University of Chicago Law School

Chicago Unbound
Journal Articles Faculty Scholarship

1999

The Case of the Speluncean Explorers: Revisited

Frank H. Easterbrook

Follow this and additional works at: http://chicagounbound.uchicago.edu/journal_articles

Part of the Law Commons

Recommended Citation
Frank H. Easterbrook, "The Case of the Speluncean Explorers: Revisited," 112 Harvard Law Review 1913 (1999).

This Article is brought to you for free and open access by the Faculty Scholarship at Chicago Unbound. It has been accepted for inclusion in Journal
Articles by an authorized administrator of Chicago Unbound. For more information, please contact [email protected].
 THE CASE OF THE SPELUNCEAN EXPLORERS:
A FIFTIETH ANNIVERSARY SYMPOSIUM

FOREWORD: A CAVE DRAWING FOR THE AGES

David L. Shapiro*

I. INTRODUCTION'

When I was a student in law school, my two favorite law review

articles were Henry Hart's famous dialogue 2 and Lon Fuller's presen-
tation of the case of the speluncean explorers.3 They still are.
Why is that so? Perhaps one never quite gets over the joy of dis-
covering a fine work of art or literature when one is young. (I still re-
vere War and Peace, which captivated me as a college sophomore,
even though I can't get past the first hundred pages any more.) But I
think more is involved here. The wonderful essays by Hart and Fuller
each combine a timely consideration of contemporaneous debates with
a timeless quality that continues to entice students and scholars to
think and write about them some half a century later - and will
doubtless engage our successors well into the next millennium.
Moreover, each of the essays takes a form that I have always admired
and that seems especially suited to the exploration of such basic ques-
tions as the nature of our federal union and the nature of law itself: an
exchange of views in which competing positions are stated as force-
fully as the author knows how. Indeed, an author's ability to make
compelling statements of contrasting views is, for me, a powerful sig-
nal of the author's worth as a scholar.
Small wonder then that when I was invited to contribute a fore-
word to this revisiting of Fuller's great work, I felt both flattered and

* William Nelson Cromwell Professor of Law, Harvard University.

1 Unlike several of the writers of the opinions that follow, I cannot resist the temptation to use
footnotes (in moderation, of course). The purpose of this footnote is to suggest to the reader that
you may well prefer, as I do, to read a foreword, if at all, only after you have read all that follows.
This practice not only tends to make the foreword more readable, but also eliminates any chance
that the views expressed in the foreword will affect your reactions to what follows. Thus, you are
in a better position to assess the merits and defects of the foreword itself. But if you must, read
on.
2 Henry M. Hart, Jr., The Power of Congress to Limit the Jurisdictionof Federal Courts: An
Exercise in Dialectic,66 HARV. L. REV. 1362 (1953).
3 Lon L. Fuller, The Case of the Speluncean Explorers, 62 HARv. L. REV. 66 (I949). The
article is reproduced below, see infra, at 1851-75, and for ease of reference, citations will be to the
article as it appears in this issue.

1834
1999] THE SPELUNCEAN EXPLORERS 1835

intimidated. If good wine needs no bush,4 and the lily is not made
more beautiful by being gilded, then what could I hope to add to such
an extraordinary achievement? No more, perhaps, than some thoughts
on just why Fuller's piece has proved so durable and so provocative,
and some effort to connect its insights with those of our contributors to
this celebration.

II. FULLER'S ACHIEVEMENT

To be sure, Fuller, like Hart and just about everyone else, was only
mortal, and he could not wholly escape the context of the times in
which he wrote. Hart was necessarily dealing with the state of consti-
tutional doctrine as it then stood,5 and casually followed the custom of
the times by using the term "wetback" when referring to a Mexican
who had illegally entered the country across the Rio Grande. 6 As for
Fuller, his hypothetical case was staffed. by justices who were all male,
and though we have little to go on, may also have been all white and
all from relatively affluent backgrounds. 7 After all, judges predomi-
4 This saying is a Shakespearean phrase I learned from Paul Freund and have always treas-
ured because it is so obscure. In olden times, a branch of ivy (a bush) was hung outside a tavern
to indicate wine for sale.
S Hart's Dialogue was reproduced in the third edition of PAUL M. BATOR, DANIEL J.
MELTZER, PAUL J. MISHKIN & DAVID L. SHAPIRO, HART AND WECHSLER'S THE FEDERAL
CouRTs AND THE FEDERAL SYSTEM (3d ed. i988), but was replete with footnotes bringing as-
pects of the text up to date. See id. at 393-423. Because this burden had become increasingly
heavy, we decided, as editors of the fourth edition, to discuss and quote liberally from the article
but not to reproduce it. See RIcHARD H. FALLON, DANIEL J. MELTZER & DAVID L. SHAPIRO,
HART AND WECHSLER'S THE FEDERAL COURTS AND THE FEDERAL SYSTEM 366-67 (4th ed.
i996).
6 See HART, supra note 2, at 1395. At this point, Hart was complaining that under then-
current Supreme Court doctrine, an alien who had entered the country illegally had all the protec-
tions afforded by the guarantees of due process, while a resident alien who goes abroad to visit a
dying parent and seeks to return with a duly issued passport and visa appeared to have none.
Now, we live in an era when - perhaps as a reaction to earlier times - one can be accused of
a racial slur and lose one's job (at least for a while) for using "niggardly," a word of Scandinavian
origin, see Jonathan Chait, Doubletalk, NEW REPUBLIC, Feb. 22, 1999, at 50, 5o, and when "he" is
no longer a politically acceptable generic pronoun. We are all caught up in our times.
7 That all the justices are male is evident from the internal references by the justices them-
selves to their colleagues. Their other characteristics are matters of conjecture, and indeed as-
sumptions about those characteristics can only be based on a guess about how people in the late
1940S thought about the judiciary, and on the failure of any of the justices to make a point about
his race, nationality, or background. To quote Professor Eskridge in his 1993 discussion of
Fuller's piece, "There is no explicit clue of any sort to the race of any participant. That is, itself,
an implicit clue. In the 194os, it went without saying that you were white if your race was not
noted." William N. Eskridge, Jr., The Case of the Speluncean Explorers: Twentieth-Century
Statutory Interpretationin a Nutshell, 61 GEO. WASH. L. REv. 1731, 175o n.Iii (i993). (Is this
assertion - that a black writer in the I94Os, writing in any context, would always refer to his race
- supported by empirical data?) Eskridge goes on to say, "The affluence of the Speluncean world
is suggested by the preppy, upper-class context of the hypothetical: the hobby is the rarefied, rela-
tively expensive one of cave-exploring. Moreover, [the case ends] up as a battleground of New-
garth's political elites (the Chief Executive and the Court)." Id. at 175o-5i n.112 (citation omit-
ted). (Was Fuller's move - from the real-life seafaring cases cited below, see infra note 9, to a
1836 HARVARD LAW REVIEW [Vol. II12:1834

nantly had those qualities when Fuller wrote. And when one of his
justices wanted to argue that it can be easy to tell that a speaker's pre-
cise language contains a slip of the tongue or an overgeneralized com-
mand, the justice pointed to the ability of the "stupidest housemaid" to
interpret her employer's words in light of their purpose8 - thus per-
haps revealing some assumptions about the nature of the employer-
employee relationship, especially when the employee is a domestic.
Other examples doubtless abound, as they do in the work of every
writer.
But Fuller was still able to write a piece that will endure - one
that posed eternal dilemmas in a remarkably lucid and accessible
fashion. Let me count the ways.
First, while the hypothetical - about the dilemma facing those
who must kill one of their number or all die of starvation - drew
loosely on two famous cases, 9 Fuller made his own case more difficult
and challenging through a variety of devices. He moved the setting to
Newgarth, a jurisdiction of which we know little except for a few mat-
ters that leak out of the opinions - for example, that it has prece-
dents, statutes, judges (all male in this case), a chief executive with the
power to pardon, and housemaids who may sometimes be stupid. And
to confirm the limits of our knowledge, the time of the relevant events
is in the fifth millennium.10
With respect to the facts themselves, Fuller enriched the knowledge
of the defendants and increased the dilemmas of the case in wondrous

