Moral Responsibility in the Age of Bureaucracy

Author(s): David Luban, Alan Strudler and David Wasserman

Source: Michigan Law Review, Vol. 90, No. 8 (Aug., 1992), pp. 2348-2392
Published by: The Michigan Law Review Association
Stable URL: https://www.jstor.org/stable/1289575
Accessed: 09-04-2025 10:03 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

The Michigan Law Review Association is collaborating with JSTOR to digitize, preserve and
extend access to Michigan Law Review

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
 MORAL RESPONSIBILITY IN THE
AGE OF BUREAUCRACY

David Luban, *
Alan Strudler, **
and David Wasserman***

The ranks of officials in this judiciary system mounted endlessly, so that

not even the initiated could survey the hierarchy as a whole. And the pro-
ceedings of the Courts were generally kept secret from subordinate offi-
cials, consequently they could hardly ever quite follow in their further
progress the cases on which they had worked; any particular case thus
appeared in their circle ofjurisdiction often without their knowing whence
it came, and passed from it they knew not whither. Thus the knowledge
derived from a study of the various single stages of the case, the final
verdict and the reasons for that verdict lay beyond the reach of these
officials.
Franz Kafka, The Trial1

I. INTRODUCTION: BUREAUCRACY AND NATURAL LAW

No twentieth-century writer has thought so deeply, or so yearn-

ingly, about natural law as Franz Kafka. Kafka's is a world in which
we seek desperately to know the natural law that is sovereign in
human affairs but find that knowledge of the law is withheld from us.
For this reason, we lead our lives in a state of, if not original sin, then
original guilt - guilt for violating the law, or perhaps guilt for not
knowing the law, despite the fact that we wish to know it.
The Trial is Kafka's greatest elaboration of this theme. Joseph K.
is arrested for a crime, but he cannot discover what his crime is. He is
convinced that the arrest is a gigantic misunderstanding, but he is un-
able to bring the inquest to a halt. In the end, he is executed, and

* Professor of Law, University of Maryland; Research Scholar, Institute for Philosophy and
Public Policy. B.A. 1970, University of Chicago; M.A. 1973, M.Phil. 1973, Ph.D. 1974, Yale. -
Ed.

** Research Scholar, Institute for Philosophy and Public Policy. B.A. 1975, Johnston Col-
lege; Ph.D. (Philosophy) 1982, J.D. 1985, University of Arizona. - Ed.
*** Research Scholar, Institute for Philosophy and Public Policy. B.A. 1975, Yale; J.D.
1978, University of Michigan; M.A. (Psychology) 1984, University of North Carolina. - Ed.
We have presented versions of this paper at the University of Baltimore and The Australian
National University, and have benefited greatly from the helpful comments we received. We are
particularly grateful to John Braithwaite, Teresa Chandler, and Jorge Garcia.
1. FRANZ KAFKA, THE TRIAL 133-34 (E.M. Butler rev. definitive ed., Willa Muir & Edwin
Muir trans., Seeker and Warburg 1956) (1925).

2348

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2349

Kafka leaves us with the distinct impression that Joseph K.'s crime is
precisely his inability to discover what crime he has been accused of.
K. never has a formal trial, yet the novel is called The Trial: evi-
dently, K.'s fruitless effort to learn what he is accused of is his trial.
His execution is therefore just, because Joseph K. lives in a state of
culpable ignorance. In the remarkable parable that his confessor re-
lates to Joseph K. in the ninth chapter, a man comes to the court of
justice but is denied admittance. He waits patiently by the gate for
years, and at the moment of his death learns from the doorkeeper that
the gate will now be closed forever, for it was never intended for any-
one other than the dying petitioner.2 The law is real, and sovereign in
human affairs, but forever withheld from us. That is the extremity of
our condition.3
Kafka found the perfect literary image for his legal and theological
theme in The Trial, and that is the image of bureaucracy. In The
Trial, the emblem of the protagonist's inability to learn the law is his
confrontation with a bureaucracy that frustrates and evades his every
effort to get to the bottom of things, to obtain clarification. Assume a
hierarchical world, a great chain of being; then take away the pinnacle
of the hierarchy, leaving only the lower orders who cannot take ulti-
mate responsibility for anything they do. That is the theological situa-
tion that Kafka ponders, but it is also the secular situation of the
bureaucracy. The everyday experience of the bureaucratic runaround
- tedious, preposterous, yet with potentially fatal consequences
gives the word Kafkaesque its common meaning. The epigraph at the
beginning of this paper encapsulates the Kafkaesque world of bureau-

2. Id. at 238-47.
3. In the grisly story "In the Penal Colony," condemned prisoners are executed on a torture
device that slowly inscribes the law they have violated on their bodies with needles. As the
message takes shape over many hours of suffering, recognition and understanding gradually grow
in the prisoners, until finally they attain a kind of spiritual redemption that comes of knowing
why they are suffering and dying. The officer in charge of the apparatus fervently believes that
this form of punishment alone satisfies the requirements of justice. Faced with the imminent
abolition of this form of execution on humanitarian grounds, the despairing officer climbs into
the apparatus himself, programming it to carve "Be Just" on his body. But the machine disinte-
grates; instead of writing "Be Just," the needles merely punch the officer to pieces, and the dead
officer's face "was as it had been in life; no sign was visible of the promised redemption; what the
others had found in the machine the officer had not found." FRANZ KAFKA, In the Penal Col-
ony, in SELECTED SHORT STORIES OF FRANZ KAFKA 126 (Willa Muir & Edwin Muir trans.,
1952).
Like The Trial, the story may be read as a comment on natural law - law that is so much a
part of the human condition that it must be harrowed into our bodies when we transgress it.
This is, evidently, an outward emblem of the fact that the law already dwells in our nature
and, perhaps, that it dwells in our nature cruelly. In Kafka's parable, it is the officer, the one
remaining believer in this law, who alone is denied the redemption it promises - redemption
that is meant to come through ultimate knowledge of natural law engraved in one's own body.
The officer represents the predicament of the natural lawyer who understands that natural law
binds us but, for whatever reason, cannot discover what it asks of him.

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2350 Michigan Law Review [Vol. 90:2348

cracy. It exaggerates, but only slightly, a phenomenon that all of us

recognize at once.
Within our collective imagination, evidently, bureaucracy and nat-
ural law are antithetical to each other; otherwise, Kafka's imagery
would not seem so perfectly appropriate. We begin our own argument
by attempting to vindicate this perception. We believe that the perva-
siveness of the bureaucratic phenomenon in contemporary life threat-
ens to make natural law irrelevant to our political and economic
institutions.
In its classical form, the key idea of natural law is that legal sys-
tems are legitimately instituted in order to promote the common good.
Propositions of natural law impose constraints on precepts of positive
law: if a precept of positive law is not instituted to promote the com-
mon good, it lacks legitimacy - it is not law.
Natural law principles are thus, in the first instance, criteria for
assessing systems of positive law. However, natural law theory also
offers important insights into the dimensions of individual moral and
legal responsibility. The natural law understanding of legal systems as
cooperative efforts to promote the common good implies a moral rela-
tion between those who govern and those who are governed. The gov-
erned bear a moral obligation to obey the law, provided that the law
aims at the common good, whereas those who govern possess author-
ity only to the extent that they undertake to promote the common
good. The two sides of this relation fit together: the citizens' obliga-
tion to obey the law depends on the rulers' undertaking to legislate for
the common good.4
This doctrine of individual responsibility on the part of rulers and
ruled alike has played a prominent role in the development of twenti-
eth-century international law. Positivism, including legal realism, was
the ascendant view among legal theorists outside the Catholic Church
for the century between the 1830s and the 1930s. World War II
changed that. If there is a single historical event that accounts for the
revival of serious secular interest in natural law, it is surely the Nu-
remberg trials. The Nuremberg Tribunal held individual Nazi officials

4. At this point, however, natural law theorists diverge. Philip Soper, for example, contends
that a good faith belief on the part of the rulers that they are legislating for the common good
suffices to establish a moral obligation on the part of the ruled to obey the law. PHILIP SOPER, A
THEORY OF LAW 79 (1984). John Finnis, by contrast, believes that positive law's inconsistency
with the common good undermines the moral obligation to obey, regardless of what the legislator
may believe. JOHN FINNIS, NATURAL LAW AND NATURAL RIGHTS 351-60 (1980). Thus, the
notion of "the rulers' undertaking to legislate for the common good" may be understood subjec-
tively (Soper) or objectively (Finnis). Elsewhere, one of us has criticized Soper's argument.
David Luban, Conscientious Lawyers for Conscientious Lawbreakers, 52 U. PITT. L. REV. 793,
806-09 (1991).

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2351

responsible for acts that positive law did not forbid at the time they
were committed - so-called "crimes against peace" and "crimes
against humanity."5 Anticipating the defendants' protest that they
were merely following official orders that carried the force of positive
law, Article 8 of the Nuremberg Charter specifically provided that
"[t]he fact that the defendant acted pursuant to an order of his govern-
ment or of a superior shall not free him from responsibility."6
That Article 8 represents a flat-out rejection of what might be
called the positivist excuse for atrocious official acts - the excuse that
the acts were licensed by positive law - is intuitively clear. As Stan-
ley Paulson has shown, the defense at Nuremberg relied extensively
upon the positivist excuse, which the Tribunal had little difficulty re-
jecting.7 The natural law argument that unjust laws lose their obliga-
tory character provides a straightforward philosophical justification
for Article 8. Similarly, appeals to natural law clearly form the most
obvious justification for criminalizing "murder, extermination, en-
slavement, deportation, and other inhumane acts ... whether or not in
violation of domestic law of the country where perpetrated."8 Such
crimes against humanity are radically inconsistent with the common
good, and any domestic legal system that permits them must violate
natural law. In addition, Article 7 of the Charter eliminated the act-
of-state defense on the part of those in command positions,9 thereby
recognizing that those who legislate bear moral responsibilities just as
surely as do those who follow orders.
Indeed, the view that the Nazi era exposes the moral deficiency of
positivism compared with natural law is a commonplace. As early as
the mid-1940s, Gustav Radbruch, an eminent pre-War German posi-
tivist, repudiated positivism and embraced natural law, arguing in sev-
eral influential essays that positivism had disarmed German jurists in
the face of Nazism. 0 Thus, international revulsion at the official crim-

5. For an argument that the criminalization of crimes against peace and crimes against hu-
manity represented a legal novelty, see David Luban, The Legacies of Nuremberg, 54 Soc. RES.
779, 797-801 (1987).
6. 1 INTERNATIONAL MILITARY TRIBUNAL, TRIAL OF THE MAJOR WAR CRIMINALS
BEFORE THE INTERNATIONAL MILITARY TRIBUNAL 12 (1947).
7. Stanley L. Paulson, Classical Legal Positivism at Nuremberg, 4 PHIL. & PUB. AFF. 132
(1975).
8. CHARTER OF THE INTERNATIONAL MILITARY TRIBUNAL art. 6(c) (defining crimes
against humanity).
9. See CHARTER OF THE INTERNATIONAL MILITARY TRIBUNAL art. 7.
10. See GUSTAV RADBRUCH, Five Minutes of Legal Philosophy, in PHILOSOPHY OF LAW
109, 109-10 (Joel Feinberg & Hyman Gross eds., Stanley L. Paulson trans., 3d ed. 1980) (1950)
(translating GUSTAV RADBRUCH, FunfMinuten Rechtsphilosophie, in RECHTSPHILOSOPHIE 335,
335-37); GUSTAV RADBRUCH, Gesetzliches Unrecht und Ubergesetzliches Recht [Legal Injustice
and Supralegal Justice], in RECHTSPHILOSOPHIE, supra, at 347. Radbruch's arguments proved

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2352 Michigan Law Review [Vol. 90:2348

inality of Hitler's regime, as manifested legally in the Nuremberg tri-

als, represents a triumph for natural law thinking. 1 At the very least,
this worldwide condemnation provisionally lays to rest the positivist
excuse in international law.12
But the excuse that "I was only following orders," or "I was only
doing my job," was not the only one offered by those implicated in
Nazi crimes. In the aftermath of World War II, the world heard with
equal frequency the cry, "I didn't know!" This is the epistemological
excuse, whose elements present the problem that will occupy our at-
tention in this article. Though it is often insincere, and seldom en-
tirely persuasive, the epistemological excuse seems to come naturally
to those who commit wrongs in a bureaucratic setting. We shall argue
(1) that bureaucracies function (often by design) to permit their func-
tionaries to truthfully plead the excuse "I didn't know!"; (2) that
traditional accounts of moral responsibility typically recognize this
epistemological excuse; and (3) that it is therefore very difficult to find
a workable account of moral responsibility within bureaucratic institu-
tions. The strength and prevalence of the epistemological excuse may
render the historic rejection of the positivist excuse an empty or very
partial victory.
We cannot overemphasize the importance of this point, for per-
haps the single most salient characteristic of the Nazi crimes was their
bureaucratic nature. They were committed, not by a lawless gang of
criminals, but by a regularly functioning state bureaucracy executing
official policies.13 Not only Nazi crimes have this bureaucratic charac-
ter - it exists as well in the misdeeds of the recently departed regimes
of the Soviet empire. The emerging democracies of Eastern Europe
if democracies are what they prove to be - are beginning to come to
grips with the bureaucratic crimes committed for generations by offi-

influential in English-language jurisprudence, largely through the prominent place they occupied
in the well-known Hart-Fuller debate. See H.L.A. Hart, Positivism and the Separation of Law
and Morals, 71 HARV. L. REV. 593, 615-21 (1958); Lon L. Fuller, Positivism and Fidelity to Law
- A Reply to Professor Hart, 71 HARV. L. REV. 630, 646-47 (1958). More recently, however,
Ingo Miiller has criticized Radbruch's view, arguing that the worst perversions of justice in the
Third Reich occurred because the Nazi judges were not good enough positivists - they were too
willing to pervert the law for political reasons. See INGO MULLER, HITLER'S JUSTICE (Deborah
L. Schneider trans., 1991). This critique, of course, does not affect the argument that Nuremberg
law represents a repudiation of positivism, nor does it render the positivist excuse acceptable.
11. For an argument that the Nuremberg Charter did not go far enough in this direction
because it safeguarded the principle of national sovereignty, the linchpin of positivism, see
Luban, supra note 5, at 780-90.
12. Regrettably, the acquittal of Oliver North suggests that the excuse lives on in American
criminal law.
13. Rhetorical references to the Nazis as gangsters, criminals, or the like often obscure this
important point. See Luban, supra note 5, at 817-20.

