposes a new policy that will save $1 million a year, but will harm the environment.  “You know me,” the Chairman laughs, “I don’t care at all about the environment -- I only care about making money!”  So they implement the policy and the environment is harmed.  Did the Chairman intentionally harm the environment?  Most people say he did.

But now consider an alternative case.  This time, the Vice-President proposes a new policy that will save $1 million and help the environment.  Again, the Chairman laughs, “ I don’t care at all about the environment -- I only care about making money!”  So they implement the policy and the environment is helped.  Did the Chairman intentionally help the environment?  Most people say he did not.

Our differing intuitions in these cases are surprising.  We don’t usually think that the moral status of an action plays a role in shaping whether we judge it as intentional or unintentional.  Some people have proposed that we judge the harmful and helpful Chairmen differently because of our different emotional reactions to each.  However, individuals with damage to brain areas involved in emotional processing show the normal pattern of judgment on these cases, distinguishing the two chairmen (Young et al 2006). 

It turns out that moral judgments shape the way we describe lots of properties of actions and events.  For instance, the psychologist Mark Alicke has shown that we are more likely to ascribe causal responsibility for harm to morally bad agents than morally good agents.  Additionally, we are more likely to describe morally bad behaviors as active, rather than passive, compared to morally neutral behaviors (Cushman et al 2008).  For instance, a woman who fails to eat the necessary vitamins to sustain her pregnancy is more likely to be described as “making” the fetus die by pro-life individuals, and is more likely to be described as “allowing” the fetus to die by pro-choice individuals.

The effect of morality on judgments of intentional action, often called the “Knobe Effect”, raises important questions about the general criteria we use to make judgments of intentionality.  I joined the philosopher Al Mele in a project aimed at sketching some of the contours of the folk concept of intentionality (Mele & Cushman 2007) and illustrating diversity between individuals in their concept of intentionality (Cushman & Mele 2008).

What does it all mean?  ❧

As a psychologist, the primary aim of my research is to describe how things are, and not to decide how they ought to be.  At the same time, I do think that psychological research has an important role to play in philosophy, law, and our everyday lives.  For example, understanding how children learn and make decisions might change how philosophers assess their morally responsible, how legislators and judges treat them under the law, and how parents and teachers raise them.  Along similar lines, understanding how and why we make certain moral judgments might have important practical implications.

One clear lesson of recent work in moral psychology is that our brains often provide multiple, contradictory answers to the same problem.  Should we punish a drunk driver who hits a tree as much as a drunk driver who hits a girl?  Should we push a man in front of a train to save five others?  In both cases, one part of our brain answers “yes”, and another answers “no”.  This means that there are important limits to our ability to provide a satisfactory moral answer to every problem just by relying on our intuitive moral sense.  Still, we have to decide what to do somehow.  Consequently, Liane Young and I have suggested that rather than looking for a single, definitive moral answer to every problem, we should treat our moral intuitions as one consideration alongside other, non-moral considerations when deciding what to do (Cushman & Young 2009).  We have also suggested developing practices and institutions that avoid conflict between moral systems in the first place.