Bator 1963

Finality in Criminal Law and Federal Habeas Corpus for State Prisoners
Author(s): Paul M. Bator

Source: Harvard Law Review, Vol. 76, No. 3 (Jan., 1963), pp. 441-528
Published by: Harvard Law Review Association
Stable URL: http://www.jstor.org/stable/1338627
Accessed: 12-02-2016 06:37 UTC
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
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].
Harvard Law Review Association is collaborating with JSTOR to digitize, preserve and extend access to Harvard Law Review.
http://www.jstor.org
This content downloaded from 192.236.36.29 on Fri, 12 Feb 2016 06:37:50 UTC
All use subject to JSTOR Terms and Conditions
VOLUME 76 JANUARY 1963 NUMBER 3
HARVARD LAW REVIEWV
FINALITYIN CRIMINALLAW AND FEDERAL

HABEASCORPUS FOR STATE PRISONERS
Paul M. Bator*
In examining the circumstances under which the habeas corpus
jurisdiction of the federal courts should be used to redetermine
the merits of federal questions decided in state criminal proceedings,
Professor Bator argues that institutional considerations support taking
jurisdiction when the state courts fail to provide a satisfactory
process for deciding federal questions, but that this justification is
not clearly present when the challenge is not to the state courts'
decisional processes, but rather to the correctness of their results.
Finding support for this contention in the history of federal habeas
corpus until 1953, the author then analyzes the decision in Brown
v. Allen, which made at least federal constitutional questions re-
determinable on habeas corpus even if fully and fairly litigated in
the state courts, and concludes that that decision has not been, and
probably cannot be, justified in terms of the proper institutional role
of the federal courts.
1 HEproblem of finality in criminal law raises acute tensions

in our society. This should not, of course, occasion surprise.
For the processes of the criminal law are, after all, purposefully
and designedly awful. Through them society purports to bring
citizens to the bar of judgment for condemnation, and those con-
demned become for that reason subject to governmental power
exercised in its acutest forms: loss of property, of liberty, even
life. No wonder that our instinct is that we must be sure before
we proceed to the end, that we will not write an irrevocable finis
on the page until we are somehow truly satisfied that justice has
been done.
But the general tendency to hesitate before pronouncinga final
judgment in a criminal case is, I think, reinforced today by some
*Professor of Law, Harvard Law School. A.B., Princeton, I95I; A.M., Har-
vard, I953, LL.B., I956.
The author wishes to express his thanks to Mr. Charles R. Nesson, of the
Harvard Law School class of I963, for his invaluable assistance in research
bearing on this article.
1 In this context "final judgment" refers to the end of all judicial proceedings
44I
442 HARVARD LAW REVIEW [Vol. 76:44I
rather special currents in our thought. Partly, our century has

peculiarly sensitized us to and made us fearful of abuses of power
exercised through the legal process; we find the claims of liberty,
of our residue of autonomy, particularly sweet in an age of dic-
tators, political prosecutions and concentration camps. More
crucial, even, is our general and deep-seated uneasiness about the
ethical and psychological premises of the criminal process itself.
The notion that a criminal litigation has irrevocably ended may
have been an acceptable one in an age with a robust confidence in
(or, if you prefer, complacency about) the rationality and justice
of the basic process itself. But no such confidenceor complacency
can be said to exist today. For some decades the purposes and
methods of our penal law have been the subject of sustained and
intensive criticism and debate. The general line of the challenge
need not be detailed here; in summary we are told that the
criminal law's notion of just condemnation and punishment is a
cruel hypocrisy visited by a smug society on the psychologically
and economically crippled; that its premise of a morally au-
tonomous will with at least some measure of choice whether to
comply with the values expressed in a penal code is unscientific
and outmoded; that its reliance on punishment as an educational
and deterrent agent is misplaced, particularly in the case of the
very membersof society most likely to engage in criminalconduct;
and that its failure to provide for individualized and humane
rehabilitationof offendersis inhumanand wasteful.2
Although many lawyers - I daresay most lawyers - would
disagree with many of these contentions, it cannot be doubted
that the challenge has caused profound concern. And the response
among many sensitive lawyers, judges and scholars has been, it
seems to me, a peculiar receptivity toward claims of injustice
with respect to the conviction of the accused. Of course the criminal process is
by no means necessarily at an end, even with such a "final" judgment of convic-
tion. Executive clemency always still remains, as well as administrative decisions
with respect to parole, etc., which bear on the severity of punishment.
2 My summary merely paraphrases the admirable survey of the situation made
by Professor Wechsler some ten years ago in The Challenge of a Model Penal Code,
65 HARV. L. REV. I097 (I952), particularly at II02-04. The literature representing
fundamental criticism of the traditional criminal law is, of course, huge; for two
random illustrations of a "popular" nature, one by a distinguished lawyer and
judge, the other by a leading psychiatrist, see Box, STAR WORMWOOD(I959), and
Karl Menninger, Verdict Guilty-Now What?, Harper's, August I959, p. 6o.
For contrary views, see Hart, The Aims of the Criminal Law, 23 LAW & CONTEMP.
PROB. 40I (I958), and MOBERLY, RESPONSIBILITY: THE CONCEPT IN PSYCHOLOGY,
IN THE LAW, AND IN THE CERISTIAN FAITH (I956). See also Allen, Criminal
Justice, Legal Values and the Rehabilitative Ideal, 50 J. CRIM. L., C. & P.S. 226
('959).
i963] FINALITY IN CRIMINAL LAW 443
which arise within the traditional structure of the system itself;

fundamental disagreementand unease about the very bases of the
criminal law has, inevitably, created acute pressure at least to ex-
pand and liberalize those of its processes and doctrines which
serve to make more tentative its judgments or limit its power.
In short, our fear (and, in some, conviction) that the entire ap-
paratus of the criminal process may itself be fundamentally un-
just makes us peculiarly unwilling to accept the notion that
the end has finally come in a particular case; the impulse is to
make doubly, triply, even ultimately sure that the particular
judgment is just, that the facts as found are "true" and the law
applied "correct."
It is thus not surprising to find that turbulence surrounds the
doctrines of the criminal law which determine when, if ever, a
judgment of conviction assumes finality. One of the areas of
acutest controversy, namely, the proper reach of the federal
habeas corpus jurisdiction for state prisoners, is the subject of this
essay. The problems created by this jurisdiction are peculiarly
difficult because underlying dilemmas with respect to finality in
criminal cases are here compoundedby the complex demands of a
federal system with its two sets of courts applying law derived
from two sovereignties. Imbedded in this often murky and tech-
nical field of law are fundamental problems about justice: what
processes and institutions in a federal system can best assure that
the exercise of the powers invoked by a judgment of conviction
will be based on premises acceptable politically and morally?
When has justice been done?
The ultimate issue I propose to treat is this: under what cir-
cumstances should a federal district court on habeas corpus have
the power to redetermine the merits of federal questions decided
by the state courts in the course of state criminalcases?
Let us briefly review the structure within which this problem
arises. A defendant is tried in a state court for an offense against
state law. A variety of federal questions may arise in the course
of the litigation; thus the accused may claim that the federal con-
stitution precludes the state from introducing a confession which,
he alleges, was forcibly extracted from him. It is, of course, the
duty of the state court conscientiously to decide this federal ques-
tion in accordance with the governing federal law, which the
command of the supremacy clause 3 makes applicable to the case.
Let us assume that the state court does so: its decision will turn on
3U.S. CONST.art. VI.
444 HARVARDLAW REVIEW [Vol. 76:44I
findings of fact as to what phenomenaexisted or occurredbearing

on the making of the confession, and on the application of federal
legal standards to these facts. Alleged error in the disposition of
the federal question may be and is, I further assume, properly
raised for appellate review in the state system. On affirmance,
the federal issue is subject to direct review by the United States
Supreme Court, usually on certiorari, sometimes on appeal. In
case of affirmance by that Court or a denial of the writ, the
judgment, in the normal operations of the legal system, becomes
final and binding. Should then a federal district court, in the face
of such a judgment, have jurisdiction in a collateral habeas corpus
proceeding to redetermine the federal question which the state
court has already decided and which the Supreme Court has had
an opportunity to review? The law today seems squarely to
answer "yes"; the reigning principle is the striking one that a
state prisoner may seek and automatically obtain federal district
court collateral review of the merits of all federal (at least con-
stitutional) questions, no matter how fully and fairly these have
been litigated in the state-court system.4
One word of reservation: I assume, in this article, that the
defendant in the state case has not forfeited his right to litigate
his federal claim by procedural default; in other words, I do not
propose to deal with the vexing question whether a state prisoner
who fails to raise his federal contentions in accordancewith state
procedural law loses his right to raise them on federal habeas
corpus.5
I. SOME GENERAL CONSIDERATIONS
A. Finality,in CriminalLitigation
The federal writ of habeas corpus has its roots in the common
law.6 Its function, in the great phrase, is to test "the legality of the
'Brown v. Allen, 344 U.S. 443 (i953), made the principle explicit. In part II
of this article I survey the question whether the pre-i953 law supports the decision
in Brown.
5 This question has recently been the subject of debate among Professors Henry
M. Hart and Curtis R. Reitz and Mr. Justice William J. Brennan. See Hart, Fore-
word: The Time Chart of the Justices, The Supreme Court, i958 Term, 73 HARV. L.
REV. 84 (I959) [hereinafter cited as Hart, Foreword]; Reitz, Federal Habeas Cor-
pus: Impact of an Abortive State Proceeding, 74 HARV.L. REV. I3I5 (I96I) [here-
inafter cited as Reitz, The Abortive State Proceeding]; Brennan, Federal Habeas
Corpus and State Prisoners: An Exercise in Federalism, 7 UTAH L. REV. 423 (ig6i).
Hopefully, some light will be shed on it in a case to be argued at the Supreme Court's
i962 Term, Fay v. United States ex rel. Noia, 3oo F.2d 345 (2d Cir.), cert. granted,
369 U.S. 869 (I962).
I
The Constitution it5elf (art, I, ? 9) guarantees against suspension of the
1963] FINALITY IN CRIMINALLAW 445
detention of one in the custody of another."7 In our constitu-

tional context this refers, of course, to detentions made illegal by
federal law. Power to issue the writ is conferred, and this limita-
tion made explicit, in section 224I of the Judicial Code:
Writs of habeascorpusmay be grantedby the SupremeCourt,
any justicethereof,the districtcourtsand any circuitjudgewithin
theirrespectivejurisdictions....
The writ of habeas corpusshall not extend to a prisonerun-

less
He is in custody in violation of the Constitutionor laws or

treatiesof the UnitedStates . . .8
When is a state prisoner held in "violation" of federal law?
I suppose that the answer that may first suggest itself is that one
is held in violation of federal law whenever the state courts have
erroneously decided a federal question bearing dispositively on
the judgment authorizing the detention. And if this is the case,
the proper reach of the jurisdiction can be tersely summarized:
the writ should test the merits of every dispositive federal ques-
tion in the case.9
I do not claim that such a result is necessarily unsound. The
conclusion that a federal court should, at some point, have the
power to decide the merits of all federal constitutional questions
arising in state criminal proceedings (with a habeas court doing
so if the Supreme Court has failed to review the issue) may be a
sound one, resting on the specific institutional and political prem-
ises of our constitutional federalism. The fourteenth amend-
ment does, after all, direct superveningcommands to the states in
"'privilege of the Writ of Habeas Corpus"; and from the time of the first Judiciary
Act certain federal courts and judges have been authorized under certain circum-
stances to issue the writ. Neither Constitution nor statute, however, defines the
term "habeas corpus," and from the beginning the Supreme 'Court has recognized
that "to ascertain its meaning and the appropriate use of the writ in the federal
courts, recourse must be had to the common law, from which the term was
drawn . ..." McNally v. Hill, 293 U.S. I31, 136 (I934). On the other hand, it
was also early made clear that the jurisdiction of the federal courts to issue the
writ must be conferred by statute and is not part of their inherent power. Ex
parte Bollman, 8 U.S. (4 Cranch) 75 (i8o7).
7 McNally v. Hill, supra note 6, at 136.
8 28 U.S.C. ? 224I (I958).
I There would seem to be no constitutional objection to any such sweeping juris-
diction if Congress should want explicitly to create it; at least Hamilton thought it
very clear that the Congress could provide for appeal from state courts to lower
federal courts with respect to questions of federal law. THE FEDERALIST No. 82,
at 5I5-i6 (Lodge ed. i888) (Hamilton).
the management of their criminal law; the Constitution does not

accept the state process as "complete." The creation of a remedial
frameworkto ensure effective implementationof these commands
is, therefore, one of the important tasks of our system. It is the
purpose of the main body of this essay to analyze habeas corpus
in terms of this task, to weigh the strength of the claim that
federal rights should be tested by federal courts.
In these introductory pages, however, I want to ignore the
special demands for a specifically federal forum to which habeas
corpus may be responsive, and direct my attention more generally
to the problem of finality as it bears on the great task of creating
rational institutional schemes for the administrationof the crimi-
nal law. More particularly, consider some of the general premises
which may underlie the demand for relitigation of constitutional
questions on habeas corpus. The fundamental assumption often
seems to be a generalizedversion of the notion adverted to above:
a prisoner is obviously held in violation of law if the decision to
detain was "wrong." It would follow that a detention may not be
considered lawful unless the proceedings leading to it were, in
some ultimate sense, free of error, unless the facts as found were
"really" true and the law "really" correctly applied. If a tri-
bunal finds that the prisoner was not whipped in order to procure
a confession, but, in fact, he actually was whipped, he is detained
illegally. If the court determines that on the facts as found the
confession was admissible, but the "correct" view of the law is
that such a confession is not admissible, he is detained illegally.
Underlying all the processes is the ultimate reality that he was
wrongly condemned.10And thus to determine the legality of the
detention one must, evidently, determine whether the committing
tribunal fell into error.
What must be noted, however, is that on this underlying pre-
mise the conclusion is inescapable that no detention can ever be
finally determined to be lawful; for if legality turns on "actual"
10 For an illustration which seems to me to reflect this line of assumptions, see
Note, 68 YALE L.J. 98, ioi n.I3 (I958):
The doctrines of collateral estoppel and res judicata subordinate the search for
truth to the policy of ending litigation. . . . The policy against incarcerating
or executing an innocent man, however, should far outweigh the desired
termination of litigation.
And herewith Professor Pollak:
[C] oncepts [of finality], like stare decisis, stem from the principle that "in
most matters it is more important that the applicable rule . . . be settled than
that it be settled right." But where personal liberty is involved, a democratic
society employs a different arithmetic and insists that it is less important to
reach an unshakable decision than to do justice.
Pollak, Proposals To Curtail Federal Habeas Corpus for State Prisoners: Collateral
Attack on the Great Writ, 66 YALE L.J. 5o, 65 (I956).
[963] FINALITY IN CRIMINALLAW 447
freedom from errors of either fact or law, whenever error is al-

leged the court passing on legality will necessarily have to satisfy
itself by determining the merits whether in fact error occurred.
After all, there is no ultimate guarantee that any tribunal arrived
at the correct result; the conclusions of a habeas corpus court, or
of any number of habeas corpus courts, that the facts were X
and that on X facts Y law applies are not infallible; if the
existence vel non of mistake determines the lawfulness of the
judgment, there can be no escape from a literally endless relitiga-
tion of the merits because the possibility of mistake always exists.
In fact, doesn't the dilemma go deeper? As Professor Jaffe
has taught us, if the lawfulness of the exercise of the power to
detain turns on whether the facts which validate its exercise
"actually" happened in some ultimate sense, power can never be
exercised lawfully at all, because we can never absolutely re-
create past phenomena and thus can never have final certainty
as to their existence:
A court cannot any more than any other human agency break
down the barrierbetweenappearanceand reality. In short, the
court can be wrong."L
Precisely the same point can be made about rulings of law. As-
suming that there "exists," in an ultimate sense, a "correct"
decision of a question of law, we can never be assured that any
particular tribunal has in the past made it: we can always con-
tinue to ask whether the right rule was applied, whether a new
rule should not have been fashioned.
Surely, then, it is naive and confusing to think of detention as
lawful only if the previous tribunal's proceedings were "correct"
in this ultimate sense. If any detention whatever is to be vali-
dated the concept of "lawfulness" must be defined in terms more
complicated than "actual" freedom from error; or, if you will,
the concept of "freedom from error" must eventually include a
notion that some complex of institutional processes is empowered
definitively to establish whether or not there was error, even
though in the very nature of things no such processes can give
us ultimate assurances:
[L]aw is not a simple concept . . . consisting as it does of rules
distributing authority to make decisions as well as rules that govern
the decisions to be made. There is a sense, therefore, in which a
prisoner is legally detained if he is held pursuant to the judgment
11 Jaffe, Judicial Review: Constitutional and Jurisdictional Fact, 7Q HARV. L.
REV. 953, 966 (1957).
or decision of a competent tribunal or authority, even though the

decision to detain rested on an erroras to law or fact.
That there must be some room for limiting conceptions of this
kind seems clear enough: the writ [of habeas corpus] cannot be
made the instrument for re-determining the merits of all cases in
the legal system that have ended in detention.12
Our analysis of the purposes of the habeas corpus jurisdiction

must, thus, come to terms with the possibility of error inherent in
any process. The task of assuring legality is to define and create
a set of arrangements and procedures which provide a reasoned
and acceptable probability that justice will be done, that the
facts found will be "true" and the law applied "correct." To use,
from another context, Professor Jaffe's illuminating formulation,
the question then is not whether the fact exists in any absolute
sense but whether the evidence is adequate to justify the exercise
of power: ultimately, whether the evidence is a sufficient moral
predicate in the sense that society will accept it as sufficient for the
exercise of the power in question.'3
Now of course none of this tells us that it is not a valid function

of a collateral jurisdiction to redetermine the merits of questions
of law or fact previously decided in a litigation. The verity that
the lawfulness of the exercise of power must eventually turn on
institutional arrangementswhich provide, through their findings
and judgments, an assurance of justice deemed acceptable by
society, does not define what these arrangementsshould be. Just
because a court of appeals cannot assure us that ultimate justice
has been done does not mean that trial court determinations
should not be reviewed. Perhaps we should say that state-court
decisions of federal questions fairly litigated in the state system
12 HART & WECHSLER,THE FEDERALCOURTS AND THE FEDERALSYSTEM I238-39
(I953). And see H.L.A. HART, THE CONCEPTOF LAW I39 (I96I): "It is impossible
to provide by rule for the correction of the breach of every rule."
I should at this point make it explicit that my argument in no sense assumes or
requires any so-called relativization of "truth." Our fallibility in perceiving
phenomena or making judgments of value does not logically mean that the
phenomena had no existence or that values may not be "absolute"; the possi-
bility of error does not belie - indeed it assumes - the existence of truth. Simi-
larly, the fact that society allocates competences to make final decisions in no
sense deprives us, from the "outside" as it were, of the power to make judgments
about the correctness of any decision. As H.L.A. Hart has told us, the umpire is
not necessarily correct just because there is no recourse from his ruling. Id. at
I38-39. The general point is an elementary one in the literature of philosophy;
it has recently been made by Popper in his essay, On the Sources of Knowledge
and of Ignorance, Encounter, Sept. i962, p. 42.
13Jaffe, Judicial Review: Questionof Law, 69 HARV. L. REv. 239, 244 (955).
i9631 FINALITY IN CRIMINALLAW 449
and subject to Supreme Court review should finally establish the

lawfulness of a detention; but there is no a priori reason why we
should not decide that the most acceptable arrangement for the
decision of such questions is that all such state-court determina-
tions should be reviewed by a federal district court on collateral
attack. What is important, however, is that the choice between
the two arrangements must be made on the basis of functional,
institutional and political considerations; and that neither ar-
rangement can be validated by the assertion that it is logically
necessary if "truth" is to be established. And does this not hint
tentatively at a further insight: that an inquiry whether an exer-
cise of power such as detention is "lawful" could meaningfully
address itself, at least initially, not so much to the substantive
question whether truth prevailed but to the institutional or func-
tional one, whether the complex of arrangements and processes
which previously determined the facts and applied the law vali-
dating detention was adequate to the task at hand?
These generalities will not, I know, give us the answer to
difficult issues of policy. But confusion about them has led to
overly easy assumptions and conclusions. This is particularly so
in talk about "constitutional rights." Prisoners have a "right,"
we are told, not to have a coerced confession introduced against
them. From this it is easy to slip into the assumption that the
right has a kind of ultimate reality or existence apart from the
institutional processes which we create to determine whether the
right has been violated in a particular case. And the next step is
to see a particular institutional arrangementas logically necessary
and validated, because it has in a particular case found that the
right has been violated. Leyra v. Denno 14 will serve as a classi-
cal illustration. The question was the admissibility of several con-
fessions made by a defendant in a New York murder trial. After
conviction, one confession was ruled inadmissible by the New
York Court of Appeals,'5 and Leyra was retried. The remaining
confessions were found voluntary and admissible, seriatim, by the
state trial court, the jury, the New York Court of Appeals, and
(after denial of certiorari) on habeas by the federal district court
and the Court of Appeals for the Second Circuit.'6 At this point
14 347 U.S. 556 (I:94).
15People v. Leyra, 302 N.Y. 353, 98 N.E.2d 553 (I95I).
16Leyra's first confession had been elicited through blatant trickery and
psychological pressure by a police psychiatrist. The question on retrial was
whether these practices infected subsequent confessions made to a police captain
and to Leyra's own business partner. The issue was, in accordance with standard
New York practice (see Stein v. New York, 346 U.S. I56 (I953)), first canvassed
the Supreme Court, 5 to 3, held the confessions inadmissible;

and, after retrial, it was held that the evidence apart from the
confessions was inadequate to support the conviction.'7
Now Professor Reitz sees the Leyra case as conclusive justi-
fication for the habeas jurisdiction; after all, Leyra's rights had
"in fact" been violated and if institutional arrangementshad per-
mitted the original judgment to stand we would have had an ex-
ample of a man convicted in violation of his constitutionalrights."8
"The purpose of the guarantees of due process and equal protec-
tion in the ultimate is to prevent conviction of the innocent," he
says. "In [Leyra and one other case] . . . that fundamental
truth is brought home in the most dramatic fashion. . . . But for
federal habeas corpus, these two men would have gone to their
deaths for crimes of which they were found not guilty." Thus, he
says, the case "demonstrate[s] the urgent necessity for the
present federal habeas corpus jurisdiction."19 Quite apart from
the confusion created by the implication that Leyra was "inno-
cent," 20 the reasoning seems to me to illustrate the fundamental
epistemologicalerror which I have been discussing. The existence
of institutions cannot be validated on the ground that in a par-
ticular case they arrived at a result correct in an absolute sense.
All kinds of considerations- in particular, those derived from
federalism- may have made it sound to relitigate Leyra's federal
claim on habeas corpus, but the relitigation cannot be supported
on the ground that habeas was "logically" necessary in order to
vindicate rights which exist as an ultimate reality.
We see, then, that if a criminal judgment is ever to be final, the
by the trial judge, who, finding the subsequent confessions not coerced, then
submitted them to the jury conditionally, instructing it to disregard the confes-
sions if it found coercion. The state affirmance of the conviction so obtained is
found in 304 N.Y. 468, io8 N.E.2d 673 (1952), cert. denied, 345 U.S. 9I8 (i953);
the denial of habeas in II3 F. Supp. 556 (S.D.N.Y.), aff'd, 208 F.2d 605 (2d Cir.
I953).
" People v. Leyra, i N.Y.2d I99, I34 N.E.2d 475 (1956).
'8See Reitz, Federal Habeas Corpus: Postconviction Remedy for State
Prisoners, xo8 U. PA. L. REV. 46I, 496-97 (I960) [hereinafter cited as Reitz,
Federal Habeas Corpus]. See also Pollak, supra note io, especially at 6o-6i.
Professor Pollak points out that but for the existence of habeas in its present
reach, Leyra "would have been dead long since." The statement is, no doubt,
accurate, but I fail to see how this justifies the jurisdiction. As Mr. Justice Jackson
once suggested, if there were a Super Supreme Court, the convictions of many
men "dead long since" which were affirmed by the Supreme Court would have been
in turn reversed; does this mean we ought to create such an institution?
19 Reitz, Federal Habeas Corpus, io8 U. PA. L. REV. 46I, 496-97 (I960).
20 The mere fact that apart from the confessions there was insufficient evidence
to convict does not, in the absence of reason to doubt the trustworthiness of the
confessions, prove Leyra's innocence of the underlying crime.
i963] FINALITY IN CRIMINALLAW 45I
notion of legality must at some point include the assignment of

final competences to determine legality. But, it may be asked,
why should we seek a point at which such a judgment becomes
final? Conceding that no process can assure ultimate truth, will
not repetition of inquiry stand a better chance of approximating
it? In view of the awesomeness of the consequences of convic-
tion, shouldn't we allow redetermination of the merits in an
attempt to make sure that no error has occurred?
Surely the answer runs, in the first place, in terms of conserva-
tion of resources -and I mean not only simple economic re-
sources, but all of the intellectual, moral, and political resources
involved in the legal system. The presumption must be, it seems
to me, that if a job can be well done once, it should not be done
twice. If one set of institutions is as capable of performing the
task at hand as another, we should not ask both to do it. The
challenge really runs the other way: if a proceeding is held to
determine the facts and law in a case, and the processes used in
that proceeding are fitted to the task in a manner not inferior to
those which would be used in a second proceeding, so that one
cannot demonstrate that relitigation would not merely consist of
repetition and second-guessing, why should not the first proceed-
ing "count"? Why should we duplicate effort? After all, it is
the very purpose of the first go-around to decide the case.
Neither it nor any subsequent go-around can assure ultimate
truth. If, then, the previous determination is to be ignored, we
must have some reasoned institutional justification why this
should be so.
Mere iteration of process can do other kinds of damage. I
could imagine nothing more subversive of a judge's sense of
responsibility, of the inner subjective conscientiousnesswhich is so
essential a part of the difficultand subtle art of judging well, than
an indiscriminate acceptance of the notion that all the shots will
always be called by someone else. Of course this does not mean
that we should not have appeals. As we shall see, important
functional and ethical purposes are served by allowing recourse
to an appellate court in a unitary system, and to a federal supreme
court in a federal system. The acute question is the effect it will
have on a trial judge if we then allow still further recourse where
these purposes may no longer be relevant. What seems so objec-
tionable is second-guessingmerely for the sake of second-guessing,
in the service of the illusory notion that if we only try hard
enough we will find the "truth."
Another point, too, should be remembered. The procedural
452 HARVARDLAW REVIEW [Vol. 76:441
arrangements we create for the adjudication of criminal guilt

have an important bearing on the effectiveness of the substantive
commands of the criminal law. I suggest that finality may be a
crucial element of this effectiveness. Surely it is essential to the
educational and deterrent functions of the criminallaw that we be
able to say that one violating that law will swiftly and certainly
become subject to punishment, just punishment. Yet this threat
may be underminedif at the same time we so define the processes
leading to just punishment that it can really never be finally im-
posed at all. A procedural system which permits an endless
repetition of inquiry into facts and law in a vain search for
ultimate certitude implies a lack of confidenceabout the possibili-
ties of justice that cannot but war with the effectiveness of the
underlying substantive commands.2' Furthermore,we should at
least tentatively inquire whether an endless reopening of convic-
tions, with its continuing underlying implication that perhaps the
defendant can escape from corrective sanctions after all, can be
consistent with the aim of rehabilitatingoffenders.22The first step
in achieving that aim may be a realization by the convict that he
is justly subject to sanction, that he stands in need of rehabilita-
tion; and a process of reeducation cannot, perhaps, even begin
if we make sure that the cardinal moral predicate is missing, if
society itself continuously tells the convict that he may not be
justly subject to reeducationand treatment in the first place. The
idea of just condemnation lies at the heart of the criminal law,
and we should not lightly create processes which implicitly belie
its possibility.
One further point should be made in this canvass of the gen-
eral policies which support doctrines of finality in the criminal
law. It is a point difficult to formulate because so easily twisted
into an expression of mere complacency. Repose is a psycho-
logical necessity in a secure and active society, and it should be
one of the aims - though, let me make explicit, not the sole aim
- of a procedural system to devise doctrines which, in the end,
do give us repose, do embody the judgment that we have tried
hard enough and thus may take it that justice has been done.
There comes a point where a procedural system which leaves
matters perpetually open no longer reflects humane concern but
21 It is of course a commonplace of classical criminal-law theory that certainty
and immediacy of punishment are more crucial elements of effective deterrence than
its severity.
22 I do not mean to imply that in fact our contemporary systems of penal
punishment do serve effectively to rehabilitate offenders. But insofar as we purport
to undertake a task of rehabilitation, the question is a relevant one.
i963] FINALITY IN CRIMINALLAW 453
merely anxiety and a desire for immobility.23 Somehow, some-

where, we must accept the fact that human institutions are short
of infallible; there is reason for a policy which leaves well enough
alone and which channels our limited resources of concern toward
more productive ends. I want to be careful to stress that I do not
counsel a smug acceptance of injustice merely because it is dis-
turbing to worry whether injustice has been done. What I do
seek is a general procedural system which does not cater to a
perpetual and unreasoned anxiety that there is a possibility that
error has been made in every criminal case in the legal system.
B. Limitations on the Policies of Finality

