concurring in part and concurring in the judgment.
I write separately because, in my view, Federal Rule of Criminal Procedure 16 does not limit a defendant’s discovery rights to documents related to the Government’s case in chief. Ante, at 462-463. The Rule says that “the government shall permit the defendant to inspect and copy” certain physical items (I shall summarily call them “documents”) “which are material to the preparation of the defendant’s defense.” Fed. Rule Crim. Proc. 16(a)(1)(C). A “defendant’s defense” can take many forms, including (1) a simple re*472sponse to the Government’s case in chief, (2) an affirmative defense unrelated to the merits (such as a Speedy Trial Act claim), (3) an unrelated claim of constitutional right, (4) a foreseeable surrebuttal to a likely Government rebuttal, and others. The Rule’s language does not limit its scope to the first item on this list. To interpret the Rule in this limited way creates a legal distinction that, from a discovery perspective, is arbitrary. It threatens to create two full parallel sets of criminal discovery principles. And, as far as I can tell, the interpretation lacks legal support.
The Court bases its interpretation upon what it says is a “perceptible symmetry,” ante, at 462, between two phrases in Rule 16(a)(1)(C) — the phrase “material to the preparation of the defendant’s defense,” and the next phrase, “intended for use by the government as evidence in chief at the trial.” To test the Court’s argument, consider these two phrases in context. The Rule says:
“Upon request of the defendant the government shall permit the defendant to inspect and copy [documents and other items]... which [1] are material to the preparation of the defendant’s defense or [2] are intended for use by the government as evidence in chief at the trial, or [3] were obtained from or belong to the defendant.” Fed. Rule Crim. Proc. 16(a)(1)(C).
Though symmetry may reside in the eye of the beholder, I can find no relevant symmetry here. Rather, the language suggests a simple three-part categorization of the documents and other physical items that the Rule requires the Government to make available to the defendant. From a purely linguistic perspective, there is no more reason to import into the first category a case-in-chief-related limitation (from the second category) than some kind of defendant’s-belongings-related limitation (from the third category).
Rule 16 creates these three categories for a reason that belies “symmetry” — namely, to specify two sets of items (the *473Government’s case in chief evidence, the defendant’s belongings) that the Government must make available to the defendant without a preliminary showing of “materiality.” The Rule’s first category creates a residual classification (items “material to the preparation of the defendant’s defense”) that require a preliminary “materiality” showing. The Committee thought, however, that “[limiting the rule to situations in which the defendant can show that the evidence is material seems unwise.... For this reason subdivision (a)(1)(C) also contains language to compel disclosure if the government intends to use the property as evidence at the trial or if the property was obtained from or belongs to the defendant.” Advisory Committee’s Notes on Fed. Rule Crim. Proc. 16,18 U. S. C. App., p. 762 (second and third categories added to specify that, without a special showing of materiality, certain items are almost always “material”) (citing 1 C. Wright, Federal Practice and Procedure §264, p. 510, n. 58, p. 513, n. 70 (1969)). Nothing in the Notes, or in the Rule’s language, suggests that the residual category of items “material to the preparation of the defendant’s defense,” means to cover only those items related to the case in chief.
The only other reason the majority advances in support of its “case in chief” limitation concerns a later part of the Rule, subdivision 16(a)(2). As relevant here, that subdivision exempts Government attorney work product from certain of Rule 16’s disclosure requirements. In the majority’s view, since (1) a defendant asserting a valid “selective prosecution” defense would likely need prosecution work product to make his case, but (2) the Rule exempts prosecution work product from discovery, then (3) the Rule must have some kind of implicit limitation (such as a “case in chief” limitation) that makes it irrelevant to defense efforts to assert “selective prosecution” defenses.
