III. Moral Agency
I have two questions to answer here, (1) What is moral agency? and (2) What conditions are necessary for its
growth and exercise? In answering these questions, I am most indebted to an analysis by Alasdair MacIntyre3. Moral agency is the idea that a person is, and ought to be held, responsible for both their actions and
the consequences of their actions. This of course implies some framework for judging the value of an action and its
consequences, and it implies the possibility of reward or punishment for certain kinds of actions. Generally, a
person is not considered to be a moral agent when it is not reasonable to hold them responsible for their actions.
Thus a small child is not a moral agent, but becomes one later in life, an example which is particularly instructive
for my purposes.
A small child is not a moral agent for several reasons:
(1) she does not yet have a concept of ‘self-identity’ with her
actions. This means that she has not developed the capacity to recognize the connection between herself as a
person who intends something, herself as a person who takes actions, and herself as the perpetrator of unintended
consequences. A person who cannot even recognize the causal connection between their intentions and their actions
certainly is not reasonably held accountable for those actions.
(2) After (or concurrent with) developing the concept
of identity with actions, the child must also develop some methods for predicting the possible outcomes of an action.
Ordinarily, a person is not responsible for the unintended consequences of their actions (their accidents), when they
could not reasonably have predicted the results of those actions. A child needs to be able to predict in order to plan,
and thus, to be held responsible.
(3) After (or concurrent with) developing the ability to predict outcomes, the child
must also learn to predict the desirability of these potential outcomes. If she could not have known that a certain
unintended consequence far outweighs the intended benefit of an action, then she could not reasonably have
decided not to act in response to that undesirability. Thus it would not be reasonable to hold her responsible for
having acted anyway. Not possessing any one of these abilities would be sufficient to determine that the child is not
fully responsible for her actions, but we have not yet found all the qualities of a full moral agent.
For example, many former members of the Nazi regime in World War II Germany were found guilty of crimes
against humanity, despite their protestations that they were merely "doing their duty". Their defense rested on the
claim that they could not be held responsible for taking actions which were required by their societal positions, for
not considering factors which their society judged to be irrelevant, and for acting to achieve results which their
society judged to be ‘good’.
But many of these people were found guilty at Nuremberg, a precedent which depends on the premise that all adult
humans (or at least, these adult humans) are capable of questioning even the most fundamental assumptions of
their society and position, and that they are culpable for failing to question these assumptions. Essentially, the
claims of these individuals that they had acted according to the standards of their society opened them up to the
question, "Why were those the best standards available?" and to the charge that they should have questioned
This ability completes the list of qualities of a full moral agent, leading to the following definition:
A moral agent has to understand herself as a moral agent, which means to hold herself responsible, and
to be held responsible by others, for (a) that in her actions which is intentional, (b) that in her actions of
which she should have been aware, (c) the "reasonably predictable" effects of her actions, and (d) the
criteria by which she decides what standards to apply in evaluating a-c.
Several comments are in order.
(1) On the apparent circularity of the definition: to practice moral agency one
must be aware of the responsibilities one holds—one cannot reasonably be held responsible by others when one
cannot hold oneself responsible. This means that the moral agent has to understand herself as a moral agent,
which, as the definition states, means that she must understand and hold herself responsible for the four aspects in
(2) the definition will serve as the ideal of moral agency for my purposes. But as will be clear
from the development of the requirements of the definition, moral agency should be seen as a quality that occurs in
degrees. As the example of child development showed, one can be taken to be responsible to differing degrees at
different times in one’s life, and to differing degrees with respect to different kinds of decisions. (e.g. A student
cannot be held responsible in the same way as the instructor for an accident in a chemistry lab. The difference in
their two situations in terms of knowledge and accountability require different degrees of responsibility for the
effects of their actions. Thus they are moral agents to different degrees in some of their decisions.)
definition places two kinds of requirements on the full moral agent: she must have information, and she must have
certain skills in evaluating that information. Naturally, these skills are complementary, as skill in evaluating
information leads to skill in acquiring information, and further acquisition of information leads to refinement of the
skills of evaluating that information. Information is required for the agent to fully understand the situation in which
she finds herself. She must know facts about possible actions and their possible outcomes, knowledge of which is
partly acquired developmentally (e.g. the skill to recognize causal connections), partly through direct effort on the
part of the agent (e.g. doing research on past attempts to perform the actions she is contemplating), and partly
through passive reception by the agent (e.g. having learned in the past through the news, schooling, membership in
various groups, etc. about similar situations).