I am aware that my argument may seem to go too far. Nobody
contends, and I would not do so either, that the legal system
would be better served if all courts of original jurisdiction had
final competence to decide all cases. The possibilities of error,
oversight, arbitrariness and even venality in any human institu-
tion are such that subjecting decisions to review of some kind
answers a felt need: it would simply go against the grain, today,
to make a matter as sensitive as a criminal conviction subject to
unchecked determinationby a single institution.24
But it is also important to note that a system of appellate re-
view serves a fundamental institutional purpose quite apart from
the correction of "error"in findings of fact or law in a particular
case. In a unitary jurisdiction appellate review provides authori-
tative and uniform pronouncements on the law of that jurisdic-
tion; similarly, recourse in our federal system to the Federal
Supreme Court provides the state courts with authoritative and
uniform pronouncements of federal law.25 (It is for this reason
23 The problem is most acute and most obvious in cases where the death
penalty has been imposed; I would be surprised if receptivity toward endless re-
opening and delay of death cases were not sometimes masking underlying hostility
to the use of the penalty in any case.
24 We should note, however, that our sensitivity in this regard is of relatively
recent origin. Until about the turn of the century, for instance, federal criminal
cases were not generally appealable. See note 75 infra.
25 [T]he national and State systems are to be regarded as ONE WHOLE.
The courts of the latter will of course be natural auxiliaries to the execution
of the laws of the Union, and an appeal from them will as naturally lie to
that tribunal which is destined to unite and assimilate the principles of
national justice and the rules of national decisions.
THE FEDERALIST No. 82, at 5I5 (Lodge ed. i888) (Hamilton). The point was, of
course, explicitly made in Martin v. Hunter's Lessee, I4 U.S. (i Wheat.) 304
(i8i6), which first upheld the power of the Supreme Court to review decisions of
state courts respecting federal rights. See also Jackson, J., concurring in Brown v.
Allen, 344 U.S. 443, 54I (I953).
that I do not view direct Supreme Court review of state cases

as in any sense "collateral.") Appeal, then, serves that aspect of
justice which tells us that the same law ought to apply to different
people similarly circumstanced,and, curiously, in this sense does
not subvert repose but creates it. True, in answering that need,
appellate courts, state and federal, do redetermine the merits of
legal issues and thus help satisfy the search for assurance of cor-
rectness which no judge or court acting alone, unchecked, can do.
Yet even here notions of institutional competence and conserva-
tion of resources play an important role. Relitigation of facts
is not usually permitted on appellate review; the explanation is
surely, at least in part, that the institutional purposes served by
such review do not call for the redeterminationof facts, so that
the mere possibility that error of fact has occurred is thought
insufficient in itself to justify repetition of inquiry. The point
illustrates that we must be careful not to use the existence of and
felt need for appellate review (including Supreme Court review
of federal questions in state cases) to validate collateral jurisdic-
tions whose existence does not serve needs served by appellate
review. Collateral attack may be needed, but the need for appeal
does not prove it; federal habeas corpus remains to be justified
even if we concede the necessity for direct Supreme Court review
of federal questions in state cases.
Let me repeat, then, that the point is not that no court should
review any other court. Nor do I suggest at this stage that a
federal habeas corpus court should not review the merits of all
federal questions litigated in the state court, that valid justifica-
tions, rooted in our federalism, of such a jurisdiction may not be
forthcoming. All I say is that it is not enough to validate the
jurisdiction to assert that some error of fact or law as to the
federal question may in fact occur in a state litigation. The point
can be put in terms of a limiting (rather than defining) principle:
if one set of institutions has been granted the task of finding the
facts and applying the law and does so in a manner rationally
adapted to the task, in the absence of institutional or functional
reasons to the contrary we should accept a presumption against
mere repetition of the process on the alleged ground that, after
all, error could have occurred.
I should like now to set forth certain limiting principles at the
other extreme: general categories where it seems to me plain that
the first go-around, whether in a unitary or a federal system,
should not count, and where relitigation serves obvious and ap-
propriateends.
1963] FINALITY IN CRIMINAL LAW 455
i. Failure of Process. - The first category has been implicit

throughout the discussion. I have said that, presumptively, a
process fairly and rationally adapted to the task of finding the
facts and applying the law should not be repeated. This suggests
that it is always an appropriate inquiry whether previous process
was meaningful process, that is, whether the conditions and tools
of inquiry were such as to assure a reasoned probability that the
facts were correctly found and the law, correctly applied. At
the very least if no opportunity at all was provided to litigate a
question which the applicable law makes relevant to the disposi-
tion of the case, it is just and useful to have a subsequent "super-
visory" jurisdiction, as it were, to furnish such an opportunity.
Similarly, if the conditions under which a question was litigated
were not fairly and rationally adapted for the reaching of a
correct solution of any issue of fact or law, there would seem
to be no reason of principle to immunize the solution reached;
that issue should be redetermined. Suppose that well-supported
allegations are made on collateral attack that the trial judge in a
criminal trial was bribed to convict (and that the prisoner did not
discover this until the time for appeal had expired). The ap-
propriateness of exercising a collateral jurisdiction would seem
beyond question: a "trial" under such circumstances is not a
rational method of inquiry into questions of fact or law, and no
reason exists to respect its conclusions. The same would be true
of a case where a mob is alleged to have dominated the trial court
and jury, and no remedy was provided on appeal to test this
allegation itself. Again, suppose a prisoner alleges that he was,
through torture, forced to plead guilty. Such a finding of "guilt"
manifestly should not "count" as a full and fair litigation which
forecloses further inquiry; and, if no remedy exists by appeal (as
it would not if the coercion was "successful" enough), it is plainly
appropriateto test it on collateral attack.
What binds these cases together is that in all of them the
inquiry is, initially directed, not at the question whether sub-
stantive error of fact or law occurred, but at whether the proc-
esses previously employed for determination of questions of fact
and law were fairly and rationally adapted to that task. Thus
note that, in all of the cases, I assumed that the basic procedural
question raised - that is, whether the judge was in fact bribed, or
the guilty plea in fact extorted by torture -was not itself can-
vassed by a fair litigation not itself subject to the alleged flaw.
If it was, if a full opportunity to test the integrity of the proc-
esses of the committing court was furnished on appeal, collateral
inquiry might again constitute mere repetition. What does seem

clear, however, is that, at the minimum, some kind of supervisory
jurisdiction should exist to test the question whether the processes
furnishedby the previous tribunalwere meaningful and rational.26
Let me now put the point specifically in terms of due process
and the federal habeas corpus jurisdiction. We have seen that it
is, initially, the duty of the state courts to determine federal ques-
tions arising in state criminal cases, and that Supreme Court
review of the merits of such determinations is justified by our
need for one authoritative voice pronouncing uniform federal
law. What further role is left for a federal collateral jurisdiction
such as habeas corpus? When should state determinations, sub-
ject to direct Supreme Court review, not be final? I suggest that
one answer, at least, fits into the very category we have been
discussing: cases where the state has, in effect, failed itself to
provide process. It is, after all, the essence of the responsibility
of the states under the due process clause to furnish a criminal
defendant with a full and fair opportunity to make his defense
and litigate his case: the state must provide a reasoned method of
inquiry into relevant questions of fact and law (including, of
course, all federal issues applicable to the case). If a state, then,
fails in fact to do so, the due process clause itself demands that
its conclusions of fact or law should not be respected: the
prisoner's detention can be seen as unlawful, not because error
was made as to a substantive federal question fairly litigated by
the state tribunals, but because the totality of state procedures
did not furnish the prisoner with a fair chance to litigate his case.
Thus if a state fails to give the defendant any opportunity at all
to test federal defenses relevant to his case, the need for a col-
lateral jurisdiction to afford this opportunity would seem to be
plain, and federal habeas is clearly an appropriate remedy: the
26 I am aware that a doctrine which would allow a subsequent tribunal to test
the integrity of the processes of a previous tribunal could itself, theoretically, lead
to a perpetual reopening of criminal convictions. Thus if the prisoner is convicted
by tribunal A, and then alleges that judge A was bribed, this question would be
determined by tribunal B; but the prisoner would then be free to allege that judge
B was himself bribed (or dominated by a mob), so that his decision should not
count; tribunal C would then have to be furnished to test the integrity of the B
proceedings; and so on. Note, however, that this kind of infinite regress is root
and branch different from the case where tribunals B and C are asked to repeat
the inquiry made by A, that is, where the question before all three is the same one.
In my situation there has to be a new allegation in each case. All I seek, in other
words, is one forum concededly unbiased using procedures concededly rational.
(As a practical matter, it is unlikely in the extreme that a prisoner will be able
to make allegations with respect to the integrity of process more than once.)
state has furnished no process, much less "due" process, for the
vindication of an alleged federal right. Similarly, if the state
furnishes process, but it is claimed to be meaningless process-
if the totality of state proceduresallegedly did not provide rational
conditions for inquiry into federal-law (or, indeed, state-law)
questions, it seems to me clear that the federal habeas jurisdic-
tion may appropriately examine the allegation.27 Thus, to revert
to my previous examples, if it is alleged on habeas corpus that
the trial judge was bribed or that a mob dominated the trial or
that the prisoner was tortured to plead guilty (or, to give another
instance, that there was knowing use of perjured testimony by
the prosecution), and the state provided no fair process, direct or
collateral, for the testing of these allegations themselves, the
habeas court should proceed to inquire into them and, if it finds
the allegations true, set aside the trial court's conclusions.
Notice that I have again assumed that the question at issue-
that is, the meaningfulness of the state's trial process -was not
itself made the subject of a meaningful inquiry by the state courts.
If a state appellate or collateral court determines on the basis of
a fair investigation that the judge was not bribed or the court not
dominated by a mob or the guilty plea not extorted by torture,
then federal collateral investigation of these questions might con-
stitute mere repetition of inquiry, and further institutional justi-
fication for such repetition would be called for.28
27 Compare the holding in Tot v. United States, 3I9 U.S. 463 (I943), that the
due process clause forbids the creation of evidentiary presumptions where there is
no rational connection between the fact proved and the fact presumed therefrom.
28 The point is made a tricky one by the fact that there are several categories
of claims of violations of federal right. Thus there are what might be called
double-level claims and single-level claims. Claims of violations of due process
because the state trial judge was bribed or dominated by a mob are "double-
level" not only in the sense that, if true, they invalidate all of the findings of the
trial court (including findings with respect to other federal constitutional rights,
e.g., admissibility of a confession), but also because the findings of the trial court
itself with respect to these very allegations should not be deemed conclusive: an
allegedly mob-dominated court's finding that it is not mob-dominated should
surely not immunize that question from subsequent inquiry; and, I submit, the due
process clause requires that the question of mob domination should be passed on
by at least one tribunal (state or federal) which is concededly free of that flaw.
On the other hand, the prisoner may claim that a confession offered by the
prosecution was coerced, or the jury discriminatorily selected. These questions
can be seen as "single-level." Nothing in the nature of these issues prevents the
state trial court from constituting an unbiased and rational tribunal with respect
to their decision, so that, in the absence of an allegation of some other procedural
flaw or absence of state remedy which prevented the fair and rational litigation
of these issues, we do not have here a failure of process. If, then, arguendo,
federal collateral inquiry is to be restricted to the category of cases exhibiting a
failure of state process, an allegation on habeas that the state violated the
One more example which seems to fit under this rubric should
be mentioned because of its cardinal importance. This is the case
where a prisoner alleges that the failure of the state to provide
counsel deprived him of a fair chance to make his defense. Many
commentatorsagree that it is the problem of assistance of counsel
which lies at the heart of the great issue of creating fair procedures
in the states' administration of criminal justice.29 Deprivation of
counsel in cases where the demands of fairness embodied in the
due process clause call for representationby counsel is, I submit,
precisely the kind of error which should deprive a state litiga-
tion of sanctity. It casts doubt on the meaningfulness of the
process provided by the state for the resolution of all the issues
in the case: we cannot say that any question in the case, state or
federal, has had a fair and full litigation, for purposes of finality,
if the defendant is found to require the assistance of counsel be-
cause in the circumstancesof the case he was incapable of making
an adequate defense himself.
Of course we must be careful. The question whether the de-
fendant was entitled to counsel (or, just as frequent, whether he
waived the right) is itself a federal question which, arguably, a
state has, for better or worse, decided, and for this determination
itself finality may be claimed. The difficultyis that, as in our other
examples, the very same failure of process which colors the rest
of the proceedings may, and often does, affect the determination
of whether there was failure of process: if it is found that fairness
required representation,a determinationthat it didn't so require
based on a litigation in which the very same flaw existed cannot
be conclusive, any more than a finding by the judge himself that
defendant's federal constitutional rights would always lead initially to this in-
quiry: viewing the state processes in totality, did the state at any time provide
meaningful process for the testing of the question whether there was such a viola-
tion? If the underlying substantive issue was litigated at trial and does not bear
on the integrity of the trial court's decision of the issue itself (e.g., admissibility
of a confession or jury discrimination), habeas would not lie. If the issue was not
fairly litigated at trial, either because the question was simply unavailable at the
time (e.g., a later discovery of prosecution perjury, or a coerced guilty plea) or
because the issue is of a type invalidating the trial court's own decision of it (e.g.,
mob domination, or bribery of the judge), but the state provided a concededly
unflawed tribunal to test it on appeal or collateral attack, habeas again would not
lie. But if the state provides no process at all (as where there is no state post-
conviction remedy to test the question whether there was prosecution perjury), or
provides only meaningless process (as where the allegation of mob domination is
not canvassed by any state tribunal concededly free of such domination), habeas
would be available.
29 See, e.g., Schaefer, Federalism and State Criminal Procedure, 70 HARV. L.
REV. I, 8 (I956).
he wasn't bribed.30In other words, collateral or other supervisory

proceedings are always needed to rule on allegations where the
allegations go to the integrity of the processes of the previous
tribunal which, inter alia, ruled on these allegations themselves.
A further ambiguity must be explored. I have said that a
federal collateral jurisdiction is necessary to give the prisoner an
opportunity to test federal claims for the vindication of which a
state afforded either no forum at all or inadequate process. In
fact, logically, there is no necessity even for this; the need for a
collateral or supervisory jurisdiction to make inquiries not pre-
viously made at all or under meaningful circumstances does not
prove that such a jurisdiction has to be a federal one. The very
failure of a state to provide adequate process could itself be
deemed "error" subject to reversal by the Supreme Court on
direct review. Thus if a state refuses to furnish a defendant with
a meaningful procedure for the litigation of federal questions at
trial, we might conclude that a sound system would void the
conviction on direct appeal, rather than provide a habeas corpus
court to inquire into the merits of such questions. Similarly, if
there are questions which could not fairly be raised at all until
after final judgment (as, for instance, may be the situation in our
case of a coerced guilty plea, or where it is later discovered that
the judge was bribed or the prosecution knowingly used perjured
testimony), there is no logical necessity for a doctrine that these
questions be tested in a federal collateral jurisdiction: again we
could say that the state itself must provide a postconviction
forum for the canvassing of these questions, and that the refusal
30 The question of representation by counsel is, in other words, typical of what
I have referred to as "double-level" issues of due process of law. See note 28

supra. Deprivation of counsel itself may be a violation of federal constitutional
rights, and also may undermine the sanctity of the state's determination with re-
spect to other constitutional rights, e.g., admissibility of a confession. Possible
problems of circularity are, however, particularly acute here. If a state has pro-
vided a fair litigation, with counsel, of the question whether due process required
the appointment of counsel at trial, it would seem clear that state corrective process
has been furnished even if the state's answer was wrong. But must the state always
provide counsel for the determination of whether counsel was needed at trial? Such
a rule would surely be foolish; indeed, it would undermine the very purpose of the
rule that counsel need not always be furnished. If it is perfectly clear that, on the
merits, the Constitution did not require the state to afford counsel at trial, the trial
court's own determination of that issue cannot necessarily be deemed to lack in-
tegrity. As I suggest later, what we have here is a situation where the habeas
court's determination as to whether the state furnished corrective process with re-
spect to the issue of the prisoner's federal right to have counsel will have to turn on
its judgment on the merits, and problems of circularity can be solved only by vary-
ing the scope of collateral review. See p. 494 and note 142 infra.
of a state to do so is simply "error" subject to reversal by the

SupremeCourt.
We must therefore be wary before assuming too easily that the
absence of state process should be cured by using the federal
habeas jurisdiction as a backstop, as it were, rather than exer-
cising a more direct control by an affirmativecommand that the
state may not imprison the defendant without itself providing
such process. What we have is a question, to be answered in light
of the political, institutional and functional considerations, as to
which route we should take.3'
One more caution should be added. Even if we conclude that
we should have a federal collateral jurisdiction to test the
integrity of previous state proceedings, this does not mean that
such a jurisdiction should assume the precise form which has been
associated traditionally with the writ of habeas corpus. Most
significantly, we should be aware that the policies we have been
discussing do not support the notion that federal collateral attack
should be available without limit of time. The fact that a forum
is called for to test an allegation of coercion of a guilty plea or
prosecution perjury does not tell us that a prisoner should be
free to wait to raise these at his pleasure. The policies of finality
and repose must surely play a role even when there has been
failure of process, though the role may be the more limited one of
placing on the prisoner the obligation to make his allegations
within a reasonable time after they have become available to him.
2. Failure of Jurisdiction.- The second category which fur-
nishes appropriatelimitation on notions of finality in the criminal
law is the most traditional one of all, involving the concept, so
vexing and difficult on its edges, yet so useful at the core, of
"jurisdiction." It is black-letter in civil as well as criminal law
that the judgment of a court without at least colorable jurisdic-
31 The problem is discussed pp. 49I-93 infra. It can, of course, arise at several
levels. Suppose that a prisoner alleges that he has recently discovered that in the
proceeding to determine the admissibility of his confession the prosecution know-
ingly used perjured testimony. We could say that a state must itself provide post-
conviction procedures for canvassing the perjury issue, so that a refusal by the
state to afford a hearing on the ground that such procedures do not exist under
state law would be simply subject to reversal on certiorari by the Supreme Court.
On the other hand, we can provide a federal habeas court to test the issue of
knowing use of perjured testimony. Now, however, on the second level, comes this
question: if the habeas court finds that there was perjury, should it then proceed
to determine the subsidiary federal question (as to the admissibility of the con-
fession) itself, or, as to this issue, should the state be forced to grant corrective
process through an unflawed adjudication? See, in this regard, the discussion of
the Rogers case, pp. 514-i6 infra.
i963] FINALITY IN CRIMINAL LAW 46I
tion or competence to deal with the issue it is purportingto decide

enjoys no immunity from collateral attack.32 The reason must be
plain: it is not that the decision may be erroneous, but that
allowing it to "count" would violate the political rules allocating
institutional competences to deal with various matters. It is thus
no surprise that it is one of the historic functions of a court on
habeas corpus to pass on the jurisdiction of the committing
tribunal; 33 and, in our federal-state context, it would be appro-
priate for the federal habeas court to deny conclusive effect to
a state judgment of conviction where the state is made wholly
incompetentby federal law to deal with the case.
Again, however, care is called for. Whether or not a court has
jurisdiction is itself a question to be determined, and one which
can turn on close issues of fact or law. In many cases, where a
strong, or even colorable case can be made for the existence of
jurisdiction, the original tribunal should be deemed competent
to decide that question itself, so that its decision, if based on full
and fair litigation of the question, should at least presumptively be
immune from collateral attack.34 On the other hand no matter
how fully the question has been litigated, if there is a gross
absence of competence no subsequent tribunal needs to accept the
decision: no one would argue that a law professor could set him-
self up to grant divorces or convict of crime, no matter how fairly
the questions at issue- including his own jurisdiction -were
argued before him. The lines obviously cannot be drawn sharply;
decision must turn on the felt appropriateness of allowing the
previous determination by the allegedly incompetent tribunal to
"count."35 And, in fact, this is how courts have tended to draw
32
RESTATEMENT, JUDGMENTS ? 7 (I942).
33See Ex parte Watkins, 28 U.S. (3 Pet.) 193 (i830); Bowen v. Johnston, 3o6
U.S. I19(I939).
34 Suppose, for instance, that a man is tried by Vermont for an offense which,
the defendant claims, occurred just over the state line in New Hampshire. If the
issue of where the act took place is fairly and fully litigated before the Vermont
court, its decision of that question should probably not be subject to collateral
attack.
35 See RESTATEMENT, JUDGMENTS ? IO (I942):
(I) Where a court has jurisdiction over the parties and determines that it
has jurisdiction over the subject matter, the parties cannot collaterally attack
the judgment on the ground that the court did not have jurisdiction over the
subject matter, unless the policy underlying the doctrine of res judicata is out-
weighed by the policy against permitting the court to act beyond its jurisdic-
tion.
(2) Among the factors appropriate to be considered in determinra tbt
collateral attack should be permitted are that
(a) the lack of jurisdiction over the subject matter was clear;
the lines: not every question of competence is automatically open

on collateral attack if it has been litigated in the original forum;
but if the question is sufficientlygrave or goes to the heart of the
political rules allocating competences, collateral attack will be
allowed.36
Let me now summarize my conclusions. It has not been my
purpose to set up hard and fast touchstones for the habeas corpus
jurisdiction. More particularly let me again warn the reader
explicitly that what has gone before does not purport to be an
analysis of the possible roles which such a jurisdiction can play
in a constitutional federalism. It may be entirely justified and
appropriate that the question whether a defendant has been de-
prived of federal constitutional rights should always be decided
by a federal constitutional court. That is the question which I
will treat in detail in the latter parts of this essay. What I have
tried to do so far is simply to canvass some of the underlying
general problems of finality, without reference to the special in-
stitutional and functional claims of federalism, and to establish a
set of limiting notions. In these terms, then, I see the need for a
tentative presumptionthat questions fairly and fully canvassed by
institutions with general competence to deal with such questions
and making use of processes fairly and rationally adapted to the
function of deciding them should not be relitigated unless felt
functional or institutional requirements call for the repetition.
On the other hand, as the proposition implies, subsequent juris-
dictions should be free to ignore decisions previously made by
institutions without colorable competence in the premises; and
should in any event be free to direct inquiry into the question
whether the totality of previous proceedings furnished the de-
fendant with a full and fair opportunityto litigate his case.
With this general introduction I turn now to a study of the
law of federal habeas corpus itself. The next section of this article
describes and analyzes the development of the jurisdiction up to
the decision in Brown v. Allen,37which explicitly enthroned the
principle that all federal constitutional questions decided in state
criminal cases may be redetermined on the merits on federal
habeas corpus. The remainderof the article is then devoted to an
analysis of the doctrine of that case.
(c) the court was one of limited and not of general jurisdiction;
* . . .
(e) the policy against the court's acting beyond its jurisdiction is strong.
36 See p. 469 and note 65 infra.
37344 U.S. 443 (I953).
II. THE DEVELOPMENT OF THE HABEAS CORPUS

JURISDICTION, I789-I952
Perhaps the most striking aspect of the current discussions of

habeas corpus is how easily it is taken for granted that it is the
proper function of the writ to provide a determination of the
merits of all federal constitutional claims arising in state criminal
proceedings, no matter how fully these have been canvassed in
the state system.38 Professor Reitz sees this in fact as a "funda-
mental purpose" of the jurisdiction,39and he characterizesa Sixth
Circuit decision,40holding that the admissibility of a confession
in a state case may not be relitigated on habeas, as "preposterous
on its face," since "to deny that so basic a deprivation of due
process as conviction by the use of a coerced confession can be
made the subject of a collateral attack is in effect to destroy the
federal habeas corpus jurisdiction."41
Yet it is, as far as I know, a unique principle in our law that
final judgments rendered by competent tribunals should be re-
opened on collateral attack.42 And it is at most doubtful whether
any such principle existed before Brown v. Allen established it in
I952 . For what was understood to be the law just before the
38 Debate has been cast in terms of the question whether state procedures are
adequate enough to justify dispensing with the jurisdiction entirely, rather than
whether habeas is justified if state process is concededly adequate. See, e.g., Pollak,
Proposals To Curtail Federal Habeas Corpus for State Prisoners: Collateral Attack
on the Great Writ, 66 YALE L.J. 50 (1956); Reitz, Federal Habeas Corpus, io8
U. PA. L. REV. 46i (I960); Schaefer, Federalism and State Criminal Procedure, 70
HARV.L. REV. i (1956); Hearings on H.R. 5649 Before Subcommittee No. 3 of the
House Committee on the Judiciary, 84th Cong., ist Sess., ser. 6 (I955).
3 Reitz, The Abortive State Proceeding, 74 HARV.L. REV. I3I5, I3I6 (I96I).
40Wooten v. Bomar, 267 F.2d goo (6th Cir.), cert. denied, 36i U.S. 888 (I959).
41 Reitz, Federal Habeas Corpus, io8 U. PA. L. REv. 46I, 462 (I960). The
opinion of the court was, of course, inconsistent with Brown v. Allen. As an
alternative ground it stated, however, that the prisoner had not exhausted his
state remedies.
42 In fact, in a different setting Professor Reitz acknowledges the extraordinary
nature of the situation:

The . . . [habeas corpus] structure of interrelated federal and state adju-
dication is unparalleled in our federal system. . . . [E]lsewhere . . . the state
courts' decisions are accepted as final, subject only to possible review in the
Supreme Court of the United States. The federal habeas corpus proceeding
alone presents to the lower federal courts questions which, in many instances,
will have been considered by a whole hierarchy of state courts.
Reitz, The Abortive State Proceeding, 74 HARV. L. REV. 1315, I330-31 (I96I).
43In I95I, the Court of Appeals of the District of Columbia, in an opinion
which surveyed with painstaking care the functions and reach of the federal habeas
jurisdiction, at least for federal prisoners, found that the courts of appeals "have
ruled uniformly" that the admissibility of confessions "is not a ground for col-
lateral attack," and held that "where the alleged error . . . is in the admission of
evidence subject to correction on appeal, and there is representation by counsel,
Brown decision, we may turn to the authority of the late Judge

Learned Hand. In Sckecktman v. Foster 44 (decided in I949) a
prisoner complained on federal habeas corpus that his conviction
was obtained by New York through testimony known by the
prosecution to be perjured. Now this is a claim of fundamental
procedural error which by its nature is almost always unknown
and cannot be litigated at trial or on appeal, so that, in the absence
of state postconviction corrective process, it should plainly con-
stitute appropriate grounds for federal collateral attack. Here,
however, the allegations had in fact been considered on the merits
in state postconviction coram nobis proceedings (in which the
defendant was assisted by counsel), and the state court had held
them insufficient. The Court of Appeals for the Second Circuit,
by Judge Hand, held that the petitioner was not held in violation
of federal law:
If the judge who denied that petition did in fact consider the
evidence as a whole, and if he decided that it was not, even prima
facie, sufficient to make out a case of deliberate presentation by
the prosecution of perjured testimony, Schechtman was accorded
the full measure of his constitutional rights. It must be remem-
bered that upon habeas corpus a federal court does not in any
sense review the decision in the state courts. Here, for example,
the District Court could not properly have issued the writ, no
matter how erroneous the judge had thought the state judge's
conclusion that the evidence did not make out a prima facie case
of the deliberate use of perjured testimony. The writ was limited
to the assertion of the relator'srights under the Fourteenth Amend-
ment; and due process of law does not mean infallible process of
law. If the state courts have honestly applied the pertinent doc-
trines to the best of their ability, they have accorded to an accused
his constitutional rights.45
habeas corpus is not the appropriate remedy." Smith v. United States, 187 F.2d
192, I97 (D.C. *Cir. 1950), cert. denied, 34I U.S. 927 (I95I). See also Note, 6i
HARV. L. REv. 657, 668 (I948):
[T]he availability of habeas corpus is determined by whether the particular
question of unfairness has previously been litigated in the light of all informa-
tion presently available. . . . Wherever all evidence of alleged unfairness was
fully litigated before a competent tribunal, subsequent collateral attack by
means of habeas corpus is not ordinarily permissible. Such complete litigation
of the alleged unfairness is itself the fairness to which the defendant was entitled.
44 I72 F.2d 339 (2d Cir. I949), cert. denied, 339 U.S. 924 (I950).
45 I72 F.2d at 34I. Judge Hand then proceeded to state as alternative grounds of
decision that the prisoner had not exhausted his state remedies, and that the Supreme
Court had denied certiorari after the state postconviction proceeding. The latter
ground would not be considered proper today in view of the holding of Brown v.
Allen, 344 U.S. 443 (I953), that denial of certiorari by the Supreme Court is not
to be taken as an expression of views on the merits.
Surely this makes clear that it was not in I949 thought to be

the task of the federal court on habeas to test for error the
disposition of all federal constitutional questions made in previous
state adjudications. The understandingseemed to be much nearer
to the guideposts set out above: a prisoner is not held in "viola-
tion" of federal law if a state court of competent jurisdiction has
through fair process - though perhaps erroneously - decided
that question on the merits. Yet a scant three years later the
Supreme Court could take for granted that a federal court's
determination that there was federal error automatically made a
state detention illegal, and we are told that this is an obvious and
necessary cornerstoneof the habeas jurisdiction.
What led up to this situation? What were the lines established
for the jurisdiction in the previous I50 years? What is the com-
mand of the authoritative statutes? To these inquiries we now
turn.
A. The Early "Federal" Cases

We look for guidance first to statutory and case law dealing not
with state but with federal prisoners. For until I867 (and with
exceptions not relevant here 46) there was no federal habeas
jurisdiction to inquire into detentions pursuant to state law.47
Further, even after the act of I867 established such a jurisdic-
tion,48 the Supreme Court could make no pronouncements in
cases of state detention because the Court's appellate jurisdiction
under the act of I867 was removed in i868 and not reestablished
until i885.49 Thus during the first century of the Constitution
the Court had no occasion to deal with the scope of the habeas
jurisdiction for state prisoners.
The original statutory authorization for habeas corpus in the
Judiciary Act of I789 did not define the writ's substantive reach;
46
Statutes passed in I833 and I842 authorized federal judges to grant the writ
to state prisoners in custody on account of acts done pursuant to the authority of
the United :States or of any f oreign power. See HART & WECHSLER, THE FEDERAL
COURTS AND THE FEDERAL SYSTEM I236-37 (1953).
4 See Ex parte Dorr, 44 U.S. (3 How.) I03 (I845).