case involving explorers - made in order to change the social class of the accused or because the
cave situation was more pliable in terms of the facts he wanted to develop? And is the institu-
tional issue he wanted to present - the issue of institutional role in a system of law - properly
characterized, in terms of either its significance or the author's purpose, as a "battleground" of
"political elites"?)
8 Fuller, infra, at 1859 (Foster, J.). To quote Professor Eskridge again, "The only appearances
of nonwealthy people in the case are demeaning .... Most revealing is the snide reference by Jus-
tice Foster - the 'nice' Justice - to the 'stupidest housemaid."' Eskridge, supra note 7, at 1751
n.112. This point is made the capstone of Professor Paul Butler's opinion on this issue. See infra,
at 1917 (Stupidest Housemaid, J.).
9 See Regina v. Dudley & Stephens, 14 Q.B.D. 273 (1884) (involving defendants, who, after
twenty days on a lifeboat, killed and then ate the youngest person on the boat - evidently with-
out any agreed-upon procedure for determining the one to be sacrificed - and who were ulti-
mately convicted of murder but had their death sentences commuted); United States v. Holmes,
26 F.Cas. 36o (C.C.E.D. Pa. 1842) (No. 15,383) (involving a defendant who was a member of the
crew of a ship that sank and who was convicted of manslaughter and sentenced to six months
imprisonment for throwing several passengers out of a long-boat so that he and the others in the
boat might survive).
10 Some think that to deal with a case fairly and fully, we must be able to explore in depth
every aspect of the context in which it arises. Cf., e.g., JOHN T. NOONAN, JR., PERSONS AND
MASKS OF THE LAw II1-51 (1976) (discussing the context of Palsgrafv. Long Island Railroad,
162 N.E. 99 (N.Y. 1928)). Of course, no hypothetical can meet such a demanding standard,
though Fuller has clearly gone beyond the standard A, B, and C of the law school classroom, and
made a concerted effort to provide enough information for full debate of the issues he wanted to
raise.
 1999] THE SPELUNCEAN EXPLORERS 1837

ways. For example, his trapped explorers find out that they cannot be
rescued in less than ten days and are assured by experts that their
chances of survival for ten days are slim to none unless they eat one of
their members. Then, most intriguing of all, they all agree to draw lots
(actually, to throw dice), to determine who shall be sacrificed, but be-
fore the lottery, Whetmore tries unsuccessfully to pull out of the
agreement. Predictably (Fuller was never a candidate for a Booker
Prize), Whetmore turns out to be the loser when the dice are cast for
him by another, and he is killed and eaten. The others survive and are
prosecuted for murder under a statute providing, in its entirety, "Who-
ever shall willfully take the life of another shall be punished by
death.""
We learn of other important matters as well - that ten members of
the rescue party died in the course of their efforts, that Newgarth's
Chief Executive was well known for his hard-nosed attitude toward
clemency, and that there were significant precedents on the books, ad-
dressing such issues as the availability of self-defense as a justification
for killing (despite the failure of the legislature to mention it), the
willingness of Newgarth's courts to construe statutes to avoid absurd
results, and the application of the anti-theft law to one who stole bread
(Valjean) because he was starving and could not afford to buy it. In
sum, as one who has often faltered in the effort to construct a flawless
hypothetical, I think that Fuller's comes about as close to perfection as
one can get.
Second, Fuller's opinions for his five justices managed to express
an extraordinary range of views, and to do so with vigor and power.
Truepenny, the Chief Justice, plays the role of narrator (a bit like the
butler who comes on stage in Act I of a drawing-room comedy to dust
the furniture and tell the audience what happened before the curtain
went up). But he goes on, briefly but eloquently, to express two impor-
tant viewpoints: first, that statutory language governs when it is free
from ambiguity (as he claims it is in this case); and second, that institu-
tionally, the role of mercy-giver in the criminal context belongs not to
2
the judiciary but to the executive in the exercise of the pardon power.'
Chief Justice 'ruepenny is followed by Justice Foster, who strongly
disagrees that the conviction must be affirmed, and in doing so puts
forward two separate (but perhaps related) 13 arguments: the defen-
dants, when they acted, were "in a 'state of nature,"' as much outside
the laws of Newgarth as if they were on the moon, and under the prin-
ciples applicable in such a state (in other words, the principles of
"naturall law'), they were guiltless; 14 and, in any event, and in a more

11 Fuller, infra,at 1853 (Truepenny, C.J.).

12 See id. at 1853-54.
13 Eskridge suggests that they are, and I agree. See Eskridge, supra note 7, at 1742.
14 Fuller, infra, at x855 (Foster, J.).
1838 HARVARD LAW REVIEW [VOL. 112:1834

traditional vein, the murder statute must be interpreted in accordance

with its purpose - namely, deterrence. 15 That purpose, he concludes,
would no more be served by upholding a conviction on the facts at bar
than in the case of the recognized justification of self-defense.16
Justice Foster is then powerfully attacked in the two opinions that
follow. Justice Tatting derides Justice Foster's first argument - ques-
tioning when one can decide that an actor has crossed over into a state
of nature and how the court acquires its authority to apply natural law
- and then heaps similar scorn on Justice Foster's "purposive" analy-
sis, in part by insisting that purposes are both difficult to ascertain
and, usually, multiple. 7 The justification of self-defense is readily dis-
tinguished, the Valjean precedent is invoked, and then Justice Tatting,
baffled by the difficulty of the case and resentful of the decision to
prosecute these hapless defendants under a statute providing a manda-
18
tory death penalty, decides to withdraw.
Justice Keen, a man of similar views but made of sterner stuff,
votes to affirm. He insists that his own view of the morality or im-
morality of the acts charged is irrelevant, and that the court must rec-
ognize the supremacy of the legislature by applying the statute as writ-
ten - not by rewriting it as the justices would like it to read through
the dodge of ascertaining some fancied "purpose" or filling some non-
existent "gap." 1" He even suggests that the courts may have erred
years earlier in recognizing the justification of self-defense instead of
leaving it to the legislature, if it wished, to spell out the precise con-
tours of such a defense.20
Finally, Justice Handy, the realist-pragmatist, scoffing at the
learned debates among the other justices, insists that the justices must
follow their own common sense and the popular will - in this case,
evidenced by a poll showing that ninety percent of the people want to
let the defendants off with little or no punishment - and reverse the
convictions. 2' He suggests achieving this result by using whatever le-
galistic device seems most adaptable ("handy"?) 22
to the occasion - in
this instance, Justice Foster's second rationale.
Given a chance to reconsider his withdrawal, Justice Tatting sticks
to his guns (if that is an apt metaphor for a coward), and partly as a

IsSee id. at 1858.

