with whom Mr. Justice Stewart joins, dissenting.
Some 30 years ago, Mr. Justice Black warned his Brethren against the “gradual process of judicial erosion which . . . has slowly worn away a major portion of the essential guarantee of the Seventh Amendment.” Galloway v. United States, 319 U. S. 372, 397 (1943) (dissenting opinion). Today, the erosion process reaches bedrock. In the past, this Court has sanctioned changes in “mere matters of form or procedure” in jury trials, Baltimore & Carolina Line, Inc. v. Redman, 295 U. S. 654, 657 (1935), and in “pleading or practice” before juries, Walker v. New Mexico & S. P. R. Co., 165 U. S. 593, 596 (1897). But before today, we had always insisted that “[wjhatever may be true as to legislation which changes any mere details of a jury trial, it is clear that a statute which destroys [a] substantial and essential feature thereof is one abridging the right.” American Publishing Co. v. Fisher, 166 U. S. 464, 468 (1897). See also Dimick v. Schiedt, 293 U. S. 474 (1935); Capital Traction Co. v. Hof, 174 U. S. 1 (1899).
Now, however, my Brethren mount a frontal assault on the very nature of the civil jury as that concept has been understood for some seven hundred years. No one need be fooled by reference to the six-man trier of fact utilized in the District Court for the District of Montana as a “jury.” This six-man mutation is no more a “jury” than the panel of three judges condemned in Baldwin v. New York, 399 U. S. 66 (1970), or the 12 laymen instructed by a justice of the peace outlawed in Capital Traction Co. v. Hof, supra. We deal here not with some minor tinkering with the role of the civil jury, but with its wholesale abolition and replacement with a different institution which functions differently, produces different *167results,1 and was wholly unknown to the Framers of the Seventh Amendment.2
In my judgment, if such a radical restructuring of the *168judicial process is deemed wise or necessary, it should be accomplished by constitutional amendment. See, &. g., Tamm, The Five-Man Civil Jury: A Proposed Constitutional Amendment, 51 Geo. L. J. 120 (1962). It appears, however, that the common-law jury is destined to expire, not with a bang, but a whimper. The proponents of the six-man jury have not secured the approval of two-thirds of both Houses of Congress and three-fourths of the state legislatures for their proposal. Indeed, they have not even secured the passage of simple legislation to accomplish their goal. Instead, they have relied upon the interstitial rulemaking power of the majority of the district court judges sitting in a particular district to rewrite the ancient definition of a civil jury.3 They have done so, moreover, in the teeth of an Act of Congress and a Federal Rule promulgated by this Court *169which, in my judgment, were designed to guarantee the 12-man civil jury. By approving this mode of procedure, the Court turns the so-called “clear statement” rule on its head. Instead of requiring a clear statement from Congress when it legislates at the limit of its constitutional powers, see, e, g., Crowell v. Benson, 285 U. S. 22, 62 (1932), my Brethren approve a departure from settled constitutional understanding despite a clear statement from Congress that it intended no such thing. I must respectfully dissent.
I
At the outset, it should be noted that the constitutional issue in this case is not settled by the prior decisions of this Court upholding nonunanimous and six-man criminal juries. See Apodaca v. Oregon, 406 U. S. 404 (1972); Johnson v. Louisiana, 406 U. S. 356 (1972); Williams v. Florida, 399 U. S. 78 (1970). This is true for at least three reasons.
First, Apodaca, Johnson, and Williams all involved state trials and, therefore, the requirements of the Fourteenth Amendment rather than the Sixth. This case is, of course, distinguishable in that it deals with a federal trial and, therefore, with Bill of Rights guarantees which are directly applicable, rather than applicable only through the incorporation process.4 Thus, neither Apodaca, Johnson, nor Williams squarely presented the Court with the problem of defining the meaning of jury trial in a federal context.5 Indeed, as *170my Brother Powell’s concurring opinion in Apodaca and Johnson makes plain, there were, as of last Term at least, five Members of this Court who thought that the Sixth Amendment required unanimous jury verdicts in federal cases. See also Johnson v. Louisiana, supra, at 395 (Brennan, J., dissenting). Mr. Justice Powell argued in that opinion that the “process of determining the content of the Sixth Amendment right to jury trial has long been one of careful evaluation of, and strict adherence to the limitations on, that right as it was known in criminal trials at common law.” Id., at 370 n. 6 He concluded that the Sixth Amendment required unanimous federal juries because “[a]t the time the Bill of Rights was adopted, unanimity had long been established as one of the attributes of a jury conviction at common law.” Id., at 371. See also Williams v. Florida, supra, at 123-125 (opinion of Harlan, J.). It is apparently uncontested that in 1791, common-law civil juries consisted of 12 men. See infra, at 177. Thus, to the extent that Sixth Amendment precedent is applicable to Seventh Amendment problems, Johnson and Apodaca would seem to cut strongly in favor of a 12-man jury requirement in federal court, rather than against such a requirement.
