United States v. United Mine Workers of America/Dissent Rutledge

Mr. Justice RUTLEDGE, dissenting.

This case became a cause celebre the moment it began. No good purpose can be served by ignoring that obvious fact. But it cannot affect our judgment save only perhaps to steel us, if that were necessary, to the essential and accustomed behavior of judges. In all cases great or small this must be to render judgment evenly and dispassionately according to law, as each is given understanding to ascertain and apply it.

No man or group is above the law. Nor is any beyond its protection. In re Yamashita, 327 U.S. 1, dissenting opinion 41, 66 S.Ct. 340, 359. These truths apply equally to the Government. When its power is exerted against the citizen or another in the nation's courts, those tribunals stand not as partisans, but as independent and impartial arbiters to see that the balance between power and right is held even. In discharging that high function the courts themselves, like the parties, are subject to the law's majestic limitations. We are not free to decide this case, or any, otherwise than as in conscience we are enabled to see what the law commands.

Mr. Justice FRANKFURTER has shown conclusively, I think, that the policy of the Norris-LaGuardia Act, 47 Stat. 70, applies to this situation. The legislative history he marshals so accurately and cogently compels the conclusion that the War Labor Disputes Act of 1943, 57 Stat. 163, not only confirms the applicability of the earlier statute, but itself excludes resort to injunctive relief for enforcement of its own provisions in situations of this sort.

That Act expressly provides the remedies for its enforcement. Beyond seizure of plants, mines and facilities for temporary governmental operation, they are exclusively criminal in character. They do not include injunctive or other equitable relief. Nor was the omission unintentional or due to oversight. It was specific and deliberate.

The Senate thoroughly considered and debated various proposals for authorizing equity to intervene in labor disputes, one by the Act's sponsor in that body. Positively, repeatedly and unwaveringly it rejected all of them. They were likewise rejected in conference, where the Senate's view prevailed over that of the House. The latter body had not been inattentive to the problem. It sought and failed to secure the very thing this Court now says, in effect, was included. That issue and that policy were indeed the main thrust and focus of the legislative struggle, and the outcome was not negative; it was positive and conclusive against using or giving the equitable remedies.

Surely we have not come so far toward complete inversion of legislative history as to write out of the law the views concerning a matter of such major policy held by the chamber which prevailed at the final stage of enactment and to write into the law diametrically opposing views of another chamber which yielded at that time. The case, as Mr. Justice FRANKFURTER demonstrates beyond any doubt, cannot be one where inattention, oversight or inaction may explain or give significance to what was done by the House of Representatives. That body was defeated, not simply silent, in the outcome. Willingly or otherwise, it acquiesced in the Senate's policy of refusing to authorize injunctive relief, and in doing so joined formally and effectively in the final act which made that policy law.

This means to me that Congress, in that action, did not simply confirm the Norris-LaGuardia Act's policy or leave it untouched with respect to situations within the War Labor Disputes Act's coverage. It means that Congress was not departing from or nullifying that policy. Rather by the later Act Congress adopted the same policy, the long prevailing national policy, for those situations.

The Senate, and at the end the Congress, were not declining expressly to authorize labor injunctions only to turn squarely about and nullify that refusal in the same breath, merely by virtue of the fact that the employees of seized plants necessarily were made subject temporarily to ultimate governmental operating direction and control. We cannot attribute to Congress an intent so duplicitous. Thus to construe the Act not only would bring the provision for temporary control into collision with its remedial provisions as the history shows they were intended to apply. It would be to find Congress guilty of using a devious method for achieving indirectly exactly the thing it expressly declined to do. The words 'governmental employee,' 'employee * *  * for the purposes of this case' or 'relationship *  *  * of employer and employee,' none of which appear in the statute, cannot be given effect consistently with our function to write into the Act, by judicial interpolation, remedial provisions which Congress flatly and finally declined to incorporate.

Whether Congress acted wisely in this refusal is not our concern. But it is not irrelevant to the Act's meaning, purpose and effect that there were good reasons, indeed strong ones, for Congress to continue to follow the Norris-LaGuardia Act's policy rather than break away from it at that crucial time. Under the statute practically every industrial or mining facility, together with many of transportation, was subject to seizure and governmental operation. Introducing the labor injunction into the Act's structure therefore would have been tantamount to repeal of the Norris-LaGuardia Act for the duration of the emergency powers, since seizure was authorized whenever the President should find, after investigation, and proclaim that there was an interruption of operations 'as a result of a strike or other labor disturbance.' § 3. Ready means thus would have been made available, if such had been the statute's purpose, for suspending the Norris-LaGuardia policy and provisions in any case where they might become operative.

Congress was thoroughly familiar with the history and effects of injunctions in labor disputes, with the long settled national policy against them, and with the universal abhorrence in the ranks of labor, however otherwise divided, toward them. In view of all these things Congress well may have felt and I think did feel, as my brother's recital of the history shows, that it was both unnecessary and unwise, perhaps would even be harmful to furtherance of the war effort, in substance to repeal the Norris LaGuardia policy for the duration of the war emergency and thus to resurrect, in that critical situation, the long disused instruments that Act had outlawed.