16 See id.
17 See Fuller, infra, at 1859-6i (Tatting, J.).
18 See id. at 1863.
19 See Fuller, infra, at 1864 (Keen, J.).
20 See id. at i868.
21 See Fuller, infra, at 1868, 1870 (Handy, J.). In a delightful passage in which Fuller perhaps
gets carried away, Justice Handy dismisses the likelihood of executive clemency on the basis of his
knowledge of the Chief Executive's character - knowledge acquired because, as it happens, "my
wife's niece is an intimate friend of his secretary." Id. at 1872.
22 See id. at 1870.
 1999] THE SPELUNCEAN EXPLORERS 1839

result of his refusal to face up to his responsibility as a judge, the con-

victions are affirmed by an equally divided vote.
Thus, Fuller managed in these five opinions to introduce just about
every dispute about the nature of law and the role of judges. As Jus-
tice Handy notes before launching into his realist critique, the prior
opinions have explored the clash between natural law and positivism,
have examined a range of approaches to statutory interpretation, and
have raised fundamental questions about the roles and limits of our le-
gal institutions.
A third virtue of Fuller's essay is that if one were unfamiliar with
his other works, one would be hard-pressed to identify his own pre-
ferred approach, although he is perhaps too cynical about the tech-
niques of the realists (as embodied in Justice Handy's opinion) to be
readily identified with that school of thought. In fact, Fuller's other
works reveal an affinity for both aspects of Justice Foster's ap-
proach. 23 Indeed, Fuller's view of the significance of purposive aualy-
sis in interpreting statutes gave rise in later years to the "legal process"
theories of Professors Hart and Sacks,24 and yet his criticisms of that
approach in the opinions of Justices Tatting and Keen are so trenchant
that future scholars have been able to add little to their arguments. As
noted earlier, this ability - to recognize and articulate the weaknesses
in one's own theories - constitutes, in my view, a hallmark of true
scholarship.
Fourth, Fuller not only used his "quintalogue" to explore some of
the burning issues of his own day, especially the effort to resolve the
challenges that natural law and positivism posed for each other; he
also hit upon a technique for articulating these problems that has suc-
ceeded in engaging students and teachers ever since - witness this
second symposium on the case in the last six years. Moreover, as I try
to show, the scholars that have followed him may have cast some fur-
ther light, but the real illumination still comes from Fuller himself.
Finally, Fuller did all this in a remarkably compact form. Al-
though it is not unusual for a present-day article on an obscure prob-
lem of, say, bankruptcy law to stretch out for a hundred dreary pages,
Fuller's five opinions consumed only thirty. And the style was not
only lucid and accessible; it was also lively and witty throughout. Er-
win Griswold, a man of simple tastes and direct speech, caught the es-
sence of Fuller's gift for avoiding pretense and obscurity when he once
introduced Fuller as "the only jurisprude I can understand."

23 See Eskridge, supra note 7, at 1737 n.38 (citing LON L. FULLER, THE LAW IN QUEST OF
ITSELF (940); Lon L. Fuller, American Legal Philosophy at Mid-Century, 6 J. LEGAL EDUC. 457
(1954); and Lon L. Fuller, Reason and Fiat in Case Law, 59 HARV. L. REv. 376 (946).
24 See HENRY M. HART, JR. & ALBERT M. SACKS, THE LEGAL PROCESS:BASIC PROBLEMS
IN THE MAKING AND APPLICATION OF LAW IIII-I38O (William N. Eskridge, Jr. & Philip P.
Frickey eds., Foundation Press 1994) (1958).
 1840 HARVARD LAW REVIEW [Vol. 112:1834

III. LATER ANALYSES AND ONSLAUGHTS

A
The first extensive return to Fuller's cave appeared as two articles
in the George Washington Law Review in 1993. The articles were enti-
tled The Case of the Speluncean Explorers: Twentieth-Century Statu-
tory Interpretation in a Nutshell 25 and The Case of the Speluncean
Explorers: Contemporary Proceedings. 2 6 Professor William Eskridge,
the architect of the project, contributed an introductory analysis of
Fuller's work, and his essay was followed by seven new opinions in
the case authored by a range of academics. Most of Eskridge's intro-
ductory analysis consisted of a scholarly exegesis of Fuller's piece,
which Eskridge described as representing "a moment in the Anglo-
American debate over the role of equity and natural law in statutory
interpretation," 27 and also as a harbinger of the "Legal Process" ap-
proach more fully developed in later years by Professors Hart and
Sacks. 28 Eskridge also noted the skill with which Justices Tatting and
Keen question both the legitimacy and appropriateness of Justice Fos-
ter's use of natural law in his first argument and of "purposivism" in
his second. Then, as part of his introduction of the opinions that fol-
low, Eskridge notes that the world of the case "and of its Justices -
and Lon Fuller's world - [is one] in which the only actors who matter
'29
are male, white, affluent, and heterosexual.
Eskridge's introduction is knowledgeable, informative, and gener-
ally respectful of Fuller's insights. My view, which is already apparent
and which may not quite jibe with his, is that Fuller's essay is much
more than a document of historical importance - that it transcends a
moment in legal history, or even several moments, and that (granting
that it cannot wholly escape the tacit assumptions and understandings
of its day) will continue to fascinate and provoke its readers as long as
it remains available.
In the seven opinions that followed Eskridge's introduction, per-
haps the most notable feature is that not one new justice voted simply
to affirm the conviction and sentence; rather, three voted to reverse the
conviction; two voted to remand for further factual inquiry relating to
- or for jury determination of - guilt or innocence; one voted to re-

25 Eskridge, supra note 7.

26 Naomi R. Cahn, John 0. Calmore, Mary I. Coombs, Dwight L. Greene, Geoffrey C. Miller,
Jeremy Paul & Laura W. Stein, The Case of the Speluncean Explorers: ContemporaryProceedings,
61 GEO. WASH. L. REv. 1754 (1993) [hereinafter Contemporary Proceedings].
27 Eskridge, supra note 7, at 1732.
28 Id. at 1743.
29 Id. at 1750-51.
 1ggg] THE SPELUNCEAN EXPLORERS 1841

mand solely on the question of the appropriate

30
sentence; and one voted
to reverse the sentence of execution.
That no one voted to affirm both the conviction and sentence is, in
part, a result of Professor Eskridge's selection of judges. As he ac-
knowledges, three were selected as representatives of feminist theory
and two as representatives of critical race theory.3 ' Of the remaining
two, one advocated a "purposive" analysis reminiscent of that espoused
by Justice Foster,3 2 while the other appeared to speak for the neo-
33
realist, "critical legal studies" approach presaged by Justice Handy.
The three advocates of feminist theory were Naomi Cahn, Mary
Coombs, and Laura Stein. Professor Cahn voted to remand for fur-
ther development of the facts in order effectively to "integrate" the
ethics of "care and justice."3 4 Professor Coombs noted that it was too
easy for judges to identify themselves with the "privileged" male de-
fendants who found themselves facing death; she then concluded (per-
haps in part because of her fear of judicial bias) that, since the record
did not establish an effective waiver of trial by jury by the defendants
themselves, the case should be retried in order to obtain a jury verdict
on the ultimate question of guilt.35 Professor Stein, after speculating
on the possible impact of the decision on the disempowered (specifi-
cally including battered women), concluded that much would be lost
by executing these defendants and that as one who "willfully would
not give [the defendants] guidance beforehand," she was "estopped
'36
from judging with hindsight.
As for the two representatives of critical race theory, Professor John
Calmore concluded in light.of his own narrative of the history of racial
30 This is my reading of the conclusions reached, but in some instances, the authors might dis-
agree with that reading.
31 See Eskridge, supra note 7, at 1751-52.
32 See ContemporaryProceedings,supra note 26, at i8oo ("We have both the right and the re-
sponsibility to interpret statutes in such a way as to serve the apparent legislative purpose - in-
distinct as that may be.. . .") (opinion of Professor Geoffrey C. Miller).
33 See id. at 18o1-o7 (opinion of Professor Jeremy Paul). Paul rejects Justice Handy's reliance
on the views of "the common man." Id. at i8o6. But in reaching his conclusion that it would be
"monstrous ... to put these defendants to death for actions we can't even agree constitute a
crime," id. at 18o7, Paul confesses his "inability to announce an overarching principle that com-
pels reversal," along with his lack of concern that this is so, id. at 1805.
34 Id. at 1763 (opinion of Professor Naomi R. Cahn).
35 See id. at 1785, 1787, 1789 (opinion of Professor Mary I. Coombs). In Fuller's hypothetical,
the jury foreman asked that the question of guilt be determined by the court on the facts as found,
and we are told that "counsel for the defendants" accepted the procedure. Fuller, infra, at 1853
(IYuepenny, CJ.).
36 Contemporary Proceedings, supra note 26, at 1811 (opinion of Professor Laura W. Stein).
Chief Justice Thuepenny's summary of the facts states only that Whetmore (speaking from inside
the cave and just before communications were cut off) "asked if there were among the party [out-
side the cave] a judge or other official" who would tell them whether it "would be advisable" to
cast lots to determine who should be eaten, but "[n]one of those attached to the rescue camp was
willing to assume the role of advisor in this matter." Fuller, infra, at 1852 (Truepenny, CJ.).
Thus, we are not told whether any judges or other authorities on the law were present at all.
 1842 HARVARD LAW REVIEW [VOL. 112:1834