The majority’s conclusion, however, does not follow from its premises. For one thing, Rule 16’s work-product excep*474tion may itself contain implicit exceptions. After all, “[t]he privilege derived from the work-product doctrine is not absolute.” United States v. Nobles, 422 U. S. 225, 239 (1975); see also 8 C. Wright, A. Miller, & R. Marcus, Federal Practice and Procedure §2022, p. 324 (2d ed. 1994) (in civil context, work product “is discoverable only on a substantial showing of ‘necessity or justification’ ”) (quoting Hickman v. Taylor, 329 U. S. 495, 510 (1947)); J. Ghent, Development, Since Hickman v. Taylor, of Attorney’s “Work Product” Doctrine, 35 A. L. R. 3d 412, 465-469, §25 (1971) (in civil context, work-product protection is not absolute, but is a “qualified privilege or immunity”). To the extent such a reading permits a defendant to obtain “work product” in an appropriate case (say, with a strong prima facie showing of selective prosecution), the Court’s problem does not exist. Of course, to read the work-product exception as containing some such implicit exception itself represents a departure from the Rule’s literal language. But, is it not far easier to believe the Rule’s authors intended some such small implicit exception to an exception, consistent with the language and purpose of the Rule, than that they intended the very large exception created by the Court?
For another thing, even if one reads the work-product exception literally, the Court’s problem disappears as long as courts can supplement Rule 16 discovery with discovery based upon other legal principles. The language of the work-product exception suggests the possibility of such supplementation, for it says, not that work product is “exemp[t]” from discovery, ante, at 462, but that “this rule” does not authorize discovery of the prosecutor’s work product. Fed. Rule Civ. Proc. 16(a)(2). The Advisory Committee’s Notes make clear that the Committee believed that other rules of law may authorize (or require) discovery not mentioned in the Rule. See, e. g., Advisory Committee’s Notes on Rule 16, 18 U. S. C. App., pp. 762, 763 (discussion of Brady v. Maryland, 373 U. S. 83 (1963), which the Rule does not cod*475ify); 18 U. S. C. App., p. 761 (“[Rule 16] is intended to prescribe the minimum amount of discovery to which the parties are entitled. It is not intended to limit the judge’s discretion to order broader discovery in appropriate cases”); see also 2 C. Wright, Federal Practice and Procedure §254, p. 81, and n. 60 (2d ed. 1982) (“Because Brady is based on the Constitution, it overrides court-made rules of procedure. Thus the work-product immunity for discovery in Rule 16(a)(2) prohibits discovery under Rule 16 but it does not alter the prosecutor’s duty to disclose material that is within Brady”) (footnotes omitted). Of course, the majority, in a sense, reads the Rule as permitting supplementation, but it does more. It goes well beyond the added (say, constitutionally related) rule supplementation needed to overcome its problem; instead, it shrinks the Rule by unnecessarily creating a major limitation on its scope.
Finally, and in any event, here the defendants sought discovery of information that is not work product. See ante, at 459. Thus, we need not decide whether in an appropriate case it would be necessary to find an implicit exception to the language of Rule 16(a)(2), or to find an independent constitutional source for the discovery, or to look for some other basis.
In sum, neither the alleged “symmetry” in the structure of Rule 16(a)(1)(C), nor the work-product exception of Rule 16(a)(2), supports the majority’s limitation of discovery under Rule 16(a)(1)(C) to documents related to the -Government’s “case in chief.” Rather, the language and legislative history make clear that the Rule’s drafters meant it to provide a broad authorization for defendants’ discovery, to be supplemented if necessary in an appropriate case. Whether or not one can also find a basis for this kind of discovery in other sources of law, Rule 16(a)(1)(C) provides one such source, and we should consider whether the defendants’ discovery request satisfied the Rule’s requirement that the discovery be “material to the preparation of the defendant’s defense.”
*476I believe that the defendants’ request did not satisfy this threshold. Were the “selective prosecution” defense valid in this case — i. e., were there “clear evidence,” United States v. Chemical Foundation, Inc., 272 U. S. 1, 14 (1926), that the Federal Government’s prosecutorial policy “had a discriminatory effect and ... was motivated by a discriminatory purpose,” Wayte v. United States, 470 U. S. 598, 608 (1985), it should have been fairly easy for the defendants to find, not only instances in which the Federal Government prosecuted African-Americans, but also some instances in which the Federal Government did not prosecute similarly situated Caucasians. The defendants’ failure to do so, for the reasons the Court sets forth, amounts to a failure to make the necessary threshold showing in respect to materiality. See 2 C. Wright, Federal Practice and Procedure § 254, pp. 66-67 (2d ed. 1982); United States v. Balk, 706 F. 2d 1056, 1060 (CA9 1983); United States v. Johnson, 577 F. 2d 1304, 1309 (CA5 1978); United States v. Murdock, 548 F. 2d 599, 600 (CA5 1977).