Marcela Herdova and Stephen Kearns examine the implications of “situationism” for our understanding of free will. This is a web only article. Please support The Philosophers’ Magazine by subscribing.
This is a tricky situation,
I’ve only got myself to blame…
–Freddie Mercury, It’s a Hard Life
On March 13, 1964, Kitty Genovese, a young woman from New York, was brutally assaulted on her way home. The attacker left Kitty with multiple stab wounds to which she succumbed during the transport to a hospital, once help finally arrived. This gruesome case motivated a large of body of work in social psychology on what has become known as the “bystander effect”. While the exact details of this case remain uncertain, it is said that approximately a dozen people witnessed this crime. The assault lasted around half an hour and yet during this time, none of the witnesses directly intervened or called the police.
As social psychologists explained later, such emergency situations have a strong and perhaps unexpected impact on our behaviour: the likelihood of intervention inversely correlates with the number of people who are part of that situation. In other words, the larger the number of people present, the less likely it is that any of these individuals will help. This has become, over the years, a well-documented phenomenon. In one bystander study, conducted by John Darley and Bibb Latané (1968), the experimental subjects overheard a staged epileptic-like attack. In those cases where participants thought they were the only one to witness this event, 85% of the subjects intervened. This starkly contrasts with a mere 31% of the subjects who believed that the attack was overheard by four other people.
The bystander experiments, and other research in social psychology, point to the fact that situations have a great impact on how we behave. What’s more, this research suggests that we are not aware of the impact these environmental factors have; that is, if we are aware of these situational aspects at all. This is the thesis of situationism: environmental cues have a great deal of influence on how we act without us being aware of this fact. In the bystander experiments, the researchers explain that people typically fail to reference the presence of other bystanders as something that would impact on their choices and actions.
Other experiments which are frequently cited in support of the thesis that circumstances exert great power over individuals’ behaviour include the famous obedience experiments conducted by Stanley Milgram (1963/1974) and the Stanford prison experiment led by Philip Zimbardo (1971). In the obedience studies, participants were led to believe that they were taking part in a learning experiment, part of which was to deliver what appeared to be electrical shocks to other experimental subjects. Participants were to use a range of levers to deliver the electrical shocks, each of which was associated with a different degree of shock. Astonishingly, roughly two-thirds of the participants continued, obeying the instructions of the experimental confederate, to deliver shocks all the way – delivering the highest degree of shock which involved pulling levers labelled as “extreme intensity shock” or “danger: severe shock”. Many participants continued their involvement despite the fact that these shocks appeared to have caused a great distress to their recipients. In the Stanford Prison experiment, a group of college students took on the roles of guards and prisoners in a simulated prison environment. Those who played guards exhibited a range of cruel behaviours towards the prisoner participants. The experiment, which was to originally last for two weeks, was then terminated in only six days. In both of these experiments, the extremely powerful situations influenced people to behave in a way that they would ordinarily hesitate to, and which clashed with some of their core values (such as not to intentionally cause physical or emotional distress).
In another famous study, by Isen and Levin (1972), subjects who found a dime in a phone booth were strikingly more willing to help out a stranger than those participants who did not find a dime. Only 4% of the participants who did not find a dime helped out but almost 88% of those who did offered help. A well-known study by Darley and Batson involved Princeton seminary students who were asked to deliver a lecture in a nearby building (1973). On their way to deliver the lecture, the seminary students came across a person who appeared to be in need of medical help. Whether the students helped or not largely depended on how much time the students were told they had: only 10% of those in a high hurry condition offered assistance while 63% of those in a low hurry condition offered help. In both of these experiments, it seems to have been the environmental and circumstantial factors which had impact on how people were going to act rather than their own values and beliefs. The second experiment is perhaps even more striking given that some of the seminary students were going to deliver a talk on the parable of the Good Samaritan (while some other students on job prospects). The content of the lecture did not seem to make a significant difference to whether one offered help or not.
The implications of these experimental studies have been discussed in psychology and philosophy alike. In philosophy, one common worry with regards to the situationist research is that it challenges the existence of free will, and, relatedly, moral responsibility. We think such a challenge is somewhat exaggerated. In what follows, we present five situationist threats to free will and our responses to them.
