Petition for rehearing denied. Separate opinion filed by Mr. Justice FRANKFURTER.
Opinion of Mr. Justice FRANKFURTER.
Further study and reflection have reinforced the conviction I expressed last June-and on even broader grounds than I then indicated-that this case should set down for reargument. Fundamental issues which have neither been argued by counsel nor considered by the Court are here involved. On such important questions, the military authorities, the bar, and the lower courts (including the Court of Military Appeals) ought not to be left with the inconclusive determination which our disposition of the case last June implies. One has a right to assume that there is greater likelihood of securing agreement of views for a Court opinion at the beginning than at the end of a term.
The CHIEF JUSTICE took no part in the consideration or decision of this case.
First. One of these problems concerns the effect of recent developments in the scope of inquiry on habeas corpus upon the relationship of the federal district courts in their habeas corpus jurisdiction to courts-martial. If the main opinion stands, matters which are open for inquiry on collateral attack upon a judgment of conviction entered in a United States District Court, a constitutional tribunal, will be foreclosed from inquiry when the judgment of conviction collaterally assailed is that of a court-martial, an executive tribunal of limited jurisdiction ad hoc in nature. This has not been the law up to now; and the assertion that 'in military habeas corpus the inquiry, the scope of matters open for review, has always been more narrow than in civil cases' 346 U.S., at page 139, 73 S.Ct. at page 1047, is, I respectfully submit, demonstrably incorrect. [346 U.S. 844 , 845] 1. The first case in this Court involving the collateral attack, by habeas corpus, on the judgment of a court-martial was Ex parte Reed, 100 U.S. 13 . Here is the test there laid down, 100 U.S. at page 23:
It was thus clearly stated that the standard for collateral consideration of judgments of courts-martial is the same as that applied on collateral consideration of judgments of other tribunals. Once 'jurisdiction' is shown to exist, the inquiry ends; the question is not whether that jurisdiction was well or wisely exercised, or whether error was committed, it is only whether there was power to act at all.
This was always the traditional scope of inquiry when the judgment sought to be examined on habeas corpus was that of a federal or territorial or District of Columbia court. E.g., Matter of Moran, 203 U.S. 96 (Oklahoma territorial court; opinion by Holmes, J.); Harlan v. McGourin, 218 U.S. 442, 448 (U.S. circuit court; opinion by Day, J.); Matter of Gregory, 219 U.S. 210 (District of Columbia court; opinion by Hughes, J., with copious citation of authority). [346 U.S. 844 , 846] And so, in the earlier cases scrutinizing military sentences by habeas corpus, it was similarly laid down that 'The single inquiry, the test, is jurisdiction.' In re Grimley, 137 U.S. 147, 150 . 'Courts-martial are lawful tribunals, with authority to finally determine any case over which they have jurisdiction, and their proceedings, when confirmed as provided, are not open to review by the civil tribunals, except for the purpose of ascertaining whether the military court had jurisdiction of the person and subject-matter, and whether, though having such jurisdiction, it had exceeded its powers in the sentence pronounced.' Carter v. Roberts, 177 U.S. 496, 498 ; Carter v. McClaughry, 183 U.S. 365, 380 -381, 187-188; Grafton v. United States, 206 U.S. 333, 347 - 348, 752. Allegations of irregularity or illegality in the composition of courts-martial were, of course, rigorously scrutinized, e.g., McClaughry v. Deming, 186 U.S. 49 ; cf. Kahn v. Anderson, 255 U.S. 1 ; but apart from this obvious amenability to judicial inquiry, the judgment of a court-martial meeting the test above quoted was unassailable even by the most extreme allegations of prejudice, unfairness, and use of perjured testimony. See Carter v. Woodring, 67 App.D.C. 393, 92 F.2d 544.
Thus, up to December 6, 1937, when the Court denied certiorari, 302 U.S. 752 , in the case last cited-it was the last of Oberlin Carter's long series of attempts at judicial review of his court-martial-the scope of habeas corpus in both military and civil cases was equally narrow: in both classes of cases it was limited solely to questions going to the 'jurisdiction' of the sentencing court.
2. Later in the 1937 Term, Johnson v. Zerbst, 304 U.S. 458 , was decided and blazed a new trial. It was held that procedural errors-what theretofore were deemed matters not going to the defined constitution of the tribunal acting [346 U.S. 844 , 847] within the scope of its power over subject matter and persons-may be inquired into collaterally on habeas corpus, if they amounted to a deprivation of constitutional right. By giving a new content to 'jurisdiction,' the case was brought within the formula that only 'jurisdiction' may be the subject of inquiry in habeas corpus. The judgment successfully assailed in that case was one entered in a United States District Court. Since 1938 the basic premise of Johnson v. Zerbst has been neither questioned nor limited in any instance involving collateral attack, by way of habeas corpus, on judgments of conviction entered by a civil court.
