concurring.
With the result I am in agreement, but I do not see how this decision and that in the Christoff el case, 338 U. S. 84, can coexist.
The Court is agreed that this defendant could rightly demand attendance of a quorum of the Committee and decline to testify or to produce documents so long as a quorum was not present. Therefore the real question here is whether, without making any demand, the issue may be raised for the first time long afterwards in a trial for contempt.
This case is the duplicate of Christoffel in this respect: in both cases defendants have sought to raise the question of no quorum for the first time in court, when they are on trial for an offense, without having raised it in any manner before the Committee while there was time to remedy it. The Court is now saying, quite properly I think, that this question must be raised at *344the time when it can be corrected, and proper records made, and cannot be kept as an ace up the sleeve to be produced years later at a trial. But in Christoffel, the majority took the opposite view and said, “In a criminal case affecting the rights of one not a member, the occasion of trial is an appropriate one for petitioner to raise the question.” Supra, at 88. If this statement of the law is to be left standing, I do not see how we can say that what was timely for Christoffel is too late for Bryan. It is plain we are not following the Christoffel decision and so I think we should candidly overrule it.
The practice of withholding all objection until time of trial is not helpful in protecting a witness’ right to a valid Committee. It prevents correction of any error in that respect and profits only the witness who seeks a concealed defect to exploit. Congressional custom, whether written or not, has established that Committee members may indulge in temporary absences, unless there is objection, without disabling those remaining from continuing work as a Committee. Members may step out to interview constituents, consult members of their staffs, confer with each other, dictate a letter, or visit a washroom, without putting an end to the Committee — but always subject to call whenever the point of no quorum is raised; that is notice that someone deems their personal presence important. This is the custom Christoffel, in effect, denied to members of Congress. A member now steps out of a committee room at risk of nullifying the whole proceeding.
It is ironic that this interference with legislative procedures was promulgated by exercise within the Court of the very right of absentee participation denied to Congressmen. Examination of our journal on the day Christoffel was handed down shows only eight Justices present and that four Justices dissented in that *345case. The prevailing opinion does not expressly indicate the Justices who joined in it, but only four nondissenting Justices were present to do so. On the record this would show only an equally divided Court, which would affirm the judgment below. The only way the four who were present and for a reversal could have prevailed was by counting for it one shown by the record to be absent. There is not even any public record to show that in absentia he joined the decision, or approved the final opinion, or considered the matter after the dissent was circulated; nor is there any written rule or law which permitted him to do so.
I want to make it clear that I am not criticizing any Justice or suggesting the slightest irregularity in what was done. I have no doubt that authorization to include the absent Justice was given; and I know that to vote and be counted in absentia has been sanctioned by practice and was without objection by anyone. It is the fact that it is strictly regular and customary, according to our unwritten practice, to count as present for purposes of Court action one physically absent that makes the denial of a comparable practice in Congress so anomalous. Of course, there is this difference: The absent Congressman was only necessary to a quorum; the absent Justice was necessary to a decision. No Committee action was dependent upon the Representatives presumed to be absent in the Christoffel case. All they could have done if present was to listen. In our own case, personal judgment and affirmative action of the absent member was necessary to make the Christoffel opinion a decision of the Court.
The ruling of the Court today seems irreconcilable with the Court’s decision in that case. True, the ink on Christoffel is hardly dry. But the principle of stare decisis, which I think should be the normal principle of *346judicial action, is not well served by failing to make explicit an overruling which is implicit in a later decision. Unless we really accede to its authority, it were far better to undo Christ off el before it becomes embedded in the law as a misleading influence with the profession. Of course, it is embarrassing to confess a blunder; it may prove more embarrassing to adhere to it. In view of the holding today, I think that the decision in the Chris-toff el case should be forthrightly and artlessly overruled.
Mr. Justice Black,with whom
Mr. Justice Frankfurter concurs, dissenting.18 U. S. C. § 3486 provides that no testimony given by a witness before any committee of either house “shall be used as evidence in any criminal proceeding against him in any court, except in a prosecution for perjury committed in giving such testimony.” The Court admits that use of such testimony in convicting Bryan for wilful failure to produce records violated the “literal language” of § 3486, but declines to give effect to that language. I dissent from the Court’s refusal to abide by this congressional mandate.
The statutory exception of “prosecution for perjury” shows that the attention of Congress was focused on whether committee testimony should be admissible in any special type of criminal prosecution. Yet the Court now reads the statute as if Congress had forbidden the use of committee testimony “except in a prosecution for perjury or for failure to produce records.” Such extensive judicial law-making is particularly questionable when used to restrict safeguards accorded defendants in criminal cases. Moreover, this statute springs from Congress’s recognition of the constitutional privilege against compulsory self-incrimination. The Court’s narrowing of the statute marks a radical departure from the principle underlying previous interpretations of other immunity legislation. *347Smith v. United States, 337 U. S. 137; United States v. Monia, 317 U. S. 424.
The reasons given by the Court for its amendment of the statute have an anomalous basis: the Court feels compelled to alter the clear language of § 3486 in order not to “subvert the congressional purpose” which it admits has already been irrevocably frustrated by the decision in Counselman v. Hitchcock, 142 U. S. 547.
Moreover, the statutory language is so clear and precise that dubious legislative history cannot contradict it. And no part of that history even tends to show that Congress meant to permit use of a witness’ testimony to convict him of any crime other than perjury. There is a justifiable reason for the perjury exception. The crime consists of the testimony itself, without which no prosecution would be possible. Not so with default in producing papers. That crime is based not on a witness’s testimony but rather on his failure to produce — conduct which can be proved by members of a committee, clerks, or spectators. There is therefore no basis for saying that application of the statute as Congress wrote it would lead to “absurd conclusions” by encouraging the “refusal of witnesses to answer questions or produce papers.”
As for other essential elements of the crime, such as power to produce, they cannot be proved by evidence extracted from a defendant under compulsion. A witness summoned to testify and produce papers is no less entitled to invoke the protection of this statute and of the Fifth Amendment’s privilege against self-incrimination than is any other defendant. One who has failed to produce certainly could not be compelled to answer questions concerning his power to produce, thereby making him a “witness against himself.” If application of the statute as Congress wrote it would lead to “absurd conclusions,” so would the Fifth Amendment.
*348The Court finds comfort in the statement that the Committee testimony of witnesses is “uniformly printed in the reports of committees recommending contempt action” to the houses of Congress. However extensive this practice may be, it would not justify the use of such evidence in a criminal trial. By its own terms 18 U. S. C. § 3486 is expressly limited to “any criminal proceeding ... in any court.”1
For these reasons the judgment should be reversed and the cause remanded for a new trial.
This distinction between criminal trials and contempt proceedings at the bar of Congress is eminently reasonable in view of the practical differences between the two. See dissenting opinion in United, States v. Fleischman, post, p. 349. For a discussion of congressional contempt procedures, see Eberling, Congressional Investigations 179 and passim, (Columbia University Press, 1928).