delivered the opinion of the Court.

Article of War 92, 10 U. S. C. (1946 ed., Supp. IV) § 1564, which, prior to the adoption of the Uniform Code of Military Justice,1 governed trials for murder or rape before courts-martial,2 contained a proviso “That no person shall be tried by court-martial for murder or rape committed within the geographical limits of the States of the Union and the District of Columbia in time of peace.”

The question for decision concerns the meaning of the words “in time of peace” in the context of Article 92.

Petitioner, while serving with the United States Army in France, was convicted by a court-martial, dishonorably discharged, and sentenced to prison for 20 years. He was serving that sentence in the custody of the Army at Camp Cooke, California, when he was convicted by a court-martial of the crime of conspiracy to commit murder. This offense occurred on June 10, 1949, at Camp Cooke. The question is whether June 10, 1949, was “in time of peace” as the term was used in the 92d Article. The question was raised by a petition for a writ of habeas corpus challenging the jurisdiction of the court-martial. Both the District Court (148 F. Supp. 23) and the Court of *230Appeals (248 F. 2d 783) ruled against petitioner. We granted certiorari, 356 U. S. 911.

The Germans surrendered on May 8, 1945 (59 Stat. 1857), the Japanese on September 2,1945 (59 Stat. 1733). The President on December 31, 1946, proclaimed the cessation of hostilities, adding that “a state of war still exists.” 61 Stat. 1048. In 1947, Senate Joint Resolution 123 was passed (61 Stat. 449) which terminated, inter alia, several provisions of the Articles of War 3 but did not mention Article 92. The war with Germany terminated October 19, 1951, by a Joint Resolution of Congress (65 Stat. 451) and a Presidential Proclamation (66 Stat. c3). And on April 28, 1952, the formal declaration of peace and termination of war with Japan was proclaimed by the President (66 Stat. c31), that being the effective date of the Japanese Peace Treaty. Since June 10, 1949 — the critical date involved here — preceded these latter dates, and since no previous action by the political branches of our Government had specifically lifted Article 92 from the “state of war” category, it is argued that we were not then “in time of peace” for the purposes of Article 92. That argument gains support from a dictum in Kahn v. Anderson, 255 U. S. 1, 9-10, that the term “in time of peace” as used in Article 92 “signifies peace in the complete sense, officially declared.” Of like tenor are generalized statements that the termination of a “state of war” is “a political act” of the other branches of Government, not the Judiciary. See Ludecke v. Watkins, 335 U. S. 160, 169. We do not think that either of those authorities is dispositive of the present controversy. A more particularized and discriminating analysis must be made. We deal with a term that must be construed in light of the precise facts *231of each case and the impact of the particular statute involved. Congress in drafting laws may decide that the Nation may be “at war” for one purpose, and “at peace” for another. It may use the same words broadly in one context, narrowly in another. The problem of judicial interpretation is to determine whether “in the sense of this law” peace had arrived. United States v. Anderson, 9 Wall. 56, 69. Only mischief can result if those terms are given one meaning regardless of the statutory context.

In the Kahn case, the offense was committed on July 29, 1918, and the trial started November 4, 1918 — both dates being before the Armistice.4 It is, therefore, clear that the offense was not committed “in time of peace.” Moreover, a military tribunal whose jurisdiction over a case attaches in a time of actual war does not lose jurisdiction because hostilities cease. Once a military court acquires jurisdiction that jurisdiction continues until the end of the trial and the imposition of the sentence. See Carter v. McClaughry, 183 U. S. 365, 383. The broad comments of the Court in the Kahn case on the meaning of the term “in time of peace” as used in Article 92 were, therefore, quite unnecessary for the decision.

Ludecke v. Watkins, 335 U. S. 160, belongs in a special category of cases dealing with the power of the Executive or the Congress to deal with the aftermath of problems which a state of war brings and which a cessation of hostilities does not necessarily dispel. That case concerns the power of the President to remove an alien enemy after hostilities have ended but before the political branches have declared the state of war ended. Hamilton v. Kentucky Distilleries & Warehouse Co., 251 U. S. 146, involves the constitutionality under the war power of a prohibition law *232passed in 1918 after the armistice with Germany was signed and to be operative “until the conclusion of the present war and thereafter until the termination of demobilization, the date of which shall be determined and proclaimed by the President of the United States.” Woods v. Miller Co., 333 U. S. 138, concerns the constitutionality of control of housing rentals promulgated after hostilities were ended and before peace was formally declared. These cases deal with the reach of the war power, as a source of regulatory authority over national affairs, in the aftermath of hostilities. The earlier case of McElrath v. United States, 102 U. S. 426, is likewise irrelevant to our problem. It was a suit for back pay by an officer, the outcome of which turned on a statute which allowed dismissal of an officer from the service “in time of peace” only by court-martial. The President had made the dismissal; and the Court held that such action, being before August 20, 1866, when the Presidential Proclamation announced the end of the rebellion and the existence of peace, was lawful, since there was extrinsic evidence that Congress did not intend the statute to be effective until the date of the Proclamation.