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2353

cials of their former regimes, many of whom will surely enter the epis-
temological excuse. Czech novelist Milan Kundera poses the problem
directly in The Unbearable Lightness of Being:
Let us concede that a Czech public prosecutor in the early fifties who
called for the death of an innocent man was deceived by the Russian
secret police and the government of his own country. But now that we
all know the accusations to have been absurd and the executed to have
been innocent, how can that selfsame public prosecutor defend his purity
of heart by beating himself on the chest and proclaiming, My conscience
is clear! I didn't know!14
One might respond with an equally rhetorical question: how can
the prosecutor be blamed if he truly did not know? Kundera's rage
clearly stems from the ready availability of the epistemological excuse
within the secretive Communist bureaucracies, but the fact that an
excuse is a bit too handy does not in itself undercut its viability. Thus,
the problem of bureaucratic irresponsibility faces post-Communist so-
cieties much as it faced post-Nazi Germany.
Bureaucratic evasion of responsibility, however, is not a pathology
confined to police states. A recent article in The Washington Post de-
scribes the loss of six million dollars of the savings of Washington area
Hispanics when the unregulated Latin Investment Corporation failed.
When depositors blamed lax regulation, a D.C. politician responded
that no one in government was to blame. "Frankly, I think the re-
sponsibility belonged to several agencies, including federal agencies
.... Thus, effectively, the responsibility belonged to no one."15 The
bureaucratic evasion of responsibility is as American as your failing
neighborhood bank.
Nor is bureaucratic irresponsibility a pathology limited to govern-
ments. All of us live our lives in the sway of many nongovernmental
bureaucracies - HMOs, business corporations, large employers, and
the like - that function as "private governments" (and often as re-
markably autocratic ones).16 Particularly now, when the erstwhile so-
cialist nations of central and eastern Europe have committed
themselves to programs of privatization and capitalism, and their mas-
sive bureaucracies prepare to shift allegiance from the state to private
owners, the danger of bureaucratic evasion inherent in private govern-
ments as well as public should be clear. The problem of individual

14. MILAN KUNDERA, THE UNBEARABLE LIGHTNESS OF BEING 177 (Michael H. Heim
trans., 1984).
15. Charlene Drew Jarvis, D.C. Council Member, quoted in Carlos Sanchez, A Year Later,
Depositors Still Waiting, Hoping, WASH. POST, Dec. 1, 1991, at B3.
16. See Stuart Macaulay, Private Government, in LAW AND THE SOCIAL SCIENCES 445
(Leon Lipson & Stanton Wheeler eds., 1986).

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2354 Michigan Law Review [Vol. 90:2348

responsibility within bureaucratic institutions straddles the divide be-

tween private and public.
Natural law theories have functioned in the twentieth century to
strip away the positivist excuse for official wrongdoing. This will be a
hollow triumph, however, if in the end official malefactors may fall
back on the epistemological excuse. As bureaucratic institutions in-
creasingly come to pervade modem life, the contrast between natural
law and positivism threatens to become moot unless a meaningful con-
ception of individual responsibility within organizational settings can
be formulated.
A more direct reason exists to explain the tension between natural
law and the bureaucratic phenomenon, which drastically expands the
availability of the epistemological excuse. Natural law, remember, in-
sists that institutions are legitimate only when they aim to promote the
common good. The common good includes, of course, the prosperity,
stability, solidarity, and liberty of the community. But that cannot be
the end of the story. Surely one central aspect of the common good
lies in what we might call the moral intelligibility of our lives. A com-
munity is worse off to the extent that its members are unable to make
moral sense of the lives that they and their fellow citizens lead. Moral
intelligibility enhances the sense of meaningfulness and mutual assur-
ance in a community, whereas moral opacity undermines it.
The Kafkaesque world of bureaucracy is morally opaque. That
Kafka's world often seems to have the character of a bad dream in its
combination of tedium, oppressiveness, and withheld meaning is no
accident. As the passage we have taken as our epigram suggests, the
horror of the bureaucratic process lies not in officials' mechanical ad-
herence to duty, but rather in the individual's ignorance of what the
fulfillment of his or her duty may entail. Everyone operates in an epis-
temological and therefore moral vacuum. Interestingly, it was the
epistemological excuse, not the positivist excuse, that captured
Kafka's imagination in The Trial Though one might interpret the
epistemological theme of The Trial as our inability to know the natu-
ral law, another interpretation is equally plausible: We all know what
the natural law commands, but because of the way we lead our lives,
and the institutions within which we live them, we never know
whether we have transgressed it. The individual actions that make up
our lives have become ciphers, action shards whose moral character
we - like Joseph K. - are unable to determine.
In its stringency, Kafka's moral vision is an optimistic one.
Though we may wish to plead the epistemological excuse, in Kafka's
world we do so to no avail. This may be wishful thinking. As

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2355

Kundera recognizes, the problem in the world we live in is not that the
epistemological excuse fails, but that it may succeed all too well.
Thus, the moral relationship between those who rule and those who
are ruled dissolves, and natural law threatens to recede into the realm
of useless abstraction. If bureaucracy indeed installs the epistemologi-
cal excuse as a standing option in our moral lives, then bureaucratic
institutions make a mockery of natural law ideals.
In this article, we shall examine the sources of the epistemological
excuse and the moral resources available for restricting its application.
Our topic, then, is the often-noticed but poorly understood fragmenta-
tion of knowledge and responsibility in large organizations, including
government, business corporations, and professional groups. We wish
to investigate the compartmentalization, mutual buckpassing, and
deniability that too often leads organizations to commit wrongs for
which no individual in the organization seems genuinely responsible.
Bureaucratic organizations parcel out morally significant knowl-
edge among various individuals along the same lines as organizational
tasks. The division of labor is equally a division of knowledge. Super-
visors may not know of wrongful actions by subordinates implement-
ing management decisions, while subordinates may believe they have
been left no discretion and no alternatives. Put these conditions to-
gether and you have a recipe for organizational wrongdoing that will
never trouble the conscience of anyone within the organization. Indi-
viduals within the organization do not know, or perhaps do not want
to know, what their actions add up to.
Recent psychological and sociological research on wrongdoing in
organizations has examined the pressures individuals feel from their
superiors, their peers, and the norms of corporate culture to engage in
wrongful conduct. This research has yielded profound insight into the
subtle but powerful coercive forces at work in organizations. It has
not paid sufficient attention, however, to the cognitive aspects of obe-
dience and conformity, the structural features of large organizations
that prevent individuals from obtaining the knowledge they need to
make informed moral decisions and resist pressure to obey and con-
form. Individuals in organizations frequently lack awareness of the
role their acts play in the larger corporate undertaking; their igno-
rance complicates both the moral assessment of and the practical re-
sponse to the wrongs they help commit. Analyzing the problem of
fragmented knowledge (as we shall call it) is critical for an adequate
philosophical analysis of organizational wrongdoing and for an effec-
tive practical response to it.
We contend that traditional accounts of morality have failed to

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2356 Michigan Law Review [Vol. 90:2348

deal adequately with the problem of fragmented knowledge, in part

because they treat the problem as peripheral rather than central, in
part because they take too narrow and episodic a view of moral deci-
sionmaking. We contend that fragmented knowledge in bureaucratic
organizations is one of the central moral problems of our time and that
an adequate response to this problem requires us to broaden the scope
of moral prescription and appraisal.

II. BACKGROUND OF THE PROBLEM

The bureaucratic fragmentation of knowledge and dilution of re-

sponsibility are pervasive phenomena in modem society. To set the
stage for our analysis, we first describe the scope of the problem and
briefly review some of the research, commentary, and debate it has
provoked. We conclude this background section by discussing the re-
search most relevant to our own concerns, the Milgram studies of de-
structive obedience to authority.

A. The Collectivization of the Workplace

Most work in modem society is done by organizations: corpora-

tions, governments, hospitals, foundations, universities, accounting
firms, armies. Even such supposedly independent professionals as
physicians and lawyers practice in large organizations to an ever-in-
creasing extent. The HMO has replaced the family physician, and the
new graduates of today's law schools join firms, of which the largest
now employ over a thousand lawyers, rather than hanging out a shin-
gle. The problems of professional and business ethics have thus be-
come the problems of supervisors and subordinates in organizational
settings. Indeed, in a culture such as ours, where our first question to
each other is often not "How do you do?" but "What do you do?", the
ethics of the workplace has enormous impact on how we think of mo-
rality in general. To a great extent, ethics in the organizational setting
has come to define ethics as a whole. We speak of team players and
loose cannons, leaders and followers, as categories of moral judgment
and not simply of social description.

B. The Organization Man and the Other-Directed Society

The transformation of the workplace appears to have wrought a

transformation in values, replacing individual responsibility and inter-
nal norms with group identification and external norms. As the post-
war American economy assumed its contemporary form, several
leading social scientists and commentators explored the psychology of

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2357

"The Organization Man," in the famous title of William H. Whyte's

book. Whyte used this term to describe "the ones of our middle class
who have left home, spiritually as well as physically, to take the vows
of organization life."17 He ascribed to them the "Social Ethic," which
includes "a belief in the group as the source of creativity" as well as "a
belief in 'belongingness' as the ultimate need of the individual."18
David Riesman described middle-class Americans as a "Lonely
Crowd," and elaborated a famous typology of characters. In Ries-
man's scheme, people of premodern societies were tradition-directed,
and the sanction for deviation was shame; in early moder societies
people were inner-directed, guided by an internal moral compass, ac-
quired in childhood, which induces guilt when one deviates. In con-
temporary society, however, we have become other-directed: our
"contemporaries are the source of direction for the individual ....
[T]he process of paying close attention to the signals from others ...
remain[s] unaltered throughout life."19 For other-directed individu-
als, the sanction for deviance has changed: "As against guilt-and-
shame controls, though of course these survive, one prime psychologi-
cal lever of the other-directed person is a diffuse anxiety."20 Sociolo-
gist Robert Jackall conducted interviews with 143 managers in several
contemporary American corporations. In the anxiety-ridden world of
middle management, "[m]anagers have a myriad of aphorisms that re-
fer to how the power of CEOs, magnified through the zealous efforts of
subordinates, affects them .... [One such maxim is] "When he
sneezes, we all catch colds" ....21 Jackall comments:
As a result, independent morally evaluative judgments get subordinated
to the social intricacies of the bureaucratic workplace. Notions of mo-
rality that one might hold and indeed practice outside the workplace ...
become irrelevant .... Under certain conditions, such notions may even
become dangerous. For the most part, then, they remain unarticulated
lest one risk damaging crucial relationships with significant individuals
or groups.22

C. Historical Perspective

The collectivization of the workplace and the threat it poses to

traditional moral values are hardly new phenomena; they have been

17. WILLIAM H. WHYTE, JR., THE ORGANIZATION MAN 3 (1956).

18. Id. at 7.
19. DAVID RIESMAN, THE LONELY CROWD 22 (1950) (emphasis omitted).
20. Id. at 26.
21. ROBERT JACKALL, MORAL MAZES: THE WORLD OF CORPORATE MANAGERS 22
(1988).
22. Id. at 105.

