with whom SELYA, Circuit Judge, joins (dissenting).
This case involves two rules. The first rule is a Massachusetts Supreme Judicial Court rule of professional conduct that says that it is “unprofessional” for a “prosecutor to subpoena an attorney to a grand jury without prior judicial approval” where the prosecutor “seeks to compel” testimony about a client. Mass.SJ.C. Rule 3:08, PF 15. The second rule is a federal district court local rule that cross-references the first rule. It says that
[a]cts or omissions ... that violate the ethical requirements and rules concerning the practice of law of the Commonwealth of Massachusetts [including S.J.C. Rule 3:08], shall constitute misconduct and shall be grounds for discipline.
D.Mass. Local Rule 5(d)(4)(B). Three members of this Court of Appeals, sitting en banc, find this second rule — the federal district court rule — lawful. In my view, however, as presently written, that rule, at least insofar as it incorporates PF 15, falls outside the district court’s rulemaking power. In particular, the procedure that the district court followed in adopting the rule violates Federal Rule of Criminal Procedure 57, which permits a district court to “make and amend rules governing its practice” only “after giving appropriate public notice and an opportunity to comment.”
The three “majority” members of the court would avoid the procedural question because the parties did not raise it in the district court or in their briefs on appeal. The majority thus follow a well-established rule that guides appellate court practice in the ordinary case. Pignons S.A. de Mecanique v. Polaroid Corp., 701 F.2d 1, 3 (1st Cir.1983); cf. Johnston v. Holiday Inns, 595 F.2d 890, 894 (1st Cir.1979) (court of appeals ordinarily will not consider argument not presented to trial court). The case before us, however, is not one of those ordinary cases; it is distinctive in several regards. First, the procedural issue was not raised in the district court opinion because it could not have been raised. It was only after the lower court’s decision that the district courts amended the original Local Rule 5(d)(4)(B) to explicitly incorporate PF 15. The act of amending raises the new issue. It is at this point that Fed.R.Crim.P. 57 requires notice and comment in order to assure a well-reasoned and thoroughly evaluated decision.
Second, the notice and comment issue, having postdated the district court case, is critical and potentially dispositive in the appellate court proceeding. A court of appeals may, of course, decide to address a critical issue not raised below. 9 to 5 Organization for Women Office Workers v. Board of Governors of Federal Reserve System, 721 F.2d 1, 6 n. 2 (1st Cir.1983), quoting Singleton v. Wulff, 428 U.S. 106, 121, 96 S.Ct. 2868, 2877, 49 L.Ed.2d 826 (1976) (“[t]he matter of what questions may be taken up and resolved for the first time on appeal is one left primarily to the discretion of the courts of appeals, to be exercised on the facts of individual cases”). That power is appropriately exercised where, as here, the case is one of “great public concern,” affecting the administration of criminal justice generally. Dean Witter Reynolds, Inc. v. Fernandez, 741 F.2d 355, 361 (11th Cir.1984) (deciding generally important legal issues not raised in district court); see Schlesinger v. Councilman, 420 U.S. 738, 743-44, 95 S.Ct. 1300, 1305-06, 43 L.Ed.2d 591 (1975) (raising sua sponte issue of propriety of district court’s intervention in court-martial proceeding).
*672Given the importance of the rule; the federal context; the state orientation of the Supreme Judicial Court’s consideration; the potential interest of (and potential effect upon) third parties; adequate ‘federal’ notice and comment would seem essential. The matter is too controversial and (in the federal context) potentially too ‘special’ to resolve simply through a cross reference to state procedure without appropriate canvassing and examination of the views of all those likely affected. Thus, the “notice and comment” issue is important enough to warrant our consideration. See United States v. Krynicki, 689 F.2d 289, 291-92 (1st Cir.1982) (examining “purely legal” issue not raised by parties where doing so would “promote judicial economy [and] aid administration of the criminal justice system”).
This court, in my view, should now order additional briefing as to whether the district court provided “appropriate public notice and an opportunity to comment.” Having failed to convince a majority to ask for briefing, I shall here simply explain why, on the present state of the record, I conclude that the district court did not provide for the full, public discussion that the nature of the controversy, and the Federal Rules, require.