and religious persecution on the planet Newgarth that the entire New-
garthian criminal justice system was suspect "because we on Newgarth
live under circumstances of racial oppression, '37 and Professor Dwight
Greene, viewing the criminal law as a "legal trap[] ... for the less
privileged," decided to affirm the conviction because he knew that the
"affluent, all-Caucasoid, male panel" would not overturn the VaIjean
case, and one could not (under a theory of neutral principles?) find
murder justifiable in order -to survive when theft was punished under
38
similar circumstances.
This is not the place to analyze each of these approaches in detail.
Suffice it to say that while I think there is much to be learned from the
neo-realists, the feminists, and the critical race theorists, I do not count
myself among any of these schools, and I am troubled by each of their
conclusions in the context of Fuller's case. Some have simply refused
to accept the case as stated and have used the opportunity to make up
a story of their own and then act on the basis of that story. (Fuller
might well respond, as I often do in class, that "It's my hypothetical.')
Others seem to me to have copped out - Tatting-like - by imagining
that more facts might help or by insisting on a trial by jury of the ul-
timate issues. 39 In sum, the opinions rendered in the 1993 Symposium
may represent much more of a relatively brief moment in legal history,
and much less of a timeless consideration of a fundamental dilemma
than Fuller's original. In any event, Fuller's work emerges, in my
view, neither bloodied nor bowed.
B
We come then to the present symposium. This time, the editors
have sought to obtain a broader range of views. Their success in this
effort is indicated by the closely divided vote. As I count, the vote to
affirm the conviction is 3-3,40 with one of the three who voted for af-
firmance voting at the same time to invalidate the mandatory death
penalty and to remand for further hearing on the issue of the appro-
priate sentence. Since, unlike Justice Tatting, I have not been assigned
a judicial role, I could not break the tie if I wanted to - and I don't.
Thus the defendants will have to serve time, but they may not have to

37 Contemporary Proceedings,supra note 26, at 1766 (opinion of Professor John 0. Calmore).

38 Id. at 1790-91 (opinion of Professor Dwight L. Greene).
39 Fuller does not tell us whether Newgarth has a constitutional guarantee of jury trial like
ours or whether it has a constitution at all. Indeed, I am sure he wanted his readers to think
about these issues free from whatever constraints a constitution might impose. Moreover, Profes-
sor Coombs seemed determined to jam a jury trial down the defendants' throats whether they
wanted one or not - an idea squarely at odds with our own constitutional precedent, see Patton
v. United States, 281 U.S. 276 (1930).
40 My count assumes that Professor Butler would reverse the conviction in its entirety. If he
would reverse only on the issue of the appropriate sentence, then the vote to affirm would be 4-2.
 1999] THE SPELUNCEAN EXPLORERS 1843

face the tribulations of death row, and worse. (I assume that there is
no higher tribunal to which a further appeal would lie.)
A look at the six opinions reveals some surprises and many in-
sights. But once again, I find myself concluding that the foundation
for all that has followed was laid by Fuller in his thirty pages, and that
while much of the subsequent filigree is entrancing, and sometimes
brilliant, both the groundwork and the structure above it can be found
in Fuller's pages.
To begin with the justices who voted to affirm, Alex Kozinski (a
federal court of appeals judge in real life) takes the "textualist" route
blazed by Chief Justice Truepenny and Justice Keen, and also em-
braces the institutional view espoused by Chief Justice Truepenny in
his reference to the possibility of relief in the "political arena."4 1 Add-
ing to Fuller's arguments, Kozinski points out that we cannot be sure
that the defendants took the wisest course - perhaps they should have
waited for one of their number to expire before diving into their ques-
tionable repast - and that judges should not engage in lawmaking by
disregarding the plain language of a statute. For example, he asks,
should the courts permit an indictment and conviction for killing a dog
("canicide') on the theory 42that the drafters of the statute have left a
gap that needs to be filled?
This opinion is an eloquent statement of the textualist view, but it
raises some concerns. Should the courts regard themselves only as
messengers when applying the broad language of a statute to a par-
ticular problem as long as the words used are "plain"? Should it mat-
ter that the legislature, in the light of centuries of experience, may have
come to expect the process of interpretation to comprise elements of
both agency (the court as applier of the legislature's mandates) and
partnership (the court as fine tuner of the legislature's general, and
sometimes overly general, proscriptions and commands)? To take the
case at hand, Kozinski manages to sidestep the problem posed for him
by the earlier precedent (in Fuller's hypothetical), recognizing a "com-
mon law" justification of self-defense. He does so by invoking other
statutory provisions, apparently not on the books when Fuller wrote,
that "define justifiable homicide" and then by chiding the defendants
for not invoking these previously unknown statutes, "doubtless be-

41 Infra, at 1878 (Kozinski, J.). Since I first set down my thoughts on the initial drafts of the
contributors to this revisiting of Fuller's case, several of those contributors have supplemented
their opinions with insightful critiques of the approaches of their colleagues. Perhaps the most
complete of these critiques is Kozinski's, whose comments sometimes overlap, sometimes improve
on, and sometimes considerably surpass, my own. But having invested the initial effort in col-
lecting my own thoughts, I am unwilling to forgo the opportunity to voice them now.
42 See id.
 1844 HARVARD LAW REVIEW [Vol. I 12:1834

cause they do not apply. '4 3 And his "canicide" example 44 is especially
ironic in view of the statutory language proscribing the killing of "an-
other." Another what?4 Living thing? Homo sapiens? The question
may not be answerable without an analysis of legislative purpose -
with whatever materials are at hand.
The next vote to affirm, cast by Cass Sunstein, may come as a bit
of a surprise to some, but the opinion is in fact a masterful application
of Sunstein's view, developed elsewhere in his writings, that it is possi-
ble to reach a result on the basis of what he has described as an in-
complete theory - one that reasons by analogy and does not resolve
the most fundamental issues of the nature of law.4 6 While recognizing
the virtues of a plain meaning approach (and indeed placing a good
deal of reliance on that aspect of the case), as well as of a purposive
analysis, Sunstein at the same time points out their weaknesses and
limitations. 4 7 For him, the problem is best approached by a compari-
son of the facts to the prototypical case at which the statute is aimed
(the killing of an innocent for selfish purposes) and to its polar opposite
(a killing to prevent the destruction of life by a wrongdoer). Following
this analogy, Sunstein concludes that this killing should not be held
justifiable, especially because Whetmore made a timely effort to pull
out of the agreement.
This analysis, in my view, is both stunning in its own right and an
illuminating example of Sunstein's broader approach to the resolution
of legal problems. But I can't resist noting that its elements were, at
least to some degree, present in the opinions of Fuller's justices, in-
cluding the critiques of textualism and purposivism, 48 the distinction
of the justifications that had been recognized in the past, 49 and the
relevance of Whetmore's effort to get out of the lottery before the dice
were thrown s o
Robin West casts the third vote for affirmance of the conviction.
After rehearsing (with some new insights) the arguments of Justices