Threat 1: The situationist literature reveals that agents are unaware of the situational factors that causally influence their behaviour. If situationism is true, agents do not know why they do what they do. Indeed, worse than this, they think they know. Agents unaware of the power of the situation confabulate plausible but false stories that explain, to their own satisfaction, why they acted in a certain way. These facts do not sit easy with our conception of ourselves as autonomous agents. Being free and responsible requires our making sufficiently informed choices, which includes being aware of the factors that might causally impact our actions. If agents lack such knowledge, they are unable to effectively combat pernicious influences and thus should not be blamed if such influences manifest in their behaviour.
Response 1: Is it plausible that the experimental subjects have no awareness of the circumstantial factors that influence their behaviour? Take the bystanders experiments. It might be that people, at least in some of the cases, confabulate reasons for behaving the way they did, not because they are genuinely ignorant of the relevant factors, but because they recognise their lack of intervention as a bad choice which they then try to rationalise to the experimenter. One rather compelling explanation of the behaviour of participants in the obedience experiments is that they are aware that someone ‘in the know’, who they believe has a better assessment of the situation, and who has authority, told them to keep going. The prison guards are presumably aware that how they are acting is at least somewhat influenced by their assigned role. The seminary students who refrain from aiding someone know they are in a hurry, and it is difficult to tell whether they are entirely ignorant of the causal influence of this fact.
Even if it is true that the subjects of these experiments are truly unaware of some of the causes of their actions, it is unclear why this should be a threat to their responsibility. It borders on trivial that we are unaware of many of the causes of our actions. We do not know exactly how our brain activity gives rise to actions, nor do we know how our upbringing affects our behaviour. Events from before we are born, and of which we know nothing about, play some causal role in what we do. Unawareness of these influences does not diminish our responsibility, so why should unawareness of those influences highlighted by the situationist literature do so? Furthermore, any retrospective confabulation concerning what we do cannot affect our responsibility for our actions because such rationalisation occurs afterwards, when the actions have come and gone.
Threat 2: Situationism seems to suggest that agents do not act on good reasons (or, indeed, on bad reasons), but rather on situational cues that are largely irrelevant to the moral goodness or badness of their options. Instead of helping the stranger because he is in need, people help because they have just found a dime. The bystanders to a crime do nothing not because they judge it unnecessary to act, and not for more sinister motives, but because they are amongst many other people. Participants of the Milgram experiment submit to the authority of the scientist instead of freely assessing the available options on their merits and demerits. Responsibility requires, however, that we are sensitive to genuine reasons for our actions. Agents are morally assessable only to the extent that they are appropriately responsive to moral considerations. If we are not, as situationism suggests, we are not morally responsible agents.
Response 2: It is entirely consistent with the data that the agents in situationist experiments do perform their actions for good (or bad) reasons. The explanations mentioned above of their behaviour do not compete with explanations in terms of reasons. People do help the stranger in part because she is in need. After all, if she weren’t in need (or rather, didn’t appear to be), no one would help her. Perhaps finding the dime boosts a person’s mood and brings to mind more readily just those good reasons to help the stranger. This hypothesis is by no means ruled out by the data. In essence, while morally irrelevant situational cues might make a difference as to how someone acts, it simply does not follow that her action is insensitive to reasons.
Furthermore, an agent’s doing something for no reason is no guarantee of exculpation. If an agent performs an action on a whim while knowing their action is wrong, and having the capacity to act on good reasons to do otherwise, such an agent can still be blameworthy. Even if the situationist literature highlights that practically irrelevant differences affect behaviour, not only does this not show that practically relevant differences don’t affect behaviour, but even if it did show this, it still does not suggest that this behaviour is unfree.
Threat 3: Situationism arguably entails that agents exercise less control than commonly believed. If situations are more powerful than we think, then we are less powerful than we think. Our practices of responsibility and our attributions of free will rest on the assumption that we exercise rather considerable control over our actions. We are the authors, the originators, of our behaviour. Situationism does away with this conception of ourselves. We are not unmoved movers imposing our will on the world, but rather the puppets of our environment. How can our actions be up to us if they are controlled to such a large extent by the situations we find ourselves in?