48Act of Feb. 5, I867, ch. 28, ? I, I4 Stat. 385.
4' The jurisdiction was removed (see Act of March 27, i868, ch. 34, ? 2, I5 Stat.
44) in order to prevent the Supreme Court from passing on the constitutionality of
reconstruction legislation in McCardle's case. See Ex parte McCardle, 74 U.S. (7
Wall.) 506 (I869), upholding the validity of and honoring the ouster of jurisdiction.
The Court continued, however, to consider original habeas petitions from prisoners
under the act of I789. Ex parte Yerger, 75 U.S. (8 Wall.) 85 (I869). The Court's
power to hear appeals in the case of state prisoners was restored by the Act of
March 3, i885, ch. 353, 23 Stat. 437.
it merely stated that the courts of the United States "shall have
power to issue writs of ... habeas corpus . ." 50 It is thus
not surprisingthat we soon find the SupremeCourt accepting the
black-letter principle of the common law that the writ was simply
not available at all to one convicted of crime by a court of com-
petent jurisdiction.5' Ex parte Watkins 52 is the great case. The
Court there refused, on habeas, to reach the merits of the allega-
tion that the prisoner was convicted pursuant to an indictment
which failed to state a crime against the United States:
A judgment,in its nature,concludesthe subject on which it is
rendered,and pronouncesthe law of the case. The judgmentof a
Courtof recordwhosejurisdictionis final, is as conclusiveon all
the world as the judgmentof this Courtwould be. . It puts
an end to inquiry concerningthe fact, by deciding it. . . An im-
prisonmentunder a judgmentcannot be unlawful,unless that
judgmentbe an absolutenullity; and it is not a nullity if the
Courthas generaljurisdictionof the subject,althoughit shouldbe
erroneous.53
The principle is clear: substantive error on the part of a court
of competent jurisdiction does not render a detention "illegal" for
purposes of habeas corpus, because, to use Chief Justice Mar-
shall's striking phrase, "the law trusts that court with the whole
subject." 54 But limitations on this principlebegan later to appear.
50Act of Sept. 24, I789, ch. 20, ? I4, i Stat. 8i. The limitation added was
that the writ "shall in no case extend to prisoners in gaol, unless they are in custody,
under or by colour of the authority of the United States .. ."
51 The principle that a person convicted by a court of general criminal jurisdic-
tion is not entitled to habeas corpus derives from the Habeas Corpus Act of i679,
3I Car. 2, C. 2, which expressly excepted "persons convict or in Execution by legal
process." It has been suggested that this act did not purport to deny habeas to
convicted persons but, rather, left their rights to be worked out through the pre-
existing common law writ, and that the contrary view derives from a curious mis-
reading of the act by early legislatures and judges due to a misplaced parenthe-
sis. See Brief of Professor Paul A. Freund for Respondent, pp. 30-32, United States
v. Hayman, 342 U.S. 205 ( I952 ).
52 28 U.S. (3 Pet.) I93 (I830). Watkinswas preceded by Ex parte Kearney, 20
U.S. (7 Wheat.) 38 (I822), where the petitioner had been judged guilty of contempt
for failure to answer questions on the witness stand. On habeas the Court refused
to reexamine the contention that the refusal was privileged under the fifth amend-
ment, since the writ is not "a proper remedy, where a party was committed for a
contempt by a Court of competent jurisdiction." Id. at 44-45. Oddly enough, the
Court did not refer to the famous Bushell's Case, Vaughan 135, 124 Eng. Rep.
ioo6 (C.P. I670) (which was cited at argument), where the Common Pleas, on
habeas corpus, discharged a juryman who had been imprisoned for contempt at the
Old Bailey on the ground that he returned an improper verdict.
53 28 U.S. (3 Pet.) at 202.
54 Id. at 206.
In Ex parte Lange, decided in I873, the Court first announced the

rule that habeas corpus may be used to reexamine,not substantive
errors going to the conviction, but alleged illegality in the sen-
tence: if the committing court imposed two sentences where only
one is authorized by the statute, and the valid sentence has been
carried out, release will be ordered.55 This rule was held to be
applicable in the case of Ex parte Wilson, where the petitioner had
allegedly been sentenced to hard labor for an "infamous" crime
within the meaning of the fifth amendmentwithout indictment by
a grand jury; the Court on habeas held the sentence illegal after
an inquiry into the merits of whether indictment was required in
the case.56 Similarly,in In re Snow 17 andagain in Nielsen, Petition-
55 85 U.S. (i8 Wall.) I63 (I873). Lange was convicted under a statute authoriz-
ing a sentence of imprisonment or a fine. The judge sentenced him to one year and
a fine of $200. He paid the fine and, having served five days of his prison sentence,
sought habeas. The sentencing judge forthwith vacated the former sentence and re-
sentenced Lange to one year's imprisonment. The Supreme Court held the second
sentence void, stating that
when the prisoner . . . by reason of a valid judgment, had fully suffered one
of the alternative punishments to which alone the law subjected him, the
power of the court to punish further was gone. . . . [A] t the moment the
second sentence was rendered . . . the prisoner had fully performed, completed,
and endured one of the alternative punishments which the law prescribed for
that offence . . . Unless the whole doctrine of our system of jurisprudence
. for the protection of personal rights . . . [is] a nullity, the authority of
the court to punish the prisoner was gone.
Id. at I76. The Court carefully emphasized its special concern with the sentence:
If a justice of the peace, having jurisdiction to fine for a misdemeanor, and
with the party charged properly before him, should render a judgement that
he be hung, it would simply be void. Why void? Because he had no power to
render such a judgment.
Ibid.
It is, of course, clear that the sentencing court's erroneous assumption that it
was entitled to vacate the previous erroneous sentence and resentence the prisoner
did not make its judgment "void" in the sense that it had no competence to deal
with this issue. Lange thus clearly goes beyond the rule laid down by Watkins.
The extension was prefigured by Ex parte Wells, 59 U.S. (i8 How.) 307 (i855),
where the Court had, on habeas, upheld on the merits the legality of a sentence of
life imprisonment, imposed under a presidential commutation of a death sentence,
against the allegation that the Constitution does not authorize the President to
grant conditional pardons.
56 II4 U.S. 417 (i885). The underlying substantive issue, whether the fifth
amendment required indictment in the case, was seen by the Court as bearing on
the legality of the sentence, rather than the judgment of conviction, because in its
view what constitutes an "infamous" crime within the meaning of the amendment
is itself a function of the type of sentence authorized: "[A] crime punishable by
imprisonment for a term of years at hard labor is an infamous crime" (id. at 429),
and indictment is therefore a condition precedent for the valid imposition of such
a sentence. Wilson was followed in Ex parte Bain, I2I U.S. I (I887), where the
Court held that a federal circuit judge has no power to amend an indictment, so
that an imprisonment on the basis of such an amended indictment is void and
the prisoner may be released on habeas corpus.
57 I20 U.S. 274 (I887). Petitioner was convicted (in the territory of Utah)
468 HARVARDLAWREVIEW [VoI. 76:441
er,58the imposition of consecutive sentences where the indictment

was found to charge but one offense was deemed illegal on habeas
and the prisoner was released after serving the legal sentence.
Another line of doctrine emerged from Ex parte Siebold,59
where a federal prisoner alleged on habeas that his conviction was
void because the statute creating the offense for which he was tried
was unconstitutional. Mr. Justice Bradley did not question the
general rule that a conviction by the judgment of a competent
court may not be tested for error on habeas; but he concluded
that, since "an unconstitutional law is void, and is as no law," "a
conviction under it is not merely erroneous, but is illegal and
void." "[I] f the laws are unconstitutional and void, the Circuit
Court acquiredno jurisdictionof the causes." 60
On the other hand, the Court continued to enforce the rule
that habeas will not test the question whether the indictment
states an offense.6' And in Ex parte Bigelow it held that the
of cohabiting with more than one woman on the basis of three indictments
charging the same offense during three separate periods of time. The Court
held that but one continuing offense was charged, and that the fifth amendment
therefore prevented the imposition of more than one sentence. In holding that
habeas was available the Court relied on an English case, Crepps v. Durden, 2
Cowper 640, 98 Eng. Rep. I283 (K.B. I777), holding that a court in a collateral
trespass action could question the validity of multiple fines imposed for one
offense.
58 I3I U.S. I76 (I889). This, too, was a case from Utah, charging the petitioner
with (i) unlawful cohabitation and (ii) adultery, both with the same woman.
Having served the sentence under the former indictment, he was released on
habeas on the ground that only one offense was charged and he could therefore
not be punished under the second indictment. Justice Bradley used expansive
language in describing the reaches of the habeas jurisdiction, see p. 472 infra,
but the holding follows squarely from Snow.
59 I00 U.S. 37I (I879). The extensions of the habeas corpus jurisdiction
announced in Siebold and Lange cannot, of course, be explained by the existence
of the new habeas corpus act of I867, since both were necessarily decided under
the old act of I789. See note 49 supra.
60 I00 U.S. at 376, 377. The Court discussed Bushell's Case, see note 52 supra,
but reserved the question of "how far this case may be regarded as law for the
guidance of this court." On the merits, the Court upheld the constitutionality of
the provisions of the Civil Rights Acts under which the indictments were brought.
61 Ex parte Parks, 93 U.S. I8 (I876). See also Ex parte Yarbrough, II0 U.S.
65I (I884), where the Court, following Watkins and Parks, refused on habeas
to inquire whether the acts alleged in the indictment (conspiracy to prevent
Negroes from voting) constituted an offense under the Civil Rights Act, but did,
in accordance with Siebold, pass on and uphold the constitutionality of the act.
The anomalous result of the rules of Yarbrough and Siebold was, of course, that
on habeas the Court had jurisdiction to pass only on the abstract constitutional
issue in the case; presumably it had to assume that the acts charged in the
indictment were in fact forbidden by the statute, even if it thought that they
plainly were not so forbidden. One may question the wisdom of so shaping a juris-
diction that the Court is prevented from passing on possibly dispositive nonconstitu-
merits of a claim of double jeopardy will not be reached where

the alleged error did not result in cumulative sentences.62 In re
Belt 63 held that the question whether a federal statute permitting
a criminal defendant to waive trial by jury was constitutional
would not be reexamined on habeas corpus; and in Matter of
Moran Mr. Justice Holmes stated that an allegation that the
defendant in a federal criminal case was forced to incriminate
himself in violation of the fifth amendmentcould not be examined
on habeas, because even if there was error, "it did not go to the
jurisdiction of the court." 64 Finally, the Court limited even the
rule that the jurisdiction of the committing court could be
examined on habeas corpus: if there was colorable jurisdiction,
and its existence turned on close questions of fact or law, the
committing court's decision as to its own jurisdiction will be final
except when reversed on appeal.65
tional questions and is compelled to decide as to the constitutionality of statutes
construed, hypothetically, in a manner deemed by the Court to be erroneous.
62 II3 U.S. 328 (I885). Fourteen indictments for embezzlement were returned
against the petitioner. The court ordered them consolidated and tried together;
but after the jury was sworn and the prosecutor made his opening, it rescinded
the order of consolidation, discharged the jury from further consideration of the
fourteen indictments, and ordered the defendant tried before the same jury on
one of the indictments. On habeas the petitioner alleged that the empaneling
of the jury and the statement by the district attorney put him in jeopardy as to
all fourteen indictments so that he could not be tried again for any of these
offenses. The Court on habeas refused to consider the merits of the claim,
holding it an ordinary matter of defense which the trial court had jurisdiction
to determine. It explicitly rejected the contention that the fact that the issue
was one of constitutional law made it automatically a jurisdictional one. And
it distinguished the Lange case on the ground that there "the prisoner, having
been tried, convicted, and sentenced . . . and having performed the sentence
as to the fine, the authority of the Circuit Court over the case was at an end,
and the subsequent proceedings were void," whereas here "no verdict, nor judg-
ment was rendered, no sentence enforced, and it remained with the trial court
to decide whether the acts on which he relied were a defence to any trial at all."
Id. at 33I.
63
I59 U.S. 95 (I895).
64
U.S. 96, I05 (I906).
203
65 The rules laid down by the Court with respect to examination on habeas
of the jurisdiction of the committing court are in accord with the rationale sug-
gested above, see pp. 460-62 supra. In Callan v. Wilson, I27 U.S. 540 (i888), the
Court passed on a jurisdictional issue, holding that the power of the police
court of the District of Columbia, which sits without a jury, "does not extend
to the trial of infamous crimes or offences punishable by imprisonment in the
penitentiary." Id. at 556. In re Coy, I27 U.S. 73I (i888) involved the question
whether a failure of Indiana officials to perform certain duties in connection with
elections constituted federal or merely state offenses, that is, whether the
federal courts had jurisdiction in the premises. The Court on habeas stated
that it need not decide this issue, since it was properly before the trial court.
It recognized that it would go too far to say that "because every court must pass
upon its own jurisdiction, such decision is itself the exercise of a jurisdiction which
It would be useless to try to show that these decisions establish

clear and sensible guidelines for the scope of the habeas corpus
jurisdiction. Once the concept of "jurisdiction"is taken beyond
the question of the court's competence to deal with the class of
offenses charged and the person of the prisoner, it becomes a less
than luminous beacon. How is one to tell which errors cause a
court to lose jurisdiction and which do not, which render a judg-
ment void and which do not? 66 Surely there is no reason of prin-
ciple, if we are to "trust" a competent trial court with "the whole
subject," why it should not be trusted to decide the issue of the
constitutionality of the statute creating the offensetogetherwith all
belongs to it, and cannot, therefore, be questioned in any other court." The cases
do not hold
that because a federal court tries a prisoner for an ordinary common law
offence . . . with no averment or proof of any offence against the United
States, or any connection with a statute of the United States, and punishes
him by imprisonment, he cannot be released by habeas corpus because the
court which tried him had assumed jurisdiction.
In all such cases, when the question of jurisdiction is raised, the point to
be decided is, whether the court has jurisdiction of that class of offences. If
the statute has invested the court which tried the prisoner with jurisdiction to
punish a well defined class of offences, . . . its judgment . . . is not review-
able on a writ of habeas corpus.
Id. at 757-58. In In re Mayfield, I4I U.S. I07 (I89I), the Court on habeas
passed on the jurisdiction of the committing court over an offense allegedly
within the exclusive jurisdiction of an Indian national court, but in Rodman v.
Pothier, 264 U.S. 399 (I924), it refused to deal with a similar issue since
"whether the locus of the alleged crime was within the exclusive jurisdiction
of the United States demands consideration of many facts and seriously contro-
verted questions of law," which "must be determined by the court where the
indictment was found." Id. at 402. The law and the considerations underlying
it are extensively reviewed in Bowen v. Johnston, 306 U.S. I9 (I939), where
the Court said:
The rule requiring resort to appellate procedure when the trial court has
determined its own jurisdiction of an offense is not a rule denying the power
to issue a writ of habeas corpus when it appears that nevertheless the trial
court was without jurisdiction. The rule is not one defining power but one
which relates to the appropriate exercise of power. It has special application
where there are essential questions of fact determinable by the trial court. ...
It is applicable also to the determination in ordinary cases of disputed matters
of law . . . . But it is equally true that the rule is not so inflexible that
it may not yield to exceptional circumstances where the need for the remedy
afforded by the writ of habeas corpus is apparent. Among these exceptional
circumstances are those indicating a conflict between state and federal
authorities on a question of law involving concerns of large importance
affecting their respective jurisdictions.
Id. at 26-27.
66The Court's own attempts to articulate the distinctions are sometimes
worthy of inclusion in Uncommon Law. Thus in Lange, the Court said: "But
it has been said that . . . the judgment under which the prisoner is now held
is erroneous, but not void; and as this court cannot review that judgment for
error, it can discharge the prisoner only when it is void. But we do not concede the
major premise in this argument. A judgment may be erroneous and not void,
and it may be erroneous because it is void. The distinctions . . . are very
nice . . . . 85 U.S. (i8 Wall.) at I75.
i963] FINALITY IN CRIMINALLAW 47I
the other questions of fact and law in the case. Similarly it is hard
to see the sense in the rule which made the availability of habeas
turn on whether the error related to the sentence rather than the
judgment of conviction itself, and even the usefulness of that
principle was undermined as soon as the Court expanded it to
pass not only on the validity of the sentence as such but to reach
questions such as the need for indictment (as in Wilson and
Bain).67
Yet I suggest that it would be a mistake to throw up one's
hands and to see the history as reflectingmerely a steady softening
and expansion of the concept of "jurisdiction" so as to allow
collateral attack on an open-ended basis wherever the courts
deemed it appropriate. Certain guideposts must be kept in mind:
i. The Court never abandoned, in this period, paying at least
lipservice to the principle that the legality of detention may not
be tested simply in terms of whether error occurred in previous
proceedings. It carefully adhered to the concept of institutional
competence as bearing on and involved in the problem of the law-
fulness of a detention. The essential touchstone continued to be
that the writ of habeas corpus was not to be used as a writ of
error, and that decisions of competent tribunals as to issues of fact
or law bearing on conviction should be final.68
2. The strict jurisdictional test in fact continued to govern
except in two categories of cases: where the allegation was that
the conviction was had under an unconstitutional statute, and
where the Court viewed the problem in terms of the illegality of
the sentence rather than that of the judgment. Now I do not
claim that these categories are easily justified today; but viewed
in a historical context they are not completely unintelligible. In
an era when the law was not "made" but "found," unconstitu-
tional statutes were thought of as "void," as nonexistent, in a
rather literal way: they created no law at all. It was an easy step
in this intellectual climate to come to the conclusion that a judg-
ment under such a statute, too, has a nonexistent quality, as if
there were no competence in the premises at all. Again, it is pos-
sible that the sanctity of a "judgment" of conviction was not
ascribed to the sentence because it was felt that the judge's
sentencing function is not "judicial" in the same sense as his
decision of the issues bearing on the substantive outcome of the
67 See note 56 supra. It is also hard to see a principled distinction between
Lange and Bigelow, both ultimately involving questions of double jeopardy.
See notes 55 and 62 supra.
68See, e.g., Harlan v. McGourin, 218 U.S. 442 (1910).
case; 69 reexaminationof the validity of the sentence was regarded

in the same light as reexaminationof pretrial judicial decisions of
questions such as the existence of probable cause to arrest or the
right to be enlarged on bail, traditionally available on habeas
corpus.70
In any event, we should not engage in the anachronism of
taking categories intelligible to their propoundersand regardedby
them as meaningful and useful, and using them today - when
we no longer accept them - in support of an indiscriminateand
open-endedexpansion of the habeas corpus jurisdiction. The fact
is that the nineteenth century judges plainly did not feel that their
distinctions were verbal fictions under the cover of which they
could produce such an expansion, and we should not regard their
cases as authorizing us to do so. Maybe it is well to abandon the
old rubrics,but it would be unhistoricalto gain comfort from cases
which were decided on the assumption that these rubrics were
meaningful and useful.
3. The cases definitely do not establish the proposition that
collateral inquiry was thought appropriate whenever the com-
mitting tribunal ruled on an issue of constitutional law. True,
passing expressions of Mr. Justice Bradley in Nielsen did hint
at such a development: "It is difficult to see," he said, "why a
conviction and punishment under an unconstitutionallaw is more
violative of a person's constitutional rights, than an unconstitu-
tional conviction and punishmentunder a valid law." Such a case
is not one of "mere error in law, but a case of denying to a person
a constitutional right. And where such a case appears on the
record, the party is entitled to be discharged . . . . And why
should not such a rule prevail in favorem libertatis?" 71 But the
facts of the Nielsen case were clearly governed by the holding in
Snow and fitted squarely into the category of unconstitutional
cumulative sentences imposed on but one offense. And in any
event, the suggestion that habeas automatically reached all con-
stitutional questions had been explicitly rejected by Mr. Justice
Miller in Bigelow 72 and was thereafter repudiated in cases such
69 This is particularly the case with respect to questions involving the
physical conditions of detention, i.e., whether the prisoner may be sentenced by a

federal judge to imprisonment in a state prison, or whether sentence to penitentiary
rather than jail is proper. See In re Bonner, II U.S. 242 (I894); In re Mills, I3
U.S. 263 (I890).
e.g., United States v. Hamilton, 3 U.S. (3 Dall.) I7 (I795); Ex parte
70See,
Burford, 7 U.S. (3 Cranch) 448 (i8o6); Ex parte Bollman, 8 U.,S. (4 Cranch)
75 (I8o7).
71 I3I U.S. at I83-84.
72
See pp. 468-69 and note 62 supra.
as In re Belt and Matter of Moran,73and also in the post-igoo

cases which held (repudiating Siebold) that the constitution-
ality of the statute defining the offense may not be tested on col-
lateral attack.74
4. Finally, one overwhelming fact must be kept in mind in
trying to rationalize the cases: throughout most of this period
federal criminal convictions were not appealable.75This, of course,
placed tremendous expansive pressure on the habeas corpus ju-
risdiction. There can be no doubt that it offends a felt sense of
justice that a single judge, no matter how wrong (or even mali-
cious), should have both first and final say on all questions before
him no matter how crucial the rights involved. Further, lack of
appeal not only fed the natural fear of error and malice, but also
deprived the country of a great institutional need for authoritative
and uniform pronouncements on important questions of federal
criminal law. The Supreme Court could not make such rulings
except through use of the habeas writ; and it must not surprise
us, then, that the writ was sometimes used for purposes (as in
passing on the constitutionality of, say, the Civil Rights Acts)
which are easily validated in the absence of appeal, but which
would be difficult to justify had the possibility of appeal existed.76
Indeed, it is striking that after appeal in federal criminal cases
was authorized the Court repudiated the doctrine of Siebold, that
73 See p. 469 supra.
7 See pp. 473-74 and note 77 infra.
7 Until I889 federal criminal cases were reviewable by the Supreme Court
only when there was a division of opinion in the circuit court on a question of
law. Act of April 29, I802, ch. 31, ? 6, 2 Stat. 159; Act of June i, I872, ch. 255,
? I, I7 Stat. I96. A writ of error to the Supreme Court was made available in
capital cases by the Act of Feb. 6, I889, ch. II3, ? 6, 25 Stat. 656. This was
extended to all cases of "infamous" crime by the Act of March 3, i89I, ch. 5I7,
? 5, 26 Stat. 827. The latter also provided for appeals in criminal cases to the
newly created courts of appeals, with review by the Supreme Court on certiorari
or in case of certification of a question of law by the court of appeals. See generally
HART & WECHSLER, THE FEDERAL COURTS AND THE FEDERAL SYSTEM I313-I7
('953).
76The influence of the fact that there was no possibility of appeal and thus
no supervision of the lower courts in their decisions on important issues of
federal law was made plain in Mr. Justice Bradley's opinion in Ex parte Siebold,
first holding that on habeas the Court will examine the constitutionality of the
statute creating the offense:
A conviction under [an unconstitutional statute] . . . cannot be a legal
cause of imprisonment. It is true, if no writ of error lies, the judgment may
be final, in the sense that there may be no means of reversing it. But
personal liberty is of so great moment in the eye of the law that the
judgment of an inferior court affecting it is not deemed so conclusive but
that . . . the question of the court's authority to try and imprison the
party may be reviewed on habeas corpus by a superior court or judge
having authority to award the writ.
IOOU.S. at 376-77. (Emphasis added.)
474 HARVARDLAW REVIEW [VoI. 76:44I
the constitutionality of the statute creating the crime is appro-

priately tested on habeas corpus: without trouble it ruled that
such a question should be tested in the ordinary course of litiga-
tion by writ of error.77 Insofar, then, as the Court in the nine-
teenth century departedon occasion from the ancient doctrine that
the merits of substantive issues decided by a court of competent
jurisdiction are not open on habeas corpus, we can fairly ascribe
the departurechiefly to circumstanceswhich are irrelevant to our
situation.
B. The Act of I867