It is important in this connection that 1943, rather than 1945 or 1946, was the year in which the War Labor Disputes Act was adopted. We were then not yet over the hump of the war. But neither had we reached the peak of labor disturbances which came only after active hostilities ceased, more than two years later. The great body of American workers was bending to the patriotic duty of peak production for war purposes. By comparison with what occurred after the fighting ended, the volume of man-days lost was about one-tenth of the later postwar peak loss. Moreover, at that time the War Labor Board, specially constituted to deal with such disturbances, was functioning with a high degree of efficiency in their settlement. There was nevertheless strong feeling that labor disputes should not be allowed to interrupt war production, regardless of cause or blame. And from this arose the demand for more effective powers to deal with them.

It was in this setting and to meet the problems it had thrown up, not the later one out of which this controversy arose, that the War Labor Disputes Act was adopted. The Act was exactly what its title indicated, a measure for dealing with labor disputes in the emergency of the war. Congress, it is true, anticipated that for a limited period after the end of fighting the same emergency powers would be needed. But this does not mean that those powers were shaped, or are now to be measured in scope, so as to meet all of the situations which since have arisen in the vastly changed circumstances; or that Congress intended them to be met by repealing the settled policy against injunctions in labor disputes in the sweeping manner now accomplished by the Court's decision. On the contrary, in June of 1943, Congress dealt with the situation then before it and refused to authorize such relief because that situation did not demand this.

In view of all these considerations, I cannot believe that Congress, in effect and by indirection, was exerting its war power to the greatest possible extent or was thereby either repealing or suspending the nation's settled policy against injunctions in labor disputes. Rather, the conclusion is inescapable that Congress was relying exclusively upon the added powers of enforcement expressly conferred by the Act, namely, the power of seizure and the force of the criminal sanction, to accomplish the needed results.

These were in themselves powerful sanctions. They carried with them the added and very great sanction of aroused public opinion which would follow not simply upon interruption of essential war production but more particularly upon such an event in any facility taken over and operated under governmental auspices. Congress, after mature deliberation, concluded that these sanctions were adequate, and for that reason made them exclusive. In no other way can its repeated and final refusals to confer the strenuously sought equitable remedies be made consistent with the legislative and general history or be given meaning and effect. To construe the Act as permitting what Congress thus so explicitly refused to allow is to go beyond our function and intrude upon that of Congress. This we have no right or power to do. If the situation presented by the facts of this case is one which goes beyond the powers Congress has conferred for dealing with it, that is a matter for Congress' consideration, not for correction by this Court.

Accordingly, upon the specific terms of the War Labor Disputes Act itself, upon the legislative history as summarized by Mr. Justice FRANKFURTER, and upon the historical setting in which the statute was enacted as defining the problems it was designed to meet, together with shaping the nature and scope of the measures required to meet them, I conclude that that Act in no way impaired but on the contrary adopted and incorporated the policy of the Norris-LaGuardia Act concerning the issuance of injunctions in labor disputes.

This conclusion substantially compels the further one that United States v. Shipp, 203 U.S. 563, 27 S.Ct. 165, 51 L.Ed. 319, 8 Ann.Cas. 265, has no valid application to the situation presented by this case.

This Court has not yet expressly denied, rather it has repeatedly confirmed Congress' power to control the jurisdiction of the inferior federal courts and its own appellate jurisdiction. Const., Art. III, § 2. Ex parte McCradle, 7 Wall. 506, 19 L.Ed. 264; Lockerty v. Phillips, 319 U.S. 182, 187, 63 S.Ct. 1019, 1022, 87 L.Ed. 1339 and authorities cited. See Warren, New Light on the History of the Federal Judiciary Act of 1789 (1923), 37 Harv.L.Rev. 49, 67ff. That power includes the power to deny jurisdiction as well as to confer it. Ibid. And where Congress has acted expressly to exclude particular subject matter from the jurisdiction of any court, except this Court's original jurisdiction, I know of no decision here which holds the exclusion invalid, or that a refusal to obey orders or judgments contravening Congress' mandate is criminal or affords cause for punishment as for contempt.

If that were the law, the result could only be to nullify the congressional power over federal jurisdiction for a great volume of cases. And if it should become the law, for every case raising a question not frivolous co cerning the court's jurisdiction to enter an order or judgment, that punishment for contempt may be imposed irrevocably simply upon a showing of violation, the consequences would be equally or more serious. The force of such a rule, making the party act on pain of certain punishment regardless of the validity of the order violated or the court's jurisdiction to enter it as determined finally upon review, would be not only to compel submission. It would be also in practical effect for many cases to terminate the litigation, foreclosing the substantive rights involved without any possibility for their effective appellate review and determination.

This would be true, for instance, wherever the substantive rights asserted or the opportunity for exercising them would vanish with obedience to the challenged order. Cf. Ex parte Fisk, 113 U.S. 713, 5 S.Ct. 724, 28 L.Ed. 1117. The First Amendment liberties especially would be vulnerable to nullification by such control. Thus, the constitutional rights of free speech and free assembly could be brought to naught and censorship established widely over those areas merely by applying such a rule to every case presenting a substantial question concerning the exercise of those rights. This Court has refused to countenance a view so destructive of the most fundamental liberties. Thomas v. Collins, 323 U.S. 516, 65 S.Ct. 315, 89 L.Ed. 430. These and other constitutional rights would be nullified by the force of invalid orders issued in flat violation of the constitutional provisions securing them, and void for that reason. The same thing would be true also in other cases involving doubt, where statutory or other rights asserted or the benefit of asserting them would vanish, for any practical purpose, with obedience.