43 Id. at 1879.
44 In this example, Kozinski asks whether it would be appropriate for a court to remedy a
"legislative oversight" or fill in what may or may not have been an inadvertent gap in the statute,
by applying the law to the killing of a dog. Id. at 1878.
45 See Eskridge, supra note 7, at 1798 (opinion of Professor Geoffrey C. Miller) ("There are
many contexts in which 'another' can mean an animal. True, we naturally read the qualification
'human being' after the word 'another,' but that is only because execution for killing an animal
seems excessive.").
46 See, e.g., Cass R. Sunstein, The Supreme Court, 1996 Term - Foreword: Leaving Things
Undecided, i io HARv. L. REv. 4, 20-21 (1996).
47 See infra, at 1884-85 (Sunstein, J.).
48 See Fuller, infra, at 186o-62 (Tatting, J.), 1864-67 (Keen, J.).
49 See Fuller, infra, at 186i (Tatting, J.) (noting the impulsive character of resisting an aggres-
sive threat to one's life).
50 See id. at 1862 (questioning whether it would have mattered if Whetmore had refused from
the beginning to participate in the plan).
 1999] THE SPELUNCEAN EXPLORERS 1845

Tatting and Keen that a statute of this kind has multiple, sometimes
conflicting and sometimes unknowable, purposes, Professor West fo-
cuses on the distinction between the case at bar and the classic justifi-
cation of self-defense. 51 She joins with Sunstein in noting that it is one
thing to resist aggression and quite another deliberately to take an in-
nocent life in order to save the lives of others. (In the course of this
discussion, she analogizes Whetmore's plight to that of a woman who
cannot be required to sacrifice her own life to save that of the fetus
within her.)52 Finally, she concludes that the mandatory death penalty
cannot withstand constitutional assault because it fails to permit con-
sideration of mitigating circumstances. 53 At least when it comes 54
to
punishment, she insists, we need not "bifurcate" justice and mercy.
Once again, the seeds of these powerful arguments were planted by
Fuller in his critique of purposive analysis, in his distinction of the
case of self-defense against an aggressor, and in his suggestion (in Jus-
tice Handy's opinion) that a formalistic separation of institutional roles
- leaving questions of "mercy" to the executive branch - was a dubi-
ous exercise.5 5 Of course, Fuller did not have the benefit (if that's
what it is) of our Supreme Court's later pronouncements on the valid-
ity of the death penalty, 6 or of its decisions dealing with the constitu-
tionality of limitations or prohibitions on abortion. Indeed, it is far
from clear that Newgarth has a constitution that bears any resem-
blance to ours5 7 or that our Supreme Court's highly controversial and
somewhat meandering interpretations of the Constitution on these is-
sues should serve as a model. And in any event, West's use of the
abortion analogy is a puzzling one since it could, in my view, be turned
completely around. Perhaps instead of analogizing Whetmore to the
woman who may not be sacrificed to save the life of the fetus within
her, we might more appropriately draw the analogy between the
mother and the defendants. After all, just as the greater good may
consist in allowing the sentient mother to preserve her health or life by
sacrificing an unborn child, so the greater good may be achieved by

51 See infra, at 1895-97 (West, J.).

52 See id. at 1896-97.
53 See id. at 1899.
54 See id. at i898.
S See Fuller, infra, at 187o-73 (Handy, J.).
56 The history of the Supreme Court's struggle with the constitutional problems presented by
capital punishment is remarkable, with respect to both the changes in the Court's approach over
time and the deep divisions within the Court at any particular time. For several decades, the
Court has grappled with a steady series of cases involving the permissible circumstances in which
capital punishment may be imposed, as well as the considerations that may, may not, or must be
taken into account. See Carol S. Steiker & Jordan M. Steiker, Sober Second Thoughts: Reflections
on 7wo Decades of ConstitutionalRegulation of Capital Punishment, io9 HARv. L. REv. 355
(1995) (describing and critiquing the Supreme Court's treatment of these issues since 1972 and
concluding that "the death penalty is, perversely, both over- and under-regulated").
s7 See supra note 39.
 1846 HARVARD LAW REVIEW [VOL. 112:1834

sacrificing one innocent to preserve the lives of many (at least if fair
procedures are followed).
When we turn to those who would reverse the convictions, Frank
Easterbrook's vote and rationale may come as something of a surprise
to those who associate him with the "textualist" approach. In con-
cluding that this case does not fall within the broad language of the
statute, Easterbrook (a once and continuing academic and a federal
appellate judge in real life) emphasizes such matters as historical con-
text, the common law function of the courts in developing defenses to
criminal charges, and the role of the courts not just as agents but as
partners of the legislature in fitting new statutes into the "normal op-
eration" of the legal system.58 His thoughtful distinction between the
VaIjean case and the case of the starving mountaineer is presented as
part of a "utilitarian" analysis of the justification of necessity.5 9 Fol-
lowing this analysis, he corcludes that acting behind a veil of igno-
rance, five explorers willing to take the risks associated with a danger-
ous expedition would rationally agree in advance to a cannibalistic
arrangement that reduced the risk of death by starvation by eighty
percent. (He analogizes such an agreement to the use of a connecting
rope by mountain climbers.)50
Easterbrook's departure from the textualist orthodoxy in this case
is not that surprising, given the sophistication of his approach to statu-
tory construction and the particular nature of this statute. While
much legislation represents a carefully-wrought compromise between
conflicting forces - a compromise that might be perverted or even
wrecked by a refusal to adhere to the text - this criminal statute is
surely more sensibly viewed as an over-general prohibition enacted by
a legislature that, at least implicitly, contemplated the necessity of ju-
61
dicial fine-tuning.
Nor should Easterbrook's view of the utilitarian nature of the "ne-
cessity" defense, which is, I believe, a major contribution to our
thinking about the problem of the case, come as a surprise to those
familiar with his academic work. Once again, though, the approach
was heralded in Fuller's piece when Justice Foster (in the "natural
law" part of his argument) said:
If it was proper that these ten lives [of members of the rescue party]
should be sacrificed to save the lives of five imprisoned explorers, why