Habeas corpus for prisonersheld under state authority was first
made generally available by the Habeas Corpus Act of I867,
which stated that
the several courts of the United States . . . within their respective
jurisdictions, in addition to the authority already conferred by
law, shall have power to grant writs of habeas corpus in all cases
where any person may be restrained of his or her liberty in viola-
tion of the constitution, or of any treaty or law of the United
States . 78
It is sometimes asserted that it was the explicit purpose of this

statute to give the federal courts jurisdiction to redetermine the
merits of all federal questions decided in a state criminal litiga-
tion, that the act compels the rule of Brown v. Allen.79 The as-
""In re Lincoln, 202 U.S. i78 Glasgow v. Moyer, 225 U.S. 420
(i906);
(I912); Henry v. Henkel, 235 U.S. 219, 229
(1914) ("disputed matters of law,
whether they relate to the sufficiency of the indictment or the validity of the
statute on which the charge is based . . . are for the determination of the trial
court"). The connection between the availability of appeal and the scope of
review on habeas was explicitly acknowledged by the Court in Salinger v. Loisel,
265 U.S. 224, 23I (I924) ("when a right to a comprehensive review in criminal
cases was given the scope of inquiry deemed admissible on habeas corpus came
to be relatively narrowed").
78 Act of Feb. 5, I867, ch. 28, ? I, I4 Stat. 385, presently codified in 28
U.S.C. ? 2241 (I958) (see p. 445 supra). The statute also prescribed the
procedure to be followed on habeas and authorized appeals from lower courts
in habeas cases.
7 The assertion is repeatedly made, though without any supporting analysis
or argumentation, in Mr. Justice Frankfurter's opinion in Brown v. Allen, 344
U.S. 443, 488 (I952) (see pp. 5oo-oi infra), and is repeated again without
any real analysis of either the language, history, or purpose of the act, by Justice
Walter V. Schaefer in his Oliver Wendell Holmes Lecture, Federalism and State
Criminal Procedure, 70 HARV. L. REV. I, 19-2I (1956). Justice Schaefer sees
the Nielsen case (see pp. 467-68, 472 & note 58 supra) as holding that the
1867 act broadened the substantive reaches of the writ from jurisdictional to
all constitutional issues. Yet the Nielsen opinion does not advert to the I867
act in any way, and contains no suggestion that the Congress at any time
sumption is that the Congress in I867 must have meant that a

person is held in "violation of the constitution . . . or [any] law
of the United States" whenever a committing tribunal falls into
error with respect to federal law. But such an assumption flies
directly in the face of the basic and well-understood principle
deeply imbedded in the law of the time that a detention pursuant
to the judgment of a competent tribunal is not illegal and does not,
for purposes of collateral attack, "violate" the "law" even if error
occurred. (I867 was six years before the Lange case even started
the process of softening this jurisdictional criterion.) Note,
further, that the suggested interpretationwould mean that habeas
corpus would lie to redetermine all legal questions in all federal
cases ending in detention as well as all federal questions in state
cases; certainly there could be no reason on such a reading to
restrict the language to constitutional questions raised in state
cases. How likely is it that the Congress by this statute should
have intended to work such a revolutionary change in the scope
of review afforded by habeas corpus? The language is, after all,
not compelling. As we have seen, a detention need not necessarily
be deemed on habeas corpus to be in "violation" of "law" merely
because a court of competent jurisdiction commits error of fact
or law in its proceedings. And it was habeas corpus that the
Congress was legislating about -a writ well known and well
understood to serve purposes quite different from appeal or writ
of error. (It should not, after all, be forgotten that the classical
function of habeas corpus was to assure the liberty of subjects
against detention by the executive or the military without any
court process at all, not to provide postconviction remedies for
prisoners.) It would, then, require rather overwhelmingevidence
to show that it was the purpose of the legislature to tear habeas
corpus entirely out of the context of its historical meaning and
scope and convert it into an ordinary writ of error with respect
to all federal questions in all criminal cases.
The strikingly sparse legislative history does not seem to me to
furnish such evidence.80 The act of I867 received only the most
authorized an expansion of the classes of questions appropriately tested on
habeas corpus; in fact it relies for support on the Lange and Siebold cases, both
decided under the old act of 1789.
80There is no clear indication what moved the Congress to take up general
revision of the habeas corpus jurisdiction (though surely the underlying concern
was the enforcement of the reconstruction legislation). In reporting what was
to become the act of i867 to the House on behalf of the House Judiciary Com-
mittee, Congressman Lawrence explained that the bill was drafted pursuant to
a resolution passed on December I9, i865. CONG.GLOBE, 39th Cong., ist Sess.
415I (i866). That resolution had instructed the Committee to report
perfunctory attention and consideration in the Congress; indeed,

there were complaints that its effects could not be understood at
all.81 Neither house made any inquiry into the scope or purposes
of review to be afforded on habeas corpus.82 And there is no
what legislation is necessary to enable the courts of the United States to
enforce the freedom of the wives and children of soldiers of the United
States under the joint resolution of Congress of March 3, i865, and also
to enforce the liberty of all persons under the operation of the constitutional
amendment abolishing slavery.
CONG. GLOBE, 39th Cong., ist Sess. 87 (I865). The reference to the joint resolution
of March 3, i865, is somewhat mysterious, for there does not appear to be any
resolution passed (or bill enacted) that day which is in any manner apposite
to the problem. The reference may have been to the Act of March 3, i863,
ch. 8i, ?? 4, 5, I2 Stat. 756, which had provided that a presidential order
would constitute a defense in any proceeding or prosecution for a wrong pursuant
to such order, and provided for removal from state to federal court of any
case where the defendant pleaded the existence of such an order. I surmise
that this might have been the intended reference because the resolution of
December ig, i865, was followed by the reporting and passage of a bill which
greatly widened the Act of March 3, i863, extending the substantive defense to
orders by all military superiors, and strengthening the removal provisions. Act
of May ii, i866, ch. 80, I4 Stat. 46. For the debates in connection with this
act, see CONG. GLOBE, 39th Cong., ist Sess. I387-90, I423-26, 1523-30, i880-82,
20I7-23, 2052-66 (i866); there was some evidence that the removal provisions
of the earlier statute were not being honored by the state courts. See, e.g., id.
at 2054. The Act of May ii, i866, was in turn amended some months later to
provide that where a case has been removed to federal court but the defendant
continues in state custody, the federal court may issue a writ of habeas corpus
cum causa. Act of Feb. 5, I867, ch. 27, 14 Stat. 385.
81The bill was reported to the House on July 25, i866, and immediately
passed with virtually no debate at all. Congressman LeBlond complained "that
it is exceedingly difficult for us to determine the scope of the bill," and stated:
"I am compelled to object, from the fact that we have no opportunity to con-
sider and understand the bills that are brought in here." CONG. GLOBE, 39th
Cong., ist Sess. 4151 (I866). When the bill was brought up again in the House
to consider a Senate amendment, this was the debate:
Mr. COOK. I move that the House concur in the amendment.
Mr. WRIGHT. I would ask whether anybody in this House, when he
gives his vote on these amendments, knows what he is voting upon? [Laughter.]
The SPEAKER. The gentleman from New Jersey is not in order. The
question is on the motion to concur.
The motion was agreed to.
CONG. GLOBE, 39th Cong., 2d Sess. 899 (I867).
82 In the House, when the bill was reported, Congressman LeBlond suggested
that it would not cover certain civilians seized by the military. In response,
Congressman Lawrence offered this explanation of the bill:
On the igth of December last, my colleague [Mr. SHELLABARGER] introduced
a resolution instructing the Judiciary *Committee to inquire and report to
the House as soon as practicable, by bill or otherwise, what legislation is
necessary to enable the courts of the United States to enforce the freedom
of the wife and children of soldiers of the United States, and also to enforce
the liberty of all persons. Judge Ballard, of the district court of Kentucky,
decided that there was no act of Congress giving courts of the United States
jurisdiction to enforce the rights and liberties of such persons. In pursuance
of that resolution of my colleague this bill has been introduced, the effect
of which is to enlarge the privilege of the writ of hobeas [sic] corpus, and
make the jurisdiction of the courts and judges of the United States coextensive
with all the powers that can be conferred upon them. It is a bill of the
indication whatever that the bill intended to change the general

nature of the classical habeas jurisdiction: that is, that it intended
to expand the classes or categories of questions which were
thought to be appropriately tested on collateral attack.
Read in the context of its time, the statute may at most be re-
garded as ambiguous with respect to what matters would be
deemed to render a detention unlawful under federal law.83 In-
deed, it is interesting as evidence of at least reasonably contem-
poraneous understanding that in I884 the House Committee on
the Judiciary reported (in connection with the bill to restore the
Supreme Court's appellate jurisdiction under the act of I867)
that in its view the act was not "contemplated by its framers or
. . . properly . . . construed to authorize the overthrow of the
final judgments of the State courts of general jurisdiction, by the
84
inferior
il Federal judges ....
largest liberty, and does not interfere with persons in military custody, or
restrain the writ of habeas corpus at all.
CONG. GLOBE, 39th Cong., ist Sess. 4151 (i866). This constituted the entire
discussion of the bill in the House.
The claim that the habeas jurisdiction was to become coextensive with "all
the powers" conferrable by article III of the Constitution seems to me to be
largely meaningless. After all, constitutionally the jurisdiction could be expanded
to make the district courts ordinary courts of appeals with respect to all federal
questions in all state cases (see note 9 supra). And the statement thus does not tell
us anything about how far Congress intended to change the general nature
of the inquiry made on habeas corpus as then understood.
In the Senate, the debates were also cursory; in the only statement about
the purpose of the bill the problem was put solely in terms of the failure of
the existing law to allow any inquiry on habeas when the detention was not
pursuant to federal law:
[T]he habeas corpus act of I789, to which this bill is an amendment, con-
fines the jurisdiction of the United States courts in issuing writs of habeas
corpus to persons who are held under United States laws. Now, a person
might be held under a State law in violation of the Constitution and laws
of the United States, and he ought to have in such a case the benefit of the
writ, and we agree that he ought to have recourse to United States courts
to show that he was illegally imprisoned in violation of the Constitution or
laws of the United States.
Statement of Senator Trumbull, CONG. GLOBE, 39th Cong., ist Sess. 4229 (i866).
For further action on the bill, see id., 2d Sess. 730, 790 (i867).
83 See Wyzanski, J., in Geagan v. Gavin, i8i F. Supp. 466, 468 (D. Mass.
I960), aff'd, 292 F.2d 244 (ist Cir. I96I), cert. denied, 370 U.S. 903 (I962):
Congress did not use language, and there was nothing in the avowed purpose
or legislative history of the i867 statute, which compelled the Supreme
Court of the United States to interpret the statute as conferring upon
United States District Judges authority to inquire whether a state court
judgment by a jurisdictionally competent court rested upon any procedural
step or substantive ruling involving a violation of the United States Consti-
tution.
84H.R. REP. No. 730, 48th Cong., ist Sess. 5 (1884). The Committee com-
plained about the extent of the authority claimed by federal judges under the
i867 act, though all the cases cited involved review of jurisdictional questions.
It concluded, however, that it would, for the present, not propose changes in the
C. The Early "State" Cases

It was not until i886 that the SupremeCourt first had occasion
to deal with the scope and purposes of the habeas corpus jurisdic-
tion for state prisoners created in I867. The case was Ex parte
Royall.A5 The petitioner had been indicted for violation of a
Virginia statute and was detained awaiting trial. Alleging that
this statute was unconstitutional, he sought habeas corpus in the
federal district court, which dismissed the writ. On appeal the
Supreme Court affirmed,announcing the famous doctrine of ex-
haustion of state remedies: though the district court has power in
advance of trial to inquire into the allegation, it need not and
should not do so pending consideration of the question in the
normal course of trial by the state court.86 The holding of Royall
thus did not address itself to the problem of what questions are
properly cognizable on habeas: it held merely that federal
courts, as a matter of sound discretion and in the absence of ex-
ceptional circumstances,8 would not assume jurisdiction before
state trial courts have had an opportunity to decide federal ques-
I867 act, since, with the restoration of Supreme Court supervision under the
act, "the true extent of the act of i867, and the true limits of the jurisdiction of
the Federal courts and judges under it, will become defined, and it can then be
seen whether further legislation is necessary." Id. at 6.
85 117 U.S. 241 (i886).
86 We cannot suppose that Congress intended to compel . . . [the federal]
courts . . . to draw to themselves, in the first instance, the control of all
criminal prosecutions commenced in State courts . . . . The injunction to
hear the case summarily, and thereupon "to dispose of the party as law and
justice require" does not deprive the court of discretion as to the time and
mode in which it will exert the powers conferred upon it. That discretion
should be exercised in the light of the relations existing, under our system of
government, between the judicial tribunals of the Union and of the States,
and in recognition of the fact that the public good requires that those rela-
tions be not disturbed by unnecessary conflict between courts equally bound
to guard and protect rights secured by the Constitution.
Id. at 251.
" Mr. Justice Harlan suggested in Royall that federal interference prior to
exhaustion of state remedies might be justified in "cases of urgency, involving the
authority and operations of the General Government, or the obligations of this
country to, or its relations with foreign nations." Ibid. For cases involving such
special circumstances see In re Loney, 134 U.S. 372 (I890); In re Neagle, I35 U.S. I
(i890); Ohio v. Thomas, I73 U.S. 276 (I899); Boske v. Comingore, i77 U.S. 459
(I900); Hunter v. Wood, 209 U.S. 205 (I908).
It soon became apparent that the "discretionary" nature of the trial courts'
power to postpone habeas was rather illusory, for whenever the trial court granted
the writ prior to exhaustion of state remedies (and where none of these so-called ex-
ceptional circumstances involving interference with the operations of federal officials
existed), the Supreme Court reversed. See, e.g., New York v. Eno, I55 U.S. 89
(1894); Baker v. Grice, I69 U.S. 284 (I898); Fitts v. McGhee, I72 U.S. 5I6 (I899);
Minnesota v. Brundage, I80 U.S. 499 (I90I); Urquhart v. Brown, 205 U.S. I79
(1907).
r963] FINALITY IN CRIMINALLAW 479
Lions at issue.88 Similarly, subsequent cases held that habeas

should be denied while a prisoner seeks to vindicate his federal
rights in the state appellate courts 89 and through state postcon-
viction procedures.90 And still another line of cases established
that even after all state remedies were exhausted, habeas corpus
should be denied and the prisoner put to his writ of error in the
United States Supreme Court.9'
Some language in the early cases does suggest that the require-
ment that state remedies be exhausted was thought of as involving
a mere postponement of habeas corpus, that it was assumed that
the writ would be available to redetermine the federal question
once it had been litigated in the state system. Thus in Royall
itself the Court said that once
the State courtshall have finallyacted upon the case, the Circuit
Court has still a discretionwhether,underall the circumstances
then existing,the accused,if convicted,shall be put to his writ of
errorfromthe highestcourtof the State,or whetherit will proceed,
by writ of habeas corpus, summarilyto determinewhetherthe
petitioneris restrainedof his liberty in violationof the Constitu-
tionof the UnitedStates.92
This seemingly equates the substantive scope of the habeas juris-
diction with ordinary review on the merits. What must be remem-
bered, however, is that Royall (and the other cases containing
such language) in fact involved issues which were concededly open
to collateral attack under the ordinary rules of habeas corpus.93
The question in Royall was the constitutionality of the statute
creating the offense. As we have seen, this was by i886 considered
88 The exhaustion requirement announced in Royall was enforced, subsequently,
in many cases. See, e.g., Cook v. Hart, I46 U.s. i83 (I892); New York v. Eno, I55
U.S. 89 (I894); Whitten v. Tomlinson, i6o U.S. 23I (I895); Baker v. Grice, I69
U.S. 284 (I898); Fitts v. McGhee, I72 U.S. 516 (I899).
89Ex parte Fonda, II7 U.S. 5I6 (i886); In re Duncan, I39 U.S. 449 (1891);
Minnesota v. Brundage, i8o U.S. 499 (I90I); Reid v. Jones, I87 U.S. I53 (I902).
90Pepke v. Cronan, I55 U.S. IOO (I894).
91 In re Frederich, I49 U.S. 70 (I893); Bergemann v. Backer, 157 U.S. 655
(I895); Tinsley v. Anderson, I7I U.S. IOI (I898); Markuson v. Boucher, I75 U.S.
I84 (i899); Urquhart v. Brown, 205 U.S. I79 (1907).
92 II7 U.S. at 253.
9 See, e.g., In re Frederich, 149 U.S. 70 (I893), where the issue was the con-
stitutionality of a state statute permitting an appellate court itself to enter the
proper verdict and sentence which the trial court should have entered. The Court,
affirming dismissal of habeas, relied on Royall in indicating that as to such a
question it is the "better practice . . . to put the prisoner to his writ of error"
rather than inquire into the merits on habeas (id. at 77-78) ; but it also stressed
that habeas "is not a proceeding for the correction of errors" and that it will issue
only, to test the jurisdiction of the committing court (id. at 75-76).
a "jurisdictional"issue open on collateral attack; and the Court,

in announcing that there was power on habeas to deal with the
case, relied heavily on the holding of Siebold that a conviction
under an unconstitutionalstatute is "void," not merely erroneous,
and thus cannot constitute a lawful cause of imprisonment.4 And
the Court also relied on and quoted from a decision by Mr. Justice
Bradley on circuit 95 where he had carefully distinguishedbetween
the power of the federal courts under the act of I867 in cases
where the state judgment is merely erroneousand where it is void
for want of jurisdiction.96Royall thus by no means indicates an
understanding that the courts on habeas corpus would, under
the I867 act, reach and relitigate a different and wider class of
questions than were supposed under the previous law to be subject
to collateral review.
Similarly, in cases of this period where the Court on habeas
corpus reached the merits of a state prisoner's federal contention,
either because it seemingly overlooked the requirementof exhaus-
tion of state remedies or because those remedies had been ex-
hausted, the contention so canvassed was one reachable on col-
lateral attack under the traditional law of habeas corpus.97
94 II7 U.S. at 248. Similarly, in Minnesota v. Brundage, i8o U.S. 499 (I90),
where habeas was ordered dismissed "without prejudice to renewal of the same
after the accused had availed himself of such remedies as the laws of the State
afforded for a review of the judgment" (id. at 500-0), the issue was the con-
stitutionality of the statute creating the offense.
95Ex parte Bridges, 4 Fed. Cas. I04 (No. I862) (C.C.N.D. Ga. I875).
96 Bridges involved a conviction by Georgia for perjury before a United States
Commissioner, an offense over which, "it was obvious," the state had not even
colorable jurisdiction under the applicable federal statutes. Mr. Justice Bradley
ruled that Bridges was properly released on habeas and did not have to raise his
contention on appeal. But he was careful to point out that:
If it were a case in which the state court had jurisdiction of the offense, the
general rule of the common law would intervene, and require that the prisoner
should be remanded, and left to his writ of error. In such a case, although the
judgment were erroneous, the imprisonment would not be in violation of the
constitution or laws of the United States. The judgment might be wrong, but
the imprisonment under it would be right until the judgment was reversed.
But, as before shown, the state court had not jurisdiction of the offense.
Id. at io6.
" In Wo Lee v. Hopkins, reported sub nom. Yick Wo v. Hopkins, ii8 U.S.
356 (i886), petitioner was discharged on habeas corpus on the ground that he was
convicted pursuant to an unconstitutional state statute; no reference was made to
the exhaustion requirement although a companion case clearly indicated that state
habeas would have been available. The constitutionality of state statutes was
similarly tested on the merits without reference to the exhaustion requirement in
Minnesota v. Barber, I36 U.S. 3I3 (I890) (disapproved in Minnesota v. Brundage,
i8o U.S. 499 (I9OI) ), and In re Rahrer, I40 U.S. 545 (I89I). In Crowley v.
Christensen, I37 U.S. 86 (I890), the Court also reached the merits of the question
of the constitutionality of the statute defining the offense, but it appears that state
remedies were in fact exhausted and there appeared to be no effective way of
1963] FINALITY IN CRIMINALLAW 48I
Finally, some cases decided during this period seem to hold ex-
plicitly that the act of I867 did not empower the federal courts
on habeas corpus to redetermine the merits of federal questions
- even constitutional questions - which did not go to the juris-
diction of the committingcourt. In re Wood 98 is the earliest99
and leading case. Wood, a Negro, was convicted by a New York
court of murder, and his conviction affirmed by the Court of
Appeals.'00 After his conviction he allegedly discovered that the
grand jury which returned his indictment and the petit jury which
tried him were drawn from lists from which Negroes were system-
atically excluded. This flaw was made the subject of a motion
for new trial, which was denied, and Wood then sought habeas
in the federal court. Dismissal of the writ was affirmed by the
Supreme Court. Mr. Justice Harlan pointed out that there was
no contention that the New York statute regulating the selection
of juries was unconstitutional. As to the practice of discrimination
in jury selections, he affirmed the established principle that it
violates the Constitution.'0' But whether such discrimination
existed here "was a question which the trial court was entirely
competent to decide, and its determinationcould not be reviewed
. . . upon a writ of habeas corpus, without making that writ serve
the purposes of a writ of error. No such authority is given to the
Circuit Courts . . . by the statutes defining . . . their jurisdic-
tion." State courts often decide questions which "involve the
construction of" the Constitution and the "determinationof rights
asserted under it." "But that does not justify an interferencewith
its proceedings . . . upon a writ of habeas corpus . . . either dur-
bringing the case to the Court on writ of error. See also McElvaine v. Brush,
I42 U.S. I55 (I89I) (upholding on habeas constitutionality of state statute provid-
ing for the solitary confinement of a prisoner pending execution of sentence of
death; no mention of need for exhaustion of or availability of state remedies).
98 I40 U.S. 278 (I89I).
9 The ruling in Wood was foreshadowed by Ex parte Crouch, I12 U.S. 178
(i884). There is, however, some obscurity about the jurisdiction of the Supreme
Court in that case. The decision was certainly not one under the old act of I789,
since the detention was pursuant to state, not federal, law. The Court implied that
the petition lay under the act of I867; but it seemed to overlook the fact that its
appellate jurisdiction under that act was still under suspension.
100 I23 N.Y. 632 (I890). It is unclear whether the appeal was decided before
or after the denial of the motion for new trial.
'o' The Court quoted from the opinions in Neal v. Delaware, I03 U.S. 370
(i88o), and Virginia v. Rives, I00 US. 313 (i879), to the effect that the fourteenth
amendment guarantees an accused selection of grand and petit jurors without
racial discrimination. I40 U.S. at 284-85.
ing or after the trial in the state court." 102 And the Court con-
tinued:
If the questionof the exclusionof citizensof the Africanracefrom
the lists of grand and petit jurors had been made during the
trial . . . and erroneouslydecided against the appellant, such error
in decisionwouldnot have madethe judgmentof convictionvoid,
or his detention under it illegal. . . . Nor would that error, of
itself, have authorized the Circuit Court . . . upon writ of habeas
corpus, to review the decision or disturb the custody . . . . The
remedy . . . was to sue out a writ of errorfrom this court . .. .103
Wood was followed by opinions denying the competence of the
court on habeas to decide the questions of inadequacy of counsel
and discrimination in selection of jurors in In re Jugiro,104 the
latter question in Andrews v. Swartz,'05 and the question of
alleged federal constitutional flaw in the indictment in Bergemann
v. Backer.'06 Similarly, the Court declined to pass on habeas on
the question of the constitutionality of Wisconsin procedure with
respect to jury verdicts on the ground that the allegation was of
"error committed in the exercise of jurisdiction, and one which
does not present a jurisdictional defect, remediable by the
writ . . . 107
102
I40 U.S. at 285-86. It may be argued that the result in Wood rests on the
fact that the petitioner failed to secure a decision of his federal claim in the state
courts and it was consequently forfeited for purposes of habeas. But the language of
the opinion is plainly and wholly inconsistent with the notion that habeas would
have been proper to test the federal issue had the state courts decided it: the entire
thrust of the opinion is that the issue is not cognizable on habeas at all.
103 Id. at 287. The Court also had to meet the allegation that the state had
failed to give the prisoner a meaningful opportunity to raise his constitutional claim
in the state courts. As to this, the Court disagreed with the petitioner's reading of
New York law (id. at 287-89), and added as an independent ground that, under
Royall, the prisoner should have tested this question (that is, the availability of
state corrective process) in the state courts. Id. at 289-90. The Court's disposition
of this point would certainly be considered questionable today in view of the fact
that Wood was not represented by counsel.
104 I40 U.S. 29I (I89I):
The alleged assignment . . of one as his counsel who . . had not been
admitted or qualified to practice . . . ; the misdescription in the indictment
. . ., and the exclusion from the list of grand and petit jurors of citizens
. . . of the same race with appellant, were all matters occurring in the course
of the proceedings and trial in a court of competent jurisdiction, proceeding
under [constitutional] statutes . . . . The errors, if any ... did not affect
its jurisdiction . . . and cannot be reached by habeas corpus.
Id. at 296-97.
105 I56 U.S. 272 (I895). This case seems to ignore the possible appropriate-
ness of testing a "nonjurisdictional" federal question on the merits if the totality
of state process fails to provide a defendant with a meaningful opportunity to do
so in the state system.
106 I57 U.S. 655 (I895).
107 In re Eckart, i66 U.S. 48I, 483 (I897). The Court relied on cases involving
federal prisoners such as In re Belt and Ex parte Bigelow.
1963] FINALITY IN CRIMINALLAW 483
None of these cases even suggests that as to these federal con-

stitutional questions denial of the writ constitutes mere discre-
tionary postponement; as the question was phrased in Harkrader
v. Wadley, it involves the lack of "authority"of the federal courts
to issue habeas to a state prisoner under custody ordered by a
competent tribunal and pursuant to a constitutional statute.'08
Plainly, then, it was not the law at this time, and not the under-
standing of the I867 act, that a detention "violates" the Con-
stitution whenever the state courts commit error as to an issue
of federal constitutional law relevant to the case.
In fact the requirement that, even as to "jurisdictional"ques-
tions, state remedies be exhausted prior to resort to habeas, itself
becomes entirely intelligible only in terms of a principle or pre-
sumption that a court which has jurisdiction of a case should be
allowed to decide all the questions in the case under all of the
applicable law, state or federal, and that such a decision should,
indeed, "count"; that in the ordinary course supervision of such
a court should be by appeal and not through independent collater-
al proceedings. It would make little sense to encourage the use of
state remedial processes through a requirementof exhaustion only
in order to ignore these processes on collateral attack. The point
is, of course, underscoredby the decisions holding that even after
all state remedies have been exhausted, a prisoner must test his
federal claim not by habeas but on writ of error in the Supreme
Court; the requirement makes sense not because such review is
part of the state's remedies but because the institutional needs
calling for federal supervision of state decisions are adequately
met through such direct review. As the Court said in Matter of
Spencer, "the rule [requiring exhaustion] would be useless ex-
cept to enforce a temporary delay if it did not compel a review of
the question in the state court and, in the event of an adverse
decision, the prosecution of error from this court." 109
D. The Doctrine of Frank v. Mangum

We can summarize the law as of I9I5 in this way: if a court
of competent jurisdiction adjudicated a federal question in a
criminal case, its decision of that question was final, subject only
to appeal, and not subject to redetermination on habeas corpus.
There existed a few classes of issues (principally the constitution-
ality of the statute creating the offense) which were labeled jur-
isdictional though they did not really bear on the competence of
108 172 U.S. 148, I63 (I898) (passing on question of state court's jurisdiction).
109 228 U.S. 652, 66o (igi3).
the committing court; these were, however, strictly limited and

their creation was probably groundedon the lack of appeal in fed-
eral criminal cases. Further, in state criminal cas-es,pursuant to
the exhaustion doctrine even "jurisdictional"issues had to be liti-
gated in the state courts and tested by the Supreme Court on ap-
peal in the ordinary course, so that, as a practical matter, there
would be no further occasion for use of the writ.110
To this situation the great case of Frank v. Mangum "I added
some crucial insights. Frank was convicted of murderin Georgia
and sentenced to death. His petition for a new trial alleged that
the proceedings had been completely dominated by a mob which
made impartial adjudication by judge and jury impossible. This
petition was denied, and appeal was taken to the Supreme Court
of Georgia. That court made an independent inquiry, based not
only on the record but on extensive affidavits, into the merits of
the question whether in fact there had been mob disorders which
affected the possibility of an impartial trial; in careful findings it
concluded that the evidence did not warrant the conclusion that
there was prejudicial interference with the course of justice.112
After various postconviction proceedings and the denial of a writ
of error by the Supreme Court, Frank unsuccessfully sought
habeas corpus in the federal district court. On appeal the Supreme
Court affirmed,Mr. Justice Holmes and Mr. Justice Hughes dis-
senting.
Mr. Justice Pitney's opinion for the Court is a highly sophisti-
cated analysis of the complex nature of the concept of due pro-
cess. He found it easy to agree that
if a trial is in fact dominatedby a mob, so that the jury is in-
timidatedand the trial judgeyields, and so that thereis an actual
interference with the course of justice, there is, in that court, a de-
parture from due process of law .. ..113
But he saw that the question whether "a trial is in fact dominated
110 If I understand him correctly, Professor Reitz is in agreement with my
view of the situation at this time. See Reitz, The Abortive State Proceeding, 74
HARV.L. REV. 1315, 1327 (I96I). I have found only two cases whose holdings
may be thought to shade these principles. In re Converse, I37 U.S. 624 (I891),
seemingly passes on the merits of a state prisoner's nonjurisdictional due process
allegations; the opinion does not discuss the problem of habeas jurisdiction at all.
The opinion in Felts v. Murphy, 20I U.S. 123 (I906), denying habeas to a state
prisoner, is a confusing and ambiguous mixture of the points that the state court
did not lack jurisdiction and that there was no denial of due process.
i 237 U.S. 309 (I9I5).
112
Frank v. State, I4i Ga. 243, 280, 80 S.E. ioi6, I034 (1914).
113
237 U.S. at 335.
by a mob" is, after all, a question, and that decision of that ques-
tion against the petitioner by a competent and unbiased tribunal
through fair process may, on collateral inquiry, itself be deemed
the process that is "due." It is, in other words, relevant to the
question of the unlawfulness of a detention for purposes of habeas
corpus whether the state has supplied "corrective process" for
the very purpose of determining the federal question raised.114
Here the prisoner's allegations were considered by the Georgia
supreme court under conditions which were concededly free from
any suggestion of mob domination, and found by that court, on
independent inquiry, to be groundless.1"5 It need not be held,
continued the Court, that the Georgia court's decision was, tech-
nically, res judicata; 116 but
this does not mean that that decision may be ignored or disre-
garded. To do this . . . would be not merely to disregard comity,
but to ignore the essential question before us, which is not the guilt
or innocence of the prisoner, or the truth of any particular fact
asserted by him, but whether the State, taking into view the entire
course of its procedure, has deprived him of due process of law.
This familiar phrase does not mean that the operations of the state
government shall be conducted without error or fault in any
particular case ....117
Therefore, said the Court,

we hold that such a determinationof the facts as was thus made by
the court of last resort of Georgia respecting the alleged inter-
ference with the trial .-. . cannot in this collateral inquiry be
treated as a nullity, but must be taken as setting forth the truth of
the matter, certainly until some reasonable ground is shown for an
inference that the court which rendered it either was wanting in
jurisdiction, or at least erred in the exercise of its jurisdiction; and
that the mere assertion by the prisoner that the facts of the matter
are other than the state court upon full investigation determined
them to be will not be deemed sufficientto raise an issue respecting
the correctness of that determination ....118
114
[I]f the State, supplying no corrective process, carries into execution a
judgment of death or imprisonment based upon a verdict . . . produced by
mob domination, the State deprives the accused of his life or liberty without
due process of law.
But the State may supply such corrective process as to it seems proper.
[bid. See also id. at 327-28, 332-33.
115 Id. at 333.
116 Id. at 333-34.

