This post has been contributed by Dr Hugh Breakey, Senior Research Fellow in the Institute for Ethics, Governance and Law (IEGL) and a member of the Law Futures Centre.

In a new paper in Philosophical Studies (Breakey 2022: available here for free download—thanks to Springer and Griffith University), I argue that when certain common types of expectations arise, it can be morally obligatory to conform to them. I argue that these expectations can create duties even in cases where commonplace routes to justifying duties (such as through direct reciprocity or tacit consent) are not present.

In this article, I explain why it’s crucial for philosophers, ethics teachers and other social scientists to attend to the normative force of reasonable expectations.

‘Reasonable expectations’

Obviously, not all types of expectations have the type of ethical significance that can generate moral obligation. In the paper, I defend the ethical significance of what I term ‘reasonable expectations’. Simplifying a little, such expectations have four main properties:

  1. Rational: The prediction generating the expectation must be based on evidence and reason: it must be rationally justified.
  2. Relied upon: Other people must be relying on the prediction; they have planned their actions around it.
  3. Morally tolerable: The predicted action must not be morally wrong or discriminatory.
  4. Known: The subject who must shoulder the obligation needs to be aware of what is expected, and able to see that the expectation is reasonable (that is, that the expectation has the above-noted three qualities: rational + relied upon + morally tolerable).

Reasonable expectations deliver important social and personal goods

When we conform to others’ reasonable expectations, this generates significant social and personal goods for others, including loss-avoidance, epistemic security, and gains in autonomy (understood as the capacity to govern one’s life through rational choice). Conforming to reasonable expectations also allows us to avoid conflicts and avoid interfering with others’ long term projects. As well, it allows others to adapt their preferences in a socially beneficial way, and facilitates their development of key virtues.

But is the generation of these social goods sufficient to justify imposing moral obligations? In the paper, I show how these goods attach to important moral values, creating obligations based on consent (where expected behaviour becomes the default condition for many interactions), fairness (where conforming to expectations allows risks, costs, and benefits to be allocated in an appropriate and predictable way), non-interference (where conformity allows us to avoid interfering in practices socially allocated to particular times and places), and more.

In such cases, I argue, expectations lead to obligations. The ‘normal’ becomes ‘normative’.

Three take-aways

There are three ways the ethical significance of expectations can inform the thinking and practice of philosophers, ethics teachers, and social scientists.

1. The role of expectations in shaping moral and immoral behaviour

In the paper, I discuss the dark sides of expectations. The Milgram experiments are often styled as telling us something about obedience, following ‘orders’, and authority. Milgram himself described them this way (Milgram, 1963). But this is not entirely right—at least, not in the ordinary sense of those words.

As is well-known, Milgram convinced his subjects—ordinary people from many walks of life—that his experiment would test the impact of pain on learning. In fact, the experiments tested how far the subjects would go in inflicting pain on another person when asked to do so by an authority. The disconcerting answer Milgram uncovered was: a very long way.

However, invoking concepts like ‘orders’, ‘obedience’ and ‘authority’ can imply a type of political or even military authority, where one party claims to give mandatory dictates to another. But the Milgram experiment was nothing like this. The ‘authorities’ involved—the scientists running the experiment and urging wavering subjects to continue—weren’t moral or political authorities. They weren’t elected representatives, or police officers, or even employers dealing with subordinates. They were in no position to command anyone.

Nor did they utter commands in any usual sense. They never invoked their own authority and said, ‘I order you to continue!’ Instead, the four ‘prods’ used by the scientists in the experiments were as follows:

Prod 1 : Please continue.

Prod 2: The experiment requires that you continue.

Prod 3: It is absolutely essential that you continue.

Prod 4: You have no other choice, you must go on.

The first prod is a polite request. The second and third emphasize the importance for the experiment of continuing; that the scientists are relying on their action. The fourth prod is the only one that looks anything like an order—and this prod was uniquely unsuccessful! No subject whose continuous objections raised to a pitch that the fourth prod was utilized was subsequently persuaded by that prod (Piovarchy, 2021).