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2358 Michigan Law Review [Vol. 90:2348

recognized, and lamented, for the past 150 years. The erosion of indi-
vidual responsibility and the evils of bureaucracy have engaged con-
servative writers since the advent of the industrial revolution. Over a
century ago, Karl Marx likewise criticized what he called "the real
mindlessness of the state." "The bureaucracy is a circle from which
no one can escape," Marx contended. "The highest point entrusts the
understanding of particulars to the lower echelons, whereas these, on
the other hand, credit the highest with an understanding in regard to
the universal; and thus they deceive one another."23 In 1932, Rein-
hold Niebuhr wrote his classic treatise Moral Man and Immoral Soci-
ety, in which he argued that
[i]ndividual men may be moral .... They are endowed by nature with a
measure of sympathy and consideration for their kind, the breadth of
which may be extended by an astute social pedagogy .... But all these
achievements are more difficult, if not impossible, for human societies
and social groups. In every human group there is less reason to guide
and to check impulse, less capacity for self-transcendence, less ability to
comprehend the needs of others and therefore more unrestrained egoism
than the individuals, who compose the group, reveal in their personal
relationships.24

Niebuhr's argument recognizes that the increasing organization of so-

ciety will be accompanied by a dilution of morality.
As we have already observed, the problems Marx and Niebuhr dis-
cussed in a theoretical vein came to life in the most horrible way possi-
ble during World War II, where ostensibly civilized human beings
tortured and slaughtered twelve million men, women, and children in
extermination camps. The names of the camps - Auschwitz, Treb-
linka, Majdanek - have become synonymous with the incomprehen-
sible willingness of ordinary human beings to do anything, no matter
how atrocious, when ordered to do so by those in authority. Here,
again, an explanation may be offered in terms of the division of respon-
sibility within groups. Consider a historian's description of the eutha-
nasia program Hitler ordered to eliminate mentally retarded,
handicapped, or genetically ill Germans (individuals Hitler called
"useless eaters"):
The euthanasia program ... demonstrated how, through fragmentation
of authority and tasks, it was possible to fashion a murder machine.
Hitler had enunciated an offhand, extralegal decree, and had not wanted
to be bothered about it again. Brandt had ordered the "scientific" imple-
mentation of the program and, like Hitler, wished to hear no complaints.
The directors and personnel of institutions rationalized that matters

23. KARL MARX, CRITIQUE OF HEGEL'S "PHILOSOPHY OF RIGHT" 46-47 (Joseph O'Malley
ed., Annette Jolin & Joseph O'Malley trans., 1970).
24. REINHOLD NIEBUHR, MORAL MAN AND IMMORAL SOCIETY xi-xii (rev. ed. 1960).

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2359

were out of their hands and that they were just filling out questionnaires
. .., though in reality each form was the equivalent of a death warrant
.... The personnel at the end of the line excused themselves on the basis
that they were under compulsion, had no power of decision, and were
merely performing a function. Thousands of people were involved, but
each considered himself nothing but a cog in the machine and reasoned
that it was the machine, not he, that was responsible.25
The horrors of Nazism are without parallel, but the bureaucratic pat-
tern of organization that fragments the knowledge required for moral
decisionmaking is common to large institutions throughout contempo-
rary society. Jackall describes the typical corporate structure in terms
not unlike those Marx used to characterize "the real mindlessness of
the state":
Power is concentrated at the top in the person of the chief executive
officer (CEO) and is simultaneously decentralized; that is, responsibility
for decisions and profits is pushed as far down the organizational line as
possible.

... [P]ushing details down protects the privilege of authority to de-

clare that a mistake has been made.... Moreover, pushing down details
relieves superiors of the burden of too much knowledge, particularly
guilty knowledge.

... [Middle managers] become the "point men" of a given strategy

and the potential "fall guys" when things go wrong.26
Hannah Arendt described the bureaucratic phenomenon as a novel
form of governance appearing alongside the classical distinction
among rule by one (monarchy), rule by "the best" (aristocracy), rule
by the few (oligarchy), and rule by the many (democracy). She wrote
of

the latest and perhaps most formidable form of... dominion: bureau-
cracy or the rule of an intricate system of bureaus in which no men,
neither one nor the best, neither the few nor the many, can be held re-
sponsible, and which could be properly called rule by Nobody. (If, in
accord with traditional political thought, we identify tyranny as govern-
ment that is not held to give account of itself, rule by Nobody is clearly
the most tyrannical of all, since there is no one left who could even be
asked to answer for what is being done. It is ... impossible to localize
responsibility and to identify the enemy ... .)27
Such rumors of the demise of responsibility may be exaggerated; yet
Arendt's description has the ring of familiarity. A graphic contempo-

25. ROBERT E. CONOT, JUSTICE AT NUREMBERG 210-11 (1983).

26. JACKALL, supra note 21, at 17, 20-21.
27. HANNAH ARENDT, ON VIOLENCE 38-39 (1970).

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2360 Michigan Law Review [Vol. 90:2348

rary analogue appeared in litigation surrounding the Dalkon Shield.

In his opinion, Federal Judge Frank Theis angrily noted:
The project manager for Dalkon Shield explains that a particular ques-
tion should have gone to the medical department, the medical depart-
ment representative explains that the question was really the bailiwick of
the quality control department, and the quality control department rep-
resentative explains that the project manager was the one with the au-
thority to make a decision on that question.... [I]t is not at all unusual
for the hard questions posed in Dalkon Shield cases to be unanswerable
by anyone from Robins [the manufacturer].28
One must not be naive, of course: often the defense of fragmented
knowledge will be entered falsely and cynically, as a form of liability
screening. Executives in the hot seat should be treated with the same
skepticism that greeted German officials who "didn't know."29 De-
spite this healthy skepticism, however, we remain convinced that frag-
mented knowledge is a genuine phenomenon that we cannot simply
dismiss as a lame excuse.

D. The Psychology of Destructive Obedience

Social scientists have labored to understand the Holocaust and to

answer the all-important question whether it could occur in other set-
tings. Stanley Milgram conducted perhaps the most significant - and
certainly the most famous - experimental studies to address this is-
sue. Milgram's experiments underscore our thesis because they illus-
trate the ways in which social and institutional pressures to obey
reinforce, and are reinforced by, the fragmentation of knowledge in
modem bureaucracies and other large organizations.
In Milgram's experiments, volunteers in a Yale University experi-
ment were ordered by the experimenter to administer gradually in-
creasing electric shocks to another "subject" (actually a confederate of
the experimenter), ostensibly to study the effect of punishment on
learning. As the "shocks" increased in intensity, the confederate dis-
played increasing discomfort, demanded that the experiment stop,
screamed with pain, complained of a heart condition, and finally fell
silent as if he were unconscious.30 In this original experiment, sixty-
five percent of the subjects went all the way, administering the highest
possible, potentially lethal, level of shock. Those subjects who admin-

28. In re A.H. Robins Co. "Dalkon Shield" IUD Prods. Liab. Litig., 575 F. Supp. 718, 724
(D. Kan. 1983).
29. John Braithwaite, Passing the Buck for Corporate Crime, AUSTL. SOCY., Apr. 1991, at 3,
3.

30. See STANLEY MILGRAM, OBEDIENCE TO AUTHORITY: AN EXPERIMENTAL VIEW 3-4

(1974).

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2361

istered the maximum shock expressed great discomfort at the cruel

task they were assigned; many of them berated the experimenter, pro-
tested, or insisted that they would not proceed with the experiment
all the while continuing to flip the switches.
Milgram conducted a number of important variations on the origi-
nal experiment, several of which suggest the role that incomplete and
fragmented knowledge may play in facilitating destructive obedience
and the abdication of individual responsibility.31 In one version, the
experiment was removed from the anxiety-relieving auspices of Yale to
a seedy-looking storefront operation in nearby Bridgeport. Less able
to reassure themselves that the experimenters knew what they were
doing, fifty-three percent of the subjects refused to go all the way.32
This suggests that compliant subordinates often believe that their
qualms are merely the result of incomplete understanding, and assume
that those in charge have good reasons for what they are doing.
The rate of compliance also declined when the subject could see
the victim, and declined even further when the subject was actually
required to hold the victim's hand on the contact-plate. In this latter
version of the experiment, seventy percent of the subjects stopped
before administering the maximum level of shock.33
While these levels of resistance are hardly comforting, they do sug-
gest that individuals will be less willing to follow immoral orders when
the consequences of obedience are less ambiguous. One striking varia-
tion on Milgram "simulated a bureaucratic hierarchy by dividing the
teacher's role among two people: a 'transmitter' and an 'execu-
tant.' "34 Executants, who administered the shock, resisted signifi-
cantly more often than transmitters, who merely informed the
executor when to deliver a shock, although the executants were under
at least as much pressure to obey.
As Kelman and Hamilton note in discussing this variation, sub-
jects' resistance to obedience is often weakened
by the ... ability to avoid seeing the connection between their own ac-
tions and the destructive consequences of those actions. The Nazi exter-
mination program was carried out by a vast bureaucracy in which many
functionaries - from Adolph Eichmann down to junior clerks - sat at
desks, shuffled papers, arranged train schedules, and carried out a vari-

31. More recently, Herbert Kelman and Lee Hamilton have offered a detailed psychological
analysis of the dynamics of obedience and the social structures that contribute to it. See HER-
BERT C. KELMAN & V. LEE HAMILTON, CRIMES OF OBEDIENCE: TOWARD A SOCIAL PSY-
CHOLOGY OF AUTHORITY AND RESPONSIBILITY (1989).
32. MILGRAM, supra note 30, at 61.
33. Id. at 35.
34. KELMAN & HAMILTON, supra note 31, at 165.

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2362 Michigan Law Review [Vol. 90:2348

ety of other tasks without having to consider the final product of their
efforts.35

Seeing the connection between one's action and its destructive conse-
quences clearly has a strong emotional component, but it also has crit-
ical cognitive significance; it removes any doubts about the effect of the
individual's obedience.
Another form of ignorance that appears to have played a signifi-
cant role in Milgram's experiments was the absence of a clear-cut mo-
ment of decision. Few subjects would have hesitated to give a mild,
tingling shock; most probably would have refused to give an initial
shock of maximum voltage. The gradual escalation of voltage was in-
sidious because it deprived subjects of an obvious stopping point, en-
couraging them to defer resistance until they saw themselves as
committed, or as compromised. This kind of slippery slope may char-
acterize many of the decisions made in contemporary organizations.
Another variant of the Milgram experiments, however, provides
some encouragement that resistance and reform may be possible in
organizational settings. In this study, the subject was assigned to a
team administering the shocks, while the other team members were
really confederates of the experimenter. Milgram discovered that
compliance was extraordinarily sensitive to peer pressure. When the
other team members refused to proceed with the experiment, only ten
percent of the subjects remained obedient to the experimenter and
"went all the way." Conversely, when a teammate rather than the
subject took charge of physically administering the shock, 92.5% of
the subjects went along with the experiment up to the maximum
shock.36 In Niebuhr's terms, we may think of moral man made less
moral by an immoral society, but more moral by a society of his
betters.
While there are obviously other factors at work, we believe that in
this experiment the compliance or resistance of others served an im-
portant cognitive function. As Kelman and Hamilton argue, "[t]he
most important effect of the confederates' disobedience ... was proba-
bly in providing the cognitive and motivational conditions that al-
lowed the participant to redefine the situation in which he found
himself."37
The Milgram studies, then, suggest the role of imperfect and frag-
mented knowledge in organizational misconduct. The less individuals
appreciate the consequences of their acts, the need to decide, and the

35. Id.
36. MILGRAM, supra note 30, at 119.
37. KELMAN & HAMILTON, supra note 31, at 160.

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2363

available alternatives, the easier it will be for them to engage in de-

structive obedience. Milgram's experiments suggest that the fragmen-
tation of knowledge promotes organizational wrongdoing by blunting
the edge of moral conflict.
Although Milgram's research focused on subordinates, parallel
problems arise for supervisors. As we have seen, bureaucratic struc-
tures serve to deny supervisors knowledge of operational details, blunt
their awareness of harsh consequences, and help them rationalize what
they cannot ignore. The result is the deep paradox of the "rule by
Nobody": when neither superiors nor subordinates may be held re-
sponsible, we face an uncanny situation in which responsibility has
seemingly been conjured out of existence.