Indeed it was because these were so often the effects, not simply of final orders entered after determination upon the merits, but of interlocutory injunctions and ex parte restraining orders, that the Norris-LaGuardia Act became law and, as I think, the War Labor Disputes Act continued in force its policy. For in labor disputes the effect of such orders, it was pointed out officially and otherwise, is generally not merely failure to maintain the status quo pending final decision on the merits. It is also most often to break the strike, without regard to its legality or any conclusive determination on that account, and thus to render moot and abortive the substantive controversy.

It is not every case therefore where substantial doubt appears, concerning either the issues in the main cause or the court's jurisdiction to issue interlocutory or other orders, in which violation will bring the so-called Shipp doctrine into play. If that were true then indeed would a way have been found to nullify the constitutional limitations placed upon the powers of courts, including the control of Congress over their jurisdiction. Then also the liberties of our people would be placed largely at the mercy of invalid orders issued without power given by the Constitution and in contravention of power constitutionally withheld by Congress. Ex parte Fisk, 113 U.S. 713, 5 S.Ct. 724, 28 L.Ed. 1117; Thomas v. Collins, supra.

Indeed the Shipp doctrine thus broadly conceived would go far toward nullifying the historic jurisdiction of this Court and others in habeas corpus, for it would do this in the many situations where the cause of commitment is violation of a doubtfully valid court order and the ground asserted for release is the court's lack of jurisdiction to enter it. Thus, in this case, if the party Lewis had been imprisoned rather than fined, the broad application now made of the Shipp decision would dictate that he could not be released by habeas corpus, even though it were now held here that the restraining orders were beyond the District Court's jurisdiction to issue. If those orders were valid, for purposes of finally and conclusively imposing punishment in contempt, regardless of the court's want of power to issue them, this would be so whether the punishment were fine or imprisonment. And it clearly would follow in cases of criminal contempt, perhaps in others, that the court's lack of jurisdiction could furnish no basis for granting relief, unless the penalty were found to be cruel and unusual or, in the case of a fine, excessive.

I cannot believe that the historic powers of our courts in habeas corpus or the rights of citizens, confirmed as these have been for so long by an unbroken line of decisions, have been or can be overthrown and subverted, merely by the fact that the question of the court's power to issue the order violated may be doubtful and not merely frivolous. Nor do I think the Shipp decision accomplished or purported to accomplish so much.

Certainly if its purpose had been to overrule the decisions so thoroughly established, and to trench so heavily upon the historic liberties they and the Constitution itself secure, some note would have been taken of that fact. So great a revolution hardly could have been wrought unanimously or without attentive recognition of what was being done. There was indeed reference in the opinion to the previous decisions. The Court stated: 'It has been held, it is true, that orders made by a court having no jurisdiction to make them may be disregarded without liability to process for contempt,' citing the Sawyer, Fisk, and Rowland cases. 203 U.S. at page 573, 27 S.Ct. at page 166, 51 L.Ed. 319, 8 Ann.Cas. 265. But there was not the slightest suggestion, by this reference or otherwise, that the Court had any purpose whatever to impair the force of those decisions, much less to overrule them. Nor in fact was this its intent. It mentioned them only to put them aside as inapplicable to the situation before it.

Indeed, in Gompers v. Buck's Stove & Range Co., 221 U.S. 418, 31 S.Ct. 492, 55 L.Ed. 797, 34 L.R.A.,N.S. 874, decided five years after the Shipp decision, a unanimous Court joined in citing Ex parte Rowland, 104 U.S. at page 612, 26 L.Ed. 861, in context consistent only with the view that its doctrine, and therefore that of others like it decided prior to the Shipp case, remained fully effective. Page 436 of 221 U.S., page 496 of 31 S.Ct., 55 L.Ed. 797, 34 L.R.A.,N.S. 874. There was no intimation, as otherwise necessarily would have been given, that the Shipp decision had reversed or modified the Rowland case, or any like it, in any way. And in Ex parte Young, 209 U.S. 123, not only the Court, p. 143, but the opposing distinguished counsel, pp. 135, 139, 28 S.Ct. 441, 447, 52 L.Ed. 714, 13 L.R.A.,N.S., 932, 14 Ann.Cas. 764, all concurred in reaffirming the Rowland ruling. Harlan, J., dissenting, retracted his former contrary view (see note 19 supra) in this respect. 209 U.S. at pages 169, 174, 28 S.Ct. at pages 458, 460, 52 L.Ed. 714, 13 L.R.A.,N.S., 932, 14 Ann.Cas. 764. And Holmes, J., who spoke for the Court in the Shipp case, joined with the Court's reaffirmation of the Rowland doctrine in both the Gompers and Young opinions.