Moreover, even if it is assumed that the holdings in Apodaca, Williams, and Johnson are readily transferable to a federal context, it still does not follow that the definitions of trial by jury for purposes of the Sixth and Seventh Amendments are necessarily coextensive. The two Amendments use different language and they guarantee different rights. Indeed, as the Williams court itself recognized, the approval of six-man juries in crim*171inal cases did not resolve “whether, for example, additional references to the ‘common law’ that occur in the Seventh Amendment might support a different interpretation.” 399 U. S., at 92 n. 30.
The Court today goes to great lengths to show that the reference in the Seventh Amendment to “Suits at common law” speaks only to the type of suit in which a jury is required, not to the type of jury which is required in such suits. However, my brethren totally ignore another textual difference between the Sixth and Seventh Amendments which I consider to be of at least equal significance. Whereas the Sixth Amendment refers only to “an impartial jury,” the Seventh Amendment states that “the right of trial by jury shall be preserved” (emphasis added). The Seventh Amendment’s additional reference to the preservation of the right strongly suggests that the content of that right is to be judged by historical standards.
Certainly, that has been this Court’s understanding in the past. In Dimick v. Schiedt, for example, the Court held that the Seventh Amendment “in effect adopted the rules of the common law, in respect of trial by jury, as these rules existed in 1791,” 293 U. S., at 487, and the dissent agreed that the purpose of the Seventh Amendment was “to preserve the essentials of the jury trial as it was known to the common law before the adoption of the Constitution.” Id., at 490. In Baltimore & Carolina Line, Inc. v. Redman, the Court held that the “right of trial by jury thus preserved [by the Seventh Amendment] is the right which existed under the English common law when the Amendment was adopted.” 295 U. S., at 657. And in American Publishing Co. v. Fisher, the Court held that what was guaranteed by the Seventh Amendment was “the peculiar and essential features of trial by jury at the common law.” 166 U. S., at 468. It should therefore be *172clear that, whereas the words of the Sixth Amendment might be read as permitting a functional approach which measures “Sixth Amendment values,” the Seventh Amendment requires a historical analysis geared toward determination of what the institution was in 1791 which the Framers intended to “preserve.” See also Slocum v. New York Life Ins. Co., 228 U. S. 364 (1913); Capital Traction Co. v. Hof, 174 U. S. 1 (1899).
Finally, it is important to note that, whereas the legislative history of the Sixth Amendment tended to support the Court’s decision in favor of six-man criminal juries, it is at best ambiguous in the Seventh Amendment context. As the Court pointed out in Williams, the Sixth Amendment as originally introduced by James Madison in the House provided “[t]he trial of all crimes . . . shall be by an impartial jury of freeholders of the vicin-age, with the requisite of unanimity for conviction, of the right of challenge, and other accustomed requisites.” 1 Annals of Cong. 435 (1789) (emphasis added). The Amendment passed the House in this form, but when it reached the Senate, that body expressly rejected the “accustomed requisites” language, see Senate Journal, Sept. 9, 1789, 1st Cong., 1st Sess., 77, and the Amendment as ultimately adopted contained no reference to the common-law features of jury trial.
In contrast, the history of the Seventh Amendment contains no express rejection of language which would fix the common-law attributes of the civil jury. Indeed, as the Court itself recognizes, the extant history of the Amendment is exceedingly sketchy. See generally Henderson, The Background of the Seventh Amendment, 80 Harv. L. Rev. 289 (1966). Undeterred by the absence of source material, however, my Brethren concoct an elaborate theory designed to demonstrate that the Framers did not intend to fix the nature of the civil jury as it existed at common law. As I read the *173majority opinion, the theory is based on the following syllogism:
1. The delegates to the Constitutional Convention considered a clause which would have protected the right to a civil jury, but declined to adopt such a provision because state practice varied widely as to the cases in which a civil jury was provided.
2. When the Seventh Amendment was passed, Congress overrode the arguments of those opposed to a constitutional jury '■ guarantee and decided to provide a federal right of jury trial despite differences between the States as to when jury rights attached.
