with whom The Chief Justice joins, dissenting.
The Court, apparently moved by concerns for intellectual tidiness, mistakenly decides that the “clear and convincing evidence” standard governing finders of fact in libel cases must be applied by trial courts in deciding a motion for summary judgment in such a case. The Court refers to this as a “substantive standard,” but I think is is actually a procedural *269requirement engrafted onto Rule 56, contrary to our statement in Calder v. Jones, 465 U. S. 783 (1984), that
“[w]e have already declined in other contexts to grant special procedural protections to defendants in libel and defamation actions in addition to the constitutional protections embodied in the substantive laws.” Id., at 790-791.
The Court, I believe, makes an even greater mistake in failing to apply its newly announced rule to the facts of this case. Instead of thus illustrating how the rule works, it contents itself with abstractions and paraphrases of abstractions, so that its opinion sounds much like a treatise about cooking by someone who has never cooked before and has no intention of starting now.
There is a large class of cases in which the higher standard imposed by the Court today would seem to have no effect at all. Suppose, for example, on motion for summary judgment in a hypothetical libel case, the plaintiff concedes that his only proof of malice is the testimony of witness A. Witness A testifies at his deposition that the reporter who wrote the story in question told him that she, the reporter, had done absolutely no checking on the story and had real doubts about whether or not it was correct as to the plaintiff. The defendant’s examination of witness A brings out that he has a prior conviction for perjury.
May the Court grant the defendant’s motion for summary judgment on the ground that the plaintiff has failed to produce sufficient proof of malice? Surely not, if the Court means what it says, when it states: “Credibility determinations . . . are jury functions, not those of a judge, whether he is ruling on a motion for summary judgment or for a directed verdict. The evidence of the nonmovant is to be believed, and all justifiable inferences are to be drawn in his favor.” Ante, at 255.
The case proceeds to trial, and at the close of the plaintiff’s evidence the defendant moves for a directed verdict on the *270ground that the plaintiff has failed to produce sufficient evidence of malice. The only evidence of malice produced by the plaintiff is the same testimony of witness A, who is duly impeached by the defendant for the prior perjury conviction. In addition, the trial judge has now had an opportunity to observe the demeanor of witness A, and has noticed that he fidgets when answering critical questions, his eyes shift from the floor to the ceiling, and he manifests all other indicia traditionally attributed to perjurers.
May the trial court at this stage grant a directed verdict? Again, surely not; we are still dealing with “credibility determinations.”
The defendant now puts on its testimony, and produces three witnesses who were present at the time when' witness A alleges that the reporter said she had not checked the story and had grave doubts about its accuracy as to plaintiff. Witness A concedes that these three people were present at the meeting, and that the statement of the reporter took place in the presence of all these witnesses. Each witness categorically denies that the reporter made the claimed statement to witness A.
May the trial court now grant a directed verdict at the close of all the evidence? Certainly the plaintiff’s case is appreciably weakened by the testimony of three disinterested witnesses, and one would hope that a properly charged jury would quickly return a verdict for the defendant. But as long as credibility is exclusively for the jury, it seems the Court’s analysis would still require this case to be decided by that body.
Thus, in the case that I have posed, it would seem to make no difference whether the standard of proof which the plaintiff had to meet in order to prevail was the preponderance of the evidence, clear and convincing evidence, or proof beyond a reasonable doubt. But if the application of the standards makes no difference in the case that I hypothesize, one may fairly ask in what sort of case does the difference in stand*271ards make a difference in outcome? Cases may be posed dealing with evidence that is essentially documentary, rather than testimonial; but the Court has held in a related context involving Federal Rule of Civil Procedure 52(a) that inferences from documentary evidence are as much the prerogative of the finder of fact as inferences as to the credibility of witnesses. Anderson v. Bessemer City, 470 U. S. 564, 574 (1985). The Court affords the lower courts no guidance whatsoever as to what, if any, difference the abstract standards that it propounds would make in a particular case.
There may be more merit than the Court is willing to admit to Judge Learned Hand’s observation in United States v. Feinberg, 140 F. 2d 592, 594 (CA2), cert, denied, 322 U. S. 726 (1944), that “[w]hile at times it may be practicable” to “distinguish between the evidence which should satisfy reasonable men, and the evidence which should satisfy reasonable men beyond a reasonable doubt[,]... in the long run the line between them is too thin for day to day use. ” The Court apparently approves the overruling of the Feinberg case in the Court of Appeals by Judge Friendly’s opinion in United States v. Taylor, 464 F. 2d 240 (1972). But even if the Court is entirely correct in its judgment on this point, Judge Hand’s statement seems applicable to this case because the criminal case differs from the libel case in that the standard in the former is proof “beyond a reasonable doubt,” which is presumably easier to distinguish from the normal “preponderance of the evidence” standard than is the intermediate standard of “clear and convincing evidence.”
More important for purposes of analyzing the present case, there is no exact analog in the criminal process to the motion for summary judgment in a civil case. Perhaps the closest comparable device for screening out unmeritorious cases in the criminal area is the grand jury proceeding, though the comparison is obviously not on all fours. The standard for allowing a criminal case to proceed to trial is not whether the government has produced prima facie evidence of guilt be*272yond a reasonable doubt for every element of the offense, but only whether it has established probable cause. See United States v. Mechanik, 475 U. S. 66, 70 (1986). Thus, in a criminal case the standard used prior to trial is much more lenient than the “clear beyond a reasonable doubt” standard which must be employed by the finder of fact.
The three differentiated burdens of proof in civil and criminal cases, vague and impressionistic though they necessarily are, probably do make some difference when considered by the finder of fact, whether it be a jury or a judge in a bench trial. Yet it is not a logical or analytical message that the terms convey, but instead almost a state of mind; we have previously said:
“Candor suggests that, to a degree, efforts to analyze what lay jurors understand concerning the differences among these three tests . . . may well be largely an academic exercise. . . . Indeed, the ultimate truth as to how the standards of proof affect decisionmaking may well be unknowable, given that factfinding is a process shared by countless thousands of individuals throughout the country. We probably can assume no more than that the difference between a preponderance of the evidence and proof beyond a reasonable doubt probably is better understood than either of them in relation to the intermediate standard of clear and convincing evidence.” Addington v. Texas, 441 U. S. 418, 424-425 (1979) (emphasis added).
The Court’s decision to engraft the standard of proof applicable to a factfinder onto the law governing the procedural motion for a summary judgment (a motion that has always been regarded as raising a question of law rather than a question of fact, see, e. g., La Riviere v. EEOC, 682 F. 2d 1275, 1277-1278 (CA9 1982) (Wallace, J.)), will do great mischief with little corresponding benefit. The primary effect of the Court’s opinion today will likely be to cause the decisions of trial judges on summary judgment motions in libel cases to be *273more erratic and inconsistent than before. This is largely because the Court has created a standard that is different from the standard traditionally applied in summary judgment motions without even hinting as to how its new standard will be applied to particular cases.