The Court in Shipp was dealing with a situation quite different from the ones presented in the previous decisions and in this case. In none of them was the action which violated the court's order such as would have defeated its jurisdiction not only to enter the order but also to proceed with the cause before it in any manner, except to deal with the matter of contempt. In them the Court was not faced with the necessity for taking action to vindicate its power to hear and determine the main controversy, as well as the incidental one arising upon the validity of the interlocutory or other order. Nor is it here.

But exactly such a situation was presented in the Shipp case. The conduct there held to be contempt not only was in itself criminal and in violation, as it turned out, of this Court's lawful order for taking the appeal in Johnson's case. It ousted this Court altogether of jurisdiction to take any action in that cause. It rendered the cause moot, thereby putting an end to any proceedings concerning it here or elsewhere. Shipp's alleged conduct constituted therefore the most serious possible interference with the due and orderly course of administering justice. It utterly destroyed the power of all courts to act. Further, the order violated was not made directly in contravention of an act of Congress, as was true in the Fisk case and, as I think, in this one. It rather was made in complete conformity with the statutes conferring authority on this Court to take jurisdiction of and hear such causes. Nothing in it violated either a congressional mandate and policy or the rights of any party.

Moreover the decision was not effective, as its doctrine is now said to be, to put Shipp to any choice of obedience on pain of certain punishment regardless of the violated order's validity or invalidity as ultimately determined on review. No such situation was presented on the facts, and no such ruling could properly have been made. Shipp had not been convicted. The case came here upon a challenge in limine, not after the event, made upon the pleadings in the contempt proceedings to their validity. The basis asserted was the invalidity of the order allowing the appeal in Johnson's case, for alleged want of jurisdiction of this Court to enter it. That contention was rejected and the order was held valid. It was in this connection only that the Court stated it had 'jurisdiction to determine its jurisdiction' in doubtful cases. That statement was not a ruling that, regardless of a violated order's ultimate validity as determined on review, punishment in contempt for violating it could be irrevocably imposed. It was merely a statement of the reason for the order's validity. The holding was that this Court had jurisdiction, as of course it does in doubtful as well as clear cases, to determine whether the federal courts-the Circuit Court and accordingly this Court also-had power to pass upon Johnson's petition for habeas corpus.

From that ruling and from it alone the consequence followed that Shipp could be held in contempt on proof, still to be made, that he had done acts in violation of the order as thus conclusively determined to be valid by the court of last resort. This was a far cry from holding that punishment in contempt can be laid irrevocably, regardless of the outcome on review concerning the order's validity. The Court by its ruling was not making void orders valid for purposes of punishment by way of contempt. Only if the Court has held its own order which Shipp violated invalid would such a question have been presented.

The Shipp decision therefore was in fact simply an application of the long established rule that punishment in contempt may be inflicted on proof of violation of a valid order of court as determined finally on review. It did not overrule, nor was it in any way inconsistent with the long prior course of decisions holding that when an order is void for want of jurisdiction it may be disobeyed with impunity pending but depending upon determination of its invalidity by appeal, habeas corpus, or other mode of review. Gompers v. Buck's Stove & Range Co., supra; Ex parte Young, supra. It was an application, in the circumstances presented, of the settled rule that one who takes it upon himself to violate an order of court he thinks void thereby takes the risk that on review he will be sustained and, in the contrary event and then only, will he be subject irrevocably to punishment for contempt. Ibid.

In my judgment this is the rule properly applicable in this case, the only one consistent with the settled and unvaried course of decision, with the commands of the War Labor Disputes Act, of the Norris-LaGuardia Act and with § 268 of the Judicial Code, 36 Stat. 1163, 28 U.S.C. § 385, 28 U.S.C.A. § 385.

Apart from immediate and other interferences with judicial proceedings not presented here, that section authorizes punishment for contempt only for disobedience of a 'lawful writ, process, order, rule, decree, or command of the said courts.' (Emphasis added.) The section by its terms, apart from the exceptions not here applicable, limits power to punish for contempt to violations of lawful orders, thereby necessarily excluding others. Nor did it purport to make lawful for that purpose interlocutory orders issued without jurisdiction as determined finally upon review.

This case, unlike the Shipp case, in no way involves interference with any of the legal proceedings or the due course of administering justice in any sense contemplated by § 268 or by the Shipp decision. No court, trial or appellate, was deprived by the defendant's conduct of jurisdiction or power to take any action in any of the proceedings, collateral or in the main suit, which existed at the beginning of the controversy. The order therefore falls exclusively within the concluding clause of § 268 and the power to punish for contempt on account of its violation depends by the command of that clause, upon the order's lawful character.

Since in my opinion the order was jurisdictionally invalid when issued, by virtue of the War Labor Disputes Act and its adoption of the Norris-LaGuardia Act's policy, it follows that the violation gave no sufficient cause for sustaining the conviction for contempt. Ex parte Fisk, supra. Lewis and the United Mine Workers necessarily took the risk that the order would be found valid on review and, in that event, that punishment for contempt would apply. They did not take the risk that it would apply in any event, even if the order should be found void as beyond the jurisdiction of the Court to enter. See the dissenting opinion in Carter v. United States, 5 Cir., 135 F.2d 858, 862. The Shipp case furnishes no precedent for such a view nor do I know of any other in this Court which does.