3. Therefore, in the words of the Court “[w]e can only conclude . . . that . . . the Framers of the Seventh Amendment were concerned with preserving the right of trial by jury in civil cases where it existed at common law, rather than the various incidents of trial by jury.”
It hardly requires demonstration that this “logic” rests on the flimsiest of inferences. It simply does not follow that because the Amendment was, at one stage rejected because of disparities among the States in the instances in which the jury right attached, its scope is therefore limited to the surmounting of these disparities. Indeed, the opposite conclusion is equally plausible. One could argue that, whereas there was dispute as to the cases in which the jury-trial right would attach, it was common ground between opponents and proponents of the measure that when it did attach, its incidents would be as at common law. Thus, whatever the meaning of the Amendment as to jury usage, the nature of the jury is, by this argument, at its core and agreed to by all parties.
Moreover, even if the Court’s chain of reasoning were correct, the argument would still fall, since it is grounded on a faulty major premise. True, the opponents of a jury guarantee at the Constitutional Convention rested *174their argument in part on the varying practice in the States as to the cases in which the right of jury trial attached. But a more detailed examination of the debates than the Court’s highly selective quotations permit makes clear that the opponents also rested on the differences in the characteristics of jury trial between the States. Thus, when a jury guarantee was first proposed, Mr. Gorham, one of the principal drafters of the Constitution, argued against the proposal, stating: “It is not possible to discriminate equity cases from those in which juries are proper. The Representatives of the people may be safely trusted in this matter.” 2 M. Farrand, Records of the Federal Convention 587 (1911) (hereinafter cited as Farrand). But when the proposal came to a final vote, Mr. Gorham made a somewhat different argument: “The constitution of Juries is different in different States.” Id., at 628 (emphasis added). Similarly, while at one stage James Wilson defended the absence of a jury requirement on the ground that “[t]he cases open to a jury, differed in different states,” 3 Farrand 101, he also made a quite different argument:
“By the constitution of the different States, it will be found that no particular mode of trial by jury could be discovered that would suit them all. The manner of summoning jurors, their qualifications, of whom they should consist, and the course of their proceedings, are all different, in the different States; and I presume it will be allowed a good general principle, that in carrying into effect the laws of the general government by the judicial department, it will be proper to make the regulations as agreeable to the habits and wishes of the particular States as possible; and it is easily discovered that it would have been impracticable, by any general regulation, to have given satisfaction to all. 3 Farrand 164.
*175Thus, it is clear that opponents of a jury guarantee were concerned not only with the differing rules for when juries were required among the States, but also with the differing content of the jury right itself.6 To the extent that anything at all can be inferred from the rejection of these arguments, it follows by the Court’s own chain of reasoning that the Framers intended to override state differences as to both the cases in which a jury right would attach and the characteristics of the jury itself.
I should hasten to add that I do not mean to embrace that chain of reasoning. In fact, as indicated above, I view the legislative history as far too fragmentary to support any firm conclusion. But I would have thought that the very uncertainty of the legislative history would support a mode of analysis which looked to the jury as it existed at the time the Seventh Amendment was written in order to determine the intent of the Framers. As Mr. Justice Harlan argued:
“[I]t is common sense and not merely the blessing of the Framers that explains this Court’s frequent reminders that: ‘The interpretation of the Constitution of the United States is necessarily influenced by the fact that its provisions are framed in the language of the English common law, and are to be read in the light of its history.’ Smith v. Alabama, 124 U. S. *176465, 478 (1888). This proposition was again put forward by Mr. Justice Gray speaking for the Court in United States v. Wong Kim Ark, 169 U. S. 649 (1898), where the Court was called upon to define the term 'citizen’ as used in the Constitution. 'The Constitution nowhere defines the meaning of these words [the Citizenship Clause]. ... In this, as in other respects, it must be interpreted in the light of the common law, the principles and history of which were familiarly known to the framers of the Constitution.’ 169 U. S., at 654. History continues to be a wellspring of constitutional interpretation. Indeed, history was even invoked by the Court in such decisions as Townsend v. Sain, 372 U. S. 293 (1963), and Fay v. Noia, 372 U. S. 391 (1963), where it purported to interpret the constitutional provision for habeas corpus according to the ‘historic conception of the writ’ and took note that the guarantee was one rooted in common law and should be so interpreted. Cf. United States v. Brown, 381 U. S. 437, 458 (1965).” Williams v. Florida, 399 U. S., at 123-124.