58 Cf. David L. Shapiro, Continuity and Change in Statutory Interpretation, 67 N.Y.U. L.

REV. 921 (1992) (making a similar argument aboutthe role of courts in dealing with statutes).
59 See infra, at 1915 (Easterbrook, J.).
60 See id. at 1915-16.
61 In a previous article, Easterbrook more fully discusses this distinction, emphasizing, inter
alia, the difference between a statute that enacts a code of rules, on the one hand, and a statute
that delegates a kind of common law interpretive function to the courts on the other. See Frank
H. Easterbrook, Statutes' Domains, 50 U. CHI. L. REV. 533 (1983).
 1999] THE SPELUNCEAN EXPLORERS 1847

then are we told it was wrong for these explorers to 62carry out an arrange-
ment which would save four lives at the cost of one?
That Fuller regarded this analysis as most relevant to a "natural
law" thesis, while Easterbrook sees it as an appropriate tool of statu-
tory interpretation, is revealing. West insists that the prohibition of
murder is about "rights," in particular the right of the innocent not to
be assaulted or killed, 63 while Easterbrook views the issue of justifica-
tion in terms of the net cost or benefit to those affected. 64 If Easter-
brook is right, don't we have to worry about how far the many can go
at the expense of the few? And why is it irrelevant that on the "ac-
tual" facts (of the hypothetical), Whetmore tried to pull out before the
drawing - a point not mentioned by Easterbrook? In view of Whet-
more's decision, wouldn't it have been both fairer and at least as
sound from a cost-benefit standpoint to exclude him from the drawing
and from the meal that followed?
Another vote to reverse is cast by Alan Dershowitz, writing under
the pseudonym of Justice De Bunker.65 Professor Dershowitz, embel-
lishing Fuller's hypothetical, posits a religious war in the third millen-
nium that culminated, at.least for the vast majority of survivors, in the
abandonment of both religious precepts and any notions of natural
law.6 6 Having eliminated one horn of Fuller's dilemma, Dershowitz
proceeds - in the first part of his analysis - to decide for the defen-
dants on the basis of his own preference (which he hopes will appeal to
others) for allowing all conduct that is not explicitly prohibited by law.
Since the murder statute, in his view, does not address the situation at
bar, his preference, derived from his libertarian principles, furnishes a
67
basis for his vote to reverse.
While Dershowitz is surely entitled to choose the positivist road, it
is a bit unfair to Fuller's hypothetical to eliminate the clash with natu-
ral law principles by assuming that society rejected the concept of
natural law a thousand years earlier. And as to allowing whatever is
not prohibited, it is hard to quarrel with that view as a general ap-
proach to interpretation - in truth, I find it very attractive - but I'm
not sure that it is helpful in this case. To be sure, there were two
widely noted cases several thousand years earlier (in other jurisdic-
tions), but both resulted in convictions under a general statute like this
one. 68 To the extent those decisions have any relevance, why isn't the
conviction of the defendants in those cases an indication that if the
62 Fuller, infra at 1856-57 (Foster, J.).
63 See infra, at 1893 (West, J.).
64 See infra,at 1914 (Easterbrook, J.).
6S Who turns out to be a "gay woman of color." Infra, at 19oi n.3 (De Bunker, J.).
66 See id. at 1899-i9oo.
67 See id. at 1904-05.
68 See United States v. Holmes, 26 F. Cas. 360, 369 (C.C.E.D. Pa. 1842) (No. 15,383); Regina v.
Dudley & Stephens, 14 Q.B.D. 273, 288 (1884).
 1848 HARVARD LAW REVIEW [VOL. 112:1834

Newgarth legislature was aware of the problem, it was quite content

with the way it had been treated in the past? 69 Indeed, if Dershow-
itz's reading of the statute is correct, can it be taken to prohibit a
murder of a kidnap victim when the ransom is not paid, or for that
matter, any killing under facts not specified in the statute itself?70
Perhaps aware of the difficulties of his "interpretation" of the New-
garth murder statute, Dershowitz goes on in what looks like an alter-
native rationale to make an argument based on "necessity.""' This ar-
gument bears a strong resemblance to Easterbrook's utilitarian
calculus, and as I have tried to suggest and others have forcefully ar-
gued, such an argument has both virtues and shortcomings. The most
important of the shortcomings, in my view, is that it poses agonizing
problems for a system of law that seeks in general to protect the inno-
cent from being sacrificed by others for the greater good. In any
event, I remain unpersuaded by Dershowitz's concluding effort to tie
together the two strands of his argument by noting that "our legisla-
ture has not explicitly spoken to this specific problem [of the nature
72
and scope of a "necessity" defense]."
All of which brings us to Paul Butler's opinion. Already known for
his article in the Yale Law Journal, advocating that black jurors prac-
tice nullification when black defendants are charged with non-violent
crimes,7 3 Professor Butler has decided to do himself at least one better.
Seizing on Justice Foster's use of a "stupid[] housemaid '74 to make a
point about purposive construction, Butler writes an opinion from the
perspective of that housemaid - and writes it in a style that is a curi-

69 Citing what is surely the more famous of these cases - Regina v. Dudley and Stephens, 14
Q.B.D. 273 (1884) - in support of his argument, Dershowitz notes that the Dudley court was di-
vided, that the result was followed by executive clemency, and that in any event, "[t]he vast ma-
jority of comparable cases - both before and after that decision - resulted in acquittal or deci-
sions not to prosecute .... " Infra, at 1904 (De Bunker, J.). The first two of these points strike me
as furnishing little support for Dershowitz's argument. Few controversial decisions are unani-
mous; what is critical is that neither the British nor the Newgarth legislature opted to reject the
result. And as for the subsequent commutation, it resembles what Chief Justice Truepenny urged
in voting to affirm; such extraordinary cases, he contended, are not appropriate for rules promul-
gated by courts without any legislative authorization, but rather, they call for the case-by-case
exercise of executive discretion focused on the particular circumstances. See infra, at 1853-54
(Truepenny, C.J.).
Finally, I am puzzled by the reference to "[t]he vast majority of comparable cases." There is
no citation of supporting authority, and I did not know that the practice of cannibalism in these
circumstances is so common that it is possible to speak of the cases in terms of a vast majority.
(Perhaps my notion of what cases are comparable is a less expansive one.) There may be a large
iceberg under the few appellate cases on the subject, but I am unaware of any empirical studies to
support its existence.
70 Kozinski's examples in support of this point, see infra, at i88o (Kozinski, J.), are a delight.
71 See infra, at 1905-09 (De Bunker, J.).
72 Id. at I9O9.
73 See Paul Butler, Racially Based Jury Nullification: Black Power in the Criminal Justice
System, io5 YALE L.J. 677 (i995).
74 Fuller, infra, at 1859 (Foster, J.).
 1999] THE SPELUNCEAN EXPLORERS 1849

ous mixture of Butler's version of ebonics, four-letter words, thor-

oughly Bluebooked legal citations, and rather elegant phrases like
"Having determined no moral culpability in the defendants' actions,"7 5
and "[I-n the last part of the twentieth century, ... Negroes ... were
difficult and expensive to rehabilitate and it was pleasurable to punish
them. '7 6 The thrusts of his opinion are that no crime deserving of
moral condemnation has been committed, 7" that there is no sense in
killing someone to prove that killing is wrong, 78 and that in any event,
there is no true rule of law because "the Supreme Court of Newgarth
ain't never gone choose law to favor the poor and colored folks ... at
least not to the point that the rich white folks' richness and whiteness
is threatened." 79 And in the peroration, the housemaid is beguiled by
the irony that after "sacrificing the lives of people of color for centu-
ries," Newgarth has come to the point where "white folks sacrifice
white lives for the greater good." 0
Granted (as Fuller recognizes in invoking the image of Jean Val-
jean),8 ' the law may appear even-handed when in fact - as Anatole
France so brilliantly put it82 - it frequently treats the poor more
harshly than the rich, and the poor in this country have often been
people of color, especially blacks. Granted too, Butler's prose has an
attention-grabbing, if disconcerting, shock value.8 3 The question still
remains whether - by operating on the assumption that Newgarth in
the fifth millennium is like twentieth-century society at its worst, and
on the more patronizing assumption that a hypothetical "stupid
housemaid" is black - Butler has treated Fuller's case with respect, or
has simply used it as a platform to sound a brassier version of a note
he has played before. Butler's challenge to the whole concept of "legal
reasoning" echoes the criticism of the legal realists in the early decades
of this century and of their intellectual successors in the critical legal
studies movement of more recent times. Fuller himself, who valued
the rule of law, may have gone overboard in suggesting, through Jus-
tice Handy, that the alternative is to take the popular pulse and act ac-