So if the Milgram experiments weren’t testing orders, obedience, and authority, what on earth was persuading normally decent people to commit serious wrongs? There are many answers to this question. Milgram’s cleverly devised experiment diffused moral responsibility, manipulated subjects’ anchoring bias, created ambiguity about consent, and more. But one of the central answers explaining the behaviour—as Milgram himself speculated—was the situation’s novelty. Milgram’s subjects were ordinary people, who would have been unfamiliar with the experiment’s environment, context, and setup. When faced with an unfamiliar situation, people will very often turn to the default of conforming: ‘When in Rome, do as Romans do.’ As I argue at length in the paper, this is normally a wise ethical practice, allowing people to fall in line with local practices long before they learn the point of those practices—and the way they might impact on issues of fairness, consent, interference, and so on.

Ingeniously, Milgram’s experimental setup provided these vulnerable subjects with only one person who was clearly familiar with the context and who had clear expectations about appropriate behaviour. This person was the scientist whose job was to calmly push the subjects to continue, and emphasize how much the experiment was relying on them to do so. This is the only ‘authority’ the scientist wielded—but in an unfamiliar situation for the subjects, it turned out to be far more powerful than anyone had predicted.

The reason it wasn’t predicted was because when we think about ethical questions, we tend to focus on abstract matters of moral principle, and we don’t attend closely enough to the complexities of human ethical life. When it comes to experiments like this, we can fail to ask: Do the respondents have any expectations going into this type of situation? Does the experimental setup impose (subtly or overtly, deliberately or accidentally) expectations on the subjects? Is the situation novel—and, if so, are there are any ‘expert’ figures whose behaviour and attitudes will be taken to communicate the expected ‘normal’?

Far from this information cuing us in to ‘non-moral’ situational features that might overwhelm ethical decision-making, it informs us about genuinely ethical rules-of-thumb that ordinary moral subjects are bound to attend to—even if those rules are leading them disastrously awry in this particular context.

2. Expectations and philosophical thought-experiments

Another way this lack of attention to the ethical significance of expectations can lead researchers astray occurs in a very different form of experiment: the thought experiment.

Just as we don’t tend to consider the role of expectations when we are imagining how people will morally behave in unfamiliar circumstances (like the Milgram experiment), we also don’t tend to consider the role of expectations when we are conjuring up philosophical thought experiments. The novelty of those situations (which are often new because they are confected in specific ways to capture idiosyncratic features of a situation) itself matters.

Consider, for example, Judith Jarvis Thomson’s famous argument on abortion. Accepting for the sake of argument that a foetus has the same rights to life as an ordinary person, Thomson analogises to a victim who is kidnapped to keep a dying violinist alive (Thomson, 1971). If we would reasonably object to being involuntarily conscripted into saving this person’s life (even if we were the only one who could save them), then—Thomson argues—shouldn’t we on pain of inconsistency accept that it is morally wrong to force a pregnant woman to carry a child to term?

It may be objected that there are morally relevant differences between Thomson’s violinist scenario and the position of a pregnant woman considering a termination. Responding thoughtfully to this charge, Thomson’s argument carefully considers a wide range of factors to ensure that the analogy tracks abortion’s morally relevant features.

However, one factor Thomson doesn’t consider is the sheer unexpectedness of what befalls the subject in her thought-experiment. Wrenched out of her normal life into a shockingly unfamiliar situation, the subject is asked to do something that no-one else has ever been asked to do, and that no-one could have any reasonable expectation she would do.

But this setup clearly differs from the workings of a stable legal environment—or even just established social norms—governing issues of pregnancy and termination. In such a case there would be settled expectations about behaviour that have been set by past behaviour. So if Thompson’s analogy is going to work, then we need to make sure that the sheer novelty of what has happened isn’t doing the work in piquing our intuitions about right and wrong in the case she presents. We need to ensure that the horror we feel for what has happened to the kidnapped subject isn’t being elicited by any one-off novelty aspect of the setup. To check this, we need a case like the violinist scenario, but one where there is a settled system and entrenched expectations.