3. The effect of Johnson v. Zerbst on judgments of conviction pronounced by a court-martial first appears to have been considered in Shapiro v. United States, 69 F.Supp. 205, 107 Ct.Cl. 650. There the Court of Claims applied Johnson v. Zerbst to invalidate a conviction by an otherwise properly constituted court-martial, on the ground that the unreasonably short time permitted the accused to prepare his defense deprived him of the effective assistance of counsel in violation of the Sixth Amendment. The court-martial was held to have lost 'jurisdiction' to proceed. For purposes of the pending Petition for Rehearing and our responsibility for adequate consideration of the issues it is pertinent that the Shapiro case was not cited to us in any of the briefs in the present case.
Later decisions in the Court of Claims, where of course collateral attack is by way of a petition for back pay resting on allegations that the assailed court martial proceedings were void, have followed the rationale of the Shapiro case. Thus, in Sima v. United States, 96 F.Supp. 932, 938, 119 Ct.Cl. 405, 406, the court said: 'From the entire record in this case, we cannot say that plaintiff was deprived of his rights under the Fifth and Sixth [346 U.S. 844 , 848] Amendments to the Constitution of the United States to the extent that the verdict of the court-martial was void.' And in Fly v. United States, 100 F. Supp. 440, 442, 120 Ct.Cl. 482, 498: 'Only when the errors committed are so gross as to amount to a denial of due process does the erring court martial lose its jurisdiction and its power to issue a valid decree. Compare Sima v. United States, with Shapiro v. United States, both supra.'
4. This Court has never considered the applicability of Johnson v. Zerbst to military habeas corpus cases. But if denial of the right to counsel makes a civil body legally nonexistent, i.e., without 'jurisdiction,' so as to authorize habeas corpus, by what process of reasoning can a military body denying such right to counsel fail to be equally non-existent legally speaking, i.e., without 'jurisdiction,' so as to authorize habeas corpus? Again, if a denial of due process deprives a civil body of 'jurisdiction,' is not a military body equally without 'jurisdiction' when it makes such a denial, whatever the requirements of due process in the particular circumstances may be?
It is true that in Hiatt v. Brown, 339 U.S. 103 , the traditional older rule on military habeas corpus was restated and applied, and that we there disapproved the tendency of some of the lower federal courts to review court-martial records collaterally as if the habeas corpus court were a statutory agency of direct military appellate review in the Judge Advocate General's office, e.g., Hicks v. Hiatt, D.C., M.D.Pa., 64 F.Supp. 238. But the present problem was never suggested and never considered by us. Neither the Government's petition for certiorari nor its briefs cited Johnson v. Zerbst; and the respondent argued the point only inferentially until after the case went against him. The case cannot be deemed authority for an important point not discussed or considered. But assuredly Hiatt v. Brown does not sustain the proposition for which it was cited in this case, 346 U.S. at page 139, 73 S.Ct. at page 1047, that [346 U.S. 844 , 849] 'in military habeas corpus the inquiry, the scope of matters open for review, has always been more narrow than in civil cases.'1
5. In coming to this conclusion, the main opinion purported to derive some comfort from the 'finality' provision of the 1948 Articles of War and of the Uniform Code of Military Justice, AW 50(h), 10 U.S.C. (Supp. II) 1521(h); UCMJ, art. 76, 50 U.S.C. (Supp. V) 663, 50 U.S.C.A. 663, both of which state in terms that court-martial proceedings, once appellate review is completed, 'shall be binding upon all departments, courts, agencies, and offices of the United States'. But the decision in Estep v. United States, 327 U.S. 114 , should serve as a caution against applying provisions of 'finality' in legislation as though we were dealing with words in a dictionary rather than [346 U.S. 844 , 850] statutory directions, to be interpreted in the light of juridical considerations. The legislative history of the Uniform Code of Military Justice strongly suggests that it was precisely in the realm of collateral judicial attack on courts-martial that the concept of 'finality' was intended not to operate. Here is what both Armed Services Committees said of Article 76, UCMJ: 'This article is derived from AW 50(h) and is modified to conform to terminology used in this code. Subject only to a petition for a writ of habeas corpus in Federal court, it provides for the finality of court-martial proceedings and judgments.' H.R.Rep. No. 491, 81st Cong., 1st Sess., p. 35; S.Rep. No. 486, 81st Cong., 1st Sess., p. 32. I have added the italics to emphasize the congressional agreement with our decision on the same point in Gusik v. Schilder, 340 U.S. 128 , 132-133, 151-153. If that case and the Committee Reports have any meaning at all, they mean that the 'finality' provision is completely irrelevant to any consideration concerning the proper scope of inquiry in military habeas corpus cases. 2
6. It is desirable to emphasize that I express no opinion whatever on whether the allegations of the petition in the case at bar are sufficient to sustain a collateral attack on the court-martial's judgment of conviction. Nor do I express any opinion on the weight which should be given by the federal district court on habeas corpus to the findings of the military reviewing authorities. These are [346 U.S. 844 , 851] matters to be canvassed on the reargument. The issue here is whether the rationale of Johnson v. Zerbst is now to be quietly discarded or whether it will be appropriately applied, as it has been by the lower courts, in the military sphere. I do not think it is asking too much to insist that we have well-focused argument and careful deliberation before enunciating the principle that a conviction by a constitutional court which lacked due process is open to attack by habeas corpus while an identically defective conviction when rendered by an ad hoc military tribunal is invulnerable. 3
Second. There is another issue of broad importance which underlies this case but which has not been considered by the Court.