Response 3: While situations might be more powerful than we recognise, situational factors do not compel us to act. In other words, it is not automatic that a certain situational aspect should lead, without any checks or constraints, to any particular action (the subjects of the above described experiments do not act in 100% accordance with those with whom they share a situation). So while situations may take away some of our agentive powers, in the sense that there are external factors – which include our present circumstances – that have impact on our behaviour, we are far from mindless puppets of our circumstances. Situational factors do not make us behave in an impulsive, knee-jerk manner: our deliberative and rational capacities responsible for how we choose to behave are not bypassed. Situationism may highlight that a certain conception of ourselves is misguided (that of the unmoved mover acting outside the causal nexus), but it does not cast doubt on the efficacy of our choices, deliberative capacities and application of our skills. A realistic account of free will should do without the need for a Cartesian mind acting without influence.
Threat 4: If situations have the kind of power that situationists claim, and we find ourselves in them for reasons largely outside of our control, then it seems that what we do is, to a large extent, due to luck. Whether I find the dime in the phone booth or not is simply a matter of fortune. And given that this is the main predicting factor concerning whether I help the stranger in need, whether I help is largely due to luck. But luck excludes responsibility. One person’s moral status as blameworthy, and another’s as praiseworthy, cannot rest simply on the latter’s getting lucky. Can those who helped the stranger sincerely blame those who did not, knowing that, had they not found the dime themselves, they very likely would have done exactly the same thing? Can we morally criticize participants in the Milgram experiments, or in the Stanford Prison Experiment, now that we know, to paraphrase John Bradford, there but for the Grace of God go us?
Response 4: While we may not be able to control what situations we find ourselves in, the fact that we are not powerless in these situations makes the situation itself irrelevant with regards to the question of free will and moral responsibility. While it may be just a matter of luck in what environment we are and what kinds of situations we encounter, how we deal with the situation is not lucky given that the situation does not diminish our agentive powers to the extent that would warrant exemption from the usual practices of moral responsibility. Though the external circumstances we find ourselves in might be a matter of luck (though even this is not always true), and thus not up to us, it is (for all we know) up to us how we react to such circumstances. Situationism simply highlights that people’s actions in the same circumstances often conform to certain patterns—it does not imply that these patterns do not arise from freely-chosen behaviour.
Threat 5: Another apparent consequence of situationism is that we lack cross-situationally stable characters. Our actions result not from personality traits or virtues or vices, but from largely uniform reactions to environmental cues. Some have suggested, however, that agents are responsible for their actions only if they reflect our moral characters. To be free is to (be able to) act in line with one’s deep self—i.e. those stable traits and values with which one intimately identifies. Freedom is self-expression. We are responsible for actions that are ours. If there is no deep self, no character, there is, then, no freedom, or responsibility.
Response 5: The death of character is somewhat exaggerated. It is implausible that situationism shows that we have no character and that what we do is determined by situations only. What it shows, at most, is that our character traits and values adapt to different situations we might found ourselves in. To see this, we must bear in mind three things. First, many of the scenarios that lend support to situationism are extremely powerful situations which are not all that common in everyday life. While it may be the case that extreme circumstances exert unusually strong influence on how we behave, without us being aware of this influence, we should be wary of over-generalising these results. Even if it is true that highly charged and stressful situations afford us less flexibility with regards to how we act, we should be cautious about extending such conclusions to more ordinary circumstances in which we might have more time to deliberate and in which we experience less pressure with regards to our decisions and actions. Second, even in the situations studied, compliance was far from 100%. Two thirds of Milgram’s subjects pulled all the levers, but this means, of course, that one third did not. Third, the situationist literature does away with only a very strict understanding of character, according to which no practically irrelevant environmental factors influence behaviour. We should simply give up on this conception of character.
When it comes to free will, this seems to be good news rather than bad news. If our character and our values were to determine us to act in certain ways irrespective of the wider context, this would make us very inflexible and rather predictable agents. If freedom is self-expression, this self-expression goes in hand with increased flexibility rather than with behaviour stemming from unchangeable and isolated character traits.
Even if we suppose that agents lack characters, it is not a simple step from this idea to the claim that agents lack responsibility. This leap relies on the controversial idea that character is necessary for responsibility. But if an agent without character exercises control of her actions, knows what she is doing, knows right from wrong, is not compelled to act a certain way, and could have done otherwise, she is a prime candidate for having free will, and being responsible, even if she lacks a character.
Situationism, free will and responsibility can coexist. Situations may influence our behaviour, but do not excuse it. With this in mind, we end with advice from Freddie Mercury’s band-mate:
Pull yourself together,
‘Cos you know you should do better
That’s because you’re a free man
–John Deacon, Spread Your Wings
(This doesn’t mean women are off the hook.)