On the contrary that view has been long rejected, and I do not think we should disturb or depart from that settled course of decision now. 'If the command of the writ (of mandamus) was in excess of jurisdiction, so necessarily were the proceedings for contempt in not obeying.' Ex parte Rowland, 104 U.S. 604, 617, 618, 26 L.Ed. 861. The power of the federal courts to issue stay orders to maintain the status quo pending appeal, like other matters affecting their jurisdiction except in the case of this Court's original jurisdiction, is subject to Congress' control. That control has been exercised, in my view, to exclude such jurisdiction in cases of this character. And, this being true, I do not think either we or any other court subject to that mandate has power to punish as for contempt the violation of such an order issued in contravention of Congress' command. Ex parte Fisk, supra.

The issues concerning the manner in which the contempt proceeding was conducted are in themselves of great moment, apart from the foregoing conclusions which I think are dispositive of the controversy. And the Court's ruling upon them are of such a character that I cannot accede by silence.

At times in our system the way in which courts perform their function becomes as important as what they do in the result. In some respects matters of procedure constitute the very essence of ordered liberty under the Constitution. For this reason, especially in the Bill of Rights, specific guaranties have been put around the manner in which various legal proceedings shall be conducted. They differentiate sharply between the procedures to be followed in criminal proceedings and in civil ones. These differences mark one of the great constitutional divides. They separate the zone of punishment for crime, with all its odious consequences, from that of giving civil relief, where no such consequences attend, not partially but completely.

In any other context than one of contempt, the idea that a criminal prosecution and a civil suit for damages or equitable relief could be hashed together in a single criminal-civil hodgepodge would be shocking to every American lawyer and to most citizens. True, the same act may give rise to all these varied legal consequences. But we have never adopted, rather our Constitution has totally rejected the continental system of compounding criminal proceedings with civil adjudications. Our tradition is exactly the contrary and few would maintain that this has had no part in bringing about the difference existing today for individual freedom here and in Europe.

I do not think the Constitution contemplated that there should be in any case an admixture of civil and criminal proceedings in one. Such an idea is altogether foreign to its spirit. There can be no question that contempt power was conferred adequate to sustain the judicial function, in both civil and criminal forms. But it does not follow that the Constitution permits lumping the two together or discarding for the criminal one all of the procedural safeguards so carefully provided for every other such proceeding.

The founders did not command the impossible. They could not have conceived that procedures so irreconcilably inconsistent in many ways could be applied simultaneously. Nor was their purpose to create any part of judicial power, even in contempt, wholly at large, free from any constitutional limitation or to pick and choose between the conflicting civil and criminal procedures and remedies at will. Much less was it to allow mixing civil remedies and criminal punishments in one lumped form of relief, indistinguishably compounding them and thus putting both in unlimited judicial discretion, with no possibility of applying any standard of measurement on review.

If this can be done in any case, it can be done in others. And that being true, if it can be done at all, not simply a loophole but a very large breach has been left in the wall of procedural protections thrown around the citizen's punishment for crime. For it is to be recalled that under the Court's ruling here upon the Shipp doctrine not merely the violation of valid judicial orders, but also the disobedience of invalid orders issued in excess of any court's jurisdiction becomes a crime and punishable as such by summary proceedings in criminal contempt, although the substantive rights involved in the litigation are wholly civil ones. The vastly expanded area of criminal conduct under this conception would afford equally wide room for dispensing with the criminal procedural protections under the unrestricted scope, otherwise than by 'judicial discretion,' which the present ruling concerning criminal or criminal-civil proceedings in contempt affords.

In my opinion, our system does not comprehend a power so unconfined anywhere within its broad borders, and it is time the large confusion about this were swept away. It is not necessary in this case to ask or decide whether all of the Constitution's criminal procedural protections thrown about all other criminal prosecutions, without suggestion of explicit exception, apply to criminal contempt proceedings. It is enough that we are sure some of them apply, as this Court has ruled repeatedly. It does not matter that some of those which incontestably are applicable may not have been put in issue or preserved for review in this case. The question cuts more deeply than the application of any specific guaranty. It affects the right to insist upon or have the benefit of any.

The case is characteristic of the long existing confusion concerning contempts and the manner of their trial, among other things, in that most frequently the question of the nature and character of the proceeding, whether civil or criminal, is determined at its end in the stage of review rather than, as it should be and as in my opinion it must be, at the beginning. Gompers v. Buck's Stove & Range Co., 221 U.S. 418, 444, 31 S.Ct. 492, 499, 55 L.Ed. 797, 34 L.R.A.,N.S., 874. And this fact in itself illustrates the complete jeopardy in which rights are placed when the nature of the proceeding remains unknown and unascertainable until the final action on review.

Not only is one thus placed in continuing dilemma throughout the proceedings in the trial court concerning which set of procedural rights he is entitled to stand upon, whether upon the criminal safeguards or only on the civil. He also does not and cannot know until it is too late, that is, until the appellate phase is ended, whether one group or the other of appellate jurisdictional and procedural rules applies. Indeed he may find that his right of review has been taken either prematurely or too late depending entirely on whether the appellate court finally concludes that the proceeding has been civil or criminal in character. See Swayzee, Contempt of Court in Labor Injunction Cases (1935) 21 22.