Skill at evaluating this information implies the ability to recognize when the information is sufficient/insufficient,
when it is trustworthy/untrustworthy, and it also implies the ability to evaluate objectives, to determine the
desirability of different outcomes. This is most clear in part (d) of the definition, which is the point at which values
enter the decision to act—the agent must consider what kinds of consequences are relevant to their position, which
consequences are more important than others, and what would constitute a ‘good’ result. There exist a myriad of
different frameworks for evaluating these possibilities, embodied in religion, law, custom, and personal principles.
Part (d) of the definition requires both that the agent consider her actions from within some ethical framework, and
also that she must actively consider and criticize different frameworks in her decisions, and that she must
understand herself as capable of such deliberation. It is through this kind of consideration that people exercise
different frameworks in different circumstances—actions which are considered ethical in the workplace are not
considered ethical at home. The killing of a person by a soldier would probably be judged as good, but a civilian
killing the same person would be condemned as a murderer. Part (d) also brings in the question of
relevance—according to some frameworks of judgement, and in some circumstances, an action which results in the
death of, say an insect, is not judged to be bad, because the death of the insect is considered irrelevant. In other
situations, and in other frameworks, that death would be the most important consideration.
These questions are exceedingly difficult to deal with—asking them, and dealing with them appropriately, requires
a certain kind of society, a society which actively encourages the questioning of values and the frameworks through
which judgements are made, to support the growth of the ability implied by part (d) of the definition. Specifically,
society must provide milieus in which it is both possible and safe to question societal values, and to compare
alternative frameworks. This requires education that encourages the critical mindset, examples of the critical
lifestyle that others might seek to emulate, and media through which people can communicate their concerns
regarding different values.
It is this last that is most important, as it is conceivable that a nonconformist genius might learn a critical mindset
alone, but to fully exercise this mindset, she would require feedback from others to ensure the correctness of her
criticism. Some might object here that it is possible to achieve correct moral results in isolation. But I suggest that
if morality and moral responsibility are to mean anything, they must also entail accountability to other members of
society. The reason is captured in the first part of my definition, "a moral agent must … hold herself responsible,
and be held responsible by others for…" Moral responsibility is formed communally, it is a relationship—between
people, their actions, and the criteria for evaluating those actions. In forming or joining a group which holds
particular values (e.g. taking a job at a particular company), one accepts to some extent the values of that group.
One’s performance as an employee is evaluated by others in the company in relation to the standards set by the job
position, which are learned through observation, questioning, and through trial and error—but always in dialogue
with other employees, managers, etc. In short, one learns the standards of their role in cooperation with those
others who share the same standards and obligations. A person only understands their role, their responsibilities,
and the consequences of their failures, in communication with others in their company. Thus a person who chooses
to abrogate the values or particular responsibilities of her social position may be justified, but she owes some
explanation to the group whose values she has violated, and to society at large for the correctness of the values
through which the judgement was made.
Thus communication and a supportive, critical community, are essential to the practice of critical moral judgements
(also see MacIntyre, 316-317). This feedback would take the form of questioning between individuals regarding
values and judgements, leading to mutual understanding of the reasons for the decision that is ultimately reached.
Such a process is not always possible, for it is unsafe to publicly question values in some societies, and in others the
means of communication are limited or impractical, and in some others, such processes are implicitly and explicitly
I have outlined a definition of moral agency, and identified two kinds of requirements for the exercise and
development of that agency: informational, and evaluational. Both of these requirements can be met in part by the
concerted attention of the agent, seeking out information and developing the ability to reflect critically on values
and judgements. But both of these requirements must also be met by society at large, for it is not reasonable to
expect the individual to root out all the relevant information for every decision—as some of that information would
be located in distant places and difficult to acquire. And it is not reasonable to expect the individual to
independently develop the critical abilities required by part (d), as such abilities presuppose rational confidence in
one’s decisions, which is only acquired through a process of collective dialogue and criticism, as I discussed above.
It is therefore a responsibility of the society, as well, to supply the informational and evaluational requirements of
the moral agent, if full moral agency is to be a realistic possibility for the members of that society.