Our problem is not controlled by those cases. We deal with the term “in time of peace” in the setting of a grant of power to military tribunals to try people for capital offenses. Did Congress design a broad or a narrow grant of authority? Is the authority of a court-martial to try a soldier for a civil crime, such as murder or rape, to be generously or strictly construed? Cf. Duncan v. Kahanamoku, 327 U. S. 304.

We do not write on a clean slate. The attitude cf a free society toward the jurisdiction of military tribunals — our reluctance to give them authority to try people for nonmilitary offenses — has a long history.

We reviewed both British and American history, touching on this point, in Reid v. Covert, 354 U. S. 1, 23-30. *233We pointed out the great alarms sounded when James II authorized the trial of soldiers for nonmilitary crimes and the American protests that mounted when British courts-martial impinged on the domain of civil courts in this country. The views of Blackstone on military jurisdiction became deeply imbedded in our thinking: “The necessity of order and discipline in an army is the only thing which can give it countenance; and therefore it ought not to be permitted in time of peace, when the king’s courts are open for all persons to receive justice according to the laws of the land.” 1 Blackstone’s Commentaries 413. And see Hale, History and Analysis of the Common Law of England (1st ed. 1713), 40-41. We spoke in that tradition in Toth v. Quarles, 350 U. S. 11, 22, “Free countries of the world have tried to restrict military tribunals to the narrowest jurisdiction deemed absolutely essential to maintaining discipline among troops in active service.”

The power to try soldiers for the capital crimes of murder and rape was long withheld. Not until 1863 was authority granted. 12 Stat. 736. And then it was restricted to times of “war, insurrection, or rebellion.” 5 The theory was that the civil courts, being open, were wholly qualified to handle these cases. As Col. William Winthrop wrote in Military Law and Precedents (2d ed. 1920) 667, about this 1863 law:

“Its main object evidently was to provide for the punishment of these crimes in localities where, in consequence of military occupation, or the prevalence *234of martial law, the action of the civil courts is suspended, or their authority can not be exercised with the promptitude and efficiency required by the exigencies of the period and the necessities of military government.”

Civil courts were, indeed, thought to be better qualified than military tribunals to try nonmilitary offenses. They have a more deeply engrained judicial attitude, a more thorough indoctrination in the procedural safeguards necessary for a fair trial. Moreover, important constitutional guarantees come into play once the citizen— whether soldier or civilian — is charged with a capital crime such as murder or rape. The most significant of these is the right to trial by jury, one of the most important safeguards against tyranny which our law has designed.6 We must assume that the Congress, as well as *235the courts, was alive to the importance of those constitutional guarantees when it gave Article 92 its particular phrasing. Statutory language is construed to conform as near as may be to traditional guarantees that protect the rights of the citizen. See Ex parte Endo, 323 U. S. 283, 301-304; Rowoldt v. Perfetto, 355 U. S. 115; Kent v. Dulles, 357 U. S. 116, 129. We will attribute to Congress a purpose to guard jealously against the dilution of the liberties of the citizen that would result if the jurisdiction of military tribunals were enlarged at the expense of civil courts. General Enoch H. Crowder, Judge Advocate General, in testifying in favor of the forerunner of the present proviso of Article 92, spoke of the protection it extended the officer and soldier by securing them “a trial by their peers.” 7 We think the proviso should be read generously to achieve that end.

We refused in Duncan v. Kahanamoku, 327 U. S. 304, to construe “martial law,” as used in an Act of Congress, broadly so as to supplant all civilian laws and to substitute military for judicial trials of civilians not charged with violations of the law of war. We imputed to Congress an attitude that was more consonant with our traditions of civil liberties. We approach the analysis of the *236term “in time of peace” as used in Article 92 in the same manner. Whatever may have been the plan of a later Congress in continuing some controls long after hostilities ceased,8 we cannot readily assume that the earlier Congress used “in time of peace” in Article 92 to deny soldiers or civilians the- benefit of jury trials for capital offenses four years after all hostilities had ceased. To hold otherwise would be to make substantial rights turn on a fiction. We will not presume that Congress used the words “in time of peace” in that sense. The meaning attributed to them is at war with common sense, destructive of civil rights, and unnecessary for realization of the balanced scheme promulgated by the Articles of War. We hold that June 10,1949, was “in time of peace” as those words were used in Article 92. This conclusion makes it unnecessary for us to consider the other questions presented, including the constitutional issues which have been much mooted.

Reversed.

The Court today holds that on June 10, 1949, the date of this capital offense, this country was “in time of peace” within the meaning of Article of War 92, 10 U. S. C. (1946 ed., Supp. IV) § 1564, and therefore that the court-martial before which petitioner was tried was without statutory jurisdiction to entertain the proceeding. Believing that the ground upon which the Court nullifies petitioner’s conviction has long been settled squarely to the contrary, and that a de novo examination of the question also requires the conclusion that the United States, on June 10, 1949, was not “in time of peace” within the meaning of Article 92, I respectfully dissent.