Precisely for these reasons this Court, when confronted in the Gompers case, supra, with a proceeding commingling civil and criminal features, such as we have here, refused to countenance such a mixture and, finding that the proceedings had been civil, held the criminal penalty of fixed terms of imprisonment to be invalid. The Court said:

'There was therefore a departure-a variance-between the procedure adopted and the punishment imposed, when, in answer to a prayer for remedial relief, in the equity cause, the court imposed a punitive sentence appropriate only to a proceeding at law for criminal contempt. The result was as fundamentally erroneous as if in an action of 'A vs. B, for assault and battery,' and judgment entered had been that the defendant be confined in prison for twelve months.' 221 U.S. at page 449, 31 S.Ct. at page 501.

Not only must the punishments be kept separate and distinct. This must be done with the entire proceedings. Punishment and civil relief must be correlated with the character of the proceeding. Procedural rights not only in matters of practice, but in others 'which involve substantial rights and constitutional privileges,'  are so distinct and in some instances contradictory that 'manifestly' they cannot be intermingled. Nor can those applicable in criminal proceedings be disregarded when criminal penalty is sought. Not only such matters as the privilege against self-incrimination, the presumption of innocence, the necessity for proof beyond a reasonable doubt, the allowance of costs, the appropriate mode of review  with attendant limitations of time and other differences, require this. What is most important, because the application and observance of all these rights and others depend upon it, is that the person charged is entitled to know from the beginning, not merely at the end or some intermediate stage, in which sort of proceeding he is involved.

This, the Court said, 'is not a mere matter of form, for manifestly every citizen, however unlearned in the law, by a mere inspection of the papers in contempt proceedings ought to be able to see whether it was instituted for private litigation or for public prosecution * *  *. He should not be left in doubt as to whether relief or punishment was the object in view. He is not only entitled to be informed of the nature of the charge against him, but to know that it is a charge, and not a suit. United States v. Cruikshank, 92 U.S. 542, 559, 23 L.Ed. 588, 593.' 221 U.S. at page 446, 31 S.Ct. at page 500.

This rule has now been incorporated also in Rule 42(b) of the Federal Rules of Criminal Procedure, and was applicable in this case. By the terms of that rule the charge of criminal contempt was required to be 'prosecuted on notice' and it was further commanded that the notice state 'the essential facts constituting the criminal contempt charged and describe it as such,' which was not done here. The rule was adopted to outlaw 'the frequent confusion between criminal and civil contempt proceedings,' following immediately a suggestion made in McCann v. New York Stock Exchange, 2 Cir., 80 F.2d 211. See also Nye v. United States, 313 U.S. 33, at pages 42-43, 61 S.Ct. 810, at page 812, 813. But it flatly incorporates the effect of the decision in the Gompers case, supra.

The language used by the Court was language of the Constitution, reinforced by citation of the Cruikshank case. Careful as it was about expressly overruling prior decisions where the Sixth Amendment's requirement  had not been observed, there can be no doubt that the Court was announcing for the future that the constitutional requirement must be complied with. And the result in the case itself accorded with what this view required.

One who does not know until the end of litigation what his procedural rights in trial are, or may have been, has no such rights. He is denied all by a hide-and-seek game between those that are criminal and those that are civil. The view which would seem to be the only one consistent with the whole spirit of the Constitution, and with the nature of our free institutions, is that all of the constitutional guaranties applicable to trials for crime should apply to such trials for contempt, excepting only those which may be wholly inconsistent with the nature and execution of the function the court must perform. As has been said, courts in performing this function are not above the Constitution; rather they are empowere to perform it in order to make the Constitution itself operative. Accordingly, not the least but the greatest possible application of it to this phase of their work is the only rule consistent with their place in the constitutional scheme. In re Michael, 326 U.S. 224, 227, 66 S.Ct. 78, 79.

Hence, whatever may be true of indictment and jury trials, I see no compelling reason whatever for not applying the other limitations of the Sixth Amendment. None of them is inconsistent with the due and proper performance of the court's function in criminal contempt. Some at the least are applicable by virtue of the due process guaranty of the Fifth Amendment. 'Due process of law, therefore, in the prosecution of contempt, except of that committed in open court, requires that the accused should be advised of the charges and have a reasonable opportunity to meet them by way of defense or explanation. We think this includes the assistance of counsel, if requested, and the right to call witnesses to give testimony, relevant either to the issue of complete exculpation or in extenuation of the offense and in mitigation of the penalty to be imposed.' Cooke v. United States, 267 U.S. 517, 537, 45 S.Ct. 390, 395, 69 L.Ed. 767. Only one case, apart from those involving indictment or jury trial, has held the Sixth Amendment inapplicable in such proceedings. Whether or not that case was a departure from our long established tradition that in criminal proceedings the defendant is entitled to be confronted with the witnesses against him, other departures should not be made.

Surely the rights to a speedy and public trial, to have compulsory process for obtaining witness in his favor, to have the assistance of counsel for his defense, and, as the Gompers case held, to be informed of the nature as well as the cause of the accusation, cannot be denied in our system to any person charged with crime, with the single exception of contempts committed in the immediate presence of the court by way of interference with the proceedings. Those guaranties are in no way inconsistent with the court's proper and complete discharge of its function in contempt. And they would seem to be essential to any conception of a fair trial as the Fifth Amendment's due process clause comprehends this.