When a historical approach is applied to the issue at hand, it cannot be doubted that the Framers envisioned a jury of 12 when they referred to trial by jury. It is true that at the time the Seventh Amendment was adopted, jury usage differed in several respects among the States. See generally Henderson, The Background of the Seventh Amendment, 80 Harv. L. Rev. 289 (1966). But, for the most part at least, these differences did not extend to jury size which seems to have been uniform and, indeed, had remained so for centuries. One authority has noted that as early as 1164, the Constitutions of Clarendon provided that “where, in the case of a layman so rich and powerful that no individual dares *177to appear against him, 'the sheriff shall cause twelve legal men of the neighbourhood, or of the vill, to take an oath in the presence of the bishop that they will declare the truth about it.’ ” Wells, The Origin of the Petit Jury, 27 L. Q. Rev. 347 (1911). As Professor Scott wrote, “At the beginning of the thirteenth century twelve was indeed the usual but not the invariable number. But by the middle of the fourteenth century the requirement of twelve had probably become definitely fixed. Indeed this number finally came to be regarded with something like superstitious reverence.” A. Scott, Fundamentals of Procedure in Actions at Law 75-76 (1922) (footnotes omitted). See also 1 W. Holdsworth, A History of English Law 324-325 (7th ed. 1956).
To be sure, not every, element of English common law was carried over without change in the Colonies. In the case of jury trial, however, “in general this venerable and highly popular institution was adopted in the colonies in its English form at an early date.” Reinsch, The English Common Law in the Early American Colonies, in 1 Select Essays in Anglo-American Legal History 412 (1907). As the Court concluded in Williams v. Florida, “[t]he States that had adopted Constitutions by the time of the Philadelphia Convention in 1787 appear for the most part to have either explicitly provided that the jury would consist of 12, see Va. Const, of 1776, § 8, in 7 F. Thorpe, Federal and State Constitutions 3813 (1909), or to have subsequently interpreted their jury trial provisions to include that requirement.” 399 U. S., at 98-99, n. 45.7
*178On the basis of this historical record, this Court has more than once concluded that the Seventh Amendment guarantees the preservation of 12-man juries.
As the Court, speaking through Mr. Justice Gray, said in Capital Traction Co. v. Hof,
“ 'Trial by jury,’ in the primary and usual sense of the term at the common law and in the American constitutions, is ... a trial by a jury of twelve men before an officer vested with authority to cause them *179to be summoned and empanelled, to administer oaths to them and to the constable in charge, and to enter judgment and issue execution on their verdict . . . . This proposition has been so generally admitted, and so seldom contested, that there has been little occasion for its distinct assertion. Yet there are unequivocal statements of it to be found in the books.” 174 U. S., at 13-14.
Cf. Patton v. United States, 281 U. S. 276 (1930); Maxwell v. Dow, 176 U. S. 581 (1900); American Publishing Co. v. Fisher, 166 U. S. 464 (1897); Springville v. Thomas, 166 U. S. 707 (1897).
The Court today elects to abandon the certainty of this historical test, as well as the many cases which support it, in favor of a vaguely defined functional analysis which asks not what the Framers meant by “trial by jury” but rather whether some substitute for the common-law jury performs the same functions as a jury and serves as an adequate substitute for one. It is true that some of our prior cases support a functional approach to an evaluation of procedural innovations which surround jury trials. The Court has in the past upheld such devices as jury interrogatories and reports of special masters as not interfering with the functioning of a common-law jury. See, e. g., Ex parte Peterson, 253 U. S. 300 (1920); Walker v. New Mexico & S. P. R. Co., 165 U. S. 593 (1897). But see Dimick v. Schiedt, 293 U. S. 474 (1935). But I know of no prior case which has utilized a functional analysis to evaluate the very composition of the civil jury.
I submit that the reason for the absence of such cases derives from the inherent nature of the problem. It is possible to determine in a principled fashion whether the appurtenances which surround a jury interfere with the essential functioning of that institution. One can *180evaluate whether additur, for example, or directed verdicts interfere with the jury’s role as it existed at common law. See, e. g., Galloway v. United States, 319 U. S. 372 (1943); Dimick v. Schiedt, supra. But the composition of the jury itself is a matter of arbitrary, a priori definition. As Mr. Justice Harlan argued “[t]he right to a trial by jury . . . has no enduring meaning apart from historical form.” Williams v. Florida, 399 U. S., at 125 (separate opinion).