I
The “notice and comment” language in Federal Rule 57 represents a serious effort to create a rulemaking procedure roughly similar to that used by administrative agencies. Cf 5 U.S.C. § 553 (1982) (imposing procedural requirements on agency rule-making). The 1985 amendment to the Federal Rules that mandated this procedure followed widespread criticism of the way in which district courts previously had promulgated local rules. See 12 C. Wright & A. Miller, Federal Practice and Procedure § 3152, at 220 (1973); Blair, The New Local Rules for Federal Practice in Iowa, 23 Drake L.Rev. 517, 520-21 (1974); Caballero, Is There an Over-Exercise of Local Rule-Making Powers by the United States District Courts?, 24 Fed.Bar News 325 (1977); Weinstein, Reform of Federal Court Rulemaking Procedures, 76 Colum. L.Rev. 905, 951-56 (1976); Note, Rule 83 and the Local Federal Rules, 67 Colum.L. Rev. 1251 (1967); Comment, The Local Rules of Civil Procedure in the Federal District Courts — A Survey, 1966 Duke L.J. 1011. The drafters of the amendment stated that its purpose was “to enhance the local rulemaking process by requiring appropriate public notice of proposed rules and an opportunity to comment on them.” Notes of Advisory Committee on Fed.R. Crim.P. 57, 1985 Amendment.
The local rule before us presents an instance that clearly calls for application of the “notice and comment” procedure that Rule 57 mandates. The rule is an important one, significantly changing prior subpoena practice. Despite its characterization by both the Supreme Judicial Court and the district court, it is not simply an “ethical” standard. It does not simply restate the established principle that, except in narrowly defined circumstances, it is improper for a prosecutor to subpoena a lawyer in order to investigate the lawyer’s client. Cf. U.S. Department of Justice, United States Attorneys’ Manual § 9-2.-161(a) (1985) (forbidding U.S. Attorneys from subpoenaing attorneys for client information unless the Assistant Attorney General of the Criminal Division first determines that the information sought is necessary, unavailable from other sources, and not protected by privilege); American Bar Association, Grand Jury Policy and Model Act, Principle No. 29 (1982); In re Grand Jury Matters, 751 F.2d 13 (1st Cir.1984) (affirming decision to quash attorney subpoenas that imposed great burdens on the attorneys and their clients and were not urgently needed at the time). Rather, the local rule is a prophylactic procedural measure aimed at forestalling in advance the perceived offending conduct.
The merits of, and need for, the local rule are strongly contested. Cf. Levesque v. Block, 723 F.2d 175, 184-85 (1st Cir.1983) (agency need not follow notice and comment procedures where public is unlikely to be interested in rule). The government points out that the Department of Justice now prohibits prosecutors from is*673suing ‘attorney subpoenas’ without prior approval by the head of the Criminal Division, which approval will be granted only upon a showing of relevance, strong need, unavailability from other sources, and lack of privilege. United States Attorneys’ Manual § 9-2.161(a). In addition, a lawyer receiving an ‘attorney subpoena’ can ask the district court to quash it. See In re Grand Jury Matters, 751 F.2d 13 (1st Cir.1984). The appellees, on the other hand, believe these safeguards inadequate, and say that an additional prophylactic procedure is necessary. A “notice and comment” procedure would allow all interested parties to address these points before the district court at one time, in one place, in detail, and with knowledge of the basic arguments of others.
Moreover, the local rule has significant implications, raising a host of further questions. For example, will the ‘initial approval’ proceeding always take place ex parte ? Will the same judge who grants approval automatically hear any later challenge to the subpoena? Under what standard is the judge to review the proposed subpoena? Will there be an avenue of appeal from a refusal to grant initial approval? Is a subpoena issued without initial approval invalid? If a prosecutor, while interrogating a lawyer/witness before the grand jury, discovers only then that he needs to ask about a client, must he recess the session and obtain an initial court approval? How does the district court foresee implementing sanctions for violations of the rule? The answers to some of these questions may be fairly obvious; the answers to others are not. The basic point of “notice and comment” procedure, however, is to permit, indeed to encourage, all interested parties to address issues such as these, creating a single administrative record that will help the agency (or court) to promulgate a rule that reflects careful consideration of the rule’s likely ramifications and that will permit the agency (or court) to refine and to tailor the rule so as to resolve these and similar questions in the manner most likely to increase the rule’s effectiveness. See NLRB v. Wyman-Gordon Co., 394 U.S. 759, 777-78, 89 S.Ct. 1426, 1436, 22 L.Ed.2d 709 (1969) (Douglas, J., dissenting) (“Agencies ... are not always repositories of ultimate wisdom; they learn from the suggestions of outsiders and often benefit from that advice.”); Batterton v. Marshall, 648 F.2d 694, 704 (D.C.Cir.1980) (noting that notice and comment procedure “enables the agency promulgating the rule to educate itself before establishing rules and procedures which have a substantial impact on those [who are] regulated”) (quoting Texaco v. FPC, 412 F.2d 740, 744 (3d Cir.1969)); 12 C. Wright & A. Miller, supra, § 3152, at 220 (commenting that notice and comment procedures ensure that federal rules “reflect the best thinking of the entire profession”).