When the assertion and securing of all other rights depends upon one, that one is the core of all. Here the right 'to know that it was a charge, and not a suit' comprehended all other procedural rights in the trial and appellate courts. Without this, none could be asserted or maintained. The denial of that right, deferring it until the decision here is handed down, is in my opinion not only a denial of all. It is a violation both of the Constitution and of Rule 42(b).

But we are told that this, and all that followed or may have followed from it, make no difference because there was no prejudice. There are at least two answers. This Court has held that the denial of constitutional guaranties in trials for crime is in itself prejudice. Kotteakos v. United States, 328 U.S. 750, 765, 66 S.Ct. 1239, 1248, and cases cited in note 19. The other, there was prejudice and in the most important thing beyond knowing the nature of the proceeding in advance of trial, namely, in the penalty itself.

Not only was the penalty against the union excessive, as the Court holds. Vice infected both 'fines' more deeply. As the proceeding itself is said to have been both civil and criminal, so are the two 'fines.' Each was imposed in a single lump sum, with no allocation of specific portions as among civil damages, civil coercion and criminal punishment. The Government concedes that some part of each 'fine' was laid for each purpose. But the trial court did not state, and the Government has refused to speculate, how much was imposed in either instance for each of those distinct remedial functions.

This was in the teeth of the Gompers and other previous decisions here. The law has fixed standards for each remedy, and they are neither identical nor congealable. They are, for damages in civil contempt, the amount of injury proven and no more, Gompers v. Buck's Stove & R. Co., supra, 221 U.S. at page 444, 31 S.Ct. at page 499, 55 L.Ed. 797, 34 L.R.A.,N.S., 874; for coercion, what may be required to bring obedience and not more, whether by way of imprisonment or fine; for punishment, what is not cruel and unusual or, in the case of a fine, excessive within the Eighth Amendment's prohibition. And for determining excessiveness of criminal fines there are analogies from legislative action which in my opinion are controlling.

The Government concedes that the Eighth Amendment's limitation applies to penalties in criminal contempt; and that in civil contempt the damages awarded cannot exceed the proven amount of injury. It also concedes, as I understand, that purely coercive relief can be no greater than is necessary to secure obedience. But in its view there was no necessity here for allocation of specific amounts in order to comply with these distinct standards. Rather punishment and damages may be lumped with a third undefined amount for civil coercion; and the whole mass sustained, without reference to the constituent elements or any of the established standards for measuring them, other than by overall application of the Eighth Amendment's limitation to the mass. And in this view it maintains neither 'fine' is excessive.

Obviously, however, when all these distinct types and functions of relief are lumped together, in a single so-called 'fine,' none of the long established bases for measurement can be applied, for there is nothing to which they can apply. We can only speculate upon what portion of each 'fine' may have been laid to compensate for damages, what for punishment, and what, if any, for civil coercion. Moreover, the District Court made no findings whatever concerning the amount of civil damages sustained, even if it could be assumed that there was evidence to sustain such findings. And on the record none of the 'fine' was made contingent, affording an opportunity for compurgation, as is required for coercive penalties.

It follows that we have no basis except our own speculative imagination by which to determine whether the so-called 'fines,' or either of them, are excessive as damages, or indeed as coercive relief looking to the future, or as penalty for past crime.

In this state of things, it is utterly impossible to perform our function of review in the manner heretofore required, even within the broad limits prescribed for cases of civil and criminal contempt. This commingling of the various forms of relief, like that of the proceedings themselves, deprives these contemnors of any possibility for having the scope of the relief given against them measured according to law.

That is no insubstantial deprivation. When hybrid proceedings can produce hybrid penalties, concealing what is for punishment and what remedial, what criminal and what civil, and in the process can discard constitutional procedural protections against just such consequences, as convenience or other wholly discretionary impulse may command, then indeed to the extent we allow this will we have adopted the continental tradition of the civilians and rejected our own. No case in this Court heretofore has ever sustained such conglomerate proceedings and penalties.

That the Government is complainant here, both as 'employer' seeking remedial relief and in sovereign capacity seeking to vindicate the court's authority by criminal penalty, does not nullify all these long established limitations or put the courts wholly at large, limited by nothing except their unconfined discretion as to the scope and character of the relief allowable. Power there is to take adequate measures when violation is clearly shown and adequate proof is made to sustain them. For proven violation, criminal penalty within the Eighth Amendment's limits as we would measure similar impositions placed by Congress, at the most; for damages proven and found, civil award commensurate with the finding; and for coercion, civil relief by way of imprisonment or 'fine,' but in either case contingent only, not final, giving opportunity for compurgation and for termination, on its being made, of further penalty for the future.

These are the limitations the law has prescribed. They apply equally when the Government is complainant, and whether in one capacity or the other, or both, as when others are. They cannot be dispensed with, separately or by conglomerating all into a single indiscriminate lump, at the suit of the Government or another, in this case or for others. To permit this would be to throw overboard the limitations prescribed by law and make the courts purely discretionary arbitrators of controversies. That cannot be done in our system.