It is senseless, then, to say that a panel of six constitutes a “jury” without first defining what one means by a jury, and that initial definition must, in the nature of things, be arbitrary. One could, of course, define the term “jury” as being a body of six or more laymen. But the line between five and six would then be just as arbitrary as the line between 11 and 12. There is no way by reference to abstract principle or “function” that one can determine that six is “enough,” five is “too small,” and 20 “too large.”8 These evaluations can only be made by reference to a hypothetical ideal jury of some arbitrarily chosen size. All one can say is that a jury of six functions less like a jury of 12 than would *181a jury of, say eight, but more like a jury of 12 than would a jury of three.9 Although I think it clear that my Brethren would reject, for example, a jury of one, the Court does not begin to tell us how it would go about drawing a line in a nonarbitrary fashion, and it is obvious that in matters of degree of this kind, nonarbi-trary line drawing is a logical impossibility.
Of course, there is nothing intrinsically wrong with drawing arbitrary lines and, indeed, as argued above, in order to resolve certain problems they are essential. Thus, this Court has not hesitated in the past to rely on arbitrary demarcations in cases where constitutional rights depend on matters of degree. See, e. g., Burns v. Fortson, 410 U. S. 686 (1973). But in cases where arbitrary lines are necessary, I would have thought it more consonant with our limited role in a constitutional democracy to draw them with reference to the fixed bounds of the Constitution rather than on a wholly ad hoc basis.
I think history will bear out the proposition that when constitutional rights are grounded in nothing more solid than the intuitive, unexplained sense of five Justices that a certain line is “right” or “just,” those rights are certain to erode and, eventually, disappear altogether. Today, a majority of this Court may find six-man juries to represent a proper balance between competing demands of expedition and group representation. But as dockets become more crowded and pressures on jury trials grow, who is to say that some future Court will not find three, or two, or one a number large enough to satisfy its unexplicated sense of justice? It should *182be clear that constitutional rights which are so vulnerable to pressures of the moment are not really protected by the Constitution at all. As Mr. Justice Black never tired of arguing, “the accordion-like qualities of this philosophy must inevitably imperil all the individual liberty safeguards specifically enumerated in the Bill of Rights.” Rochin v. California, 342 U. S. 165, 177 (1952) (Black, J., concurring). See also Duncan v. Louisiana, 391 U. S. 145, 169 (1968) (Black, J., concurring).
Since some definition of “jury” must be chosen, I would therefore rely on the fixed bounds of history which the Framers, by drafting the Seventh Amendment, meant to “preserve.” I agree with Mr. Justice Powell’s observation in the Sixth Amendment context that determining the content of the right to jury trial should involve a “careful evaluation of, and strict adherence to the limitations on, that right as it was known ... at common law.” Johnson v. Louisiana, 406 U. S., at 370 n. 6 (separate opinion). It may well be that the number 12 is no more than a “historical accident” and is “wholly without significance 'except to mystics.’ ” Williams v. Florida, supra, at 102. But surely there is nothing more significant about the number six, or three, or one. The line must be drawn somewhere, and the difference between drawing it in the light of history and drawing it on an ad hoc basis is, ultimately, the difference between interpreting a constitution and making it up as one goes along.
II
The arbitrary nature of the line which must be drawn in determining permissible jury size highlights another anomaly in the Court’s opinion. Normally, in our system we leave the inevitable process of arbitrary line drawing to the Legislative Branch, which is far better equipped to make ad hoc compromises. In the past, we *183have therefore given great deference to legislative decisions in cases where the line must be drawn somewhere and cannot be precisely delineated by reference to principle. This Court has involved itself in the sticky business of separating cases along a continuum only when the Constitution clearly compels it to do so and when the legislature has plainly defaulted.
Today, the Court turns this practice inside out. It rejects what I take to be a clearly articulated legislative decision — a decision, incidentally, which is fully consonant with constitutional requirements — in order to draw its own arbitrary line. It does so, moreover, without any explanation for why it finds the legislative determination unsatisfactory and, indeed, with barely any explanation at all.
A
Title 28 U. S. C. § 2072 requires that the Rules of Civil Procedure promulgated by this Court “shall preserve the right of trial by jury as at common law and as declared by the Seventh Amendment to the Constitution.” As the Court recognizes, this requirement is made applicable to local rules of procedure by 28 U. S. C. §2071, which requires that “[s]uch rules shall be consistent with Acts of Congress and rules of practice and procedure prescribed by the Supreme Court.”