E. The Inadequacy of Ethical Tradition and Philosophical Theory

Most moral theories take as their central cases the moral choices
confronting individuals deliberating alone and in full knowledge of the
morally relevant circumstances - including, indeed, that the choice
they face is morally charged. Standard moral theories may be viewed,
in effect, as structures of three concentric circles. At the center lie the
core precepts, which presuppose full knowledge on the part of agents.
Limiting these precepts are principles of mitigation, including mitiga-
tion when full knowledge is absent. Qualifying these, in turn, is an
account of culpable ignorance. The phenomenon of culpable igno-
rance lies at the periphery of concern in such a moral theory, and is
typically restricted to a small range of exceptional cases.
Reflection on organizational settings suggests that knowledge and
ignorance must play a more central role in moral theory. We believe
that the specter of fragmented knowledge, divided responsibility, am-
biguous orders, and unknown consequences is inadequately addressed
in the moral discourse of Western societies. Virtually every approach
to normative ethics, from the Ten Commandments to the latest wrin-
kles in philosophy journals, focuses primary attention on moral
problems in which four knowledge conditions are satisfied - knowl-
edge conditions that are frequently absent in individual decisionmak-
ing and almost never found in organizational settings.
First, the decisionmaker recognizes that he or she has come to a
fork in the road: The decisionmaker knows that a decision must be
made. Do I or don't I cheat on the examination? Do I or don't I
protest when I hear an acquaintance tell an anti-Semitic joke? Situa-
tions such as these are readily identifiable as moral decisions. Typi-
cally, when we face one of these questions, we know that we face it.
Second, the decisionmaker recognizes that he or she must make the

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2364 Michigan Law Review [Vol. 90:2348

choice in a fairly short, distinct period of time: The decisionmaker

knows when a decision must be made, or at least by when it must be
made. The examination is tomorrow; I must confront the acquain-
tance about his anti-Semitic joke now or never.
Third, the decisionmaker confronts a small number of well-defined
options: The decisionmaker knows what choices are available. For
example, a lawyer, learning that her client is using her services to per-
petrate a fraud, can quickly catalogue her options: do what the client
asks, try to talk the client out of the plan, blow the whistle on the
client, or resign.
Fourth, the decisionmaker has the information needed to make the
decision: The decisionmaker knows what is needed to make the
choice. Even in situations of radically incomplete information, theo-
ries of rational decisionmaking under uncertainty allow us to assign
probability-estimates to these various outcomes in order to generate a
recommendation, though that recommendation may be merely to flip
a coin.
These, then, are the four knowledge conditions of moral decision-
making: we know that a decision must be made, when a decision must
be made, what choices are available, and what is needed to make the
choice. From the Biblical "Thou shalt not steal" and the Golden
Rule, to Kant's categorical imperative and the utilitarian injunction to
achieve the greatest good for the greatest number, the core precepts of
the major systems of ethical thought are directed to agents who satisfy
the knowledge conditions. If the conditions are not satisfied, ethical
systems generally respond with mitigation or even immunity: forgiv-
ing those who "know not what they do" is basic to Western under-
standings of moral responsibility.
Ignorance can, of course, be culpable; but most philosophers and
legal theorists who acknowledge the phenomenon of culpable igno-
rance have implicitly confined it to a small range of exceptional cases:
conspirators who attempt to preserve their deniability, or drunks who
have wilfully stupefied themselves. The possibility that the modern
workplace may place millions of ordinary individuals in a state of cul-
pable ignorance throughout their careers has never, to our knowledge,
been explicitly addressed in moral theory.
In an organizational setting, one or more of the four knowledge
conditions typically fails at a critical juncture. As the Milgram experi-
ments illustrated, individuals in bureaucratic settings may not fully
appreciate that a decision must be made, understand when it should be
made, realize what choices are open to them, and comprehend what
the consequences of different choices will be. A law firm associate

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2365

asked to research a small point of law or a junior architect asked to

design a detail may have no idea that the project as a whole raises deep
questions of professional ethics. Even if they have their suspicions, it
is often impossible to pinpoint a moment of truth when the decision
must be made. No clear list of options, or even clear understanding of
who to speak with, may exist, and the subordinate may never believe
she has sufficient information to fashion a solution.
The failure of these knowledge conditions is created or maintained
by organizational structure. Typically, supervisors parcel out subtasks
to a number of subordinate employees. None of the subordinates may
have more than the most general idea of what the entire project is
about, while the supervisor may know nothing about the details of
each subordinate's subtasks. No member of the organization might
recognize a moral problem, because the problem arises not from what
any one member of the team is doing, but rather from all their actions
put together. The fact that each is merely a member of a team lulls
them into a sense of security, so that they feel no pressing need to find
out more about what is going on. Though they may resolve not to be
"good Germans" at the moment of truth, the moment of truth never
arrives.
This, then, is the central philosophical question that the problem
of fragmented knowledge raises: Is it possible to formulate satisfactory
principles of individual responsibility when any or all of the four knowl-
edge conditions presupposed by standard moral theories fail?

III. RESPONSIBILITY WITHOUT KNOWLEDGE: FOUR

APPROACHES TO THE PROBLEM

In this Part, we compare and assess principles of responsibility for

individuals in organizational settings that may complement existing
moral systems that presuppose the four knowledge conditions. At bot-
tom, four approaches exist to the problem of "deeds without doers."
First, we can simply accept as a tragic fact of modern existence that
organizational wrongs may be committed for which no one - neither
individuals nor the organization - can rightly be held responsible.
More optimistically, we can either hold the organization itself morally
responsible for the wrongdoing or hold all the individuals affiliated
with the organization strictly liable. Finally, we can extend standard
principles of culpable ignorance to explain why individuals in organi-
zations may be held responsible for their actions even though the
knowledge conditions fail. This is the approach we will defend.

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2366 Michigan Law Review [Vol. 90:2348

A. The Skeptical Approach

Some might argue that "deeds without doers" are a tragic fact of
modern life. We really do live under the "rule by Nobody" discussed
by Hannah Arendt, so assigning responsibility to some individual -
whether a natural or organizational person - is simply an anachro-
nism. This we call the skeptical approach: "skeptical," because it de-
nies that the problem of fragmented knowledge has a solution.
Various postmoder philosophers and psychologists have suggested
that the very idea of individual responsibility rests on a discredited
picture of autonomous subjects or selves making rational choices.
Thus, Kenneth Gergen asserts in The Saturated Self: "To be sure,
individuals break the law, but from the postmodern perspective, such
actions should be attributed not to the individual alone but to the ar-
ray of relationships in which he or she is a part."38 Unlike a corpora-
tion or government, an "array of relationships" does not look even
superficially like an entity we could meaningfully praise or blame. We
are left where we began, with deeds that lack doers.
Under this skeptical view, we can no more condemn organizational
wrongdoing on moral grounds than we can cancer or AIDS. Suppose
that an epidemic of an unknown disease rendered the entire human
race psychotic and homicidal. A great deal of death would result, but
no one could truly be held accountable. The skeptical, postmodern
view analogizes bureaucratic organization to this imaginary epidemic.
The analogy suggests two problems with the skeptical view. First,
the argument overstates the difficulty of imputing responsibility to in-
dividuals or the organizations to which they belong or help to create.
Individuals in an organization may be ignorant, confused, anxious, or
intimidated, but they are not in the grip of an irresistible malady. Bu-
reaucratic organization is not a renegade virus: it results from con-
scious human decisions, and it possesses some of the features of a
conscious entity that make it susceptible to praise and blame.
Second, even if the analogy were descriptively correct, the skeptic's
prescription of passivity or acquiescence would hardly follow. The
theoretical difficulty of ascribing responsibility for egregious miscon-
duct to individuals or organizations might instead provide a compel-
ling reason for overhauling the basic structure of organizations in
order to restore accountability. Our working hypothesis amounts to
what might be dubbed a law of conservation of responsibility: moral
responsibility for organizational actions cannot and should not be con-

38. KENNETH J. GERGEN, THE SATURATED SELF: DILEMMAS OF IDENTITY IN CONTEM-

PORARY LIFE 244 (1991).

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2367

jured out of existence. It must lie with the organization itself, with the
individuals in the organization, or with the institutions that permit
responsibility to be endlessly circulated.

B. Replacing Individual with Organizational Responsibility

As one alternative, the organization itself could be held responsible

to precisely the degree that an organizational structure of fragmented
knowledge absolved individuals within the bureaucracy from responsi-
bility. This approach raises deep philosophical questions about the re-
ality of metaphysical fictions such as organizations and the cogency of
assigning them moral agency and responsibility. An organization may
not possess all the characteristics necessary to ascribe agency and re-
sponsibility to it. As Baron Thurlow, Lord Chancellor of England,
asked in an often quoted question, "Did you ever expect a corporation
to have a conscience, when it has no soul to be damned, and no body
to be kicked?"39
Several philosophers address Baron Thurlow's question, proposing
different methods for understanding organizations as the kind of entity
to which responsibility can be ascribed.40 Peter French, for example,
argues that, because corporations possess personality and decision-
making ability, we must regard them as "members of the moral com-
munity, of equal standing with the traditionally acknowledged
residents: human beings."41 As French sees it, corporations that
cause harm are not ordinarily mere causes of harm: they are instead
persons that may be morally responsible in the same way as human
beings. Indeed, French holds open the possibility that a corporation
may be responsible for a harm even though neither its employees nor
directors have any responsibility for that occurrence.42 Although
French's view does not imply that organizations must be held respon-
sible to precisely the degree that individuals are absolved from respon-
sibility by an organizational structure of fragmented knowledge, his
approach at least makes this possibility intelligible.
Should we agree with French that organizations are morally re-

39. THE OXFORD DICTIONARY OF QUOTATIONS 550 (3d. ed. 1979).

40. For an excellent survey of these discussions, see LARRY MAY, THE MORALITY OF
GROUPS: COLLECTIVE RESPONSIBILITY, GROUP-BASED HARM, AND CORPORATE RIGHTS
(1987). See also ELIZABETH WOLGAST, ETHICS OF AN ARTIFICIAL PERSON: LOST RESPONSI-
BILITY IN PROFESSIONS AND ORGANIZATIONS (1992). Wolgast's book is one of the best-sus-
tained philosophical efforts to deal with the issues of individual and organizational responsibility
we are concerned with. She addresses the issue through the concept of role-differentiation rather
than fragmented knowledge, but her approach and conclusions are in many respects congenial to
ours.

41. PETER A. FRENCH, COLLECTIVE AND CORPORATE RESPONSIBILITY 32 (1984).

42. Id. at 152, 162.

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2368 Michigan Law Review [Vol. 90:2348

sponsible persons? This question really has two components: first,

whether organizations are moral persons; second, even if they are not
moral persons, whether they are still the kind of entity that can be
morally responsible. If an organization is a moral person, then by defi-
nition it is the kind of entity to which responsibility can be ascribed.
If, on the other hand, it is an entity other than a person, then the issue
of whether we can ascribe responsibility to it - and what making that
ascription would mean - becomes more complex.
As French initially presented his position, all that is required to be
a moral person is that one be able to engage in a complex kind of
decisionmaking that involves making a deliberate choice. A corpora-
tion, he maintained, engages in such decisionmaking when its board of
directors or managers renders decisions on the corporation's behalf.43
French's critics responded that many entities that are not moral per-
sons apparently have the ability to engage in intentional decisionmak-
ing.44 A cat may choose after deliberation to kill a mouse, yet it is not
a moral person. To avoid this problem, French now refines his defini-
tion of moral personhood to require the ability not only to engage in
intentional decisionmaking but also to change one's conduct in re-
sponse to moral criticism.45
We do not believe, however, that even French's revised criterion
suffices to show that organizations are moral persons. There is more
than one sense in which one might respond to moral criticism, and not
all these senses are relevant to the issue of whether one is a person
morally speaking. Suppose that, observing a cat toying with a mouse,
you yell, "Fiend!" and the cat runs. In this case the cat would in some
sense be responding to moral criticism: it responds to your yell, which
voices a moral criticism.
Yet plainly the cat would not be responding to the criticism in a
way that lent credence to the idea that it appreciated the moral signifi-
cance of the criticism. Indeed, one might say that rather than re-
sponding to moral criticism, the cat responded to the simpler

43. Peter A. French, The Corporation as a Moral Person, 16 AM. PHIL. Q. 207, 213-14
(1979).
44. THOMAS DONALDSON, CORPORATIONS AND MORALITY 20-23 (1982); see also Thomas
Donaldson, Personalizing Corporate Ontology: The French Way, in SHAME, RESPONSIBILITY,
AND THE CORPORATION 101, 101-12 (Hugh Curtler ed., 1986). For a cogently argued interpre-
tation of choice and deliberation that would place animals and corporations outside the realm of
entities that can choose, see ALAN DONAGAN, CHOICE: THE ESSENTIAL ELEMENT IN HUMAN
ACTION (1987). For a view interestingly similar to Donagan's, see Manuel G. Velasquez, Why
Corporations Are Not Morally Responsible for Anything They Do, 2 Bus. & PROF. ETHICS J.,
Spring 1983, at 7. Velasquez argues that, in order to be a person, one must have the unity of
consciousness that would allow one to initiate an action, and that corporations lack this charac-
teristic. Id. at 8-9; see also WOLGAST, supra note 40, at 79-95.
45. See FRENCH, supra note 41, at 166.