Further, the answers to these questions may affect the lawfulness of the rule. At present, for example, this circuit, unlike some others, does not require a particular showing of “need” or “relevance” before a court may enforce an ‘attorney subpoena.’ In re Grand Jury Proceedings (Hill), 786 F.2d 3, 5 n. 2 (1st Cir.1986); see also In re Grand Jury Subpoena Served Upon Doe, 781 F.2d 238 (2d Cir.1985) (rejecting “need and relevance” standard), cert. denied sub nom. Roe v. United States, 475 U.S. 1108, 106 S.Ct. 1515, 89 L.Ed.2d 914 (1986); Matter of Klein, 776 F.2d 628, 632-34 (7th Cir.1985) (same); cf. In re Grand Jury Proceedings (Schofield), 486 F.2d 85 (3d Cir.1973) (requiring preliminary showing of propriety and relevance before enforcing subpoenas), after remand, 507 F.2d 963 (3d Cir.), cert. denied sub nom. Schofield v. United States, 421 U.S. 1015, 95 S.Ct. 2424, 44 L.Ed.2d 685 (1975). Does the district court intend a stricter standard of review? If so, the argument that the local rule falls outside the district court’s rulemaking authority would become stronger. See Miner v. Atlass, 363 U.S. 641, 646-52, 80 S.Ct. 1300, 1303-07, 4 L.Ed.2d 1462 (1960) (restricting power of district courts to effect “basic” changes through local rules); In re Grand Jury Proceedings, 558 F.Supp. 532, 535-36 (W.D.Va.1983) (urging restraint in adopting local rules touching on important policy issues). But, if not, how is the local rule supposed to prevent abuses? Are there, in fact, in the federal courts, *674significant numbers of ‘attorney subpoenas’ that attorneys might successfully quash under present standards but that they do not challenge? Is an ex parte proceeding likely to provide a judge with enough information to identify an abusive subpoena and thus relieve the attorney/witness of the burdens associated with a motion to quash? To what extent can the Bar Association achieve its objective (stopping certain perceived abuses) without a change in standards? And, without such a change, is the game worth the candle? Answers to these questions would seem helpful not only to the district court, but also to an appellate court reviewing the rule both for consistency with other legal provisions and for reasonableness. Cf. Frazier v. Heebe, — U.S.-, 107 S.Ct. 2607, 96 L.Ed.2d 557 (1987) (exercising the Supreme Court’s supervisory authority to overturn local rule requiring that members of a district court bar reside or maintain an office in the state). The record created by “notice and comment” rulemaking helps to provide answers to questions such as these both to those promulgating rules and to those reviewing their legality. See Weyerhaeuser Co. v. Costle, 590 F.2d 1011, 1030-31 (D.C.Cir.1978); Rodway v. Department of Agriculture, 514 F.2d 809, 816-17 (D.C.Cir.1975) (noting importance of administrative record to reviewing court).
II
A review of the record (without benefit of briefing) suggests that the district court did not give “appropriate public notice and an opportunity to comment” before adopting the local rule in question. On October 1, 1985, the Massachusetts Supreme Judicial Court added PF 15 to the “Disciplinary Rules Applicable to Practice as a Prosecutor or as a Defense Lawyer” contained in its Rule 3:08. At that time, the federal district court was in the process of revising a number of its local rules, including Rule 5(d)(4)(B). On October 7, the district court circulated proposed revised rules for comment, including a revised Rule 5(d)(4)(B) that was ambiguous as to whether or not it incorporated S.J.C. Rule 3:08. (It simply prohibited “[ajcts or omissions ... that violate the Code of Professional Responsibility or Rules of Professional Conduct of the Commonwealth of Massachusetts.”) The only commenter to discuss Rule 5(d)(4)(B) was the United States Attorney, who asked the court to make clear that Rule 5(d)(4)(B) did not incorporate S.J.C. Rule 3:08. On January 3, 1986, the Massachusetts Association of Criminal Defense Lawyers wrote to the district court, stating that the “United States Attorney ... has applied ... for an exemption for federal prosecutors from S.J.C. Rule 3:08 PF 15” and asking “that the Government’s request for an exemption be made public and that the Court establish a period of time to receive comments as provided by the Rule.” Three days later, the chief judge of the district court said that the court would not issue any public notice because the U.S. Attorney’s filing of this lawsuit (on December 31, 1985) “effectively mooted any need for consideration of Local Rule 5(d)(4)(B)” until after a ruling in the lawsuit.