The Court’s treatment of this statutory requirement is, to say the least, peculiar. When explicating the Seventh Amendment, my Brethren hold that the Framers intended to govern only the types of trials in which the jury right attaches rather than to fix the common-law characteristics of the jury. Their reason for reaching this conclusion is that the Seventh Amendment, by its terms, guarantees the right to a jury trial “[i]n suits at common law” and not as it existed at common law. This language, the Court says, “is not directed to jury *184characteristics, such as size, but rather defines the kind of cases for which jury trial is preserved, namely, 'suits at common law.’ ” Ante, at 152. This argument from the language of the Seventh Amendment is fair enough, although for the reasons given in the preceding section, I find it ultimately unpersuasive. But what, then, are we to say when interpreting a provision which guarantees jury trials, not "in suits at common law,” but “as at common law”? By the Court’s own reasoning, it would seem that this phrase should be read to guarantee the preservation of jury characteristics as they existed at common law.
Uninhibited by the seeming restraints of its own logic, however, my Brethren proceed to read this phrase to preserve juries in cases tried at common law in the face of the merger of law and equity. But if we are again to take the Court at its own word, this is precisely the result achieved by the Seventh Amendment of its own force. There is, of course, a well-recognized canon of construction which requires courts to read statutory provisions so that, when possible, no part of the statute is superfluous. See, e. g., 2 J. Sutherland, Statutes and Statutory Construction § 4705 (3d ed. 1943), and cases cited therein. Yet the Court’s reading of this statute creates not just a redundancy, but a double redundancy. If the framers of § 2072 had intended merely to preserve jury trials in cases at common law, then no statute at all would have been necessary since, as the Court recognizes, the Seventh Amendment by itself is sufficient to accomplish this purpose. Yet Congress not only passed a statute — it adopted a provision securing trial by jury both “as declared by the Seventh Amendment” and “as at common law.” If one accepts for the moment the Court’s premise that the Seventh Amendment preserves only the right to juries in common-law cases, *185Congress’ addition of the phrase “as at common law” is explicable only if the legislature also intended to protect jury characteristics from change.
My Brethren chose to reject this clear meaning of the statute and to read it instead in a manner which not only makes it redundant but also, as demonstrated in the previous section, raises the gravest constitutional questions. Yet the only argument I can discern for reaching this result is the Court’s stated reluctance to “saddle archaic and presently unworkable common-law procedures upon the federal courts.” With all respect, I had not thought it our function to determine which statutory requirements are “archaic” and “unworkable” and to enforce only those which we find to be efficient and up to date. The Court asserts that “ [i] f Congress had meant to prescribe . . . common-law features [for juries] . . . ‘it knew how to use express language to that effect.’ ” But I, for one, would be hard pressed to think of language which more expressly guarantees the jury’s common-law features than the statement that the right of trial by jury shall be preserved “as at common law.” So long as this is the command of Congress, I had thought it our duty to obey, no matter how “archaic” and “unworkable” the statutory requirement.
B
Nor is the statute the end of the matter. Federal Rule Civ. Proc. 48 provides in relevant part that “[t]he parties may stipulate that the jury shall consist of any number less than twelve.” It hardly need be demonstrated that this provision is flatly inconsistent with local Rule 13 (d) (1). The number 11, for example, falls within the class of “any number less than twelve,” so that Rule 48 requires that the parties be permitted to stipulate to a jury of 11. Yet the local rule, which requires that “[a] *186jury for the trial of civil cases shall consist of six persons” clearly would not permit a jury of 11, even if the parties stipulated to such a jury.
The Court’s contention that Rule 48 “deals only with a stipulation by ‘[t\he parties’ ” and “does not purport to prevent court rules which provide for civil juries of reduced size,” ante, at 164, therefore passes my understanding. It is true enough that Rule 48 deals with stipulations by the parties, but it expressly says that the court rules must permit such stipulations so long as the number stipulated is “any number less than twelve.” Since the numbers seven through 11 are numbers less than 12, and since the local rule does not permit stipulations of these numbers, the two rules are in conflict and the local rule must therefore fall. See 28 U. S. C. § 2071; Fed. Rule Civ. Proc. 83.
Of course, Rule 48 does not on its face guarantee a jury of 12. That function is arguably performed by Rule 38 (a) which provides that “[t]he right of trial by jury as declared by the Seventh Amendment to the Constitution or as given by a statute of the United States shall be preserved to the parties inviolate.” But as the Court itself recognizes, the framers of Rule 48 clearly presupposed a jury of 12 in the absence of stipulation. Indeed, there is no way to make sense of a provision which permits stipulations of any number less than 12 unless one assumes that in the absence of a stipulation, the jury would consist of 12. I am thus once again at a loss to understand why the Court strains to escape the plain intention of the Rule’s drafters in order to wrestle with grave constitutional questions that could easily have been avoided.