So let us imagine a case where the expectations work differently. Consider a scenario where a lethal disease randomly affects one in ten adolescents as they enter puberty, and can only be treated by being intravenously hooked up to a non-relative with an appropriate blood match for nine months. Imagine that an isolated community facing this horrifying disease institutes a custom where lots are randomly drawn amongst viable adult donors (who have not already performed the task) to be the donor in any given case. Imagine further that this practice has been consistently followed by all community-members for the last couple of decades, saving many lives, and relieving the desperate anxiety of parents and children alike.

Within this established context, an aspiring young violinist is struck down with the disease, and Thomson’s subject is selected randomly by the lottery to be her life support system. The question arises: Does our subject have a prima facie moral obligation to perform her expected role and save the young musician? The relevant parameters are similar to Thomson’s scenario: she is faced with the question of whether to make a major sacrifice to save the life of a stranger. But there is difference too. In this case, there is a system in place. A system that, perhaps, could have benefited the subject herself, if she had been afflicted with the disease when she was an adolescent. Moreover, the longstanding system will have created established expectations. Perhaps accommodations are made for people who are asked to endure the sacrifice. Perhaps adults can plan for this contingency, working out ahead of time how their affairs could continue if the lottery chooses them. Perhaps the subject’s expectations that she would be saved (if she had been afflicted as an adolescent) had been a source of solace for her when she herself was entering puberty. Perhaps other community members have relied on the system when making various life-decisions, including interacting with our subject on the assumption that she was a committed member of the community who would play her part in their system if required.

In this case, do we feel that the subject would have an obligation to play her role in this system saving the lives of the afflicted teenagers? My intuitions, at least, are far more conflicted in this case than in the original violinist scenario. The fact of the descriptive norm, the expectations attending it, and the consequences of breaching it, seem to me to rightly impact on our emotional and cognitive evaluation of what should be done in this case.

My point here is not to assert that this issue of novelty delivers a fatal objection to Thompson’s argument—only to show that we must be very careful stipulating how expectations and established practices are at work in our thought experiments. There can be no ‘gap’ between the expectations that would attend the issue in the imagined intuition-pump, and those that would be present in its actualization in the real world.

Ethical guidance

Moral philosophers and other (e.g. professional) teachers of ethics can have a role in giving guidance on ethical decision-making to their students. This requires more than speaking about universal principles of justice and benevolence. It is critical to give ethical guidance that makes people direct their attention to expectations, including others’ expectations of their behaviour, and their own expectations about others, as they might arise in a given case. Are other people relying on them? If so, are their expectations reasonable?

This explicit focus on expectations is crucial for two reasons. First, these expectations may have ethical weight attached to them: there might be a reason (or several reasons) why people’s reliance on our conformity has moral weight. Attending to the way consent, fairness and non-interference link with expectations may crucially inform our ethical judgment. No doubt it is morally informative to know, in the abstract, what the best Kantian or rule-utilitarian norm would be to deal with a generic problem. However, once decisions must be made about practical action or reform efforts in a specific context, it is critical to enquire into the existence of local practices and the expectations they have engendered.

Second, because expectations often have ethical weight attached to them, we need to be aware we might be misled into giving a false weighting to expectations in a given case. Because conformity has delivered ethically solid responses in prior cases, we will intuitively want to conform. This will often be our default, fallback stance—and it is usually a good idea. But as the Milgram case showed, we can be beguiled into behaving wrongfully by being unwilling or unable to stand up against another’s expectations, especially if they are confident and unwavering.

As such, a key part of ethical decision-making guides needs to direct people to think seriously about the expectations that are placed on them in a given scenario. Are they clueing us in to important but subtle ethical factors in the situation? Or is our tendency to fall in line with others’ expectations leading us morally awry? These are crucial practical questions for a moral agent to consider.

Breakey, Hugh (2022) ‘When Normal is Normative: The Ethical Significance of Conforming to Reasonable Expectations’ Philosophical Studies.

Milgram, S. (1963). Behavioral Study of Obedience. Journal of Abnormal and Social Psychology, 67(4), 371-378.

Piovarchy, A. (2021). Situationism, capacities and culpability. Philosophical Studies.

Thomson, J. (1971). A Defense of Abortion. Philosophy and Public Affairs, 1(1), 47-66.