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2369

phenomenon of angry yelling. In order to respond to moral criticism

in the relevant sense, one must respond in a way that confirms one's
understanding of the criticism. At the very least, one's response must
be animated or informed by an appropriate moral emotion like guilt or
shame.
This returns us to Baron Thurlow's objection. If the corporation,
or any other organization, cannot feel guilt or shame or other suitable
moral emotions, how can it appreciate the significance of moral criti-
cism? Because the organization lacks heart and soul, it literally feels
nothing. Lacking the capacity for moral feeling, the organization's re-
sponse to moral criticism may do nothing more to distinguish the or-
ganization as a person than does the reaction of the cat.
One might respond that the controlling members of the organiza-
tion, perhaps acting in their official capacity, might feel guilt or shame
on the organization's behalf, and that the possibility of this vicariously
felt shame should allow us to attribute moral responsibility to the or-
ganization, to blame it, and to certify its credentials as a person. In-
deed, it is not hard to imagine the individual sense of shame being
deepened by a mutual understanding among members of the organiza-
tion of the shame each person feels.
Under this view, the fact that the organization feels shame would
consist in the fact that relevant members of the organization experi-
ence appropriate sentiments of shame for what the organization has
done. One might thus defend French against Baron Thurlow by hold-
ing that the organization need not have moral sentiments apart from
those experienced by humans associated with it in order to be a mor-
ally responsible person, any more than it need have conscious exper-
iences of decisionmaking in order to be the kind of entity that makes
decisions.
This defense, however, overlooks the difference between the claim
that the organization feels shame and the claim that the organization
is a person that feels shame. One can plausibly say that an entity other
than a person can feel shame. A family, for example, may feel shame
for what one of its members has done. For it to do so, its members
must feel shame in a way that shows mutual understanding of the im-
portance of family ties. Is an organization's feeling shame more like
the family's feeling shame (or some other collective entity's feeling
shame) than it is like an individual person's feeling shame? We think
that the answer is yes, and thus that the organization is best under-
stood as a collective entity, not a person.
To earn the moral status of a person, one must do certain things
for oneself that an organization must have done for it by others. Con-

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2370 Michigan Law Review [Vol. 90:2348

sider what having a thought means. Unless a person herself has a

thought, she does not think. One person does not have a thought sim-
ply because some other person has a thought. Similarly, outside the
legal fictions of agency law, one person does not render a decision sim-
ply because another person makes a decision for her, and one person
does not feel shame simply because another person feels shame for her.
Because an organization can think, decide, or feel shame only by
having someone else do these things for it, the organization cannot
think, decide, or feel shame the way that a person can. At most, the
organization engages in these mental acts not as a person but as a
collective entity like a family. Indeed, if the organization were a per-
son, all collective entities that had decisionmaking abilities, like fami-
lies, bridge clubs, and law school faculties, would be persons. But it
would be absurd to suppose that a family that made collective deci-
sions consisted of five persons: the mother, father, two children, and
the family itself.46
Perhaps the organization is a morally responsible entity other than
a person. Several writers suggest the possibility that many nonper-
sons, including mobs, crowds, and nations, are morally responsible.47
An organization, conceived of as an entity other than a person, might
be regarded as morally responsible in at least two ways. Each method
involves a metaphysical interpretation of the organization.
First, one might think of the organization as a collective entity
constituted by its employees and officials,48 who stand in a distinctive
relation to one another, somewhat analogous to the way a brick wall is
made of bricks or a family is constituted by its members. This first
view holds that the organization is not an entity that can exist apart
from the individual human beings associated with it, and regards
blaming the organization as just an indirect way of blaming those
individuals.
This view naturally raises the question, why bother? Why not sim-
ply blame the individuals? One obvious answer is that often, when
organizational activity results in harm, identifying the responsible or-
ganization is much easier than identifying the responsible individuals
within. Holding the organization responsible may be a way of holding
responsible individuals within the organization whose identities we
cannot determine. This tactic makes sense, however, only when we

46. While we call this implication absurd, there is no reason to think that French would
agree.

47. For an illuminating analysis of these issues, see Joel Feinberg, Collective Responsibility,
65 J. PHIL. 674 (1968).
48. A sophisticated version of this view is developed in MAY, supra note 40.

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2371

are reasonably confident that some individuals or other in the organi-

zation really are at fault. It makes sense, that is, only when we are not
confronted with a case where no individuals are at fault because all of
them may truthfully plead the epistemological excuse. Thus, blaming
the organization simply because it is a convenient surrogate for the
responsible employees makes practical sense only when the problem of
fragmented knowledge is not the primary issue. This pragmatic ap-
proach consequently ignores the problem of fragmented knowledge
rather than solving it.
Second, one might think of the organization as an entity whose
existence is distinct from that of the human beings associated with it.
For example, rather than following French in thinking of the corpora-
tion as some weird variety of person, one might, following John Ladd
and Meir Dan-Cohen,49 think of the corporation as an entity like a
machine. Alternatively, one might follow Virginia Held and regard
the corporation to be something intermediate between a person and a
machine.50 If one accepts the view that an organization is a meta-
physically distinct entity, then one might also follow Held in suppos-
ing that the organization can act even after the humans associated
with it perish.
This second view seems odd on the surface for reasons raised by
Baron Thurlow. To attribute moral responsibility to something is to
blame it, which involves subjecting it to moral criticism. Yet morally
criticizing an entity that has no capacity for moral feeling seems point-
less. Some authors nonetheless suggest that certain situations force us
to recognize the moral responsibility of organizations apart from the
individuals who comprise them.51 These are cases in which organiza-
tional activity causes harm; no individual associated with the organi-
zation can properly be blamed for the harm, and there is therefore
nobody but the organization to blame - that is, precisely those cases
that form our central topic.
One situation in which the facts supposedly require us to acknowl-
edge the necessity of blaming the organization without blaming indi-
viduals associated with it involves the 1979 crash of an Air New

49. MEIR DAN-COHEN, RIGHTS, PERSONS, AND ORGANIZATIONS: A LEGAL THEORY FOR
BUREAUCRATIC SOCIETY 46-51 (1986); John Ladd, Morality and the Ideal of Rationality in For-
mal Organizations, 54 THE MONIST 488 (1970). Ladd, however, denies that corporations can be
morally responsible. Id. at 498-500.
50. See Virginia Held, Corporations, Persons, and Responsibility, in SHAME, RESPONSIBILITY,
AND THE CORPORATION, supra note 44, at 161-81.
51. See FRENCH, supra note 41, at 145-63; Paul B. Thompson, Why Do We Need a Theory of
Corporate Responsibility?, in SHAME, RESPONSIBILITY, AND THE CORPORATION, supra note 44,
at 115-35.

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2372 Michigan Law Review [Vol. 90:2348

Zealand jet. French52 and Paul Thompson53 argue that while perhaps
no single individual was responsible for this crash, the airline corpora-
tion was responsible; indeed, these commentators suggest that the cor-
poration's responsibility may arise from facts that, in our terms,
involve fragmented knowledge.
The facts are as follows.54 Air New Zealand flew large jets over
Antarctica as part of a sightseeing package. The jets would fly at low
altitude so that the passengers could see the terrain. Apparently, Air
New Zealand used strange procedures to disseminate information
about the flight. These procedures fragmented knowledge about the
details of the flight plan and the obstacles to flight. Air New Zealand's
flight plans were programmed into on-board computers from a central
ground computer. There was no practice of including flight crews in
discussions of changes in the flight plan; hence, pilots were not in-
formed of a flight plan change before the flight occurred. Yet these
changes put the flight on a collision course with a mountain. The col-
lision killed everyone on the jet.
Since they believe that Air New Zealand employees acted consci-
entiously, French and Thompson maintain that the responsibility for
the crash should not be attributed to any individual employee or man-
ager. Instead, they argue, the corporate policy itself was at fault for
the crash. These authors thus approve of a government report which
concluded that the cause of the air disaster had an "organizational
nexus," and that it was wrong to attribute moral responsibility to any
individual associated with Air New Zealand.55
It is difficult to believe that no individual associated with Air New
Zealand did anything wrong. If the accident occurred because of bad
corporate policy, then we must wonder whether the person or persons
responsible for the policy should also be blamed for the accident. One
wonders whether those in charge of designing the communication pro-
cedures properly discharged their duties, and if they did not, whether
doing so would have prevented the accident. In many ways, to sup-
pose that the accident might have occurred despite the fact that all
involved properly discharged their duties seems a piece of science
fiction.

These doubts about the French-Thompson description of the case

52. See FRENCH, supra note 41, at 145-63.

53. See Thompson, supra note 51, at 119-20.
54. We rely on the accounts provided in FRENCH, supra note 41, at 145-54, and Thompson,
supra note 51, at 119-20.
55. See FRENCH, supra note 41, at 152; Thompson, supra note 51, at 119-20 (implicitly con-
curring with government report).

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2373

highlight an important point about the epistemological excuse for or-

ganizational wrongdoing: though we have argued that the excuse may
be genuine, it is also clear that embattled employees will frequently
seize on it whether it is genuine or not. This suggests a practical rea-
son to be concerned about shifting blame to the organization. To the
extent that we assign responsibility to the organization itself rather
than to its managers, employees, or directors, we protect those individ-
uals from blame. This may encourage managers, employees, and di-
rectors to resort too readily to the epistemological excuse "I didn't
know" and may even lead them to resurrect the positivist excuse "I
was only doing my job." The effect may be to diminish the responsi-
bility individuals take for what they do in the firm and to encourage
individuals to acquiesce in bad and even harmful policy. On the other
hand, to the extent we hold responsible the individual employee or
official but not the organization itself, we run a complementary risk:
we may encourage organizations to treat individual employees as
scapegoats. When harm is caused by corporate structure that has
careened out of control, something must be done to change that struc-
ture. Picking out individual managers to hold responsible may dis-
tract us from the more important task of institutional reform.
Let us, then, accept for the sake of argument that even though
individuals in Air New Zealand bear no responsibility for the crash,
Air New Zealand itself was responsible. How might this happen?
Suppose as a bottom-line fact that even though all individuals involved
worked very hard at getting the relevant information, the structural
complications were so overwhelming that they were doomed to fail.
Nobody could have anticipated these complications, so nobody was to
blame for them. We might say that the accident was somebody's fault
but not, even to some small degree, the fault of any individual. There-
fore, the corporate structure must have been at fault: the corporate
structure, and hence the corporation itself, caused the crash. Had that
corporate structure not been present, a pilot would have been aware of
the risk of crash, and could have acted to avoid colliding with the
mountain. But if we concede that the corporation, not an individual,
caused the crash, are we also forced to admit that the corporation was
morally responsible for the crash?
Here again Baron Thurlow enters the picture. That causation is
not enough for moral responsibility is elementary. Lightning may
cause a fire in the forest, but it is only causally responsible, not morally
responsible, for the fire. In the case of lightning, our inability to attri-
bute moral responsibility is of no concern. We are able to understand
and control the connected phenomena - to protect ourselves from

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2374 Michigan Law Review [Vol. 90:2348

being struck, for example - so long as we understand the relevant

causal facts. When humans cause harm, the situation is different. To
get people to avoid harm, we must appeal to their sense of decency and
duty, to their desire to do nothing for which they should feel guilt or
shame. Even when we cannot hope to affect the behavior of a respon-
sible person - because, for example, he caused the harm in the past
and is now dead - we may still benefit from developing explanations
of what he did in terms of the moral concept of responsibility. Such
explanations help us hone our understanding of the distinctively
human causal web.
While it is plainly important to observe how organizations and or-
ganizational structure cause harm, even to attribute causal responsibil-
ity for a particular harm to organizational structure, we gain nothing
by going further and talking about the moral responsibility of the
metaphysically distinct organization. We do not alter an organization
by appealing to its sense of decency or duty; we do so by appealing to
the sensibilities of its employees and officials, or, in extreme cases, by
placing it in receivership. Because the language of moral responsibil-
ity is lost on the metaphysically distinct organization, to use that lan-
guage on the organization cheapens it.
Hence, even if we resist the temptation to blame the organization
morally instead of its employees or officials, we need to find another
variety of responsibility for the organization, a nonmoral variety. In
finding that the organization or its structure is causally responsible for
the accident, we need not morally blame it as a metaphysically distinct
entity, because we need not assert that the organization should have
acted with a sense of decency or duty or felt guilt or shame for what it
did. Instead, we cite the organizational structure as a causal factor
that contributed to the accident and that merits our attention for
change. Thus, by blaming the organization, we can make at least one
valid point: that, in solving the problem that led to this crash, we
must do more than admonish or even fire Air New Zealand employees
who were involved in this crash. We must also change the organiza-
tional structure that led these employees to behave in a manner that
allowed the crash to occur. This blame, however, is not moral blame.
It is causal blame, or, more simply, causal citation.
Beyond organizational structure, another aspect of the organiza-
tion may be a morally important cause of harm: organizational cul-
ture. The difference between structure and culture is straightforward.
Two organizations with the same structure - that is, the same organi-
zational flow chart and the same written policies - may behave differ-
ently because they have different cultures, different patterns of