Ill
It might appear to some anomalous after Williams to hold that 12-man civil juries are constitutionally required in federal cases. As Judge Wisdom has argued, “[w]hat-*187ever one considers the role of a civil jury and whatever importance attaches to that role, ... no one has ever contended that the function of the civil jury is more important than that of the criminal jury.” Cooley v. Strickland Transportation Co., 459 F. 2d 779, 781 (1972).
There is, of course, force to that point and a certain rudimentary logic to the proposition that if a man is entitled to a jury of only six when his very liberty is at stake, he should not be entitled to more when mere property hangs in the balance. But our function is limited to interpreting the Constitution. We are not empowered to decide as a matter of policy the cases in which 12-man juries should be guaranteed. As argued above, our prior decision on jury size arose in the state context and involved interpretation of a different constitutional provision. That decision simply does not require that we approve six-man federal juries in civil cases. As Mr. Justice Sutherland observed almost 40 years ago when the common-law jury was under attack from a different source, “this court in a very special sense is charged with the duty of construing and upholding the Constitution; and in the discharge of that important duty, it ever must be alert to see that a doubtful precedent be not extended by mere analogy to a different case if the result will be to weaken or subvert what it conceives to be a principle of the fundamental law of the land.” Dimick v. Schiedt, 293 U. S., at 485.
I find that response dispositive. The Constitution is, in the end, a unitary, cohesive document and every time any piece of it is ignored or interpreted away in the name of expedience, the entire fragile endeavor of constitutional government is made that much more insecure. This observation is as pertinent to the Seventh Amendment as it is to the First, or Fourteenth, or any other part of the Constitution. Indeed, as the Dimick court held, “[maintenance of the jury as a fact-finding body is of *188such importance and occupies so firm a place in our history and jurisprudence that any seeming curtailment of the right to a jury trial should be scrutinized with the utmost care.” Id., at 486. In my judgment, my Brethren have not given this curtailment of the jury right the careful scrutiny which the problem demands. I must, therefore, respectfully dissent.
Although I consider it ultimately irrelevant to the constitutional issue, see infra, at 180, it is still of some interest that variations in jury size do seem to produce variations in function and result. It is, of course, intuitively obvious that the smaller the size of the jury, the less likely it is to represent a fair cross-section of community viewpoints. What is less obvious but nonetheless statistically demonstrable is that the difference between a 12-man and six-man jury in this respect is quite dramatic and likely to produce different results Professor Zeisel, perhaps our leading authority on the civil jury, has demonstrated this fact through use of a model in which he assumes that 90% of a hypothetical community shares the same viewpoint, while 10% has a different viewpoint. Of 100 12-man juries picked randomly from such a community, 72 would have at least one member of the minority group, while of the 100 six-man juries so selected, only 47 would have minority representation. Moreover, the differences in minority representation produce significant differences in result. Professor Zeisel posits a case in which the community is divided into six groups of equal size with respect to the monetary value they place on a given personal injury claim, with one-sixth evaluating the claim at $1,000, another sixth at $2,000, etc. He also assumes that the damages a jury will award lie close to the average assessment of the damages each individual juror would choose. If one accepts these hypotheses, “[i]t is easy to see that the six-member juries show a considerably wider variation of 'verdicts’ than the twelve-member juries. For instance, 68.4% of the twelve-member jury evaluations fall between $3,000 and $4,000, while only 51.4% of the six-member jury evaluations fall in this range. Almost 16% of the six-member juries will reach verdicts that will fall into the extreme levels of more than $4,500 or less than $2,500, as against only a little over 4% of the twelve-member juries. The appropriate statistical measure of this variation is the so-called standard deviation. The actual distribution pattern will always depend on the kind of stratification that is relevant in a particular case but, whatever the circumstances, the six-member jury will always have a standard deviation that is greater by about 42%. This is the result of a more general principle *168that is by now well known to readers of such statistics as public opinion polls — namely, that the size of any sample is inversely related to its margin of error.” Zeisel, . . . And Then There Were None: The Diminution of the Federal Jury, 38 U. Chi. L. Rev. 710, 717-718 (1971).
See infra, at 176-177.
Even in the absence of constitutional difficulties, I view this course as an improper use of the local rulemaking power. In Miner v. Atlass, we held that the statutory procedures surrounding the rulemaking process were “designed to insure that basic procedural innovations shall be introduced only after mature consideration of informed opinion from all relevant quarters, with all the opportunities for comprehensive and integrated treatment which such consideration affords.” 363 U. S. 641, 650 (1960). We therefore declined to construe the local rulemaking power as extending to such innovations. Ibid. The Court seeks to escape the force of this precedent with the assertion that “the requirement of a six-member jury is not a ‘basic procedural innovation.’ ” I find this statement startling to say the least. Whatever one’s view of the constitutionality of six-man juries, surely it cannot be doubted that this shift in a practice of seven hundred years’ standing, likely to affect the outcome of hundreds of cases, see n. 1, supra, and infra, at 177, constitutes a “basic procedural innovation.”