117 Id. at 334.
118Id. at 335-36. The Court pointed out that the doctrine of exhaustion of
state remedies is intelligible only upon such a basis:
It follows as a logical consequence [of the exhaustion principle] that where, as
The Frank case is often seen as restrictive with respect to the

habeas jurisdiction."19Yet the importance of the case derives as
much from what the Court said the federal courts could do on
habeas as what it held they could not do. For the first time the
Court explicitly added a crucial weapon to the arsenal of the
habeas corpus court: if that court finds that a state tribunal has
failed to supply "correctiveprocess" with respect to the full and
fair litigation of federal questions, whether or not "jurisdictional,"
here, a criminal prosecution has proceeded through all the courts of the State,
including the appellate as well as the trial court, the result of the appellate re-
view cannot be ignored when afterwards the prisoner applies for his release on
the ground of a deprivation of Federal rights sufficient to oust the State of its
jurisdiction to proceed to judgment and execution against him. This is not a
mere matter of comity, as seems to be supposed. The rule stands upon a much
higher plane, for it arises out of the very nature and ground of the inquiry
into the proceedings of the state tribunals, and touches closely upon the re-
lations between the state and the Federal governments.
Id. at 329.
119 See, e.g., Reitz, The Abortive State Proceeding, 74 HARV.L. REV. I3I5,
I329 (I96I): "Frank v. Mangum . . . earlier threatened to stifle growth of the

fourteenth amendment by an amalgam of the exhaustion requirements and the
substance of the concept of procedural due process of law." Does this suggest, per-
haps, that Professor Reitz sees Frank as a holding with respect to the meaning of
due process even for purposes of direct review, that is, as a decision that even on
direct review the Supreme Court may not correct state-court errors as to federal
questions if the state has supplied a "corrective process" in the form of a fair ad-
judication of the question? Although not without its ambiguities, I do not think
the Court's opinion will bear such a reading. Mr. Justice Pitney makes clear that
his entire reasoning is in the context of habeas corpus, which he carefully differen-
tiates from ordinary appeal. (See 237 U.S. at 325-26.) And his statement of the
holding is directed in terms at the scope of the review "in this collateral inquiry."
Id. at 336. Certainly any holding that on direct review the Supreme Court does
not have plenary jurisdiction to review on the merits all dispositive questions of
federal law (including, as to federal issues, the sufficiency of the evidence and the
inferences to be drawn from the facts, cf. Norris v. Alabama, 294 U.S. 587 (i935) ),
would have been a startling reversal of the law established by Martin v. Hunter's
Lessee, I4 U.S. (i Wheat.) 304 (i8I6). The important point to remember about
Frank is its insight that there comes a point at which previous determinations,
themselves fairly arrived at, settle the question whether due process was in fact
afforded, and that in our legal system that point is usually where a judgment has
become final and immune from direct review. There is, in other words, nothing
radical about the notion that a judgment has accorded due process even though
there remains a theoretical possibility - which possibility, it is the whole thrust of
the doctrine, will not be explored - that error has occurred as to a due process
question.
Similarly, I see no basis in Frank v. Mangum for an argument that it applies
only to findings of fact made by the state courts; the opinion is wholly inconsis-
tent with the notion that, if a state supplies adequate corrective process, its rulings
as to the application of federal standards to the facts are nevertheless reviewable
on habeas. In fact Frank itself presented a question of federal law -whether the
facts as found by the state appellate court were consistent under federal constitu-
tional standards with a fair trial; the Supreme Court did not even hint that re-
view on habeas was open as to this question, though there can be no doubt that
on direct appeal it would have been the central issue before the Court.
in a state criminalproceeding,a court on habeas may appropriately

inquire into the merits in order to determine whether the deten-
tion is lawful.120From this aspect of Frank v. Mangum, I suggest,
derive all the great and beneficient expansions of the writ we have
witnessed in the past fifty years. (Of course expansion created
problems of its own. As long as habeas was narrowly restricted to
jurisdictional issues, the fact that it was available without limit
of time did not create intolerable delays in the administration of
justice by the states. But the effect of the widening of the writ in
Frank was to import the time problem into a much larger category
of cases, and this has made the problems of delay in the criminal
process much more acute.)
On the other hand the Frank opinion does, concededly, state
what I conceive to be simple common sense but which others may
regard as restrictive: that the fact that an unbiased court of
competent jurisdiction has previously adjudicated, through a full
and fair litigation, the merits of whether a defendant's federal
rights were violated is crucially relevant to the question whether
his detention may on habeas corpus be considered unlawful be-
cause he was denied due process of law. I regard this as common
sense because it directs the inquiry on habeas corpus to the mean-
ingful question whether the totality of state process assures us
of a reasoned probability that justice was done, rather than
whether in some ultimate sense the truth was in fact found.'21
120 This "expansive" aspect of Frank v. Mangum was made possible pro-
cedurally by the fact that the new habeas corpus procedure created by the act of
I867 for the first time plainly authorized the federal courts to inquire dehors the
state-court record; previously the prisoner had to show on the face of the record
that the committing court lacked jurisdiction. See Frank v. Mangum, 237 U.S. at
329-3Q. As the Court stated in Frank, these procedural changes "substitute for the
bare legal review that seems to have been the limit of judicial authority . . . a
more searching investigation, in which the applicant is put upon his oath to set
forth the truth of the matter respecting the causes of his detention, and the court,
upon determining the actual facts, is to 'dispose of the party as law and justice
require."' Id. at 330-3I. The adequacy of a state's corrective process will, of
course, often raise questions not determinable on the face of the state record; such
an inquiry was thus first made possible by the new act. The expansive procedure
authorized by the statute may in turn also explain the widened substantive scope
the Court was ascribing to the writ.
121 Mr. Justice Holmes' dissent in the Frank case seems to me striking pre-
dominantly because of his refusal to come to grips with the issue posed by Mr.
Justice Pitney. He stated that the "single question . . . is whether a petition
alleging that the trial took place in the midst of a mob savagely and manifestly
intent on a single result, is shown on its face unwarranted . . . ." and declared that
it is the "duty" of the Court on habeas to declare "lynch law as little valid where
practiced by a regularly drawn jury as when administered by one elected by a
mob intent on death." 237 U.S. at 349, 350. And he argued:
Whatever disagreement there may be as to the scope of the phrase "due
Was Frank v. Mangum "discredited"a mere eight years later

in Moore v. Dempsey? 122 Professor Hart has so stated; 123 and
Mr. Justice McReynolds' dissent in Moore 124 accuses the majority
of overrulingFrank.'25 On the other hand, Mr. Justice McKenna
and the meticulous Mr. Justice Van Devanter, who had been with
the majority in Frank, joined the Holmes opinion for the majority
in Moore."26 And in fact that opinion neither purports to be nor
necessarily needs to be interpreted as a departure from Frank.
Moore, like Frank, was a case involving allegations of mob dom-
ination of trial court and jury. A motion for new trial on this
ground had been overruled, and an appeal taken to the state su-
preme court, which affirmedthe convictions.'27 What is striking is
that in Moore, unlike in Frank, the state supreme court did not
conduct any proceeding or make any inquiry into the truth of the
process of law," there can be no doubt that it embraces the fundamental con-
ception of a fair trial, with opportunity to be heard. . . . We are not speaking of
mere disorder . . . but of a case where the processes of justice are actually
subverted. In such a case, the Federal court has jurisdiction to issue the writ.
The fact that the state court still has its general jurisdiction and is otherwise a
competent court does not make it impossible to find that a jury has been sub-
jected to intimidation in a particular case.
Id. at 347. But as Mr. Justice Pitney saw, this is really beside the point:
It is argued that if in fact there was disorder such as to cause a loss of jurisdic-
tion in the trial court, jurisdiction could not be restored by any decision of
the Supreme Court. . . [This] begs the question of the existence of such
disorder . . . - which should not be assumed, in the face of the decision of
the reviewing court, without showing some adequate ground for disregarding
that decision.
Id. at 336. Mr. Justice Holmes in fact provides no reason why it should be dis-
regarded, other than the general proposition that "when the decision of the ques-
tion of fact is so interwoven with the decision of the question of constitutional
right that the one necessarily involves the other, the Federal court must examine
the facts." Id. at 347. But this proposition, too, is beside the mark if, as Mr.
Justice Pitney holds, the question on habeas is not whether error of fact or law
occurred as to a federal question but rather whether the state supplied a fair
and rational process for its litigation.
122 26I U.S. 86 (I923).
123
Hart, Foreword, 73 HARV.L. REV. 84, 105 (I959).
124
26I U.S. at 92. The dissent (which Mr. Justice Sutherland joined) seems
to me to exhibit a complete failure to understand Frank v. Mangum in its "ex-
pansive" aspect: that the federal habeas court does have the duty to inquire
whether the state did supply "corrective process."
125 Professor Reitz asserts that the dissent, in arguing that Frank was being
overruled, "recognizes the realities." Reitz, The A bortive State Proceeding, 74
HARV. L. REV. I3I5, I329 n.53 (I96I).
126 These two, together with Holmes (who dissented in Frank and wrote
Moore) and McReynolds, were the only members of the Frank court still sitting
when Moore was decided.
127 Hicks v. State, I43 Ark. I58, 220 S.W. 308, cert. denied, 254 U.S. 630
(I920). For a detailed analysis of the facts and proceedings in Moore v. Dempsey,
see the two articles on the case by Waterman and Overton, The Aftermath of
Moore v. Dempsey and Federal Habeas Corpus Statutes and Moore v. Dempsey,
both published originally in I933 and both reprinted at 6 ARK. L. REV. I (I95I).
allegations of mob domination, and made no findings with respect

to them; 128 its opinion refers to the charge that fair trial was
impossible in the trial court solely by commenting, laconically,
that it could not say "that this must necessarily have been the
case." 129 On these facts Mr. Justice Holmes, allowing that under
Frank v. Mangum "the corrective process supplied by the State
may be so adequate that interference by habeas corpus ought not
to be allowed," held that "if the case is that the whole proceeding
is a mask . . . and that the State Courts failed to correct the
wrong," then "perfection in the machinery for correction" will
not bar habeas jurisdiction. And the Court, reciting the course
the case took in the state courts, concluded:
We shall not say more concerning the corrective process afforded
to the petitioners than that it does not seem to us sufficient to
allow a Judge of the United States to escape the duty of examining
the facts for himself when if true as alleged they make the trial
absolutely void.130
Though the opinion is admittedly far from clear, all Moore v.
Dempsey may be saying, then, is that a conclusory and out-of-
hand rejection by a state of a claim of violation of federal right,
without any process of inquiry being affordedat all, cannot insu-
late the merits of the question from the habeas corpus court: if
the state's findings are to "count," they must be reasoned findings
rationally reached through fair procedures. So viewed, the case is
entirely consistent with Frank.131 And so, I suggest, is Mooney
128 It is in fact possible that under Arkansas practice the state court felt it
was precluded from examining the merits of the contention on appeal. See Water-
man & Overton, supra note I27, at I4.
129 The state court continued: "The trials were had according to law, the jury
was correctly charged . . . and the testimony is legally sufficient . . . . We can-

not, therefore, . . . assume that the trial was an empty ceremony . . .." I43
Ark. at I62, 220 S.W. at 3I0. Of course nobody asked that the court "assume"
that there had been no fair trial; the question was whether any unbiased forum
would make an inquiry into the question.
130 26I U.S. at 9I, 92. The entire thrust of the last paragraph of the opinion is
directed at the inadequacy of the process furnished by Arkansas for determining
whether a fair trial had been had.
131 The contrary view, that Moore held that the habeas court must auto-
matically redetermine the merits of the contention of mob domination, derives

from Mr. Justice Holmes' statement that, though corrective process may be so
adequate as to preclude habeas, "if the case is that the whole proceeding is a mask
- that counsel, jury and judge were swept to the fatal end by an irresistible wave
of public passion, and that the State Courts failed to correct the wrong, neither
perfection in the machinery for correction nor the possibility that the trial court
and counsel saw no other way of avoiding an immediate outbreak of the mob can
prevent this Court from securing to the petitioners their constitutional rights." Id.
at 9I. The crucial question is, of course, what the Justice meant by the phrases
v. Holohan7132which established the substantive principle that a

conviction procured through testimony known by the state au-
thorities to be perjured results in a deprivation of liberty without
"failed to correct the wrong" and "perfection in the machinery for correction." If
these are read to mean that if in fact there was mob domination and if the state
failed to correct the wrong by correctly finding that fact, and if perfection in'
the machinery for correction is irrelevant even if in this very case there was cor-
rective process in the sense of a full and fair inquiry, then Moore surely does
overrule Frank and compel the habeas court automatically to inquire into the
merits. But it is possible, and far more consistent with the rest of the opinion
(though admittedly not with the dissent in Frank), to read the phrases to mean
that "failure" to "correct the wrong" means not failure correctly to discover the
ultimate truth but failure to conduct a fair and rational inquiry with regard
to the facts; and that perfection in the "machinery" for correction is irrelevant if,
as in this very case, the machinery never went into operation: that is, if there
was no fair state inquiry into the merits of the federal claim.
Professor Reitz disagrees with my reading; he sees Moore as holding that the
habeas court "could not escape the duty of examining the facts for itself since, if
true, they entitled the applicants to relief. . . . Instead of being merely a second-
ary alternative to the state-court channel of litigation, federal habeas corpus was
thereby recognized as an available sequel. A case which had completely run the
state channel could then be taken into federal court for independent reexamination
of the federal contentions previously passed upon by the state courts." Reitz, The
Abortive State Proceeding, 74 HARV. L. REV. I3I5, 1329 (I96I).
Professor Reitz acknowledges that "the opinion of the Court does not discuss
the radical change being worked in the role of federal habeas corpus," but he as-
cribes the revolution assertedly wrought by Moore directly to the fact that in i9i6
Congress abolished appeal as of right to the Supreme Court from state decisions
involving federal questions and substituted a discretionary certiorari jurisdiction;
presumably he sees Moore radically extending habeas to provide automatic federal
review in view of its withdrawal from the Supreme Court. Id. at 1328-29. The
trouble with this analysis is that it is awfully hard to extrapolate it from Mr.
Justice Holmes' short and ambiguous little opinion, an opinion which in its lan-
guage - and on the facts of the case - does not have to be read as making a
"radical change" at all, but indeed can be viewed as entirely consistent with
previous law. And even if one can see Mr. Justice Holmes as having intended to
accomplish, sub silentio, the rather highhanded tour de force of expanding the
habeas jurisdiction to replace the Supreme Court's lost jurisdiction on writ of
error, it is hard to visualize judges as meticulously aware of the duty of the Court
to state what it is doing and the reasons therefor as Mr. Justice Brandeis and
Mr. Justice Van Devanter joining him.
I also find the notion that Mr. Justice Holmes intended in Moore to overrule
Frank and make habeas automatically available to redetermine federal constitu-
tional questions decided in state criminal cases hard to reconcile with the opinion
of the Court, two years after Moore, in Knewel v. Egan, 268 U.S. 442 (I925), which
Mr. Justice Holmes joined, and where the Court held:
It is the settled rule of this Court that habeas corpus calls in question only the
jurisdiction of the court whose judgment is challenged [citing, inter alia, Frank
v. Mangum].
A person convicted of crime by a judgment of a state court . . . may
proceed by writ of habeas corpus on constitutional grounds summarily to
determine whether he is restrained of his liberty by judgment of a court acting
without jurisdiction. . . . But . . . he may not use it as a substitute for
a writ of error. . ..
[T3he ultimate question presented is whether the procedure established by
the statutes of South Dakota . . . is a denial of a constitutional right. With
due process of law. Therefore, stated the Court in a habeas case

involving such an allegation, the state is "required"under Frank
and Moore to afford corrective process, and in its absence federal
habeas will be available.'33
The suggestion in Mooney that a state is "required"to afford
corrective process raises a problem to which I have referred
before,'34and which the Supreme Court did not really consider in
its elaboration of the reaches of habeas corpus in Frank v. Man-
gum and the subsequent cases. I have suggested that it would be
logically possible to have a system whereby a state's failure to
provide corrective process would itself be deemed "error" sub-
ject to reversal on direct review by the Supreme Court; that is,
the due process clause could be interpreted as imposing on the
states the affirmative duty to provide a full system of remedies
for the meaningful litigation of all federal questions relevant to
the case. Thus in Moore v. Dempsey the Supreme Court could
have held on direct review that the failure of the state courts to
conduct a full inquiry into the allegation of mob domination was
itself federal error which would lead to reversal of the conviction
rather than investigation of the allegation on habeas corpus. And
if Mooney is taken literally in the language that corrective process
is "required,"the result would be, not that the failure of the state
to afford a collateral postconviction forum would lead to federal
inquiry into the allegation of knowing use of false testimony, but
a reversal on direct review on the ground that the state is itself
required to furnish such a forum.'35
Of course there are situations where a state's refusal to furnish
process would clearly be considered simple error. I take it as
common ground that if a state court should hold that a defendant
in a state trial may not at any time question the admissibility
respect to that question, we hold . . . that the judgment of state courts in
criminal cases will not be reviewed on habeas corpus merely because some
right under the Constitution of the United States is alleged to have been
denied to the person convicted. The proper remedy is by writ of error.
Id. at 445-47. (Emphasis added.)
See also Mr. Justice Holmes' own opinion for the Court in Ashe v. United States
ex rel. Valotta, 270 U.S. 424 (1926).
132 294 U.S. I03 (I935).
133The actual holding of the Court in Mooney was that petitioner must seek
habeas corpus in the state courts before resorting to federal habeas, so that leave
to file the petition was denied, "but without prejudice." Id. at II5.
134
See pp. 459-60 supra.
135 There would be no procedural obstacle to the enforcement of such a
principle; if a prisoner sought state postconviction relief on the basis of a federal
constitutional claim not raisable at trial or appeal and the state failed to grant
relief on the ground that there is no remedy under state law, that very denial, if
final, would be a judgment subject to Supreme Court review.
of a confession on the ground that it was coerced, it would be

subject to reversal in the ordinarycourse (this would be a classical
example of an "inadequate"state ground), so that the principle
of Frank v. Mangum would not of itself justify a federal collateral
jurisdiction simply to furnish a forum for the litigation of that
question. At the other extreme, however, we encounter obvious
difficulties. If a state litigation has ended, and historically the
state has never provided a system of remediesfor raising questions
(federal or state) on postconvictionproceedings,and further given
the fact that statutory authorization for federal collateral pro-
ceedings does exist (the question would be far more acute if it
didn't) it is far from clear that the sound solution calls for the
rather sweeping supervision and molding of state procedures
required by a doctrine that every state must itself provide a full
system of remedies. Historically the question may never have
been explicitly canvassed at the time of Frank and Moore and
Mooney because state postconviction processes were so rudi-
mentary that the notion of deeming their very existence to be part
of "due" process would never seriously have presented itself.
But beyond that, it may be that there are political and institution-
al considerations which counsel conservatism in requiring the
states themselves always to furnish process. It is, after all, one
of the advantages of a dual system of courts that we can deal
with the abrasions possible in a federal system in a flexible man-
ner. Is there not political wisdom in using a federal collateral
jurisdiction as a "backstop" for inadequacies of state process,
rather than having the Supreme Court on direct review undertake
the task of forcing the states to fashion procedural systems ad-
equate to the reasoned litigation of all federal questions? The
federal habeas jurisdiction has the added convenience that the
very question of the adequacy of the state's process can be can-
vassed in a trial court which can conduct a hearing as to the facts
and otherwise clarify situations clouded by inadequate records
presented on direct review. And once the habeas court determines
that state corrective process has been inadequate, it is the less
abrasive (if the less heroic) path to proceed to a determination
of the federal claim (and, if it is discovered to be groundless, to
leave the state detention undisturbed) rather than to force the
state to undertake positive rectificationon its own.
In any event, the cases (without ever providing a rounded and
systematic analysis of the problem) have proceeded on the as-
sumption that a state's failure to provide corrective process will
often be left undisturbed on direct review and the federal ques-
tion determined on federal collateral review; 136 only occasionally

are there intimations, as in Mooney, that the state may itself be
"required"to afford corrective process.'37
E. From Mooney to Brown v. Allen

The decisions of the three decades following Frank, Moore, and
Mooney did not produce major changes in the general principles
of the habeas jurisdiction. We review first a trio of cases dealing
with federal prisoners, for these continued to be relied on in the
state prisoner cases. Johnson v. Zerbst 138 has been denominated
"pathbreaking."139 There the Court held that the sixth amend-
ment "withholds from federal courts, in all criminal proceedings,
the power and authority to deprive an accused of his life or liberty
unless he has or waives the assistance of counsel," 140 and ordered
136 See, e.g., White v. Ragen, 324 U.s. 760 (1945) (state failed to reach merits
of claims of denial of counsel and knowing use of perjured testimony; certiorari
dismissed with opinion stating that federal habeas is available if no corrective
process is afforded by the state) ; Woods v. Nierstheimer, 328 U.S. 2II, 2I7 (I946)
("if the State of Illinois should at all times deny all remedies . . . the federal
courts would be available to provide a remedy"). Compare cases such as New
York ex rel. Whitman v. Wilson, 3I8 U.S. 688 (I943); Loftus v. Illinois, 334 U.S.
804 (1948); and Jennings v. Illinois, 342 U.S. I04 (Ij9I). In all of these the
Court sought clarification from the state as to whether the refusal to afford state
postconviction relief was grounded on the merits of the federal question or the un-
availability of the remedy under state law. In the last-named case the Court made
clear that "where the state does not afford a remedy, a state prisoner may apply
for a writ of habeas corpus in the United States District Court . . .." Id. at iii.
Of course if a state court denies a hearing on a properly presented federal
question, not on the ground that the state lacks a remedy, but on the basis of an
erroneous holding as to the federal law, it is subject to reversal on direct review.
See Smith v. O'Grady, 3I2 U.S. 329 (I94I) (reversing a holding on state habeas
that an allegation of a coerced plea of guilty does not state a claim under the
fourteenth amendment); Williams v. Kaiser, 323 U.S. 47I (I945); Hawk v. Olson,
326 U.S. 27I (I945).
137 See Young v. Ragen, 337 U.S. 235 (I949); Woods v. Nierstheimer, 328
U.S. 2II See generally HART & WECHSLER, THE FEDERAL COURTS AND
(I946).
THE FEDERAL SYSTEM 395-99, 474-77, 5I2-I7 (I953); Note, State Enforcement of
Federally Created Rights, 73 HARV. L. REV. 55I (I960); Note, Effect of the Federal
Constitution in Requiring State Post-Conviction Remedies, 53 COLUM.L. REV.
II43 ('953).
The extent to which the Court on direct appeal will normally defer to the
state's remedial law in testing a federal question (leaving to habeas the question
whether the state's corrective remedies are adequate) is exemplified in Carter v.
Illinois, 329 U.S. I73 (I946). Here the Court on direct review limited its inquiry
(into the question whether counsel should have been provided) in accordance with
the state rule that review is on the so-called common law record; it stated that
facts not appearing in that record could be considered on federal habeas corpus.
138 304 U.S. 458 (I938) .
139 Hart, Foreword, 73 HARV. L. REV. 84, I04 (I959).
140 304 U.S. at 463.
the federal court on habeas corpus to determine on the merits

whether the defendant had made an effectual waiver of his right
to counsel. The reason was made plain in the very first sentence
of the Court's opinion dealing with the habeas question:
The purposeof the constitutionalguarantyof a rightto counselis
to protect an accused from convictionresulting from his own
ignoranceof his legal and constitutionalrights,and the guaranty
would be nullifiedby a determinationthat an accused'signorant
failureto claimhis rightsremovesthe protectionof the Constitu-
tion.141
The Court's language makes clear that in these counsel cases

the habeas court's jurisdictionmust necessarily turn on an adjudi-
cation of the merits. If it is found that the defendant did not, due
to ignorance, youth or the like, make an effective waiver of coun-
sel, surely the habeas court cannot hold that a previous determina-
tion that he has waived his right, made in circumstancessubject to
the very same proceduralflaw, is binding. Similarly,if the defend-
ant has unwittingly failed to seek counsel, the very same ignorance
will normally prevent him from testing the effectiveness of such
a "Cwaiver" on appeal, so that the theoretical right to appeal from
the judgment of conviction will rarely serve as adequate corrective
process with respect to this question. Habeas turns out to be, as
the Court insists, the only legal remedy for the safeguarding of
the underlying right.142 The same is true, as was held in Walker
v. Johnston 143 and Waley v. Johnston,'44of the allegation that the
defendant was coerced to plead guilty at trial. The whole point of
such coercion, if it "works," is to prevent all further inquiry at
trial and on appeal - the more successful it is, the less likely
that the original proceeding or direct appeal therefrom will pro-
141 Id. at 465. The Court in Johnson buttressed its holding with language to
the effect that failure to provide counsel ousted the trial court of "jurisdiction."
142
I do not mean to indicate that I would automatically disregard a previous
adjudication that a prisoner has waived his right to counsel merely because in that
adjudication itself the prisoner had no counsel: any such doctrine would really
undermine the very possibility of waiver. The problem of circularity must, it
seems to me, be solved by varying the scope of review. If the original record
makes a rather clear showing that waiver was intelligently based on fair choice,
and no reason is suggested why that record ought to be suspect, I would see no
reason why the court on habeas should proceed to a de novo inquiry: it could take
the facts as found in the original court. But if the record itself indicates great doubt
whether waiver could fairly be implied, I would think it unfair to take the matter
as concluded by a proceeding where the defendant did not have counsel. I see, in
other words, no escape from the need to make some judgment about the merits.
See pp. 458-59 and note 30 supra.
1433I2 U.S. 275 (I94I).
14 3I6 U.S. IOI (I942).
vide a fair forum for the litigation of any question whatever, much
less the issue of coercion itself. It is thus not surprising that the
Court held that such an allegation must be litigated in postconvic-
tion collateral proceedings if it is ever to be litigated. The basis
of the decisions was made clear in Waley, which finally dispensed
with the fiction of "jurisdiction" as applicable to this kind of case:
The issue here was appropriately raised by the habeas corpus
petition. The facts relied on are dehors the record and their effect
on the judgment was not open to consideration and review on
appeal. In such circumstances the use of the writ in the federal
courts to test the constitutional validity of a conviction for crime
is not restricted to those cases where the judgment of conviction
is void for want of jurisdiction of the trial court to render it. It
extends also to those exceptional cases where the conviction has
been in disregard of the constitutional rights of the accused, and
where the writ is the only effective means of preservinghis rights.145
The notion that the basic function of the habeas jurisdiction