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2375

informal institutional practice and social relations. One organization

might stress upright behavior, honoring the spirit of laws as well as
abiding by their letter, putting safety before profits, and frowning on
backbiting and internal politicking among middle management. An-
other might stress the bottom line, encourage employees to get away
with all they can, and applaud personal advancement through dirty
tricks. We can cogently ascribe moral terms to these cultures. We can
find one decent and the other reprehensible.
Does this distinction imply that the culture is a moral person or
some other kind of entity that we may suitably hold responsible for
wrongdoing? From the moral point of view, is organizational culture
less like organizational structure and more like a person? Referring
back to French's criterion for moral personhood, the answer would
seem to be no. A culture can experience moral emotions like guilt and
shame, if at all, only vicariously. Still, one may be drawn to impute
moral responsibility to organizational cultures more than to organiza-
tional structures because cultures are more suffused with human char-
acter than are structures: we may blame a culture that encourages
Smith to lie and cheat because that culture expresses the character of
- perhaps even seems part of - many individuals within the culture
who lie and cheat. We can more easily anthropomorphize an organi-
zational culture than an organizational structure.
Yet the fact that we may correctly use some terms of moral ap-
praisal when talking about an organization does not imply that the
whole range of moral terms applies to it. Some things are subject to
moral appraisal only in very limited ways. If Smith starts a fight by
making a nasty remark, I may assess the remark as morally reprehen-
sible while also judging Smith to be reprehensible for making the re-
mark. On the other hand, it makes better sense to blame Smith, not
the remark, for causing the fight. How the remark itself could be at
fault or blameworthy is not plain. Hence we must hold open the possi-
bility that organizational culture, while subject to moral appraisal,
cannot be held responsible for wrongdoing.
Moreover, even if organizational cultures may be partly responsi-
ble for individual wrongdoing, that would not exculpate the individ-
ual. We do not exonerate mobsters merely because they live their lives
in a culture of violence. Nor can an organizational culture act except
through individuals, who ultimately bear the praise or blame for the
lives they craft out of the raw material the culture provides.
We have now seen that facts such as those of the Air New Zealand
case do not force us to attribute moral responsibility to the organiza-
tion as a metaphysically distinct entity. But another factor, not em-

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2376 Michigan Law Review [Vol. 90:2348

phasized by French or Thompson, might discourage us from insisting

that the responsibility of an organization ultimately collapses into the
responsibility of individuals. If we say that humans acting as individu-
als are the ultimate bearers of moral responsibility, we seem to deny
the moral significance of an individual's organizational involvement.
Yet when organizational activity causes harm, not merely individuals
acting in their capacities as participants in the organization are to
blame for the outcome. The relationship between individuals and the
organization matters morally. In the case of organizational crime, for
example, we want to do more than blame each executive involved in
the relevant decision or perhaps restructure the organization. Our
blaming practices must somehow reflect the fact that these executives
engaged in criminal activity as part of a larger enterprise. To ac-
knowledge the moral relevance of the organization, we might blame it
while also blaming the individuals who form its membership.
Consider conspiracy law by way of analogy. According to tradi-
tional criminal jurisprudence, stealing something is bad, but agreeing
with someone else to steal that same thing may be worse, even if
neither you nor your coconspirator actually steals it.56 Involvement
with a group for illicit purposes has distinctive moral significance. It
does not follow, however, that the group as an entity is responsible in
some way that transcends the responsibility of the individuals.
Rather, the idea is that those who act in groups may impose greater
risks on society, and thus deserve greater blame when they act
wrongly.
Joining a legal organization is not an evil like joining a conspiracy,
but both legitimate organizations and criminal conspiracies pose more
of a threat to social welfare than do their individual members, a threat
which has implications for the responsibility of those members. One
of the principal rationales for punishing conspiracy as a separate of-
fense is "the special danger incident to group activity."57 As one oft-
quoted commentary argues, "[I]t is more difficult to guard against the
antisocial designs of a group of persons than those of an individual
.... The advantages of division of labor and complex organization
characteristic of modern economic society have their counterparts in
many forms of criminal activity."58

56. See generally WAYNE R. LAFAVE & AUSTIN W. SCOTT, JR., CRIMINAL LAW ? 6.4, at
530 (2d ed. 1986).
57. MODEL PENAL CODE ? 5.03 comment on criminal conspiracy (Tent. Draft No. 10,
1960).
58. Note, The Conspiracy Dilemma: Prosecution of Group Crime, 62 HARV. L. REV. 276,
283-84 (1948) (footnotes omitted).

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2377

Obviously, we do not wish to condemn "modem economic soci-

ety" because dangers arise from its "division of labor and complex
organization." But we may wish to impose on individuals who partici-
pate in that society obligations commensurate with the dangers. After
all, the converse of the passage we have just quoted is equally true:
many forms of criminal activity have their counterparts "in the divi-
sion of labor and the complex organization characteristic of modem
economic society."
While this comparison highlights the special risks of organiza-
tional activity, it also reminds us that the explicit agreement to do
wrong found in conspiracies is rarely the source of individual responsi-
bility for organizational wrongdoing. Individual blameworthiness lies
rather in the failure to respond adequately to the heightened risks of
unintended and inadvertent wrongdoing.
In sum, we have identified three distinct reasons that the organiza-
tion must be taken seriously in our practices of fixing blame and attrib-
uting responsibility. First, the corporate structure may be a locus of
events causally relevant to morally significant harm. Second, organi-
zational culture may serve as both a causal locus for harm and an
object of moral condemnation. Third, an individual's participation in
organizational activity involves special risks and imposes special obli-
gations, factors that may deepen responsibility for unintended and in-
advertent wrongs.
None of these reasons, however, suggests that organizations are
moral agents of some metaphysically unique type, nor that we can
blame organizations for their wrongdoing in such a way as to replace
individual responsibility. Organizational structure may mitigate or
even excuse individual conduct, but organizational structures cannot
be the subject of moral blame. Organizational culture may be subject
to blame in a certain sense, but blaming the culture does not excuse or
mitigate the conduct of the individuals who are part of it. Finally,
recognition of the special risks that organizations share with conspira-
cies serves to heighten, not reduce, individual responsibility for organi-
zational wrongdoing.
Our approach splits the difference between antireductionist and re-
ductionist views of organizations. With the antireductionists, we ac-
knowledge that praising and blaming organizations is not merely a
shorthand way of praising and blaming individuals. Organizations are
real and irreducible to their component individuals. With the reduc-
tionists, however, we insist that organizations are not metaphysically
distinct seats of moral responsibility - persons or otherwise - that
can absorb the blame from individuals who "know not what they do."

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2378 Michigan Law Review [Vol. 90:2348

Individuals alone have souls to damn, and if the epistemological ex-

cuse relieves them of moral responsibility, then nothing in the world is
morally responsible.

C. Strict Moral Liability

Let us return to Kundera's reflections on the guilt of Czech Com-

munists for the crimes that they committed because they were
deceived by the Russians. Kundera's protagonist Tomas observes that
"In the end, the dispute narrowed down to a single question: Did they
really not know or were they merely making believe?"59 To Tomas,
however, this seemed a shallow and wrongheaded question:
It was in this connection that Tomas recalled the tale of Oedipus:
Oedipus did not know he was sleeping with his own mother, yet when he
realized what had happened, he did not feel innocent. Unable to stand
the sight of the misfortunes he had wrought by "not knowing," he put
out his eyes and wandered blind away from Thebes.
When Tomas heard Communists shouting in defense of their inner
purity, he said to himself, As a result of your "not knowing," this coun-
try has lost its freedom, lost it for centuries, perhaps, and you shout that
you feel no guilt? How can you stand the sight of what you've done?
How is it you aren't horrified? Have you no eyes to see? If you had eyes,
you would have to put them out and wander away from Thebes!60
In Sophocles's drama, a plague visits Thebes because it is harboring
someone polluted by the twin crimes of parricide and incest. The gods
do not care if Oedipus did not know that the man he killed at the
crossroads was his father or that Jocasta was his mother; the gods did
not care that he had no guilty intentions. Oedipus Rex portrays a
world of strict moral liability. Kundera wishes us to understand
moral responsibility among the apparatchiks through the categories of
this world.
By analogy, one may approach the problem of fragmented knowl-
edge through a conception of strict moral liability by holding organi-
zational functionaries morally responsible for the wrongs they
participated in despite the fact that they "didn't know" and thus had
no wrongful intentions.
The notion of strict moral liability seems at first blush unpromis-
ing, simply because we typically bridle at blaming someone morally
for ignorant or unintended wrongs. In this, no doubt, we differ from
Sophocles and his culture, which placed less emphasis on intentions,
and indeed on any interior mental states, than we do. Yet even the

59. KUNDERA, supra note 14, at 176.

60. Id. at 177.

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2379

Greeks recognized the excuse, "I didn't realize what I was doing!"
When Agamemnon apologizes to Achilles in book 9 of the Iliad, he
attempts to mollify Achilles by explaining that a blindness had come
over him when he seized Briseis.61 It is hard to imagine any culture
for which the agent's intention is simply irrelevant to the moral classi-
fication of acts; as Holmes observed, even a dog knows the difference
between being tripped over and being kicked.
A.W.H. Adkins has argued that it was principally in the context of
wrongful acts involving the religious notion of pollution that the
Greeks regarded the agent's intentions as irrelevant.62 The notion of
pollution, however, seems to have no place in a world that does not
believe in demons, Furies, or vengeful ghosts howling for blood; the
Greek conception of strict moral liability, so closely tied to cultic and
chthonic religions, appears to hold little promise for us. Reinforcing
this conclusion is the fact that the analogous notion of strict liability
for criminal offenses is widely regarded as a jurisprudential aberration
with no place in a civilized legal system.
Yet the argument cannot be quite so quick or straightforward. For
one thing, the concept of pollution has by no means vanished from
contemporary, secular moral discourse, as the postwar debate over the
dimensions of German "metaphysical guilt" makes clear.63 Anthony
Appiah has noted that many of those advocating Western divestment
of South African holdings until the end of apartheid base their argu-
ments explicitly on the idea that, regardless of the practical efficacy of
divestment, possessing South African holdings amounts to a form of
pollution.64 Moreover, in recent years moral philosophers have paid
increasing attention to the phenomenon of moral luck,65 the possibility
that matters beyond our control may affect our moral blameworthi-
ness. Philosophers who believe in moral luck must be prepared to
weaken the requirement of evil intent in attributions of moral blame.
Finally, as George Fletcher points out, the concept of pollution has

61. HOMER, THE ILIAD bk. IX, lines 136-45 (Robert Fugles trans., 1990).
62. ARTHUR W.H. ADKINS, MERIT AND RESPONSIBILITY: A STUDY IN GREEK VALUES
92-104 (1960). Some acts, such as murder, rendered the perpetrator loathsome in the eyes of the
gods. One thus polluted could not perform prayers and sacrifices without first undergoing expia-
tion and ritual purification.

63. See KARL JASPERS, THE QUESTION OF GERMAN GUILT (E.B. Ashton trans., 1947).
64. Anthony Appiah, Racism and Moral Pollution, 18 PHIL. F. 185 (1987), reprinted in COL-
LECTIVE RESPONSIBILITY: FIVE DECADES OF DEBATE IN THEORETICAL AND APPLIED ETHICS
219 (Larry May & Stacey Hoffman eds., 1991).
65. See, e.g., THOMAS NAGEL, Moral Luck, in MORTAL QUESTIONS 24 (1979); MARTHA C.
NUSSBAUM, THE FRAGILITY OF GOODNESS: LUCK AND ETHICS IN GREEK TRAGEDY AND
PHILOSOPHY (1986); BERNARD WILLIAMS, MORAL LUCK (1981).

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2380 Michigan Law Review [Vol. 90:2348

survived in our criminal law in forfeiture provisions.66 Forfeiture of

goods involved in criminal activity is a residue of the law of deodands,
which grew out of a tradition of regarding the physical instruments of
wrongdoing as polluted objects.67 Deodands, it is true, are a thing of
the past; but forfeiture remains. All this suggests that the notion of
strict moral liability may not be as anachronistic, and as bound to
superstition, as it seems.
We nevertheless continue to find it an unpromising approach to
the problem of fragmented knowledge. To begin with, the notion of
moral pollution simply cannot substitute for the concept of moral
blameworthiness. Proponents of the concept of pollution face a di-
lemma. If they base their argument, as they seem to, on the fact that
the concept of pollution persists in our actual practices of moral criti-
cism, they must acknowledge that those practices do not restrict it to
cases where the agent plays a causal role in wrongdoing. In the para-
digmatic instances of moral pollution - metaphysical guilt, or the
sense of taint felt by the close relatives of notorious murders - the
agent's causal role in the wrongdoing is as irrelevant as his intentions.
German metaphysical guilt, for example, is metaphysical precisely be-
cause it applies even to Germans who played no causal role in the
Holocaust.
But decisive reasons exist to reject the concept of noncausal moral
pollution. Recently, enraged citizens of Auckland, New Zealand
burned the garage of a murderer's parents. In their anger they evi-
dently regarded the parents as polluted by their son's crime; surely,
however, such thinking is a terrible moral error. The U.S. Constitu-
tion thus justifiably proscribes punishment by corruption of blood, a
holdover from the ancient jurisprudence of pollution. The promiscu-
ous use of the noncausal concept of moral pollution is suspect; and it is
a concept that invites promiscuous use because it simply seems impos-
sible to cabin within precise, or even imprecise, boundaries. If, on the
other hand, we restrict the concept of pollution to cases where the
agent played a causal role in wrongdoing, we can no longer base the
argument on "the normative force of the actual" - the fact that the
concept of pollution persists in actual practice. In that case, propo-
nents of the concept must justify it.
Moreover, an analogous objection arises even in many cases where
the agent did play a causal role in wrongdoing. The secretary who
printed the address labels for shipments of Dalkon Shields, the clerk