75 Infra, at 1918 (Stupidest Housemaid, J.).

76 Id. at 1919.
77 See id. at 1918.
See id.
78
Id. at 1920.
79
Id. at 1923.
80
See Fuller, infra, at x862 (Tatting, J.).
81
"[Tjhe majestic equality of the law ...forbids the rich as well as the poor to sleep under
82
bridges, to beg in the streets, and to steal bread." ANATOLE FRANCE, LE Lys ROUGE 111-23
(1894), quoted in THE OXFORD DICTIONARY OF QUOTATIONS 292 (Angela Partington ed., 4 th
ed. 1992).
1 must admit that in spite of myself, I couldn't help smiling at Butler's Handy-like trashing
83
of the opinions of the other new justices, as well as of his own, and at the allusion (advertent, I'm
sure) to Henny Youngman's most famous one-liner.
1850 HARVARD LAW REVIEW [VOL. 112:1834

cordingly 8 4 But I am left wondering whether Butler's critique carries

us beyond these earlier contributions. As Kozinski states forcefully in
his opinion, Butler's approach contains its own puzzling inconsisten-
cies and leaves us in the dark about how a better world might apply
the rule of law in a case like this. The difficulty may well lie in But-
ler's insistence on viewing the explorers' case as a parable about race
and class.
A Jewish colleague of mine - one of the participants in this project
who shall remain nameless - told me years ago that when he was
young, he would come home from a baseball game and announce
proudly to his grandmother that "The Dodgers won!," to which his
grandmother would reply, "So, is it good for the Jews?" Probably not,
but it wasn't bad for them either.

In raising some questions about the new opinions assembled for

this project, I do not mean either to deny the many insights in these
opinions or even remotely to suggest that I could have done better. I
am quite sure that I could not. But - as the reader must be tired of
reading by now - I am convinced that one proposition is established
by the continuing debate: Lon Fuller has posed a problem as chal-
lenging for those who worry about the law and legal institutions as is
the origin and ultimate fate of the universe for astronomers.

84 See Fuller, infra, at 1870 (Handy, J.).

1999] THE SPELUNCEAN EXPLORERS 1913

an open mind, untrammeled by the "natural" and "supernatural"

myths of the past.

EASTERBROOK, J.* "Whoever shall willfully take the life of another

shall be punished by death." N. C. S. A. (N. S.) § 12-A. Defendants
killed and ate Roger Whetmore; they did this willfully (and with pre-
meditation, too). Were the language of the statute the end of matters,
the right judgment would be straightforward, as Justices Keen and
Kozinski conclude. See supra, at 1864 (Keen, J.); supra, at 1876 (Koz-
inski, J.). Then when the hangman had finished implementing the
judgment, he too would be doomed, for the executioner takes life will-
fully; likewise we would condemn to death the police officer who shot
and killed a terrorist just about to hurl a bomb into a crowd. Yet
throughout the history of Newgarth such officers have been treated as
heroes, not as murderers - and not just because the Executive de-
clines to prosecute.
Language takes meaning from its linguistic context, but historical
and governmental contexts also matter. Recall the text: "Whoever
shall willfully take the life of another shall be punished by death."
"[W]illfully take the life of another," not "be convicted of willfully tak-
ing the life of another." Yet the latter reading is one that all would
adopt: in our political system guilt is determined in court, not by the
arresting officer or the mob. The statute is addressed in part to
would-be killers and in part to judges, who in adjudicating a charge
apply the complex rules of evidence that may make it impossible to
prove beyond a reasonable doubt the guilt of someone who actually
committed a murder. No one believes that N. C. S. A. (N. s.) § 12-A
oyrrides the rules of evidence, the elevated burden of persuasion, the
jury, and other elements of the legal system that influence whether a
person who committed a killing will be adjudicated a murderer. Like
other criminal statutes, N. C. S. A. (N. S.) § 12-A calls for decision ac-
cording to the legal system's accepted procedures, evidentiary rules,
burdens of persuasion - and defenses.
For thousands of years, and in many jurisdictions, criminal statutes
have been understood to operate only when the acts were unjustified.
The agent who kills a would-be assassin of the Chief Executive is jus-
tified, though the killing be willful; so too with the person who kills to
save his own life. Only the latter is self-defense; the case of the agent
shows that self-defense is just one member of a larger set of justifica-
tions. All three branches of government historically have been entitled
to assess claims of justification - the legislature by specifying the

* Judge, United States Court of Appeals for the Seventh Circuit; Senior Lecturer, The Law
School, the University of Chicago.
1914 HAR VARD LA W RE VIE W [Vol. 11I2:1834

prohibition and allowing exceptions, the executive by declining to

prosecute (or by pardon after conviction), and the judiciary by devel-
oping defenses. As a result, criminal punishment is meted out only
when all three branches (plus a jury representing private citizens) con-
cur that public force may be used against the individual. The legisla-
ture might curtail the role of the judiciary by enacting a closed list of
defenses to criminal charges, but it has not done so. New statutes fit
into the normal operation of the legal system unless the political
branches provide otherwise. N. C. S. A. (N. S.) § 12-A does not provide
otherwise. Our legislature could write a law as simple as N. C. S. A.
(N. S.) § 12-A precisely because it knew that courts entertain claims of
justification. The process is cooperative: norms of interpretation and
defense, like agreement on grammar and diction, make it easier to leg-
islate at the same time as they promote the statutory aim of saving life.
The terrorist example proves the point.
"Necessity" is the justification offered by our four defendants. Af-
ter the first landslide, all five explorers were in great peril, and the res-
cuers outside the cave confirmed that all were likely to starve by the
time help came. The choice was stark: kill one deliberately to save
four, or allow all five to die. The death of one was a lesser evil than
the death of five, and it was therefore the path that the law of justifi-
cation encouraged. Military commanders throughout time have under-
stood this equation and have sent squads and platoons on missions
from which they were not expected to return, so that a greater number
might be saved.
Like all of the lesser-evil justifications, necessity is openly utilitar-
ian. Self-defense may reflect uncertainty about the ability of the law
to affect conduct by those in imminent fear of death, as Justice Tatting
supposes, see supra, at 1862 (Tatting, J.) - though if this is so one
wonders why the force used. must be the least necessary to defeat the
aggression, a restriction that makes sense only if the object of ag-
gression is capable of rational thought and susceptible to influence of
legal subtleties. But other lines of justification assume that the actor
(our police officer, for example) is calculating and alert. The question
is: what shall the law lead him to include or exclude from the calcula-
tion?
Allowing a defense of necessity creates a risk that people may act
precipitately, before the necessity is genuine. Thus if the law allows a
starving mountaineer to break into a remote cabin as a last resort to
obtain food - if, in other words, necessity is a defense to a charge of
theft - it creates a risk that wanderers will break doors whenever
they become hungry, even though starvation is far in the future. The
parallel risk is that a hungry and poor person surrounded by food may
decide to bypass the market and help himself to sustenance. These
risks are addressed by the rule that the evil must be imminent and the
means, well, necessary; the departure from the legal norm must be (as
1999] THE SPELUNCEAN EXPLORERS 1915