Indeed, the Seventh Amendment is one of the few remaining provisions in the Bill of Rights which has not been held to be applicable to the States. See, e. g., Hardware Dealers Mutual Fire Ins. Co. v. Glidden Co., 284 U. S. 151, 158 (1931); Wagner Electric Mfg. Co. v. Lyndon, 262 U. S. 226, 232 (1923).
The author of this opinion believes that the Fourteenth Amendment was intended to incorporate fully Sixth Amendment guarantees. *170See Duncan v. Louisiana, 391 U. S. 145 (1968). Nonetheless, the fact remains that this Court has yet to decide the issues posed by majority verdicts and six-man juries in a purely Sixth Amendment context.
See also George Washington’s contemporaneous explanation in a letter to Lafayette for the absence of a jury guarantee (“[I]t was only the difficulty of establishing a mode which should not interfere with the fixed modes of any of the States, that induced the Convention to leave it, as a matter of future adjustment”) 3 Farrand 298; and Edmund Randolph’s explanation to the Virginia Convention (“I will risk my property on the certainty, that [Congress] will institute the trial by jury in such manner as shall accommodate the conveniences of the inhabitants of every state: the difficulty of ascertaining this accommodation, was the principal cause of its not being provided for”) 3 Farrand 309
I do not mean to suggest that isolated experiments with juries of different sizes cannot be found in colonial history. Indeed, when one considers the number of jurisdictions and the span of time involved, it would be surprising if there were no aberrations. Some scholars have argued from the few cases involving juries consisting of more or less than 12 that there was no common-law requirement *178as to jury size in the Colonies. See, e. g., Fisher, The Seventh Amendment and the Common Law: No Magic in Numbers, 56 F. R. D. 507 (1973). In fact, however, the cases cited for this proposition seem to constitute no more than the exceptions which prove the rule.
Fisher, for example, bases his thesis on the fact that Maryland used a jury of 10 in one case in 1682 and a jury of 11 in another case that year and that Delaware used juries of 11, 7, and 13 in three cases tried between 1676 and 1705. See id., at 530. But when one remembers that thousands of civil and criminal cases were tried during the prerevolutionary period, these five apparently isolated instances prove virtually nothing. Similarly, South Carolina’s provision for a jury of less than 12 in the “Court for the Trial of Slaves and Persons of Color,” ibid., was obviously limited to the peculiar circumstance of persons who, at that time, were considered to be without civil rights of any kind. Fisher’s reliance on petitions from the citizens of Anson, Orange, and Rowan Counties for juries of less than 12, ibid., is unaccountable since these petitions were in fact rejected and the smaller juries never impaneled. See id., at 530-531, n. 87.
Fisher’s final example is particularly revealing. Just prior to the Revolution, New Jersey passed an act providing for six-man juries in small-court cases. Id., at 531. The law was challenged in the case of Holmes v. Walton, in 1780, in which the defendant argued “the jury sworn to try the above cause and on whose verdict judgment was entered, consisted of six men only, when by the laws of the land it should have consisted of twelve men.” Id., at 532 n. 88. The New Jersey Supreme Court rejected this argument and upheld the verdict. A scant month later, however, the New Jersey Legislature reversed this decision and reinstituted the right to 12-man juries. See ibid.
The Court asserts that “[w]hat is required for a 'jury' is a number large enough to facilitate group deliberation combined with a likelihood of obtaining a representative cross section of the community.” See ante, at 160 n, 16. We can bypass for the moment the intriguing question of where the majority finds this requirement in the words of the Seventh Amendment. For our purposes, it is sufficient to note that, upon examination, this “test” turns out to be no test at all. It may be that the ideal jury would provide “enough” group deliberation and community representation. But the question in this case is how much is “enough.” Obviously, the larger the jury the more group representation it will provide. See n. 1, supra. Merely observing that a certain level of group representation is constitutionally required fails to tell us what that level is. And, more significantly, it fails to tell us how to go about deciding what that level is.
It thus will not do to argue, as has my Brother White, that one “can get off the 'slippery slope’ before he reaches the bottom. . . Williams v. Florida, 399 U. S. 78, 91 n. 28 (1970). This begs the question how one knows at what point to get off — a question for which the Court apparently has no answer.