On March 19, 1986, the U.S. Attorney again asked the court to clarify Rule 5(d)(4)(B) by specifically excluding PF 15. On April 7, the presidents of the Massachusetts and Boston Bar Associations sent the court a letter calling the U.S. Attorney’s request “premature,” but adding that, if the district court were nonetheless to act on the request, it should grant “an ample amount of time for preparation of comments on PF 15.” The bar associations cited, among other matters, the importance of the issue, the need for time to compile relevant empirical data, and the need to discuss what standards, if any, the district court’s rule should include. The authors concluded that they “look forward to the [future] opportunity to present to the Court the comments of our respective bar associations on this important issue.” The Massachusetts Association of Criminal Defense Lawyers also sent the court a short letter briefly endorsing the position of the Massachusetts and Boston Bar Associations and stating that the Association “await[s] advice as to a schedule for the Court’s receipt of comments.”
*675Copies of these letters were forwarded to the judges of the court. Nonetheless, the Court issued no further notice and set no additional period for comment on Rule 5(d)(4)(B). Rather, the court issued its final revised rules on June 27,1986, effective July 1. These final rules included a revision in Rule 5(d)(4)(B) from the proposed rule of a year earlier explicitly stating that “[t]he ethical requirements and rules concerning the practice of law mean those canons and rules adopted by the Supreme Judicial Court of Massachusetts, embodied in Rules 3:05, 3:07 and 3:08 of said court.” Thus, the final rules, unlike the proposed revised rules, explicitly incorporate Rule 3:08 and hence PF 15. And, the court adopted this language without having had the benefit of widespread comment and a thorough discussion of the relevant issues, as intended by the “notice and comment” language of Rule 57.
The record before the Supreme Judicial Court cannot substitute for the record that Rule 57 requires both because the issues before the Supreme Judicial Court were somewhat different (largely involving state prosecutions and state procedure) and because the Supreme Judicial Court had no occasion to address a federal district court’s legal authority to create and to control federal procedure. Nor is there any indication that the Supreme Judicial Court considered many of the questions that we have identified as likely relevant here. In any event, the federal district court did not purport to rely on the Supreme Judicial Court record as a basis for promulgating its own rule.
In sum, the procedural rule here at issue is too important, its ramifications too complex, its contours too uncertain, for the district court to adopt it simply by means of a cross-reference, without the “appropriate public notice and an opportunity to comment” upon which Rule 57 insists. That being so, in my view, Rule 5(d)(4)(B), insofar as it simply cross-references and thereby adopts Mass.S.J.C. Rule 3:08, PF 15, is invalid.
Ill
The question remains whether the U.S. Attorney is nonetheless bound directly by the Massachusetts rule. That rule has two significant, but different, effects: (1) it tells prosecutors to apply to a court for prior approval of an attorney subpoena, and (2) it tells trial courts to create a “prior approval” procedure. Regardless of the state’s power to impose the first requirement, I think it clear that the supremacy clause, U.S.Const. Art. VI, cl. 2, forbids a state from ordering a federal court to adopt such particular procedures. Cf. Cooper v. Aaron, 358 U.S. 1, 78 S.Ct. 1401, 3 L.Ed.2d 5 & 19 (1958) (state cannot obstruct implementation of injunction entered by federal court); Thompkins v. Stuttgart School District No. 22, 787 F.2d 439, 441-42 (8th Cir.1986) (state cannot restrict federal court’s exercise of pendent jurisdiction over state claim). There is no indication in the Massachusetts rule or its underlying record that the Supreme Judicial Court ever considered its power to impose procedures upon a federal court. Furthermore, it seems doubtful that Massachusetts could (or would) find that a federal prosecutor has acted unethically in failing to use a procedure that does not exist. Cf. Woods v. New York Life Insurance Co., 686 F.2d 578, 581 (7th Cir.1982) (state cannot punish lawyer for mailing to potential class members notice that was approved by federal court). Thus, unless and until the federal district court, or the Supreme Court, validly promulgates a rule creating a “prior approval” procedure, it seems to me that federal prosecutors cannot and need not use it.
For these reasons, I believe that a declaratory judgment similar to that prayed for by the appellants should issue.