with self-defense) the very least that will avert the evil. United States
v. Bailey, 444 U.S. 394 (i98o), employs this understanding to conclude
that a prisoner under threat of (unlawful) torture by the guards may
defend against a charge of escape by asserting that the escape was
necessary to avert a greater evil, but the prisoner loses that defense if
he does not immediately surrender to a peace officer who will keep
him in safe custody.
Allowing a defense of necessity creates a second hazard: the very
existence of the defense invites extensions by analogy to situations in
which criminal liability should not be defeated. That risk is met by
the rule that all lawful or less hazardous options must first be ex-
hausted. A prisoner must report his fears to the warden before escap-
ing; and if the warden does nothing, the prisoner must escape rather
than harm the guard. United States v. Haynes, 143 F.3 d 1o89 (7th Cir.
1998), which held that a prisoner who poured boiling oil over his tor-
mentor rather than trying to flee could not assert a defense of neces-
sity, illustrates this approach. The difference between the mountaineer
case, in which breaking into a cabin is permitted, and Commonwealth
v. Vajean, which held that a poor person may not steal a loaf of bread
from a grocer, is that the poor person could negotiate with the grocer,
or get a job, or seek public or private charity. A mountaineer who
lacks other options to find food, and cannot negotiate with the cabin's
(missing) owner, may break into the cabin because that is the last re-
source; theft is a lesser evil than death, though not a lesser evil than
working.
Negotiation, actual or potential, offers a good framework with
which to assess defenses based on utility. If a defense actually pro-
motes public welfare, then people who are not yet exposed to the peril
would agree that the defense should be entertained. Suppose the five
speluncean explorers had stopped on the way into the cave to discuss
what to do in the event they became trapped. Doubtless they would
have undertaken to wait as long as possible for rescue; and it does not
stretch the imagination to think that they would have further agreed
that if starvation appeared before rescuers did, they would sacrifice
one of their number to save the rest. Each would prefer a one-fifth
chance of death, if calamity happened, to a certainty of death. Al-
though they might find the prospect so revolting that they would
abandon their journey rather than reach such an agreement, the alttr-
native - entering the cave under a set of rules that required all five to
starve if any did - would be even worse in prospect. We know that
they did enter the cave, and did so under a legal regimen that some
members of this Court believe condemned all to starve; it follows that
they would have preferred an agreement in which each reduced that
risk by eighty percent.
Hypothetical contracts are easy to devise; perhaps this accounts for
endless philosophical debate about how people negotiate behind a veil
1916 HARVARD LAWREVIEW [VOL. I112: 1834

of ignorance. Judges should subject these speculations to a reality

check. What do actual contracts for risk-bearing provide? I refer not
to agreements reached after a disaster (such as the explorers' initial
plan to cast dice on the twenty-third day, a plan that Whetmore later
abjured in favor of waiting some more), but to agreements made be-
fore the fateful venture begins - agreements that encompass all of the
relevant options, including the option of avoiding the risk altogether.
Before going underground, spelunkers, like their above-ground
comrades the rock climbers, agree to rope themselves together when
scaling or descending walls and chimneys. If one loses his grip, the
rope may save a life by stopping the fall - but the rope also creates a
risk, for the falling climber may take the others down with him. By
agreeing to rope up, each member of the group exposes himself to a
chance of death because of someone else's error or misfortune. In ex-
change he receives protection against his own errors or misfortunes.
Each accepts a risk of death to reduce the total risk the team faces,
and thus his portion of the aggregate risk. Each agrees, if only implic-
itly, that if one person's fall threatens to bring all down, the rope may
be cut and the others saved. What happened in the cave after the
landslide was functionally the same: one was sacrificed that the others
could live. That Whetmore turned out to be that one is irrelevant; the
case for criminal culpability would have been equally strong (or weak)
had any of the others been chosen. The explorers' ex ante agreement
did not cover the precise form that the risk would take, or the precise
way in which total loss would be curtailed, but it established the prin-
ciple of mutual protection by individual sacrifice. Securing the reci-
procity of advantage ex ante justifies the fatal outcome ex post for an
individual team member. Society should recognize this agreement, and
the way in which it promotes social welfare, through the vehicle of the
necessity defense. To reject the defense is to reject the agreement it-
self, and to increase future loss.
To accept the necessity defense (that is, the risk-sharing agreement)
in principle is not necessarily to accept that a given death is within its
scope. Rock climbers who cut a dangling comrade's rope prematurely,
without exhausting the options to save all, commit murder. Cicero
opined that if two sailors were cast adrift on a plank adequate to sup-
port only one until rescue came, each could try to be the survivor
without criminal liability. But what if they were mistaken, and the
plank would support two for long enough? What if all five explorers
could have survived until rescue (on day thirty-two), or could have
found another exit by further exploration rather than encamping near
the cave mouth? Ancient mariners consented to the practice of sur-
vival cannibalism in principle, but a broad defense of necessity would
have led them to kill a comrade too quickly. Reports were remarkably
consistent in relating that the youngest or most corpulent survivor
drew the short straw. See A.W. Brian Simpson, Cannibalism and the
 1999] THE SPELUNCEAN EXPLORERS 1917

Common Law 124, 131 (1984). To prevent a lesser-evil defense from

becoming a license to perpetrate evil, the necessity must be powerful
and imminent - again following the self-defense model. But the
prosecutor did not argue that the speluncean explorers should have
looked for another exit from the caverns, and the jury found that a
committee of medical experts had informed the men trapped in the
cave that if they did not eat, then there was "little possibility" of their
survival until day thirty. The danger that a necessity defense would
lead people to magnify (in their own minds) the risk they are facing,
and to overreact, did not come to pass. On the facts the jury found, all
five very likely would have died had they passively awaited rescue.
They acted; four lived. Putting these four survivors to death would be
a gratuitous cruelty and mock Whetmore's sacrifice. The judgment of
conviction must be reversed.

STUPIDEST HOUSEMAID, J.*

No superiorwants a servant who lacks the capacity to read between

the lines. The stupidest housemaid knows that when she is told "to
peel the soup and skim the potatoes" her mistress does not mean what
she says.

Supra, at 1858-59 (Foster, J.)

I. THE TRUTH
"O'yeah, O'yeah, O'yeah." Now comes the "stupidest housemaid"
to clean up the mess the white folks have made. Of course the convic-
tions should be reversed. The stupidest housemaid don't know nothin'
'bout the rule of law. Of all the pretty things she's seen in the Big
House she ain't never run cross that. But she knows what she thinks
is right. That is the basis of her judgment. As it is the basis of all the
other judgments as well. The housemaid the onliest one stupid enough
to admit it. Maybe 'cause she got the least to lose.
They call these things opinions for a reason. In the stupidest
housemaid's opinion, the government should not stand a person on a
platform, tie a rope around his neck, and then kick the platform out
from under him. And invite guests to watch him vomit blood. In the
first place, who but the stupidest housemaid gone be left to scrub the
blood out the city square? She good at cleaning up white folks' ugly
messes, but it hard work and it take a long time.

* Paul Butler, Associate Professor, George Washington University Law School. I am grateful
to Sherrilyn Ifill, Chip Lupu, William Rubenstein, and Jonathan Siegel for their insightful com-
ments.