is to afford a remedy where there is no other way to vindicate a
prisoner's constitutional rights is also emphasized in the cases
dealing with state prisoners. In the famous per curiam in Ex parte
Hawk,146 which formulated governing procedural rules for the
guidance of state prisoners, the Court made this quite explicit.
Stating that federal habeas may not be sought before the prisoner
has exhausted presently available state remedies, it continued:
Where the state courts have considered and adjudicated the
merits of . . . [the prisoner's] contentions, and this Court has
either reviewed or declined to review the state court's decision, a
federal court will not ordinarily re-examine upon writ of habeas
corpus the questions thus adjudicated. . . . But where resort to
state court remedies has failed to afford a full and fair adjudication
of the federal contentions raised, either because the state affords no
remedy . . . or because in the particular case the remedy afforded
by state law proves in practice unavailable or seriously inadequate
. . . a federal court should entertain his petition for habeas corpus,
else he would be remediless. In such a case he should proceed in the
145 Id. at 104-05. (Emphasis added.) See also Adams v. United States ex rel.
McCann, 3I7 U.S. 269 (I942), and United States ex rel. McCann v. Adams, 320
U.S. 220 (I943), which held habeas available to determine whether a defendant
may and did waive his right to trial by jury where special circumstances frustrated
the possibility of litigating these questions on direct review. The cases illustrate
the flexibility of the writ in assuring that a defendant should have a meaningful
opportunity to test his federal contentions at some point. Note how meticulously
the Court in the latter case made sure that in fact there had not previously been
a chance to raise the question whether there was waiver.
146 321 U.S. II4 (I944).
496 LAWREVIEW
IIARVARD [Vol. 76:441
federal district court before resorting to this Court by petition for
habeas corpus.'47
And in House v. Mayo,'48 where the Court held that federal

habeas will lie to determinewhether the state should have afforded
the petitioner counsel and whether a plea of guilty was coerced,
the opinion is meticulous in emphasizingthat these questions may
be reached only in the absence of state corrective remedies.149
Similarly, in White v. Ragen,150where the Court dismissed a peti-
tion for certiorarito the Illinois supreme court on the ground that
the judgment rested on an adequate state ground, namely, the
unavailability of state habeas corpus, the Court's opinion makes
clear that the availability of federal habeas turns on the failure
of the state itself to provide a fair process for the litigation of
federal claims; if a state does fail to provide a remedy, the pris-
oner may proceed without more (notably without having to seek
certiorari from the Supreme Court to review the decision denying
the remedy) to apply for federal habeas, because
the allegations of fact in the petitions [as to denial of counsel and
knowing use of perjured testimony] are sufficient to make out
prima facie cases of violation of these constitutional rights of
petitioners, sufficient to invoke corrective process in some court,
and in the federal district court if none is afforded by the state.'5'
I do not mean to give a picture of the law of this time which is

neater than it actually was. The Court did not, after Frank, give
any rounded consideration to the reaches and purposes of the
habeas jurisdiction. Most of the cases of the period are explicitly
concernednot with the problem of relitigation of federal questions
already canvassed in state courts, but with the complications cre-
147
Id. at ii8. The Court cited Moore v. Dempsey as an example of a case
where "the remedy afforded by state law proves in practice unavailable or seriously
inadequate."
148
324 U.S. 42 (I945).
"' The district court in House denied the writ on the ground that the federal
questions had been litigated in the state courts. The Supreme Court pointed out
that in fact all the state decisions were grounded not on the merits but on the un-
availability of state remedies, so that the case met the test of Waley, namely, that
habeas is the only effective means of preserving the prisoner's rights. The Court
added: "It is true that where a state court has considered and adjudicated the
merits of a petitioner's contentions, and this 'Court has either reviewed or declined
to review the state court's decision, a federal court will not ordinarily reexamine
upon writ of habeas corpus the questions thus adjudicated. ... But that rule is
inapplicable where, as here, the basis of the state court decision is that the par-
ticular remedy sought is not one allowed by state law ..." Id. at 48.
150
324 U.S. 760 (i945).
151 Id. at 764. See also Hawk v. Olson, 326 U.S. 27I, 276 (i945).
ated by the exhaustion doctrine and with the vexing question

whether a prisoner must seek direct Supreme Court review of a
state judgment as a condition of the right to seek habeas corpus.'52
Several times the Court stressed that technically res judicata does
not apply in habeas proceedings; 153 and it will have been noticed
that its dicta state that state-court adjudications of the merits of
federal questions reviewed or left undisturbed by the Supreme
Court will not "ordinarily" be redetermined on habeas.'54 And
there are some opinions which could be taken to intimate that
the writ automatically reaches the merits of all federal constitu-
tional questions.155Furthermore, there can be no doubt that, by
152See Ex parte Botwinski, 314 U.S. 586 (1942); Ex parte Davis, 3I7 U.S. 592
(1942); Ex parte Williams, 317 U.S. 604 (1943); Ex parte Abernathy, 320 U.S. 219
(I943); Ex parte Hawk, 321 U.S. 114 (I944); Marino v. Ragen, 332 U.S. 56I
(1947) (especially the concurring opinion of Mr. Justice Rutledge, id. at 563);
Dowd v. United States ex rel. Cook, 340 U.S. 206 (1951); Jenningsv. Illinois, 342
U.S. I04 (I951); Sweeney v. Woodall, 344 U.S. 86 (1952).
Problems as to the requirement of seeking certiorari were canvassed in White v.
Ragen, 324 U.S. 760 (I945); Wade v. Mayo, 334 U.S. 672 (1948); and Darr v.
Burford, 339 U.S. 200 (I950). The latter case disapproved the softening of the
requirement which had been announced in Wade.
It should again be mentioned that these requirements themselves become at
least partly unintelligible if considered to be merely timing devices, if securing
decision of the federal question in the state system and subjecting it to supervision
by the Supreme Court on direct review is seen as a mere prelude to collateral in-
quiry automatically available to redetermine the issue. True, the exhaustion doc-
trine does allow a state to correct its own errors before federal interference is
authorized; but is this a sufficient rationale? If the state makes a conscientious
effort to do so by providing full and fair corrective process, should the result be
ignored? And what is the point of forcing the litigant to apply for certiorari if
there is automatically to be federal review through habeas corpus in any event?
153See Salinger v. Loisel, 265 U.S. 224 (1924), stating that a refusal to dis-
charge on habeas is not, technically, res judicata, but that prior refusals to issue the
writ may be given weight in the sound discretion of the court. See also Darr v.
Burford, 339 U.S. 200, 214-15 (1950).
154 See the passages cited from Ex parte Hawk and House v. Mayo, p. 495 and
note I49 supra. The Court never explained what circumstances would be sufficiently
extraordinary to justify collateral inquiry even after the question has been decided
by the state courts and review declined or afforded by the Supreme Court; nothing
suggests that the Court was not referring to questions of jurisdiction traditionally
subject to relitigation on collateral attack.
"I Some of the language in Mr. Justice Reed's opaque opinion in Darr v.
Burford, 339 U.S. 200 (1950), suggests that he assumed that this was the function
of the writ.
In Wade v. Mayo, 334 U.S. 672 (1948), the Court held habeas available to test
the question whether the defendant had been denied the assistance of counsel at
trial in violation of the Constitution, and this in spite of the fact that the question
was litigated in a state habeas proceeding in which the defendant was in fact
represented. The Court's opinion makes clear, however, that at the time the state
habeas petition was denied it was wholly unclear whether the decision was grounded
on the merits of the federal question or on a holding that state habeas was un-
available in view of the defendant's failure to appeal from his conviction. (The
498 HARVARD LAW REVIEW [VoI. 76:441
I952, the integrity and continuing authority of the doctrine of

Frank v. Mangum had been endangered, as it were, on several
occasions. They were endangered first by the ambiguities intro-
duced by the Moore case, which had never been explained or
clarified. They were endangered further by the abandonment,in
Waley, of the language of "jurisdictional"error; psychologically,
at least, this may have served as an invitation to further widen-
ings of the writ. Finally, we should note that the Court never
not even in the course of its procedurelesson to state prisoners in
Ex parte Hawk - strengthened the Frank doctrine by requiring
prisoners on habeas to plead explicitly that the state's corrective
processes had been inadequate to the task of protecting their
federal rights.158
On the other hand, what is equally clear is that at no time did
the Court hold that the habeas jurisdiction is appropriately exer-
cised in cases where a federal question has been fully considered
by the state, where there has been adequate corrective process.
And the essential purpose of the writ as affordinga forum where
state supreme court wrote no opinion.) If the state denial was based on the latter
ground, under Johnson v. Zerbst it would seem clear that federal habeas would be
available since the uncounseled defendant's theoretical right to appeal on the issue
of denial of counsel is plainly not adequate corrective process. Further, although
later Florida cases were deemed by Mr. Justice Murphy to show that under
Florida law the state habeas case must have been decided on the merits, the am-
biguity of the decision at the time it was rendered was held to excuse the prisoner
from the requirement of making a probably fruitless attempt to seek direct Su-
preme Court review of the state habeas decision. But the same ambiguity may
arguably be seen as rendering the state corrective process inadequate under Frank
v. Mangum, because in effect it precluded direct Supreme Court review of the
state holding as to the federal question. (This problem, that is, the bearing on the
"4adequacy" of state corrective process of ambiguous state decisions which effectively
foreclose the possibility of direct Supreme Court supervision, is discussed in detail
at pp. 5I9-2I infra.)
Dicta in Jennings v. Illinois, 342 U.S. I04 (i95i), can also be read as indicating
that habeas corpus is automatically available to redetermine a federal question
(here the admissibility of a confession) after it has been litigated in the state sys-
tem, without reference to the adequacy of the state's corrective process. But here,
too, the "adequacy" of state corrective process with respect to this question may
be doubted in view of the fact that defendant's indigence made it impossible for
him under Illinois law to raise the question on appeal. Nevertheless, it would
seem that, if the state denies to indigents corrective process equally as effective as
it grants to those with funds, under Frank v. Mangum the correct result may be,
not to make federal habeas available to the indigent but, on direct review, to deny
the state the power to make such a distinction. Eventually this is the result
which was reached in Griffin v. Illinois, 35I U.S. I2 (I956).
158 It may be significant that the author of one of the opinions of the Court
in Brown v. Allen, Mr. Justice Frankfurter, in his dissent in Jennings v. Illinois,
supra note I55, stressed the requirement of such a pleading (see note I70 infra);
he may have regarded the failure of the Court to accept this in Jennings as a
justification for Brown.
the state fails to provide such process was, as we have seen,

stressed again and again. There can be no doubt that when Brown
v. Allen reached the Court in I952, the central thrust of the law
was as Judge Learned Hand described it: 157 for purposes of ha-
beas corpus a detention was not to be deemed "unlawful"if based
upon the judgment of a competent state court which had afforded
full corrective process for the litigation of questions touching on
federal rights.
III. Brown v. Allen AND ITS JUSTIFICATIONS
A. The Decision and Its Problems

Brown v. Allen 158 reached the SupremeCourt in the I95I Term
and was decided in 1953. It involved three proceedings: Brown's,
Speller's, and Daniels'; only the first two need concern us here.
Brown was convicted of rape by a North Carolina court and sen-
tenced to death. In his appeal to the state supreme court he
claimed his conviction violated the federal constitution because
of the admission of a coerced confession and racial discrimination
in the selection of grand and petit juries. These issues had been
fully litigated, with the aid of counsel, in the trial court through
procedures not themselves alleged to have been in any way un-
fair. The state supremecourt affirmedthe conviction, rejecting the
defendant's federal contentions on the merits in a reasoned opin-
ion.159 Certiorari was denied.160 The prisoner then sought fed-
eral habeas corpus. The district court denied the writ without
holding a hearing, and the court of appeals affirmed.161
Speller, also a capital case, likewise involved a claim of racial
discrimination in jury selections which had been fully litigated
in the state system. (Indeed, the state courts had twice pre-
viously reversed Speller's convictions, first on the ground that
such discriminationwas shown, and then after retrial because his
counsel had not been furnished adequate opportunity to test the
question; 162 this was, therefore, Speller's third trial.) The state
supreme court rejected the allegation on the merits, and certiorari
was denied.163 Now Speller, too, sought habeas. The district
157
See pp. 464-65 supra.
158
344 U.S. 443 (I953).
159 State v. Brown, 233 N.C. 202, 63 S.E.2d 99 (i95i).
160 34I U.S. 943 (I 95I ).
161
I92 F.2d 477 (4th Cir.), affirming 98 F. Supp. 866 (E.D.N.C. I95I).
162 State v. Speller, 229 N.C. 67, 47 S.E.2d 537 (1948); 230 N.C. 345, 53
S.E.2d 294 (949).
163
23i N.C. 549, 57 S.E.2d 759, cert. denied, 340 U.S. 835 (5950).
court took testimony on the discrimination issue in addition to

considering the state record, and denied the writ. Again, the
court of appeals affirmed.164
The SupremeCourt granted certiorariin both cases and affirmed
the convictions. It did so, however, not on the basis of Frank v.
Mangum, that the state had provided adequate corrective process,
but by reaching and rejecting on the merits the federal claims
presented which had been previously adjudicated by the state
courts. The Court did so without any explicit discussion of the
question of jurisdiction or. any apparent understanding of how
radical this step was: with only Mr. Justice Jackson disagreeing,
eight of nine Justices assumed that on habeas corpus federal dis-
trict courts must provide review of the merits of constitutional
claims fully litigated in the state-court system. As Professor
Hart says, the decision thus "manifestly broke new ground." It
seems to say
that dueprocessof lawin the caseof stateprisonersis not primarily
concernedwith the adequacyof the state'scorrectiveprocessor of
the prisoner'spersonalopportunityto avail himselfof this process
. . .but relates essentiallyto the avoidancein the end of any
underlyingconstitutionalerror....165
And ever since Brown v. Allen the Supreme Court has continued
to assume, without discussion, that it is the purpose of the federal
habeas corpus jurisdiction to redetermine the merits of federal
constitutional questions decided in state criminal proceedings.166
What is the basis of Brown v. Allen? The opinions do not cast
much light on that question. Mr. Justice Frankfurter, in one of
the two opinions of the Court in the case, repeatedly asserts,
without discussion, that the act of i867 compels the conclusion
that state consideration "cannot foreclose" federal determination
of the merits of constitutionalclaims; 167 he says that as to all legal
issues (such as the admissibility of a confession) the district
judge "must" exercise his own judgment,'68that Congress has
164
192 F.2d 477 (4th Cir.), affirming 99 F. Supp. 92 (E.D.N.C. I95I).
165
Hart, Foreword, 73 HARV.L. REV. 84, io6 (I959).
166
See, e.g., Irvin v. Dowd, 366 U,S. 7I7 (I96I); United States ex rel.
Jennings v. Ragen, 358 U.S. 276 (I959); Cicenia v. Lagay, 357 U.S. 504 (1958);
Thomas v. Arizona, 356 U.S. 390 (1958); Leyra v. Denno, 347 U.S. 556 (I954).
167 344 U.S. at 500 ("else the State court would have the final say which
the Congress, by the Act of I867, provided it should not have").

168Id. at 507. The Justice continues:
Although there is no need for the federal judge, if he could, to shut his eyes
to the State consideration of such issues [of federal law], no binding weight
is to be attached to the State determination. The congressional requirement
is greater. The State court cannot have the last say when it, though on
"seen fit" to "give to the lower federal courts power to inquire into
federal claims, by way of habeas corpus," and that it would be
"inadmissible to deny the use of the writ merely because a State
court has passed on a federal constitutional issue." 169 But these
are mere statements of conclusion; they ignore completely that
such a purpose cannot be derived automatically from either the
language or history of the I867 act, and that the act was not so
understood in the first eighty years of its history; they ignore,
too, the explicit ruling in In re Wood 170 that at least the question
of jury discrimination in violation of the Constitution is not a
question open on collateral attack if opportunity to litigate it was
afforded in the state system and on direct"review.
Oddly enough, much of the discussion in the two majority opin-
ions in Brown deals not with the question why a state's adjudica-
tion of the law should be disregarded, but why the state's ad-
judication of the facts should not necessarily be disregarded. Mr.
Justice Reed states that no new hearing as to the facts is necessary
if the district court is satisfied that the state process has given
fair consideration and what procedurally may be deemed fairness, may
have misconceived a federal constitutional right.
Id. at 5o8.
169 Id. at 508-09, 5I3. Mr. Justice Frankfurter asserts that the rule (as he
states it in Brown) is less liberal than the English practice; under the federal
statute (28 U.S.C. ? 2244 (1958) ) a judge may take into account previous
refusals to issue the writ, whereas in England the prisoner may, in the famous
phrase, "go from judge to judge" repeatedly, any one decision to discharge being
final. 344 U.S. at 509. Professor Pollak, too, sees the federal writ as narrow
compared to the "extraordinary liberality" of the English practice. Pollak, Pro-
posals To Curtail Federal Habeas Corpus for State Prisoners: Collateral Attack on
the Great Writ, 66 YALE L.J. 50, 65 (I956). But this ignores the crucial fact that,
though it is true that in England if a prisoner has a claim cognizable on habeas
corpus, he may then make repeated applications, the category of such cognizable
claims is extremely narrow compared to our law as defined in Brown v. Allen.
It is thus misleading to speak of the English practice as more liberal than ours
in the context of the rule of Brown v. Allen. (In fact, the right under English
law to go from judge to judge was itself later somewhat curtailed by the judgment
in In re Hastings (No. 2)', [I959] i Q.B. 358 (1958), and substantially restricted
by the Administration of Justice Act, I960, 8 & 9 Eliz. 2, c. 65, ? I4(2).)
170 I40 U.S. 278 (I89I); see pp. 481-82 supra.
Mr. Justice Frankfurter's opinion in Brown is particularly surprising in light
of his own dissent, two years before, in Jennings v. Illinois, 342 U.S. I04, 112
(I95I). There the Justice had argued that a writ of certiorari to the Supreme
Court of Illinois should be dismissed for want of a properly presented federal
question:
It is true that petitioners allege they were convicted on the basis of
coerced confessions . . . in violation of the Fourteenth Amendment. But so
far as appears from the record, these issues were fully litigated and determined
at the trials. Until the cases came to this Court, no showing was made, or
sought to be made, that circumstances were such as to warrant a new and
independent inquiry into those determinations as a matter of federal right.
Id. at II5.
them "fair consideration"and resulted in a "satisfactory conclu-

sion," if no "unusual circumstances" are present and the record
"affords an adequate opportunity" to test the merits of the
claim.'7' The taking of new evidence by the district court is "in
its discretion."172 Mr. Justice Frankfurter says that as to the
facts, the state's conclusions "may" be accepted by the district
judge "unless a vital flaw be found in the process of ascertaining
such facts." 173 These ambiguous formulations have created vex-
ing problems as to the scope of review on habeas corpus which I
need not canvass here.'74 I do suggest, however, that the basic
reason why the courts have had such difficulties in defining the
scope of review on habeas is that the Court in Brown did not pro-
vide a principled rationalization of the purpose being served by
affording the federal court the right to review the determination
of the state court in the first instance. If it is the purpose of the
habeas jurisdiction to assure that no error has been made, then
there is no reason why the state courts' determinations of fact
should be any more sacred than their conclusions of law. In view
of the function of appellate courts to make pronouncements of
law, it is manifestly sensible to restrict their review to issues of
law. But district court jurisdiction on habeas certainly does not
serve that purpose; and if the purpose is to assure "correct"
determinations,that purpose should not be disregardedwhen the
allegation is that the state court has erred in finding the facts
bearing on a constitutional claim. On the other hand, if meaning-
ful process serves as an adequate guarantee of the probability of
the correctnessof factfindings,we are entitled to some explanation
why it does not satisfy us with respect to legal conclusions.'75
171 344 U.S. at 463.
172Id. at 478.
173 Id. at 506.
174 The problems of the scope of review of

questions of fact under the rule
of Brown v. Allen is the subject of a comprehensive and wholly admirable paper
written for my seminar in Remedies Against the Government by Mr. Richard
Posner, LL.B. I962, entitled The Trial of Fact De Novo in Federal Habeas
Corpus Proceedings, April i962, on file in the Harvard Law School Library. See
also Note, 53 Nw. U.L. REv. 765 (I959); Note, 68 YALEL.J. 98 (I958).
175 The point is really implicit in one of Mr. Justice Reed's formulations in
Brown:
As the state and federal courts have the same responsibilities to protect per-
sons from violation of their constitutional rights, we conclude that a federal
district court may decline, without a rehearing of the facts, to award a writ
of habeas corpus to a state prisoner where the legality of such detention
has been determined, on the facts presented, by the highest state court with
jurisdiction, whether through affirmance of the judgment on appeal or
denial of postconviction remedies.
344 U.S. at 465. On its face this sounds like an argument for Frank v. Mangum.
There is another aspect of this question which should be no-

ticed. The Court in Brown made much of the notion that the
integrity of a state's process with respect to factfinding should
be respected. But it made clear that this is solely a matter of dis-
cretion, that the district judge may redetermine the facts even
where there is no "unusual circumstance" or "vital flaw" which
casts doubt on the adequacy of those findings; and the Court went
out of its way to reaffirm this grant of discretion a few years
later.'76 But if the federal judge has an impeccable record before
him, that is, he has no reason to believe that the factfinding proc-
esses of the state courts were in any way inadequate to the task
at hand, on what principle should he decide whether to exercise
such a discretion? Note that if he does exercise it, he would seem
to be completely free to disregard the state-court findings even
though the issue may turn on an assessment of the veracity of
witnesses; in such a case we have the startling result that a state
prisoner is deemed to be held in violation of the Constitution of
the United States because a state judge believed the prosecution's
witnesses and the federal judge believes those of the defendant.
Nor do the opinions in Brown v. Allen deal adequately with the
grave problems of federalism created by the doctrine of that case.
It is fashionable today to dismiss the resentments created in the
states by the existence of an indiscriminate federal habeas juris-
diction; we are told that complaints about intrusion by federal
habeas courts into state criminal processes are disingenuous, di-
rected not at the remedy but at the substantive due process doc-
trines enforced thereby; 177 we are further assured that such
Yet it is ignored in Mr. Justice Reed's own exhaustive review of the merits
of the state courts' legal conclusions.
176 In United States ex rel. Rogers v. Cummings, 154 F. Supp. 663 (D. Conn.
I956), the district judge on habeas took testimony with respect to the volun-
tariness of a confession admitted into evidence in a state criminal trial. The
court of appeals reversed, holding that the district court could redetermine the
facts only if the state proceeding was tainted by a "vital flaw." United States
ex rel. Rogers v. Richmond, 252 F.2d 807 (2d Cir. 1958). The Supreme Court
denied certiorari but took the unusual step of accompanying the denial with an
opinion, which read as follows:
We read the opinion of the Court of Appeals as holding that while the
District Judge may, unless he finds a vital flaw in the State Court pro-
ceedings, accept the determination in such proceedings, he need not deem
such determination binding, and may take testimony.
357 U.S. 220 (I958). For devastating and wholly justified comment on this
mode of proceeding, see Brown, Foreword: Process of Law, The Supreme Court,
1957 Term, 72 HARV. L. REV. 77, 92-93 (I958). For the eventual disposition of
the case, see Rogers v. Richmond, 365 U;S. 534 (I96I) (see pp. 5I4-I6 Mnfra).
177 See, e.g., Reitz, Federal Habeas Corpus, io8 U. PA. L. REV. 46I, 5I6 (I960);
Pollak, supra note I69, at 66 ("The articulate premises of the attack on the
complaints are in any event beside the mark since very few pris-
oners are actually released by the federal courts.'78 But the very
unanimity of the resentmentamong state law-enforcementofficials
and judges,179 many of them, surely, as conscientious in their
adherence to the Constitution and as intellectually honest as their
critics, counsels, not against the jurisdiction, but against its in-
discriminate expansion without principled justification. Note
further that it was not state officials but the Judicial Conference
of the United States, headed by Chief Justice Warren, which in
I955 adopted a report of a committee consisting of Circuit
Judges Parker, Phillips and Stephens and District Judges Hooper,
Vaught and Wyzanski, which stated that the expansion of the
habeas jurisdiction has
greatly interferedwith the procedureof the State courts,delay-
ing in many cases the proper enforcementof their judgments.
Whereadequateprocedureis providedby State law for the han-
dling of such matters,it is clearthat the remedyshouldbe sought
in the State courts with any review only by the Supreme
Court
and recommended, inter alia, statutory preclusion of the writ
where a federal question was "raisedand determined"in the state-
court system.180
habeas corpus writ are grounded in the supposed first principles of judicial
finality and of the sound management of a federal union. Just below the surface,
however, lurks the less plainly articulated but perhaps more deeply felt belief
that the Supreme Court has grievously erred in the sequence of great cases
which utilized habeas corpus to probe trial records for fundamental error.").
178 See Brennan, Federal Habeas Corpus and State Prisoners: An Exercise
in Federalism, 7 UTAH L. REV. 423, 440-41 (I96I); Frankfurter, J., in Brown

v. Allen, 344 U.S. 443, 5IO (I953) ("reliable figures . . . showing that during
the last four years five State prisoners, all told, were discharged by federal
district courts, prove beyond peradventure that it is a baseless fear, a bogeyman,
to worry lest State convictions be upset by allowing district courts to entertain
applications"); Statement of Thurgood Marshall, Hearings on H.R. 564g Before
Subcommittee No. 3 of the House Committee on the Judiciary, 84th Cong., ist
Sess., ser. 6, at 8o (I955) [hereinafter cited as Hearings].
179 See Hearings, in which are recorded not only the statements but also the
major documents expressive of such resentment, including resolutions adopted by
the Conference of Chief Justices, reports to such Conference by its habeas corpus
committees, and resolutions adopted by and statements in behalf of the National
Association of Attorneys General. The attorneys general of not less than forty-
one states joined in the brief attacking the constitutionality of the Habeas Corpus
Act of I867 (surely a hapless cause), rejected by the Court of Appeals for the
Third Circuit in United States ex rel. Elliott v. Hendricks, 2I3 F.2d 922 (3d Cir.),
cert. denied, 348 U.S. 85I (1954). As far as I know the only voice raised in favor
of the present habeas jurisdiction among state judges has been that of Justice
Schaefer, in Federalism and State Criminal Procedure, 70 HARv.L. REV. i (I956).
180 See Hearings 89,
go. The proposed statute (H.R. 5649, 84th Cong., ist
The point is not that it is unseemly for a federal district judge