66. GEORGE P. FLETCHER, RETHINKING CRIMINAL LAW ? 5.1, at 347-49 (1978).

67. Id.

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2381

who affixed these labels to the boxes, the trucker who delivered them,
the stevedore who loaded them onto a ship, the UPS driver who deliv-
ered them to physicians' offices, and the physician's receptionist all
played a causal role in the wrongful distribution of the product. Yet it
seems outrageous to insist that their causal role in wrongdoing morally
polluted all these individuals.
This suggests that if the concept of pollution is to be used at all, it
demands fine-grained distinctions of degree. One might distinguish,
for example, between investors (and employees' relatives) who are pol-
luted because they unknowingly benefited from organizational wrong-
doing, employees who are polluted merely by virtue of working for an
organization involved in wrongdoing, employees who are polluted be-
cause they played an unintentional causal role in the wrongdoing, em-
ployees who are polluted because they knew of the wrongful act and
did nothing (though they did not participate in it), and employees who
are polluted because they knowingly or intentionally participated in
the wrongful act. Ordinarily, we would regard the last category as a
worse form of pollution than its predecessors; the list proceeds in
roughly ascending order of pollution.
If that is the strategy, however, the concept of pollution does not
solve the problem of fragmented knowledge - it merely restates that
problem. True, the concept of pollution provides a resource by which
we can in some sense condemn nonintentional involvements in wrong-
doing. But under conditions of bureaucratic fragmentation of knowl-
edge, the worst form of moral pollution - pollution by intentional
wrongdoing - seems to disappear, so that bureaucracy seems able to
conjure it out of existence. Thus, even when the concept of pollution
fortifies our moral repertoire, bureaucratic fragmentation of knowl-
edge blocks the severest and most characteristic form of moral criti-
cism. That is the problem of fragmented knowledge.
To summarize, we may either resolve the problem of fragmented
knowledge by employing an undifferentiated notion of moral pollution
that is hopelessly overbroad or distinguish gradations of pollutions, in
which case the problem of fragmented knowledge has merely been re-
stated rather than resolved.
Let us return to Kundera's example. The most striking fact about
the Czech Communists he discusses is that they were not simply igno-
rant of what they were doing in the way Oedipus was ignorant when
he married Jocasta. They were "enthusiasts convinced they had dis-
covered the only road to paradise. They defended that road so val-

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2382 Michigan Law Review [Vol. 90:2348

iantly that they were forced to execute many people."68 They had
embarked on a course of political action that involved recognizable
moral risks. They knowingly collaborated with a powerful foreign
power in order to bring their own country under its dominion, bank-
ing on the Soviets' claim to represent the party of humanity. If they
did not know what they were doing, they should have found out.
Thus, to analyze Kundera's problem correctly, we should not hold
the Communists strictly morally liable, like Oedipus. Rather, we must
understand that they took voluntary part in an organizational enter-
prise engaged in risky business, giving them special and stringent obli-
gations to comprehend the import of their actions. Tomas is right to
find Did they know or did they not? a "second-rate" question69 - not
because, as with Oedipus, knowledge is irrelevant, but because the
right question is Should they have known? Accordingly, we propose to
give the concept of culpable ignorance a more central role in assigning
responsibility for organizational wrongdoing.

D. Extending Individual Responsibility

The simplest way of extending the concept of culpable ignorance to

situations in which the knowledge conditions fail is to invoke an anal-
ogy to drunk driving. By the time a driver has had six drinks, he may
no longer have the reflexes or judgment to avoid an accident, and thus
in one sense he is not responsible for what he does behind the wheel.
Obviously, though, we do hold him responsible. Why? The answer
seems simple enough: although we agree that once he became drunk
he lost effective control of his actions, we blame him for becoming
drunk in the first place. Though he was not fully responsible at the
time of the accident, it was his own fault that he was not responsible.
In Aristotle's words, "when one has once let go of a stone, it is too late
to get it back - but the agent was responsible for throwing it, because
the origin of the action was in himself."70
Analogously, we may agree that individuals in organizational set-
tings often do not know enough to be held responsible for organiza-
tional wrongdoings and yet we insist that they should have known.
They were wilfully blind. Thus, for example, if an SS officer claimed

68. KUNDERA, supra note 14, at 176.

69. Id. at 178.
70. ARISTOTLE, NICOMACHEAN ETHICS bk. III, ch. V, 1114a, at 124 (J.A.K. Thomson
trans., rev. ed. 1976). For discussion, see T.H. Irwin, Reason and Responsibility in Aristotle, in
ESSAYS ON ARISTOTLE'S ETHICS 117 (Amelie O. Rorty ed., 1980). For a proposal similar to
ours, see DENNIS F. THOMPSON, POLITICAL ETHICS AND PUBLIC OFFICE 60 (1987). Thomp-
son's discussion of what he calls "the moral responsibility of many hands," id. at 40-65, is one of
the best recent efforts to address the problem of fragmented knowledge.

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2383

that he did not know about the SS's murderous activities, we may wish
to insist that his ignorance is blameworthy. He should have known
what he was joining.
The drunk driving analogy suggests that we can hold people re-
sponsible for getting into the very predicament that at first glance
seems to relieve them of responsibility. If we focus on the act of drink-
ing, this suggests that an employee is responsible for the predicament
he gets into by joining an organization that fragments relevant knowl-
edge. Except in the case of outlaw groups like the SS, however, we do
not want to treat the act of joining an organization, like the act of
heavy drinking, as suspect or presumptively wrong.
We would do better to focus on the driving aspect of the drunk
driving analogy. Driving is a valuable activity, and our licensing pro-
cedures are designed to make it widely available. Because of the lethal
potential of the automobile, however, the privilege of driving hinges on
an exercise of alertness, caution, and self-restraint that we do not re-
quire of pedestrians or passengers. We allow pedestrians and passen-
gers to impair their reflexes and judgment with alcohol, but we treat it
as a legal and moral offense for a driver or prospective driver to do so.
Analogously, because of the great potential for harm arising from
the division of labor and fragmentation of knowledge in a corporate or
bureaucratic organization, employees may acquire duties far more de-
manding than doing no evil. They must look and listen for evil and
attempt to thwart it if they discover it. These duties, however, are not
as limited and well-defined as those imposed on the driver. We expect
prospective drivers to "just say no," but we cannot expect organiza-
tional employees to know everything about the operation in which
they are involved. While drinking is a gratuitous impediment to driv-
ing, fragmented knowledge inheres in the structure of the organization
itself. We are left with a question that does not arise in the context of
drunk driving: What and how much precaution do we require of the
individual employee?
We cannot answer this question definitively, but we can begin by
suggesting several obligations that arise from the specific risk of orga-
nizational enterprise: the risk that an individual will do or contribute
to great harm without knowing it.
1. Obligations of investigation. The first, most obvious, possibility
is to hold individuals in organizational settings morally responsible for
discerning the nature of their own projects and for discovering what
other employees are doing with their work products. The idea is obvi-
ous because it remedies the absence of knowledge in the most straight-

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2384 Michigan Law Review [Vol. 90:2348

forward way: by demanding that individuals do their best to acquire

the knowledge they lack.
2. Obligations of communication. A second possibility is to hold
individuals who possess troublesome knowledge morally responsible
for communicating it to others in the organization. Obviously, com-
munication may be a risky course of action: supervisors treasure their
"deniability," and shooting the messenger is often their knee-jerk re-
sponse. Yet riskiness does not distinguish this from other moral re-
sponsibilities: we often believe that people have moral obligations to
act against their self-interest.
3. Obligations of protection. The previous suggestions imply that
supervisors may have moral obligations to protect their subordinates
from adverse consequences of investigation and communication. For
example, they may be morally responsible for protecting
whistleblowers from retaliation.
4. Obligations ofprevention. Those in management positions may
have moral obligations to forestall wrongdoing by setting up structures
that avoid the problems we have been examining. Such preventive
mechanisms might include ombudsmen, incentive structures that re-
ward moral action, channels for anonymous information about
problems, and so on. Interestingly, the American Bar Association's
1983 ethics code requires supervisors in law firms to take measures
that ensure that their subordinates behave ethically.71
5. Obligations of precaution. In some cases, we may be able to
analogize the act of joining an organization to the act of heavy drink-
ing: the individual knows or should know that once she becomes in-
volved, her discretion and knowledge will be so constricted that she
cannot be held responsible for wrongdoing that, in broad outline, she
can reasonably anticipate. The fatal misstep is involving herself in the
first place. More often than not, would-be employees of organizations
have some prior sense of the organization's values and culture. We
may therefore hold individuals responsible for joining the organization
in the first place, as we might hold an individual German responsible
for joining the SS.
That individuals in an organization have obligations like these, and
that their breach provides a basis for assigning the individuals respon-
sibility for wrongs done in ignorance, seems plausible. But this ap-
proach to extending individual responsibility for organizational
wrongdoing raises two critical questions. First, how demanding are

71. MODEL RULES OF PROFESSIONAL CONDUCT Rule 5.1 (1983); see also John Braithwaite,
Enforced Self-Regulation: A New Strategy for Corporate Crime Control, 80 MICH. L. REV. 1466
(1982).

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2385

these obligations? There are some things we could not have learned,
communicated, or prevented even by the most vigilant inquiry, and if
should have does not imply could have, we are back to strict liability.
Second, how much vigilance is enough? The standard for liability can
be onerous even if it is not strict.
As we have described them, the moral obligations of the individual
employee seem to fall somewhere between perfect duties like not kill-
ing, with fairly precise boundaries, and imperfect duties like charity or
self-improvement, that require only some indeterminate effort. If we
attempt to make these duties perfect, through more precise formula-
tion, we risk defining them too narrowly; if we attempt to make them
imperfect, by demanding a "reasonable" effort, we risk making them
too vague.
Similar questions arise in settings where the law recognizes an af-
firmative duty to prevent harm: where the agent has a special duty to
care for the victim, has contributed to the victim's peril, has initiated
rescue, or has a responsibility to control the threatened harm. In these
settings, we hold the agent culpable for not trying to prevent harm,
but not strictly liable for its occurrence. This leaves the question of
how much of an attempt to prevent harm the agent must make. Legal
norms are hard to discover, because few courts actually recognize a
duty to aid, and in the cases where they do, the defendant has usually
failed to make even a token effort to render assistance.72
This may be more problematic when assigning criminal than moral
responsibility, since the notice requirement of criminal law compels us
to draw a clear line between required and merely permitted assist-
ance.73 One might think that moral appraisal could tolerate more
vagueness. Why should a conscientious agent need a sharp line be-
tween required and supererogatory precautions rather than an indefi-
nite duty to be as cautious and vigilant as possible? If we treat the
fulfillment of these duties as a matter of degree, however, we raise the
problem of deciding how the degree of fulfillment affects the individ-
ual's responsibility for organizational wrongs that greater effort might
have averted.
This raises the second question: What is an individual responsible
for in failing to fulfill these preemptive duties, as we shall call the obli-
gations of investigation, communication, protection, and precaution?

72. See FLETCHER, supra note 66, ? 8.3.7.

73. When the criminal law imposes liability for unintended harm, it usually requires gross
negligence or recklessness - conscious disregard of a risk. See, e.g., MODEL PENAL CODE
? 2.02(d) (1985) (culpability requires negligence constituting "a gross deviation from the stan-
dard of care that a reasonable person would observe in the actor's situation").

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2386 Michigan Law Review [Vol. 90:2348

Should we blame her only for the breach of the preemptive duty, or
should we blame her for the resulting offense as if she had known all
the relevant facts (and thus hold her immune from censure if no harm
results)? Is the employee responsible just for failing to investigate (at
the time she fails to do so), or does she lose her excuse of ignorance
with respect to any facts the investigation would have yielded?
Neither approach seems fully satisfactory. In limiting responsibil-
ity to the breach of preemptive duties, we impose the same blame or
punishment regardless of what wrongs result; in withholding the ex-
cuse of ignorance for the offense, we treat a negligent employee as if
she were responsible for intentional wrongdoing. And if we assign re-
sponsibility only when a wrong is actually done, we fail to censure
those lucky enough to ignore their special obligations without adverse
effect.