The Court seemingly recognizes this, in part, in the revision it makes of the District Court's penalties. Lewis' fine is affirmed in amount but wholly changed in character. Instead of composite relief as the District Court made it, the Court makes that fine wholly a criminal penalty, thus in effect increasing the amount of his criminal imposition. The union's fine, though held excessive and 'reduced,' by what standard is not apparent, is replaced by a flat criminal fine of $700,000 plus a contingent penalty of $2,800,000 said to be entirely for civil coercion, although the strike was ended in December. Any award for civil damages allegedly sustained apparently is eliminated.

The Court thus purports to make separate the distinct items of relief commingled in the District Court's action. But in doing so, in my opinion, it wholly disregards the established standard for measuring criminal fines and its own as well as the District Court's function relating to them. If Lewis and the union had been convicted on indictment and jury trial in a proceeding surrounded by all the constitutional and other safeguards of criminal prosecution for violating the War Labor Disputes Act, the maximum fines which could be applied by that Act's terms would be $5,000 for each. In addition, Lewis could have been imprisoned for a year.

In my opinion, when Congress prescribes a maximum penalty for criminal violation of a statute, that penalty fixes the maximum which can be imposed whether the conviction is in a criminal proceeding as such for its violation or is for contempt for violating an order of court to observe it temporarily. Gompers v. United States, 233 U.S. 604, 612, 34 S.Ct. 693, 696, 58 L.Ed. 1115. If the fine or other penalty in such a case can be multiplied twice or any other number of times, merely by bringing a civil suit, securing a temporary restraining order and then convicting the person who violates it of criminal contempt, regardless of the order's validity and of any of the usual restraints of criminal procedure, the way will have been found to dispense with substantially all of those protections relating not only to the course of the proceedings but to the penalty itself.

But it is in relation to the flat criminal fine of $700,000 against the union that the Court's disregard of the constitutional and other standards is most apparent. By what measuring rod this sum has been arrived at as the appropriate and lawful amount, I am unable to say, unless indeed it is simply by a rough estimate of what the union should be forced to pay on all counts. Never has a criminal fine of such magnitude been heretofore laid and sustained, so far as I am able to discover. And only for treason, with one other possible exception, has Congress authorized one so large. Moreover, the Court's enumeration of factors to be taken into account indicates expressly, as I read the opinion, that one is the coercive effect of the imposition for the future, though it is thoroughly settled that in contempt criminal punishment is to be l id only for past conduct. Gompers v. Buck's Stove & R. Co., supra, and authorities cited.

Thus, the Court in effect imposes double coercive penalties, in view of the additional contingent award of $2,800,000 for that specific and sole purpose. I think the criminal fine of $700,000 not only constitutionally excessive, far beyond any heretofore sustained for violation of any statute or order of court. It is also an unlawful commingling of civil coercive and criminal penalties, without the essential contingent feature in the coercive phase, under our prior decisions.

Moreover, it is the District Court's function, not ours, in the first instance to fix the amounts of criminal fines. In equity proceedings for coercive relief, appellate courts including this one have power to revise and fix awards for such purposes, and if damages also are sought to review amounts awarded for this purpose for consistency with the proof. Gompers v. Buck's Stove & R. Co., supra. But in a criminal proceeding which is at law even in contempt, ibid., our function is not in the first instance to fix the fines ourselves. That function is the District Court's. Ibid. We can only determine whether those imposed by it are excessive under the Eighth Amendment.

In its revision of the penalties therefore the Court in my opinion not only fails to unscramble the coercive and criminal elements, as the prior decisions here require to be done. It imposes grossly excessive criminal penalties, determined in amount by wholly arbitrary estimate related to no previously established standard legislatively or judicially fixed. And in doing so, it usurps the District Court's function. All this flows in part at least from its basic error, which is its failure to follow the rule of the Gompers and other cases that not only civil and criminal penalties, but also civil and criminal proceedings are altogether different and separate things, and under the Constitution must be kept so.

Much more is involved in this controversy than the issues which have been discussed. The issues in the main suit have not been determined and it would be beyond our function to intimate opinion concerning them now. But beyond this controversy as a whole lie still graver questions. They involve opposing claims concerning the right to strike and the power of the Government, as against this, to keep the nation's economy going. Those are indeed grave matters.

No right is absolute. Nor is any power, governmental or other, in our system. There can be no question that it provides power to meet the greatest crises. Equally certain is it that under 'a government of laws and not of men' such as we possess, power must be exercised according to law; and government, including the courts, as well as the governed, must move within its limitations.

This means that the courts and all other divisions or agencies of authority must act within the limits of their respective functions. Specifically it means in this case that we are bound to act in deference to the mandate of Congress concerning labor injunctions, as in judgment and conscience we conceive it to have been made. The crisis here was grave. Nevertheless, as I view Congress' action, I am unable to believe that it has acted to meet, or authorized the courts to meet, the situation which arose in the manner which has been employed.

No man or group is above the law. All are subject to its valid commands. So are the government and the courts. If, as I think, Congress has forbidden the use of labor injunctions in this and like cases, that conclusion is the end of our function. And if modification of that policy is to be made for such cases, that problem is for Congress in the first instance, not for the courts.

Mr. Justice MURPHY joins in this opinion.