to reverse the action of the highest court of the state, but that
it is unseemly for him to do so without principled institutional
justification for his power. This justification was simply not
provided by the opinions in Brown v. Allen. Mr. Justice Frank-
furter did attempt to deal with the problem by stating that
insofaras this [habeas]jurisdictionenablesfederaldistrictcourts
to entertainclaims that State SupremeCourtshave deniedrights
guaranteedby the United States Constitution,it is not a case of
a lower courtsitting in judgmenton a highercourt. It is merely
one aspectof respectingthe SupremacyClauseof the Constitution
wherebyfederallaw is higherthan State law.181
But this does not meet the issue. Of course federal law is higher
than state law. But that does not automatically tell us that it
is better for federal judges to pronounce it than state judges,
much less that once a state judge has done so on a fair and
rational investigation, this should be disregarded and done over
again by a federal judge. Nothing about the substantive superior-
ity of federal law tells us why, in Leyra v. Denno,'82the admissi-
bility of a confession should have been redeterminedby a federal
court after the question was exhaustively canvassed pursuant to
the applicable federal law by the state courts in as conscientious
a manner as possible and opportunity had on a full and adequate
record to secure review in the United States Supreme Court.
The problem of federalism created by Brown v. Allen should
not be seen in terms of the possible irritation of state judges at
Sess. (I955)) received the unanimous support of the Judicial Conference of the
United States, the Conference of Chief Justices, the Association of Attorneys
General, and the section of judicial administration of the ABA; it was also
supported by the Justice Department. See id. at 7. See also Clark, J., concurring
in United States ex rel. Caminito v. Murphy, 222 F.2d 698, 706 (2d Cir. I955):
I doubt the utility for dignified or effective law enforcement of review and
overturn by any federal judge of the reasoned conclusions reached by a
whole hierarchy of state tribunals. . . . [0] ur sphere of superintendence
should not extend to state police activities; there the state courts should have
the burden, subject only to certiorari by the Supreme Court in the few
cases where needed. Consequently the pending legislation to that end .
seems wise policy.
The bill was passed by the House in both the 84th and 85th 'Congresses (see 102
CONG. REC. 940 (1956); I04 CONG. REC. 4675 (I958)), but died both times in
the Senate. I might say that I do not mean to intimate approval of this particu-
lar bill; in fact it was so drafted that it would probably have created more
problems than it solved. The point is only that the wide support the bill gained
is expressive of the very real and profound concern which the present habeas
jurisdiction arouses.
181344 U.S. at 5IO.
182
347 U.S. 556 (I954); see pp. 449-50 supra.
being reversed by federal district judges. The crucial issue is

the possible damage done to the inner sense of responsibility, to
the pride and conscientiousness, of a state judge in doing what
is, after all, under the constitutional scheme a part of his busi-
ness: the decision of federal questions properly raised in state
litigation. And the problem must be further analyzed in terms
of its effect on the integrity and effectiveness of the substantive
criminal law of the states. Certainly the complaints on the part
of law-enforcement officers that this effect is harmful are vocif-
erous. Some of these complaints may be tendentious or even
intellectually dishonest. But the contention- that endless and
inordinate delays in the imposition of criminal sanctions, brought
about by the indiscriminate reopening of cases on collateral
attack, have interfered intolerably with effective enforcement of
the criminal law - cannot strike any of us living with the mem-
ory of the Chessman case as absurd. And thus the very existence
of the complaints should, at least, give us pause; the burden of
proof, as it were, should be on the proponent of the proposition
that Brown v. Allen constituted a justified expansion of the
habeas jurisdiction.
Finally, the doctrine of Brown v. Allen must be assessed in
light of the strains put on the federal judicial system itself by the
ever increasing flood of habeas petitions from state prisoners.183
It is, of course, notorious that most of these petitions are frivo-
lous. And as Justice Schaefer has said, that these "have depreci-
ated the writ of habeas corpus cannot be doubted." 184 I have
suggested before that this matter should be seen not only in
terms of time and money but in terms of husbanding the intel-
lectual and moral energies and intensities of our judges. Let us
remember Justice Jackson's charge:
[T]his Courthas sanctionedprogressivetrivializationof the writ
until floodsof stale, frivolousand repetitiouspetitionsinundatethe
docket of the lower courts and swell our own. . . . It must
prejudicethe occasionalmeritoriousapplicationto be buriedin a
flood of worthlessones. He who must search a haystackfor a
needleis likely to end up with the attitudethat the needleis not
worththe search.'85
183
In fiscal 1950 there were 56o habeas petitions filed by state prisoners in
the federal district courts. In I955 there were 66o; in I958, 755; in I960, 87I;
and in I96I, 906. DiR. ADM. OFF. U.S. COURTS ANN. REP. ii6 (I960); id. at
i65 (I96I).
184
Schaefer, supra note I79, at 25.
185 Brown v. Allen, 344 U.S. 443, 536-37 (I953) (concurring in judgments of
affirmance but dissenting with respect to habeas jurisdiction).
I do not mean to suggest that the flooding of district court

dockets with frivolous habeas cases would be cured if the doc-
trine of Brown v. Allen were overthrown. But the facts compel
caution. We should not encouragethe flow of petitions by expand-
ing the jurisdiction unless there is a felt need for such expansion.
B. The Right to a Federal Forum

Professor Hart sees as a possible justification for Brown v.
Allen "the principle that a state prisoner ought to have an oppor-
tunity for a hearing on a federal constitutional claim in a federal
constitutional court, and that, if the Supreme Court in its dis-
cretion denies this opportunity on petition for certiorari, it ought
to be available on habeas corpus in a federal district court." 186
Professor Reitz assumes that such a "right" exists,187and Mr.
(now Judge) Thurgood Marshall, testifying about habeas before
the Congress, claimed it in extreme terms:
I do not see how any State can claim the right to determine Federal
questions. Federal questions should be determined by the Federal
judiciary . . 188
The proposition is not easy to assess. Surely it is plain that

there exists no constitutional right to have the merits of a federal
question determined by a federal constitutional court; this would
seem to be implicit in the power of the Congress over the appellate
jurisdiction of the Supreme Court and over the very existence
of lower federal courts.189 In civil cases, and in criminal cases
which do not eventuate in detention, it is accepted without ques-
tion that state courts exercising original jurisdiction have full
authority to pass on federal questions arising therein, subject
only to the discretionary certiorari jurisdiction of the Court; 190
if no such right is thought to be appropriate even for constitu-
186
Hart, Foreword, 73 HARV. L. REV. 84, io6-07 (I959)-
187
Reitz, Federal Habeas Corpus, IO8 U. PA. L. REv. 46I, 464 (I960): "[T]he
ultimate right of a state criminal defendant to have a federal court and federal
judges pass upon his federal contentions cannot be secured solely by the Supreme
Court's power to review state judgments. The sheer volume of cases is enough
to preclude it."
188 Hearings 85.
189 U.S. CONST. art. III. See HART & WECHSLER, THE FEDERAL COURTS AND
THE FEDERAL SYSTEM 3I2-40 (1953).

190 Of course the fact that civil cases may in certain instances be removed
entirely to federal courts does not qualify the point that if the state court is left
to adjudicate the case, it is allowed to adjudicate all the questions, federal and
state, in it, and the litigants have no right to a federal-court disposition of the
merits of the federal issues therein.
tional questions in such cases, why should there be one for

cases involving imprisonment? Of course it might be argued that
constitutional rights in criminal cases have a particular sanctity
and importance so as to deserve consideration as of right by a
federal court. But again we must be cautious in speaking about
these "rights." After all, the "existence" of the right often turns
on the narrowest kind of difference arising between judges in
highly particularistic assessments of evidence or in judgments
as to the proper application of the law to the evidence, and it is
hard to see the result as automatically crucial to justice merely
because that difference in opinion is formulated as a holding as
to constitutional rights. Take Leyra's case, again, as an example:
the admissibility of the confessions turned on a subtle, difficult
and closely balanced assessment of a highly peculiar set of psy-
chological facts and their legal significance. That the result
eventuates, formally, as a ruling of constitutional law about
constitutional rights does not logically tell us anything about the
underlying importance of the actual issue canvassed.19' And to
revert to the instance previously given, if a federal judge releases
a prisoner because his opinion about the credibility of a witness
differs from that of the state judge, that episodic difference of
opinion cannot be sanctified into a great issue merely because the
result is characterizedas a ruling as to constitutional rights. Liti-
gation about constitutional rights may raise issues of peculiar
sensitivity and importance, but will not necessarily do so. The
issue may involve simply the application of well-settled and well-
understood principles to a highly particular and closely balanced
set of facts with no further elaboration of the principles them-
selves being involved, and review may consequently consist of
nothing more than second-guessingthe ultimate leap of judgment
involved. On the other hand, if the issue in the particular case
really is important,it is the very purpose of the certiorarijurisdic-
tion to provide direct Supreme Court review.
In fact the result of the rather wooden differentiationwe now
91 Leyra may fruitfully be compared with a case such as !Sunal v. Large, 332
U.S. I74 (I947). Formally, the former involved an issue of constitutional law
bearing on a constitutional "right." Formally, the latter did not. But I cannot
persuade myself that the issue in Leyra-whether in fact coercion of a prior
confession "infected" later ones-was anywhere near as important, both with
respect to the general principles involved and with respect to whether justice
was done in the particular case, as the issue in Sunal, where the question was
whether in a prosecution for failure to submit to the draft the defendant should
have been permitted to show that he was unlawfully denied exemption as a
minister of relgion,
make between constitutional and nonconstitutional questions is

not without its ironies. Why is it, for instance, that we go so
far to allow relitigation of constitutional questions (even where
the particular issue is closely balanced and technical) and yet
do not allow any relitigation of the fundamental question of the
factual guilt or innocence of the accused? If a state prisoner
claims that he confessed after he was interrogated for six hours,
not (as the state court found) for four, the law says he may
relitigate the issue and, perhaps, gain release as a consequence,
even though the evidence of guilt may be overwhelming. But if
a defendant is convicted of murder and ten years later another
person confesses to the crime, so that we can be absolutely certain
that the defendant was innocent all the time, the law says that
he must rely on executive clemency. Why? Why should we pay
so little attention to finality with respect to constitutional ques-
tions when, in general, the law is so unbending with respect to
other questions which, nevertheless, may bear as crucially on
justice as any constitutional issue in the case?
In any event, even if we assume that issues bearing on consti-
tutional rights are necessarily "important," this does not auto-
matically validate the claim to a federal forum. Why should a
federal court hear cases even if they do involve these important
rights? The answer is, of course, clear where state courts will
not - and are not compelled to - hear them at all or under
fair circumstances. But what if a state court has done so? Is
there any sense in which the federal courts will, in the abstract, be
more '"correct"with respect to issues of federal law than state
courts? Surely not. There is no intrinsic reason why the fact
that a man is a federal judge should make him more competent,
or conscientious, or learned with respect to the applicable federal
law than his neighbor in the state courthouse. The federal judge
is more "correct" under the present system only because our
institutional arrangements make him authoritative. As Justice
Jackson pointed out about the Supreme Court itself:
[R]eversal by a higher court is not proof that justice is thereby
better done. There is no doubt that if there were a super-Supreme
Court, a substantial proportion of our reversals of state courts
would also be reversed. We are not final because we are infallible,
but we are infallible only because we are final.'92
The Justice's language reminds us that no institutional arrange-

ment, no court, can guarantee a result "correct" in an ultimate
192 Brown v. Allen, 344 U.S. 443, 540 (i953).
5IO HARVARD LAW REVIEW [Vol. 76:44I
sense. Why, then, is a fair state-court adjudicationnot a sufficient

guarantee (always subject, of course, to the reviewing power
of the Supreme Court, made necessary by the need for uniform,
authoritative pronouncementsof federal law) ?
I do not wish to overstate the argument. Important values may
be served by having federal judges pass on federal issues. Even
in a very general sense a federal judge, operating within a dif-
ferent system and with a differently defined set of institutional
responsibilities, may bring to bear on such issues an objectivity,
a freshness and insight which may have been denied to the state
judge, no matter how conscientious, whose perspective will be
subtly shaped by implicit assumptions derived from his responsi-
bilities within the state institutional framework, who stands
within that system.'93 More particular considerations may be
mentioned too. The federal judge is independent by constitu-
tional guarantee; the state judge may not be. The difference
surely does bear on conditions necessary for principled judging;
it is, at least, a common assumption-perhaps implicit in the
Constitution itself - that state courts may be more responsive
to local pressures, local prejudices, local politics, than federal
judges.'94 And there is, too, the fear that state officials, including
judges, will somehowbe less sympathetic or generouswith respect
to federal claims raised by state prisoners than federal judges.
There are, in other words, perfectly sound and honest con-
siderations which do buttress the claim that justice will be
better served if a federal constitutional judge is allowed to pass
on the merits of federal claims arising in state criminal proceed-
ings. Yet the claim, stated so largely, does not allay doubts.
From the beginning it was one of the central features of our
federalism that federal law is a part of the state law, that deciding
193 See Schaefer, supra note I 79, at 7:
The Supreme Court's position at the summit also gives it a different
perspective from that of state courts. The horizons of the state court will
be relatively limited, whereas the Supreme Court can more effectively con-
trast the local procedures, perhaps defined by statute, with practice else-
where throughout the country. The "insulated chambers afforded by the
several States" are sometimes an advantage. But they may be too well
insulated. Someone once wisely said that the basic trouble with judges is
not that they are incompetent or venal beyond other men; it is just that
they get used to it. And it is easy indeed to get used to a particular procedural
system. What is familiar tends to become what is right.
Though directed to review by the Supreme Court, justice Schaefer's remarks
are, to a more limited extent, applicable to collateral review by federal district
courts.
194 It was, in part, this assumption that led the framers to authorize Congress
to create federal courts of original jurisdiction with cognizance over federal-

question and diversity cases. See THE FEDERALISTNos. 8o & 8i (Hamilton).
i963] FINALITY IN CRIMINAL LAW 5II
federal questions is an intrinsic part of the business of state

judges.195 (This is, of course, particularly true - as Mr. Justice
Story saw in Martin v. Hunter's Lessee 196 - in that federal
questions often come up by way of defense to claims arising
purely under state law.) Implicit in the whole structure is the
need for confidence that the state courts will conscientiously
apply to a case the whole of the applicable law, including the
federal law. One tends, therefore, to resist the notion that our
procedural system should be built on the opposite premise, that
state-court judgments about federal rights bearing on state cases
should be automatically ignored on the basis of rather nebulous
and open-ended assumptions about their inadequacy. We must
always remember that it is common ground that where there
are, in a particular case, allegations with respect to the fairness
of the state's process in trying the federal questions, it is clearly
open to the federal habeas judge to inquire into that issue. In
other words, we already have accorded to federal judges a large
power to supervise the fairness of the methods by which the
state adjudicates claims of federal right. The issue, then, is the
narrow one whether we should go beyond this, whether the federal
court should redetermine the facts and the law in cases where
there is no reason to suspect failure on the part of the state to
provide a full and conscientious adjudication of the federal
claim, and this merely on the general premise that federal courts
195I hold that the State courts will be divested of no part of their primi-
tive jurisdiction, further than may relate to an appeal; and I am even
of opinion that in every case in which they were not expressly excluded
by the future acts of the national legislature, they will of course take
cognizance of the causes to which those acts may give birth. This I infer
from the nature of judiciary power, and from the general genius of the
system. The judiciary power of every government looks beyond its own
local or municipal laws, and in civil cases lays hold of all subjects of litiga-
tion between parties within its jurisdiction, though the causes of dispute
are relative to the laws of the most distant part of the globe. Those of
Japan, not less than of New York, may furnish the objects of legal discus-
sion to our courts. When in addition to this we consider the State govern-
ments and the national governments, as they truly are, in the light of kindred
systems, and as parts of ONE WHOLE, the inference seems to be conclusive,
that the State courts would have a concurrent jurisdiction in all cases aris-
ing under the laws of the Union, where it was not expressly prohibited.
TEE FEDERALIST No. 82, at 514 (Lodge ed. i888) (Hamilton). For other
celebrated statements of the point, see Martin v. Hunter's Lessee, I4 U.S. (I
Wheat.) 304, 340-4I (i8i6) ("From the very nature of their judicial duties they
[state judges] would be called upon to pronounce the law applicable to the case
in judgment. They were not to decide merely according to the laws or constitu-
tion of the state, but according to the constitution, laws and treaties of the
United States-'the supreme law of the land."'); Robb v. Connolly, iii U.S.
624, 637 (I884) ("upon the State courts, equally with the courts of the Union,
rests the obligation to guard, enforce, and protect every right granted or
secured by the Constitution of the United States").
196
14 U.S. (i Wheat.) 304, 341-42 (i8i6).
are institutionally capable of providing better justice than the

state courts.
Notice, too, that our question - whether on collateral attack
federal courts should redetermine federal questions already ad-
judicated by state courts and subject to Supreme Court review
- is quite different from the question whether an entire law-
suit involving federal issues should originally be adjudicated
by a federal or state court. I do not, in other words, mean to
be taken as questioning the propriety or wisdom of the general
federal-questionjurisdiction, original or removal, for the vindica-
tion of federal rights.'97 Supporting it are the factors already
mentioned: the independenceof federal judges and their objective
perspective with respect to federal claims. In addition, federal-
court adjudication provides specialized and knowledgeable tri-
bunals with procedural and remedial tools which may be more
effective in enforcing federal rights than those available in the
state system.198 But the calculus which tells us that these ad-
vantages justify giving litigants the option to have their lawsuits
tried entirely in federal court is surely a different one than where
the question is whether a suit which is and must be tried in state
court should then be reopened to allow the redetermination of
federal questions by a federal judge. The whole point is that in
the latter situation we have already made the fundamental deci-
sion that we do want the state courts to decide the case.'99 And
it is this decision that creates the special problems of waste of
resources, strain in federal-state relations and damage to the
fabric of criminal law which bear so acutely on the decision
whether we should superimposecollateral review on the Supreme
Court's direct supervisory jurisdiction.
' See the illuminating discussion of the federal-question jurisdiction in
Wechsler, Federal Jurisdiction and the Revision of the Judicial Code, I3 LAW &
CONTEMP.PROB. 2I6, 223-34 (1948). Professor Wechsler suggests that even
the original jurisdiction over claims of federal right "in cases where the claim is
that state action-legislative or administrative -violates the federal Constitu-
tion or conflicts with national law" be abolished where a "plain, speedy and
efficient remedy" is available in state court: "application of the federal authority
to invalidate the action of a state is best accomplished when the issue finds its
way to the Supreme Court after it has had examination in the state courts."
(Professor Wechsler excepts from his suggestion suits under the civil rights
laws.) Id. at 227, 229, 230.
198 In fact the states could legitimately protest if they had to assume the
entire burden and expense of administering justice in cases where the rights and
duties involved are wholly creations of federal law.
199 I do not suppose it would be seriously argued that all state criminal
cases in which the defendant raises a federal defense should be made removable
to federal court.
There is another aspect of the claim to a "right" to a federal

forum which must, however, be taken into account, one which
derives not from the notion that federal courts will provide
better justice than state courts, but from a feeling that perhaps
the structure of the SupremeCourt's certiorarijurisdiction creates
inequalities in the treatment of state prisoners which are, from
the viewpoint at least of those prisoners, unfair. As the Court
has so often pointed out, it will not grant certiorari in a case
merely because it thinks the case was erroneously decided: the
purpose of the jurisdiction is to have the Court
resolveconflictsof opinionon federalquestionsthat have arisen
amonglowercourts,to pass upon questionsof wide importunder
the Constitution,laws, and treatiesof the United States, and to
exercisesupervisorypower over lower federalcourts. If we took
every case in which . . . our prima facie impressionis that the
decisionbelowis erroneous,we couldnot fulfill the Constitutional
placeduponthe Court.200
andstatutoryresponsibilities
The Court thus will often not review an "unimportant"case
which, on the merits, might have been reversed because deemed
to be "wrong." On the other hand, once it decides to review an
"important" case, or one which creates a conflict as to the law,
it of course does pass on the merits and correct error. But from
the individual prisoner's point of view, whether his case is "im-
portant," or was decided contrary to other lower court rulings,
is entirely adventitious -whether justice was done in his case
does not turn on the character of the issue decided. May we not,
then, justify Brown v. Allen by seeing it as aimed at equalizing
the chances of state prisoners for federal review on the merits:
where the Supreme Court does not exercise its discretionary
power to have a federal question decided "right," a lower federal
court on habeas should do so. Why, in short, should a state
prisoner's opportunity to have his federal question decided
"right" turn on the happenstance of whether he is involved in
an "important"litigation?
The argument assumes, of course, that the federal courts have
a patent on "right" decisions of federal questions. If we assume
that there is, intrinsically, no reason why such federal-court
decisions should be more "correct" in an ultimate sense than
state-court decisions, if a conscientious state-court litigation of
200IChief Justice Vinson, Work of the Federal Courts, Address to the Ameri-
can Bar Association, Sept. 7, I949, printed in 69 Sup. Ct. V, VI. See generally
HART & WECHSLER, THE FEDERAL COURTS AND THE FEDERALSYSTEM I394-1422
('953).
5I4 HARVARD LAW REVIEW [Vol. 76:44I
a federal question without federal review is as likely to come

to a just result as a conscientious state-court litigation followed
by federal review, the mere fact that institutional considerations
compel us to furnish such review in some cases does not auto-
matically mean that justice requiresus to furnish it in other cases
where these institutional considerationsare not operative. If the
purpose of the certiorari jurisdiction is to settle conflicts and
decide important cases, the mere fact that in the process the
Supreme Court (infallible only because it is final) has authority
to make decisions which will be deemed right does not prove that
its jurisdiction is also necessary for the doing of justice, and
therefore does not tell us that without federal review the proba-
bilities that the states will do justice are unacceptably low.
What it comes down to, really, is that the state prisoner in an
important case does have one more chance to persuade a set
of judges - and judges, perhaps, of superior objectivity and
fresher perspective with respect to matters of federal law - to
see the case his way than the prisoner in an unimportant case.
And this is a real difference. The sense of inequality created by
the discretionarycertiorarijurisdiction does seem to me a weighty
and important consideration. Whether it serves in itself to justify
Brown v. Allen, in view of all the problems created, I am not at
all sure. But at least it helps rationalize and explain it.
The notion that in Brown v. Allen the Supreme Court was
creating an alternative to certiorari in order to equalize the op-
portunities of state prisoners to be heard in federal forums may
serve, at least in part, to explain the puzzling holding of the
Court recently in Rogers v. Richmond.201In that case the Court,
reviewing an application for federal habeas by a state prisoner,
found that at the state trial the judge had incorrectly assumed
that under federal law the trustworthiness of a confession is
relevant to its admissibility, that he had, in other words, tested
the question of admissibility under an erroneous constitutional
standard. The Court ruled, further, that the federal court on
habeas could not now itself decide the question of admissibility
on the basis of the state record, because the trial judge's miscon-
ception as to the law may have colored his findings of fact: the
state record was, in other words, flawed. Instead of ordering the
201365 U.S. 534 (ig6i). For an admirable comment on this case see The
Supreme Court, I960 Term, 75 HARV.L. REV. 40, I73 (1I96I). The analysis of
the case made in the text was first prompted by suggestions in Mr. Richard
Posner's paper cited in note I74 supra.
I963] FINALITY IN CRIMINALLAW 5I5
federal habeas court to hold a hearing on the facts surrounding

the confession and then to decide the admissibility issue, how-
ever, the Court held that the proper mandate is to order the
prisoner's release subject to the state's power to grant him a
new trial under proper constitutional standards. What the Court
did, in other words, was just what it would have done on direct
review had it found an improperassumption in the state's adjudi-
cation of the confession issue and a record on which it could not
itself rule on admissibility - reverse and remand for new trial.
At first glance such a disposition on habeas is baffling. If it
is the function of the habeas court in any event to redetermine
the federal issue on the merits, it would seem unnecessary to have
the state retry the petitioner in order to force it to go through
the gesture of ruling on the matter under proper standards of
constitutional law -at some stage the federal court will substi-
tute its judgment for the state's in any event. Why should the
federal court not do so now? At least in such a case if the
habeas court, taking evidence to cure the flaws in the record,
finds that the confession was admissible, the state judge's errone-
ous assumptions of law will be deemed harmless and the deten-
tion left untouched. To force the state to retry the petitioner's
guilt 202 without first using the habeas court to make sure that
in fact there was error in admitting the confession seems to be
a sacrifice of one of the few advantages the habeas court has
over the Supreme Court on direct review - that is, its fact-
finding processes, which can illuminate the question whether the
state court's erroneous assumption as to the law in fact led to
an erroneous disposition of the admissibility question. We may
ask further: if Rogers is now retried and the confession judged
admissible by the state courts on the basis of an impeccable state
record and correct assumptions about the law, and if certiorari
is thereafter denied, will a subsequent district court on habeas
have the power, not only to redetermine the ultimate question of
admissibility but also, in its discretion, to redetermine the facts?
If so, Rogers v. Richmond surely makes no sense at all.
But assuming that Rogers means that the facts relating to a
confession must always be found in the state courts, with the
202
It may be that in a state where the admissibility of confessions is a
matter solely for the court, it would be permissible for the state court in a case
like Rogers' to redetermine the admissibility of the confession, without retrial
of the merits unless it is found that under proper standards the admission of
the confession was erroneous. But in a state where the admissibility issue is also
passed on by the same jury which determines guilt, such a fragmentation of the
case would appear to be impossible.
5I6 HARVARD LAW REVIEW [Vol. 76:441
federal courts, whether on direct review or on habeas, restricted

to testing the issue of admissibility on the basis of such facts,
the case is explainable- if by no means necessarily justifiable-
in terms of a purpose to make the habeas jurisdiction a replica
of the Supreme Court's appellate jurisdiction over these cases:
the federal district court will review them as if it were a purely
appellate court. Such a reading of Rogers would also serve to
explain the rather puzzling expressions in the Court's opinion
that to have the federal habeas court determine the facts sur-
rounding the confession would be inconsistent with the state
prisoner's right "to have all issues which may be determinative
of his guilt tried by a state judge or a state jury under appropriate
state procedures" and the state's "interest in having valid
federal constitutional criteria applied in the administrationof its
criminal law by its own courts and juries." 203 If there is such
an interest in state, as against federal, factfinding, it makes
sense to satisfy that interest not only in those cases where the
Supreme Court provides direct review but also where review
is provided by the habeas court.
It may be added that the proposition that Brown v. Allen and
Rogers v. Richmond make sense in terms of providing prisoners
whose cases were not reviewed on certiorari an equivalent of such
review on habeas corpus is made particularly plausible in the
context of these coerced confession cases by the rather peculiar
role the Supreme Court itself has assumed in such cases. For
even on direct review the Court, in passing on the admissibility
of confessions, has departed strikingly from its normal role
of authoritative "law-giver"; in fact the Court has been signally
unsuccessful in articulating general legal standards or rules for
these cases. Instead of "pronouncing"the law, it has exercised
a highly particularistic case-by-case supervision, carefully evalu-
ating the particular facts and quite explicitly second-guessingthe
ultimate judgments of lower courts about admissibility. In effect
the Court has seen its own function in these cases in terms of
making sure that elusive federal standardsare "correctly"applied
in particular cases. But if this is the proper institutional function
of federal supervision in this area (an assumption by no means
to be taken for granted), then it becomes plausible to allocate this
function, in part at least, to the district courts (thereby relieving
the dockets of the Supreme Court); this role can be carried out
by habeas courts as well as (perhaps even better than) by the
203 365 U.S. at 547, 548.
i9631 FINALITY IN CRIMINALLAW 517
Supreme Court itself. In other words, once we decide (as the

Supreme Court seems to have decided) that the purpose of
reviewing confession cases in federal courts is, frankly, to check
up on the correctness of state courts in particular cases, then,
on that assumption, a habeas replica of Supreme Court review
not only serves the purpose of equalization but may be justified
on functional grounds as well.
On the other hand, the notion that prisoners whose cases the
Supreme Court chose not to review should be given an oppor-
tunity to have such review in a habeas court is not without its
own difficulties. Note that such a system in fact leaves the
prisoner on habeas with one overwhelming advantage over the
man whose case is given direct review: there is no time limitation
on his claim, so that he is free to maneuver the timing of his
case to his best advantage, perhaps benefiting by loss of evidence
or absence or death of witnesses, which may not only help estab-
lish his story with respect to the confession but effectively bar
retrial of the case. If, as in Rogers, it is determined on federal
review that the state judge passed on the admissibility of a con-
fession under an erroneous standard of constitutional law, so as
to call for retrial, it may make all the difference to the state
whether that judgment comes soon after the state proceeding or
five, ten, or twenty years later. If our purpose is to equalize the
chances of those whose cases are not reviewed by the Supreme
Court, we would surely be justified in imposing a statutory time
limit on the writ (running from the time that certiorari is denied)
at least in those cases where no other considerations justify
delay.204
Further, more fundamental questions, could be raised. Brown
v. Allen creates a district court replica of Supreme Court review
only for state litigants in criminal cases, and among these only
for those presently under detention.205If equality of treatment
204
Of course I would not suggest imposing such a limit in cases where the
whole point of affording a collateral jurisdiction is to enable prisoners to raise
federal claims not previously available to them.
205This is the case unless we take seriously the astonishing holding of the
district court in Goss v. Illinois, 204 F. Supp. 268 (N.D. Ill. I962). Goss was
convicted by Illinois of contempt over his contention that the proscribed conduct
was protected as an exercise of free speech by the fourteenth amendment. The
conviction was affirmed by the Supreme Court of Illinois on a full consideration
of the federal question, and certiorari was denied. The sheriff now holding a
"mittimus" for his arrest and confinement, Goss sought a declaratory judgment
in the federal district court under the Civil Rights Act, advancing precisely the
same claim as was adjudicated on the merits in the state criminal proceeding. The
district court held that it had jurisdiction and proceeded to redetermine the
5I8 HARVARD LAW REVIEW [Vol. 76:44I
is the aim, should it not hold for civil litigants with federal ques-
tions in their cases which the Supreme Court has chosen not to
review, and similarly for state criminal defendants who have
been fined rather than jailed? Further, what about the claims
for equality of treatment of federal prisoners (or, indeed, federal
litigants) whose petitions for certiorari have been denied? For
if the rationale of Brown v. Allen is that one has a right to test
federal constitutional questions in at least one federal constitu-
tional court, then prisoners convicted by federal courts would
plainly not be permitted on collateral attack to relitigate issues
previously determinedby the committing courts 206 (thus making
the category of questions cognizable on collateral attack turn
on whether one is held pursuant to state or federal law). Yet
these federal prisoners, too, could claim that they are badly
treated compared to those who have had their cases reviewed
by the Supreme Court.
Finally, do these possible anomalies not raise still a further
question? Was it institutionally sound for the Court to assume
such a forward role in the complicated and sensitive task of
making "as prudent use as we can . . . of the important national
resources represented by the federal courts"? 207 Is the creation
of a replica for certiorari in the case of state prisoners derivable
federal question, deciding in favor of Goss and declaring the mittimus in the
hands of the sheriff to be null and void. I take it that the implication of the
decision is that every federal constitutional question involving "rights, privileges
and immunities" which has been fully adjudicated by state courts in civil or
criminal proceedings may be redetermined by a federal district court on collateral
attack under the Civil Rights Act and the federal-question jurisdiction.
206 The Supreme Court has not passed on the question whether Brown v.
Allen is applicable to petitions for release by federal prisoners brought under

the statutory substitute for federal habeas corpus (section 2255 of the Judicial
Code, 28 U.S.C. ? 2255 (I958) ).; Nor have the implications of Brown for section
2255 cases been discussed by the lower federal courts. The issue was potentially
before the Supreme Court in Hodges v. United States, where the Court of
Appeals for the District of Columbia, without referring to Brown, had reaffirmed
its pre-Brown holding in the Smith case (see note 43 supra), that "admission
[of a coerced confession] alone does not result in the denial of a constitutional
guaranty so long as the error is subject to correction on appeal and there is no
indication of any deterrent to appeal, such as lack of counsel." Hodges v. United
States, 282 F.2d 858, 865 (D.C. Cir. I960). The Supreme Court, after granting
certiorari and hearing argument, dismissed the writ as improvidently granted
because the records "conclusively show" that the petitioner was not entitled to
relief. 368 U.S. I39 (I96I) (with Warren, Black and Douglas, JJ., dissenting).
The Hodges case is complicated by the fact that the Court may have conceived
that what was at issue was not the scope of review of questions previously litigated
but the problem of forfeiture where a litigant fails to appeal. See also Jordan v.
United States, 352 U.S. 904, reversing 233 F.2d 362 (D.C. Cir. I956).
207 Wechsler, supra note I97, at 2I8.
i963] FINALITY IN CRIMINAL LAW 5I9
even from a generous and open-endedreading of the act of i867?