This dilemma is hardly peculiar to the problem of organizational

settings, however. We display much the same ambivalence about as-
signing blame for the variable consequences of individual agency. Le-
gally and morally, we are torn between a desire to make agents with
the same mental state equally culpable and an urge to consider the
consequences in assigning blame. Moreover, while we condemn those
who fail in their deliberate efforts to bring about wrongs almost as
severely as those who succeed, we condemn those who risk doing
wrong without intending it far more severely when the wrong is actu-
ally done.
But denying the excuse of ignorance to those who do not fulfill
their preemptive duties in organizations raises special problems. By
denying employees the epistemological excuse, we rest their fault on
events beyond their control; we effectively treat them as if they inten-
tionally engaged in serious wrongdoing of which they were in fact ig-
norant. If this is harsh legally, it is unacceptable morally. One can
hardly imagine how any level of negligence in failing to investigate the
possibility of wrongdoing could be the moral equivalent of intention-
ally doing the wrong.
Moreover, other agents will often do the wrong the employee fails
to avert. If we held an employee responsible for others' wrongdoing
he could have prevented, just as if he had intentionally approved or
encouraged that wrongdoing, we would greatly increase the scope of
his responsibility and the sway of moral luck. We would impose on
him a moral responsibility as great as the legal responsibility we now
reserve for co-conspirators, based on their explicit agreement to pur-
sue a criminal venture.
Finally, as we noted earlier, the fact that the obligations we have

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2387

described are indeterminate in their extent and stringency exacerbates

the tension. To say to a manager that she should have done more to
prevent her employees' wrongdoing, or to an employee that she should
have blown the whistle on her boss despite the risk, is one thing. It is
quite another thing to say that as a result of having failed to do
enough, or of having caved in to an insufficiently grave threat, the
individual is now responsible for serious wrongdoing as if she had in-
tentionally engaged in it.
Several writers on criminal jurisprudence have suggested that in
order to assign responsibility appropriately, we need to view the indi-
vidual's actions in a consistently broad "time-frame" - to look not
only at the actus reus defined by statute but also at the actions and
decisions leading up to it, particularly those that promote, mitigate,
excuse, or justify the act's commission. In the most fully developed
proposal of this kind, Paul Robinson has argued that the law can rec-
ognize a person's responsibility for getting into a predicament without
denying her the excuse created by that predicament.74 The critical
inquiry is whether, in acting in a way that creates an excuse, the indi-
vidual is at fault for the offense excused.
To return to the drunk driving analogy, a driver who drank until
he was no longer able to appreciate the hazards of the road would not
lose the excuse of ignorance if an accident occurred, as if he had know-
ingly caused the accident. Rather, he would be guilty of reckless
homicide for disregarding the risk of a fatal accident; if he should have
known of the risk but remained ignorant (e.g., because he did not
bother to find out that the punch he was drinking was highly alco-
holic), he would be guilty of negligent homicide.75
Applied to organizational wrongdoing, an approach like Robin-
son's would ground the employee's responsibility for the harm in her
preemptive duties to investigate, prevent, and so forth. By breaching
these duties, the employee has played a role in causing or contributing
to the commission of the offense. The key question would not be
whether the employee deliberately, recklessly, or negligently breached

74. Paul H. Robinson, Causing the Conditions of One's Own Defense: A Study in the Limits
of Theory in Criminal Law Doctrine, 71 VA. L. REV. 1 (1985).
75. Similarly, if a person negligently provoked another to use deadly force, he would not lose
his right to kill in self-defense. If he exercised that right, however, he would be guilty of negli-
gent homicide for creating the need for self-defense. If he provoked another in order to give
himself a legal justification for killing that person, he would be guilty of intentional homicide.
To bear any responsibility for the ultimate offense, the agent must be at fault for creating the
specific excusing or justifying condition he relies on. If he negligently provoked someone without
reason to expect that she would resort to deadly force, he would not be guilty of homicide in
creating the need for self-defense. He would be responsible at most for negligent provocation and
the use of nondeadly defensive force.

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2388 Michigan Law Review [Vol. 90:2348

her duty but whether, by failing to perform it, she intentionally, reck-
lessly, or negligently facilitated the wrongdoing.
Thus, if an employee should have known that by investigating a
new project before it commenced, she could ascertain whether it in-
volved exporting toxic substances, and there was some reason to sus-
pect that it might, she would be responsible for negligently exporting
those substances, even if, by the time she exported them, the most
diligent inquiry would not have revealed their toxic character. Had
she actually known that an investigation could have revealed that in-
formation, but failed to investigate, she would be responsible for reck-
lessly exporting toxic substances, even if, by the time she exported
them, she was no longer able to ascertain their toxic character.
In some cases, this approach may lead us to hold employees who
act in ignorance responsible for intentional wrongdoing. If the em-
ployee deliberately insulated herself from knowledge about the ex-
ports, intending to export toxic substances without being told that
specific exports were toxic, she would be responsible for their inten-
tional export despite her ignorance at the time she exported them. But
if she deliberately insulated herself from such knowledge only because
she hated confrontations, she would be responsible for no more than
reckless export, since she did not know the character of the substances
or intend or hope that they might be toxic. This approach, then,
avoids the harshness of denying an employee an excuse for organiza-
tional wrongdoing if she has any fault for creating the excuse. It treats
her as responsible for wrongdoing only to the extent that she is at fault
for excusing or justifying its commission.
Applying this approach to organizational wrongdoing, however,
involves complexities. In the examples Robinson discusses, a person
gives himself an excuse of ignorance by an action that impairs his abil-
ity to obtain information critical to the safe performance of his task: a
driver drinks himself insensible, a police officer leaves his glasses home
when he goes on duty. In such cases, the risks of acting in the im-
paired condition seem obvious; to regard the driver or police officer as
reckless in disregarding those risks makes sense. But the risks of pro-
ceeding in ignorance may not be so obvious in an organizational set-
ting. Those risks depend on what the organization does, what it can
or might do wrong, and how the employee's own action can or might
contribute to those wrongs. Acquiring this background knowledge
may itself require extensive inquiry, and the question then becomes
what should trigger such an inquiry. The strength of the employee's
preemptive duties will depend on the risks they serve to preempt, but

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2389

the employee may not be able to appreciate those risks without having
already fulfilled those duties.
When we consider this problem less abstractly, we do not face a
regress or a bootstrap. We merely face the need to condition the du-
ties we impose on what the (prospective) employee knows or should
know at the outset about the organization, its work, and her role in it.
For example, an employee's duty of investigation and precaution will
be much greater if she has heard rumors that the import-export busi-
ness to which she has applied is a CIA front, or if she knows that
similar businesses in that part of the world have often been used as
CIA fronts. Because so much depends on what the employee knew or
should have known, suspected or should have suspected, at various
times, we must be very cautious in assigning individual responsibility
for organizational wrongdoing. Frankly, we will often lack the infor-
mation necessary to do so. But such epistemic limits to faultfinding
are hardly unique to the organizational setting.
The devices we have suggested for extending individual responsi-
bility may be too demanding because they require a debilitating level
of caution, wariness, and investigation in routine organization work.
Alternatively, they may fail to justify the ascription of responsibility to
individuals in many common organizational settings. But until the at-
tempt has been made to broaden and adapt the notion of individual
responsibility, news of its death will be premature.

CONCLUSION

The preemptive obligations we are proposing have in recent years

become widely accepted in both public and private bureaucracies.
Governmental agencies typically have ombudsmen, and state and fed-
eral governments have enacted protections for whistleblowers. In the
wake of several incidents, the Exxon Corporation has enacted regula-
tions requiring employees who notice possible misconduct or danger-
ous situations to notify their superiors in writing; the superiors, in
turn, are required to respond in writing, and if no written response is
forthcoming, the employee must jump the chain of command and in-
form higher-level executives.76
Regarded as public policy proposals, our preemptive obligations
are already found in corporate manuals and memoranda. We are not
offering a proposal for regulations whose time has come, however, but
an account of individual moral responsibility. That is, we argue not
only that bureaucratic organizations should institute policies along the

76. We owe our information about Exxon's regulations to John Braithwaite.

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2390 Michigan Law Review [Vol. 90:2348

lines we have suggested, but also that individual executives within the
organizations are morally blameworthy for failing to implement such
policies and that individual employees and executives are morally
blameworthy for violating preemptive obligations even in the absence
of policies implementing them.
The latter point is particularly significant, because it implicates the
moral culture of organizations as well as their formal policies. One
could imagine an organization enacting Exxon-style regulations while
preserving a moral culture that discouraged the whistleblowing that
the regulations require; in that case, employees would fear to go over
their supervisors' heads, and the regulations would function primarily
to insure that lower- or middle-level employees took the fall when
things went wrong.77 To insist as we do that the preemptive obliga-
tions form part of our moral responsibilities is to insist on the spirit as
well as the letter of preemptive regulations.
Thus, our proposal amounts to a reconfiguration of the dimensions
of individual moral responsibility as well as a reconfiguration of the
structure and culture of bureaucratic organizations. Earlier, we ar-
gued that familiar moral theories cannot handle the problem of frag-
mented knowledge. Not only do their first-order precepts presuppose
the four knowledge conditions, but they typically marginalize the phe-
nomenon of culpable ignorance. We conclude with some brief reflec-
tions on how our proposal differs from these familiar moral theories.
The account we propose retains the concentric-circles structure of
familiar moral theories: a core of first-order precepts; a surrounding
layer of mitigations and excuses; and an outer layer of conditions de-
feating those mitigations and excuses. In effect, the modification we
propose relocates the notion of culpable ignorance from the category
of conditions defeating mitigations and excuses to the first-order
precepts. That is, we place preemptive obligations among the basic
moral duties of the individual in bureaucratic settings. In one sense,
then, our proposal may be regarded as a minor modification, or
friendly amendment, to familiar accounts of morality.
In another sense, however, the change is fundamental, for it in-
volves a different model of the moral life. Traditional moral theories
presuppose an episodic or picaresque model: the moral life consists of

77. John Braithwaite has reported that American pharmaceutical company executives re-
peatedly told him in interviews that someone's job was "vice-president responsible for going to
jail": his duty was to take the fall for his superiors, and after several years in this position he
would be rewarded by moving to a well-compensated "safe" job. See Braithwaite, supra note 29,
at 3. In such an institutional culture it is hardly farfetched to assume that Exxon-style regula-
tions might be instituted for purposes of liability screening rather than avoiding fragmented
knowledge.

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
August 1992] Moral Responsibility 2391

a progression of discrete trials or encounters, in which the individual

confronts a temptation or a difficult choice and marshals her virtue
and willpower to meet the challenge. Between these episodes of
heightened moral tension and challenge, life proceeds more tranquilly,
in a state of comparative moral neutrality. From the labors of Hercu-
les to the voyage of Odysseus to the progress of Bunyan's Pilgrim, the
basic stories of the West have pictured the moral life as a kind of er-
rantry. Roberto Unger has described this view of moral encounter as
the "Christian-romantic view of the self,"78 by which he means to sug-
gest not only that Christianity adapted the pagan literary form of the
romance but that this form is the deepest expression of how we mor-
ally conceive of our lives. Plainly, the episodic form of the moral life
requires the four knowledge conditions, whose satisfaction permits us
to recognize that the next moral test has come upon us. Our argu-
ment, however, has been that discrete and bounded episodes of moral
challenge arise in a moral world far different from that of the bureau-
cratic organizations in which moder men and women lead their lives.
Moving the preemptive obligations to the core precepts of morality
does more than reblend the wine in old bottles; it changes how we
understand the moral life. We now see ourselves not as individual
souls, progressing from one moral episode to the next, but as members
of organizations in a continuous but changing network of involve-
ments. We now see moral encounters not simply as tests of will and
virtue in the face of known and defined obstacles, but as a part of the
continual effort to position ourselves in a state of clarity where knowl-
edge and definition at last become possible. The comprehension that
we stand at a moral crossroads is not given, but acquired through con-
siderable effort, courage, and even luck. Obtaining that knowledge is a
moral challenge coequal to, or perhaps greater than, the challenge of
responding to it.
Indeed, the organizational setting changes the contours of moral
challenge and thus of moral virtue. Because of the division of respon-
sibility in organizations, responding to a clearly presented choice may
be less demanding than the episodic view suggests, for others are there
to share our burden. Recall that in the Milgram experiments it proved
easier to resist the wrongful demand when others resisted with the
subject. However, because of the fragmentation of knowledge in orga-
nizations, coming to understand that one faces a clearly presented
moral choice may be more demanding than the episodic view suggests;
it may require considerable personal courage and persistence. Our

78. ROBERTO M. UNGER, PASSION: AN ESSAY ON PERSONALITY 24 (1984).

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms
2392 Michigan Law Review [Vol. 90:2348

moral trials may be less intense but more protracted; if we are heroes,
we will be heroes by virtue of the quantity rather than the quality of
our efforts. The institutionalization of everyday life may demand
fewer of the virtues of a soldier or saint and more of the virtues of a
detective, a scientist, or a trial lawyer aiming to reconstruct shards of
evidence into a coherent theory.
That takes us back to Kafka. In its form, The Trial seems like a
picaresque novel in which Joseph K. progresses from one episodic en-
counter to the next. At each stage, however, the moral test in the
encounter is the same: can Joseph K. fend off distraction and obfusca-
tion to learn what he is accused of and what he can do about it? His
trial, on our account, consists precisely of this repeated test, and his
crime consists of his failure to meet the challenge. His obligation is to
find his way through a nightmare of fragmented knowledge: to reas-
semble the misshapen pieces of his experience into a picture that tells
him at last where he stands in the order of guilt and innocence. Under
our view of morality, Joseph K.'s predicament is ours. His failure is
our failure as well.

This content downloaded from 104.28.88.131 on Wed, 09 Apr 2025 10:03:37 UTC
All use subject to https://about.jstor.org/terms