Both petitioners, alleging confinement in Japan (R. 1, 9) and American citizenship (id.), sought habeas corpus in the District of Columbia.
Thus there is raised squarely the question, thus far reserved by us, Ahrens v. Clark, 335 U.S. 188 , 192, note 4, 1445; Johnson v. Eisentrager, 339 U.S. 763 , 790-791, 949-950, whether an American citizen detained by federal officers outside of any federal judicial district, may maintain habeas corpus directed against the official superior of the officers actually having him in custody.
This question was originally answered squarely in the negative by the highest court of the District of Columbia. McGowan v. Moody, 22 App.D.C. 148 (detention on Guam, writ sought to be directed against the Secretary of the Navy). That precedent was followed as late as 1948 without question. Ex parte Flick, D.C.D.C., 76 F.Supp. 979, reversed on other grounds sub nom. Flick v. Johnson, 85 U.S.App.D.C. 70, 174 F.2d 983. It may have been, and probably was, overruled by Eisentrager v. [346 U.S. 844 , 852] Forrestal, 84 U.S.App.D.C. 396, 174 F.2d 961, which we in turn reversed for other reasons in Johnson v. Eisentrager, supra.
Petitioners have not discussed the question of jurisdiction, and the Government appears disinclined to argue it.
We should not permit a question of jurisdiction as far-reaching as this one to go by concession, or decide it sub silentio. I express no view on how we should determine the issue, or on what grounds, but I think that we should frankly face it, even at the risk of concluding that a legislative remedy is necessary. Cf. Wolfson, Americans Abroad and Habeas Corpus, 9 Fed.Bar J. 142, 10 id., at 69. It is particularly important that we do so at this time when thousands of our citizens in uniform are serving overseas.
[ Footnote 1 ] The direction of the opinion may well have been influenced by the following assumption regarding this Court's relation to military law: 'This Court has played no role in (the development of military law); we have exerted no supervisory power over the courts which enforce it; ....' 346 U.S. at page 140, 73 S.Ct. at page 1047. Of course it is true that we have no direct appellate jurisdiction over military courts. But it disregards both history and the statute books to say that our decisions have played no role in the development of military law. The pages of Winthrop are witness to the extent that the 'Blackstone of American military law' (as General Crowder, Judge Advocate General of the Army from 1911 to 1923, called him) considered himself bound by this Court's pronouncements. Since 1920, Article of War 38, 10 U.S.C. (1926-1946 eds.) 1509, has provided that the 'modes of proof' in courts-martial cases shall conform as nearly as practicable to the rules of evidence applicable to criminal cases in the United States district courts. Those rules of course are prescribed by this Court. In 1948 this language was expanded to include 'principles of law' as well as rules of evidence, 10 U.S.C. (Supp. II) 1509, and our decisions have been frequently cited by the military, as indeed they were in this very case. The same broad language is now in Art. 36, UCMJ, 50 U.S.C.A. 611, and the judges of the U.S. Court of Military Appeals apparently consider themselves bound by what we say. See Brosman, The Court: Freer Than Most, 6 Vand.L.Rev. 166, 167.
[ Footnote 2 ] It is noteworthy, though it was not referred to in the briefs, that as a matter of administrative recognition this 'finality' provision has not been read with dictionary literalness. See 41 Op.Atty.Gen., No. 8, Dec. 29, 1949, which holds that AW 50(h) of 1948-the very provision involved in the present case-did not bar the reopening of a record of conviction by court-martial by a Departmental Board for the Correction of Records functioning pursuant to 207 of the Legislative Reorganization Act, 5 U.S.C. 191a, 275, 5 U.S.C.A. 191a, 275. The action of those boards required approval by the Secretary concerned.
[ Footnote 3 ] I say 'ad hoc,' not in any derogatory sense, but merely to put the matter in its proper setting. See Winthrop, Military Law and Precedents ( 2d ed. 1896) *54.