Was it the business of the Supreme Court under the general man-
date of that act to accomplish such a tour de force? In view of the
inequalities created by the Court's attempt to equalize, is not
mediation here appropriately a task for the Congress?
C. The Problem of Inadequate Supreme

Court Supervision
Another line of argument to support the present structure of
the habeas corpus jurisdiction rests on the proposition that
direct review by the Supreme Court provides inadequate super-
vision of the state courts' adjudications with respect to state
defendants' federal constitutional rights. Of course, whether
supervision is "adequate" turns on how one defines the purposes
being served by such supervision. Insofar as we are told that
habeas corpus is necessary because the overcrowding of the Su-
preme Court's docket makes it impossible for the Court to con-
sider the merits of every case where a prisoner alleges that he has
been deprived of a constitutional right,208this is but another
version of the assertion that a federal court should decide the
merits of every such case. But there does exist another, rather
different, problem. It stems from the well-known fact that the
typical certiorari petition from a state prisoner, and the record, if
any, that accompanies it, are often wholly inadequate to inform
the SupremeCourt whether the case should be reviewed. The peti-
tions, we are told, drafted usually without a lawyer, are frequently
unintelligible and rarely clear; certified records are "almost un-
known" and
the numberof casesin whichmostof the papersnecessaryto prove
what happenedin the State proceedingsare not filed is striking.
Whetherthere has been an adjudicationor simply a perfunctory
denialof a claimbelowis rarelyascertainable.Seldomdo we have
enoughon whichto base a solid conclusionas to the adequacyof
the State adjudication.Even if we are told somethingabouta trial
of the claims the applicantasserts,we almostnever have a tran-
scriptof these proceedingsto assist us in determiningwhetherthe
trialwas adequate.209
208
See Reitz, Federal Habeas Corpus, IO8 U. PA. L. REv. 46I, 464 (i960);
Letter From Judge Jerome Frank to Congressman Celler, printed in Hearings i6.
209Opinion of Mr. Justice Frankfurter, Brown v. Allen, 344 U.S. 443, 493-94
(0953) S
And these messy and often inscrutable petitions come before

the Court in very large numbers indeed.210
Effective adjudication by an appellate court is peculiarly de-
pendent on the delineation of issues and focusing of facts which
can be provided only by reasonably clear and competent plead-
ings, records and briefs. Simply put, the root of the trouble in
these cases is that the Supreme Court is often unable to tell
what happened below. How then can it judge wisely whether the
case merits review?
The problem is not too troublesome with respect to many,
maybe even the great majority, of these cases. For it will be
discovered on habeas corpus that the inadequate record is a
symptom of inadequate process, that the prisoner has not been
accorded a chance to test his federal claim in a meaningful litiga-
tion in the state system, so that the habeas court would be free
to turn to the merits under the conventional doctrine of Frank
v. Mangum. But a differentkind of case may be more troubling.
Suppose a state prisoner has had a fair opportunity to litigate
a federal question in the state system, but no effective record was
made and his petition for certiorari does not intelligibly set forth
his story, so that the Supreme Court cannot really grasp the
import of the case. Certiorariwill surely be denied (unless there
is some special circumstance to call the Court's attention to the
situation). On habeas, once the district judge has satisfied him-
self as to the availability of corrective process, should review on
the merits be denied? The problem here is not whether one
should have a federal review of the merits as of right, but
whether one should not at least have an honest chance to obtain
such review. If the condition of the recordprecludes the Supreme
Court from exercising a considered judgment whether the case
deserves review, does not fairness call for deciding it on the
merits on habeas corpus-not the least in order that the Su-
preme Court will have a renewed chance to decide on the basis
of the habeas corpus record whether it wishes to review the case?
After all, the inability to exercise supervision in such cases not
only creates inequality but damages the purposes served by
the certiorarijurisdiction itself: if it should turn out that hidden
in the cloudy record is a question of importance,or a decision con-
210
In its I96I Term the Court disposed of 1,282 cases on the Miscellaneous
Docket. This included 7I2 denials of petitions for certiorari to state courts. See
tables I and II, The Supreme Court, I96I Term, 76 HARV. L. REV. 54, 8i, 82
(i962).
flicting with other holdings, there should be some way, eventually,

to subject it to Supreme Court disposition on the merits.2"'
Thus one certainly could not be scandalized by a decision that
habeas should issue in such cases. Indeed, it can be rationalized
in terms of Frank v. Mangum itself: we could view as part of
the corrective process which the state must supply the creation
of a record-whether in the form of a transcript, or findings, or
opinion - sufficiently explicit and clear so as to give the Supreme
Court a fair opportunity to decide whether it wishes to review
the case.212 (The reasoning would not, of course, work for in-
competent petitions - but a good record will cure most flaws in
the petition.) On the other hand, no such doctrine would in itself
validate habeas in a case where the state has afforded corrective
process and the SupremeCourt has had a meaningful opportunity,
on the basis of a full record, to determine whether it wishes to
review the case.
D. Some Cautions: The Inadequate State,Process

The existence, notorious and oft-exhibited, of grave inadequa-
cies in the states' criminal procedures, both original and post-
conviction, makes the federal habeas corpus jurisdiction a present
necessity. On this bedrock proposition I agree with the writ's
defenders. These inadequacieshave been explicitly acknowledged
by distinguished state judges, including a committee of the Con-
211 Compare the suggestion made by Judge Wyzanski in Geagan v. Gavin, i8i
F. Supp. 466, 469 (D. Mass. I960), aff'd, 292 F.2d 244 (ist Cir. I96I), cert.
denied, 370 US. 903 (i962):
The Supreme Court [in Brown v. Allen] recognized that if procedural and
substantive federal constitutional questions with respect to state cases could
be reviewed only upon the often skimpy state records . . . filed in the Supreme
Court of the United States those federal constitutional questions would fre-
quently receive inadequate attention. . . . To expand the role of federal
district judges in habeas corpus cases made it possible for the Supreme
Court to treat such district judges virtually as its delegates or masters to
make such findings and rulings as were necessary, all subject to ultimate
review by federal courts of appeal and the Supreme Court ....
And see the suggestion of Judge Frank in his letter to Congressman Celler, supra
note 208, that if the habeas jurisdiction is to be restricted, the Congress should
provide that in cases of petitions for certiorari from state courts, the Supreme
Court be permitted to "refer the matter to the appropriate Federal district court
to hear and report the facts, and recommend legal conclusions ...."
212 It is interesting to note that the Senate and House reports in the 85th
Congress accompanying the proposed habeas corpus statute which would have
denied use of the writ where the federal question was determined or still
could be determined in the state courts (see note i8o supra) both stated that it is
a purpose of the bill to retain the jurisdiction where it "is not possible to make
a State court record which could serve as a basis for review by the Supreme
Court of the United States." S. REP. No. 2228, 85th Cong., 2d Sess. 3 (I958)
H.R. REP. No. I293, 85th Cong., 2d Sess. 3 (I958).
ference of Chief Justices, whose report, recommendingmany im-

provements in state procedures, states that
responsibility for the unfortunate conditions prevailing in habeas
corpus litigation rests upon the State as well as upon the Federal
judicial systems, and . . . the evils presently prevailing can be
reducedsubstantially by action taken at the State level.213
If any doubt remains as to this point, it should be allayed by

Professor Curtis Reitz's admirable article in the University of
Pennsylvania Law Review,2"4in which he reports on a painstak-
ing canvass of some thirty-five cases from a ten-year period
involving state prisoners in which habeas corpus was ultimately
granted by a federal court. These cases show beyond a doubt
that the states frequently fail to provide a fair and rational
setting for the litigation of claims of federal constitutional right,
so that habeas turns out to be the only available remedy for
the vindication of such claims.
What is, however, so hard for me to grasp is why the existence
of habeas to cure failures of state process justifies its present
reach to cases where there has not been such a failure of process.
Inadequacies of state procedure do not validate the status quo.
The issue is not whether the jurisdiction should be abolished but
whether its expansion to cases where there is no reasoned basis
to suspect failure to provide a rational trial of the federal ques-
tion, before an unbiased tribunal and through fair procedures, is
justified. Professor Reitz's thirty-five cases show that there is
a need for habeas corpus, but do not satisfy me that there was a
need for it in all thirty-five, do not satisfy me that it was needed
in Leyra (or Brown itself), where the state provided a full trial
of the federal question and opportunity for direct Supreme Court
review on a record so impeccable that the habeas court based its
own judgment on it.
In fact, if the principal problem is inadequacy of state pro-
cedures for the vindication of federal constitutional rights, and
the principal aim is the cure of these inadequacies,215 then Brown
213 Report of the Special Committee on Habeas Corpus to the Conference
of Chief Justices, June I953, printed in Hearings 92-93. The report and its appendix
recommended many improvements in state criminal procedure, both in trial
courts and in postconviction proceedings, see id. at 93, I03-08, which were
unanimously approved by the 'Conference.
214
Reitz, Federal Habeas Corpus, io8 U. PA. L. REV. 46I (I960).
215 Both Mr. Justice Brennan and Justice Schaefer have argued that federal
habeas serves to stimulate the states to devise adequate postconviction pro-

cedures. Brennan, supra note I78, at 44I; Schaefer, supra note I79, at 24. This
would of course be true insofar as the writ does probe the question of adequate
v. Allen would seem to be a step in the wrong direction. In

effect it tells the states that not much will turn on whether or
not they provide corrective process: no matter how conscien-
tiously and fairly they apply themselves to the consideration of
the merits of federal claims, whether presented at trial or on
postconviction process, they will nevertheless automatically be
second-guessed by federal district courts as to their conclusions
of law and, possibly, factfindings too. Furthermore, the institu-
tional needs advanced to justify this doctrine are likely to be
most unpersuasive to the very group which has the power and
the responsibility to improve and reform state procedure: state
judges, legislators, and law-enforcement personnel. I do not,
therefore, see the present system as conducive to state procedural
reform. It was the "corrective process" doctrine of Frank v.
Mangum which really created the stimulus for such reform;
Brown v. Allen has merely blunted that stimulus.
IV. CONCLUSION:THE ROLEOFDISCRETION

What should we conclude about Brown v. Allen? I do not
pretend to find the answer easy, nor the claims for federal super-
vision unweighty. Their strength derives from our own historical
experience. The last twenty-five years have seen rapid and tre-
mendous expansion and movement in the substantive doctrines
derived from the due process clause which limit the power of the
states in their administration of criminal justice. It is natural
that, in an era of such rapid growth in the substantive federal
law, there should be a demand that the remedial system keep
pace, that federal supervision be expanded to make sure that the
states receive the new doctrines hospitably. And there is, of
course, the underlying suspicion that in fact the states have
not done so, that if we do not keep a sharp eye out, federal rights
will be subtly eroded, verbal respect paid to the principles but
the substance robbed of meaning through astringent and un-
sympathetic application. (The suspicion is surely fed by the
knowledge that a substantial proportion of those accused of
crime, particularly in the Southern States, will be Negroes.)
Yet we must remember that the remedial system we construct
must be tailored for tomorrow as well as today. It is not fanci-
ful to suppose that the law of due process for criminal defendants
state process; but how conceivably does it justify use of the writ where there
has been, by hypothesis, full corrective process?
524 HARVARDLAW REVIEW [VoI. 76:44I
will, in the foreseeable future, reach a resting point, will become

stabilized. Wolf v. Colorado is already overruled; 216 Betts v.
Brady will be reargued shortly217 and may not be far behind.
There must soon come a time in this field when it will be felt that
the great battles have been won, that we should return from molar
to molecular motion. And if there is to be a stabilization of the
law, we should be wary about constructing a remedial system
premised on unceasing and revolutionary change.
Similarly, I resist the notion that sound remedial institutions
can be built on the premise that state judges are not in sympathy
with federal law. Again we must think in terms of tomorrow as
well as today. Hopefully we will reach the day when the suspicion
will no longer be justified that state judges - especially Southern
state judges - evade their responsibilities by giving only the
appearance of fairness in their rulings as to state defendants'
federal rights. The unification of the country is, after all, in
progress; the day when Southern justice is like Northern justice,
justice for the Negro like justice for the white, is no longer out
of sight. And our remedial system ought to take account of
this motion.
The crucial point, it is worth reiterating, is that the question
before us is not whether federal supervision of the states' admin-
istration of criminal justice is necessary. The question is whether
such supervision is inadequate if limited to the very sweeping
powers of the Supreme Court on direct review, and of the district
courts on habeas to inquire into the fairness of the state's process.
Is more than this needed? Must there be further supervision if
the Supreme Court has had a chance to review the case and has
chosen not to, and if the federal district court finds that the
state has afforded fair process for the litigation of federal rights?
True, it will be argued that such supervision cannot ensure
that in each case the federal right has in fact been conscientiously
protected; it does not guard against cases where there has been
the appearance of fairness but not its inner essence. But is it
so clear that it is the function of judicial review to give us this
ultimate guarantee? Can the legal system assure us in any event
of this ultimate inner conscientiousness? And may one not specu-
late that perhaps conscientiousness is a by-product not only of
216 Mapp v. Ohio, 367 U.S. 643 (i96I), overruling Wolf v. Colorado, 338
U.S. 25 (I949).
217 In Gideon v. Cochran, the Supreme Court granted certiorari to the Supreme
Court of Florida and asked counsel to discuss this question: "Should this Court's
holding in Betts v. Brady, 3I6 U.S. 455, be reconsidered?" 370 U.S. 908 (i962).
i963J FINALITY IN CRIMINAL LAW 525
supervision from above but also of responsibility, that it is not

the best way to assure honest respect for federal law to relieve
the state judge of responsibility?
In sum, it seems to me that the proper verdict on the case made
for Brown v. Allen is "not proven"- and I do think the burden
of proof is on the proponents. I rest partly on the federalist
premise, that the abrasions and conflicts created by federal inter-
ference with the states' administration of criminal justice should
be avoided in the absence of felt need, where the institutional
necessities are as dubious as they seem to me to be in this case.
I also reason from the very real claims which the need for finality
and repose seem to me to make on the criminal process, claims
particularly strong in view of what I consider to be philosophi-
cally faulty premises about justice which are often at the heart
of the demand that we repeat inquiry endlessly to make sure
that no mistake has been made. And, finally, Brown v. Allen
seems to me to be unresponsive to (and even subversive of)
what should be our central aim: encouragingreform and improve-
ment in state criminal procedures.
Nevertheless, I cannot pretend that the case the other way
is weightless. Our traditional doctrines of judicial review do rest
on the premise of good-faith judging. Whenever good faith is
questioned, strain is put on the ordinary rules of review; and
maybe untraditional and extraordinary accommodations should
therefore be made. There is surely appeal in the notion, and
perhaps it makes sense at a time when there still is a justified
suspicion and distrust of state-court rulings as to federal consti-
tutional rights, to have a jurisdiction with a large and roving
commission "to prevent a complete miscarriage of justice"; 218
maybe
it is well that a writ the historicpurposeof which is to furnish
"a swiftandimperativeremedyin all casesof illegalrestraint". .
shouldbe left fluid and free from the definitenessappropriateto
ordinaryjurisdictionaldoctrines.219
Of course we should not forget that there is already in existence
an instrument in the administration of criminal justice whose
very purpose is to assure relief where there has been a miscarriage
of justice: the power of executive clemency supposedly already
218Learned Hand, J., in United States ex rel. Kulick v. Kennedy, I57 F.2d
8ii, 813 (2d Cir. 1946), rev'd sub nom. Sunal v. Large, 332 U.S. I74 (I947).
- 219 Frankfurter,J., dissenting in Sunal v. Large, 332 U.S. 174, I87 (I947).
See also Rutledge, J., dissenting in Ahrens v. Clark, 335 U.S, 1 T93 (I948).
gives us a roving commission, usually free of technicalities and

jurisdictional limitations, to seek out and right injustice. In fact
it is striking how often we lose sight of the pardon as an integral
part of the administration of criminal justice; in many cases
where judicial relief is sought by way of collateral attack, the
question that leaps to mind is: why is not pardon the obvious
and sound solution here? Why wasn't executive clemency exer-
cised? Surely it is a sorry thing that in so many states the pardon-
ing power has been allowed to atrophy, and is reserved for highly
extraordinary (usually death) cases. And again we have the
striking phenomenon that the states' own failure to provide,
through the pardon, an effective instrumentof justice has prompt-
ed a search for a federal substitute, in this case in the form of
habeas corpus.
In any event, Brown v. Allen may be justified in terms of a
need for an extraordinaryroving jurisdiction to make sure once
more that there has not been, with respect to constitutional rights,
a miscarriage of justice; perhaps we do need to grant federal
courts the power to redetermine the merits to assure that covert
unfairness does not lurk behind the appearance of fairness.
But if this is the theory of the jurisdiction, then what is called
for, I submit, is a decision, based on all the facts, whether jus-
tice really does call for release in the particular case. Suppose,
for instance, that a federal district judge finds (i) that the
admissibility of a confession has been fully and fairly litigated in
the state courts, on an impeccable record giving the Supreme
Court a full opportunity to review if it so chooses; (2) that on
the merits he disagrees with the state judges as to admissibility;
and (3) that the evidence in the record apart from the confession,
or, indeed, the trustworthiness of the confession itself, makes it
absolutely clear that the defendant was guilty as charged. It
is (today's) hornbook law that on direct review there would be
automatic reversal, that neither the trustworthiness of the con-
fession nor the other evidence of guilt can cure the error; let us,
further, assume the soundness of this doctrine. But is it abso-
lutely clear that the reasons which have moved us to so astringent
a rule of law on the merits, as it were, should necessarily apply
in a jurisdiction which is concededly extraordinary,the existence
of which is justified by the need to make sure once more that
justice in the particular case has indeed been done? Remember
that I assume that full corrective process has been afforded in
the original litigation: the defendant has had a fair opportunity
to establish his federal claim. We now give him one more
chance, in the fear that justice has miscarried. But has it mis-
carried if he has had one meaningful go-around and the district
judge can be morally assured that he was guilty? I acknowledge
that the purpose of the rule of automatic reversal is to deter
uncivilized police behavior and to keep the courts' own processes
unsullied; but the claims for deterrence do not strike me as so
inexorable that we must honor them in cases where our funda-
mental purpose, in fact our justification for inquiry, is, after all,
to do our best by the defendant; nor can I take too seriously the
idea that the integrity of the legal process will be sullied if a
court in effect finds that for purposes of the extraordinaryhabeas
jurisdiction the admission of the confession - previously tested
by fair process - was harmless error.220
I continue to resist, in sum, the notion that the inquiry on
habeas should be mere repetition, an exact replica, of what has
gone before. I do not see that as institutionally justified. If we
wish to have an ultimate recourse, if we want to grant the federal
courts a roving extraordinarycommission to undo injustice, then,
it sems to me, all the factors which bear on justice should be
put on the scales. If Brown v. Allen is to remain the law, it
should be modified to make clear that where a federal constitu-
220
The approach I suggest might have a further value -it might be a start
to the solution of a problem of collateral attack which I have not discussed at
all and which needs separate and rounded treatment in an independent article.
This is the baffling problem of the case where there has been a change or
reversal in constitutional doctrine between the time of conviction and the time
when habeas is sought. If a man was convicted, under the regime of Wolf v.
Colorado, with the aid of evidence illegally obtained from him, is his detention
"unlawful" for purposes of habeas corpus now that Wolf has been overruled?
If Betts v. Brady dies, will state prisoners who under the doctrine of that case
were not entitled to counsel be deemed to be illegally detained? Surely it would
be wholly impossible to administer a doctrine which extended Brown v. Allen
by telling the states not only that it is their duty under the due process clause
to decide every federal question "right" but that it is their further duty to
decide it "right" in accordance not only with today's law but that of tomorrow
too. On the other hand, a rigid rule that collateral attack is never proper
where there has been a change in the applicable federal constitutional doctrine
will often offend our sense of justice. May we not find an acceptable accommoda-
tion by according the district judge a large discretion to determine whether
justice in the individual case would be served by release? This would involve
not only an assessment of the guilt or innocence of the prisoner but also
the question whether the new constitutional doctrine has as its principal purpose
the safeguarding of the innocent or rather a different social purpose: i.e., deter-
rence of unlawful police conduct.
The "retroactive" nature of the ruling in Mapp v. Ohio has been discussed
recently in two admirable articles: Traynor, Mapp v. Ohio at Large in the
Fifty States, [i962] DUKE L.J. 3I9; Bender, The Retroactive Effect of an
Overruling Constitutional Decision: Mapp v. Ohio, I IO U. PA. L. REV. 650
(I962).
tional question has been fully canvassed by fair state process,

and meaningfully submitted for possible Supreme Court review,
then the federal district judge on habeas, though entitled to re-
determine the merits, has a large discretion to decide whether
the federal error, if any, was prejudicial, whether justice will be
served by releasing the prisoner, taking into account in the
largest sense all the relevant factors, including his conscientious
appraisal of the guilt or innocence of the accused on the basis of
the full record before him.

Bator 1963

Uploaded by

Copyright:

Available Formats

Bator 1963

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Bator 1963

Uploaded by

Copyright:

Available Formats

Finality in Criminal Law and Federal Habeas Corpus for State Prisoners

Author(s): Paul M. Bator

HARVARD LAW REVIEWV

FINALITYIN CRIMINALLAW AND FEDERAL

1 HEproblem of finality in criminal law raises acute tensions

rather special currents in our thought. Partly, our century has

which arise within the traditional structure of the system itself;

3U.S. CONST.art. VI.

findings of fact as to what phenomenaexisted or occurredbearing

I. SOME GENERAL CONSIDERATIONS

detention of one in the custody of another."7 In our constitu-

The writ of habeas corpusshall not extend to a prisonerun-

He is in custody in violation of the Constitutionor laws or

the management of their criminal law; the Constitution does not

freedom from errors of either fact or law, whenever error is al-

or decision of a competent tribunal or authority, even though the

Our analysis of the purposes of the habeas corpus jurisdiction

Now of course none of this tells us that it is not a valid function

and subject to Supreme Court review should finally establish the

the Supreme Court, 5 to 3, held the confessions inadmissible;

notion of legality must at some point include the assignment of

arrangements we create for the adjudication of criminal guilt

merely anxiety and a desire for immobility.23 Somehow, some-

B. Limitations on the Policies of Finality

that I do not view direct Supreme Court review of state cases

i. Failure of Process. - The first category has been implicit

inquiry might again constitute mere repetition. What does seem

he wasn't bribed.30In other words, collateral or other supervisory

I have referred to as "double-level" issues of due process of law. See note 28

of a state to do so is simply "error" subject to reversal by the

tion or competence to deal with the issue it is purportingto decide

the lines: not every question of competence is automatically open

37344 U.S. 443 (I953).

II. THE DEVELOPMENT OF THE HABEAS CORPUS

Perhaps the most striking aspect of the current discussions of

nature of the situation:

Brown decision, we may turn to the authority of the late Judge

Surely this makes clear that it was not in I949 thought to be

A. The Early "Federal" Cases

4 See Ex parte Dorr, 44 U.S. (3 How.) I03 (I845).

In Ex parte Lange, decided in I873, the Court first announced the

er,58the imposition of consecutive sentences where the indictment

merits of a claim of double jeopardy will not be reached where

It would be useless to try to show that these decisions establish

case; 69 reexaminationof the validity of the sentence was regarded

physical conditions of detention, i.e., whether the prisoner may be sentenced by a

as In re Belt and Matter of Moran,73and also in the post-igoo

the constitutionality of the statute creating the crime is appro-

B. The Act of I867

It is sometimes asserted that it was the explicit purpose of this

sumption is that the Congress in I867 must have meant that a

perfunctory attention and consideration in the Congress; indeed,

indication whatever that the bill intended to change the general

C. The Early "State" Cases

Lions at issue.88 Similarly, subsequent cases held that habeas

a "jurisdictional"issue open on collateral attack; and the Court,

None of these cases even suggests that as to these federal con-

D. The Doctrine of Frank v. Mangum

the committing court; these were, however, strictly limited and