In Kahn v. Anderson, 255 U. S. 1, 10, this Court unanimously held that the term “in time of peace” in Article 92 “signifies peace in the complete sense, officially declared.” See also Givens v. Zerbst, 255 U. S. 11, 21. The Court now dismisses this square holding as “dictum” and as “quite unnecessary for the decision,” pointing out that the statement of facts in Kahn shows that the capital offense for which petitioner there was tried was committed before the Armistice which resulted in the termination of active hostilities in World War I, and that the court-martial which tried him was also convened before the Armistice. I think that Kahn can hardly be dismissed so lightly. The conclusion there as to the meaning of “in time of peace” might have been regarded as unnecessary to decision only had the Court, proceeding on a theory entirely different from that which it actually adopted, relied on the date of the offense or of the beginning of trial as dispositive. But plainly the Court did not proceed on any such basis. Rather, it accepted at least arguendo petitioner’s contention that the court-martial which had tried him did not have jurisdiction *238to continue “in time of peace” even a trial previously-begun. It is thus not sound to say that the holding that “peace” in Article 92 “signifies peace in the complete sense, officially declared,” was unnecessary to the decision in Kahn. Given the ground upon which the Court chose to decide the case it was quite indispensable. The idea that the ground on which a court actually decides a case becomes dictum because the case might have been decided on another ground is novel doctrine to me.

I think that Congress, and the military authorities charged with the implementation and enforcement of the Articles of War, should be able to rely on a construction given one of those Articles by a unanimous decision of this Court. The conclusion in Kahn was not reached lightly without full consideration, as is shown by the fact that nearly two pages of the summary of counsels’ argument contained in the report of the case are devoted to a discussion of the question, and another two pages to the Court’s expression of the reasoning underlying its decision on the point. In 1948, 27 years after Kahn and a single year before the prosecution here involved, Congress re-enacted Article 92 without change in the relevant language. The Court now holds that between 1921 and 1949 the meaning of the statute underwent an inexplicable change, and that the authority under the statute then confirmed must now be denied. I see no warrant for thus speculating anew as to the motives of Congress in enacting and re-enacting the phrase “in time of peace” in Article 92.1

*239Entirely apart from Kahn, I think today’s decision is demonstrably wrong. This Court has consistently for nearly 100 years recognized, in many contexts, that a cessation of active hostilities does not denote the end of “war” or the beginning of “peace” as those or similar terms have been used from time to time by Congress in legislation. In McElrath v. United States, 102 U. S. 426, there was before the Court a statute of Congress prohibiting summary dismissal by the President of military officers “in time of peace.” Although I venture to say that almost as many reasons could be conjured up for construing the term loosely in that context as in that now before us, the Court unanimously held that July 1866 was not “in time of peace” although active hostilities between North and South had long since ceased, and that “peace, in contemplation of law” did not exist until the Presidential Proclamation of August 20, 1866. See also United States v. Anderson, 9 Wall. 56. In Ludecke v. Watkins, 335 U. S. 160, 168-169, this Court in construing a statute recognized that “ ‘The state of war’ may be terminated by treaty or legislation or Presidential proclamation. Whatever the mode, its termination is a political act.” See also Woods v. Miller Co., 333 U. S. 138; Knauff v. Shaughnessy, 338 U. S. 537, both expressly recognizing that the state of war between this country and the Axis powers was not terminated by either the Presidential Proclamation of 1946 or the Joint Resolution of July 1947.

The Court says that “Congress in drafting laws may decide that the Nation may be ‘at war’ for one purpose, and ‘at peace’ for another.” Of course it may. But the Court points to no case, and I know of none, which has *240construed statutory language similar to that found in Article 92 to mean anything but “peace in the complete sense, officially declared.” Under these circumstances, and given McElrath and Kahn, the conclusion seems to me unmistakable that Congress intended that “peace” in Article 92 mean what we have always, until today, held it meant in this and other congressional legislation. When Congress has wished to define “war” or “peace” in particular statutes as meaning something else, it has explicitly done so. See, e. g., War Brides Act, 59 Stat. 659: “For the purpose of this Act, the Second World War shall be deemed to have commenced on December 7, 1941, and to have ceased upon the termination of hostilities as declared by the President or by a joint resolution of Congress.”

Today’s decision casts a cloud upon the meaning of all federal legislation the impact of which depends upon the existence of “peace” or “war.” Hitherto legislation of this sort has been construed according to well-defined principles, the Court looking to “treaty or legislation or Presidential proclamation,” Ludecke v. Watkins, 335 U. S., at 168, to ascertain whether a “state of war” exists. The Court, in an effort to make a “more particularized and discriminating analysis,” has apparently jettisoned these principles. It is far from clear to me just what has taken their place.2

*241The Court does not reach petitioner’s contention that he could not constitutionally be tried by court-martial because he was not a member of the armed forces at the time this offense was committed. It is sufficient to say that this contention is also squarely foreclosed by Kahn v. Anderson, supra, and that in my opinion nothing in Toth v. Quarles, 350 U. S. 11, or in Reid v. Covert, 354 U. S. 1, impairs the authority of Kahn on this score.

I would affirm.