This case grows out of an investigation by the National Institutes of Health's Office of Scientific Integrity into alleged scientific misconduct by Dr. James Abbs, a professor of neurology at the University of Wisconsin and the director of the national shared x-ray microbeam laboratory there. The laboratory studies the neurology of human speech with the aid of techniques that, we are told without contradiction, make it the only facility of its kind in the world. Abbs's research is supported by the National Institutes of Health (a branch of the Public Health Service of the Department of Health and Human Services) to the tune of a million or more dollars a year. Scientific fraud, on which see Patricia K. Woolf, "Deception in Scientific Research," 29 Jurimetrics J. 67 (1988), is much in the news these days; and this case, the government advises us in its brief, "is of far-reaching national significance, as it affects investigations into suspected scientific misconduct in research funded by the federal government. Essentially, it asks whether the policies and procedures governing such investigations comply with the Administrative Procedure Act and the due process clause of the Fifth Amendment."

In 1987 a predecessor body to the Office of Scientific Integrity received a tip that graphs in a then-recent article coauthored by Abbs had been traced from graphs in a previous publication rather than generated by data from a study supported by an NIH grant, as the article claimed. The University of Wisconsin investigated the allegation (superficially, in the view of the Office of Scientific Integrity) and exonerated Abbs. Nothing further happened until January 1990, when the Office of Scientific Integrity began to investigate the allegation. The first thing it did, however--an act not strictly investigative--was to place in the Public Health Service's "ALERT" system a notice that Dr. James Abbs of the University of Wisconsin was being investigated for scientific misconduct. The ALERT system distributes such notices to all agencies of the Public Health Service that make research grants. Next, the Office of Scientific Integrity retained outside experts to work with its staff in the investigation of the alleged misconduct. The investigative panel wanted to interview Dr. Abbs and at first he agreed, but he changed his mind when he discovered (in June 1990) that he would not have complete access to the investigative file, would not be allowed to attend interviews with other witnesses, and would not be entitled to a full evidentiary hearing before a finding of misconduct was made.

At that point Abbs filed this suit. It charges that the investigative procedures used by the Office of Scientific Integrity are invalid because they were not adopted in conformity with the requirements of the Administrative Procedure Act for administrative rulemaking and because they do not provide due process of law. Section 493(b) of the Public Health Service Act, 42 U.S.C. § 289b(b), directs NIH to establish a "process" for responding to complaints of scientific fraud. Pursuant to this directive, but without notice or opportunity for public comment, the predecessor of the Office of Scientific Integrity had announced "Policies and Procedures for Dealing with Possible Misconduct in Science" in the July 18, 1986, issue of an NIH publication called the NIH Guide for Grants and Contracts. The Office of Scientific Integrity adopted the procedures when it was established. Under them, the Office, when it receives a report of possible scientific misconduct (defined as "serious deviation, such as fabrication, falsification, or plagiarism, from accepted practices in carrying out research or in reporting the results of research"), is to conduct a preliminary inquiry to determine whether a formal investigation would be appropriate. If, as in Abbs's case, the office determines on the basis of this inquiry that there should be a formal investigation, it places the name of the scientist (and perhaps of his institution as well, although this is unclear) who is under investigation in the ALERT system and at the same time notifies the scientist and his institution that they are being investigated. The notice must describe the issues that are the focus of the investigation. The scientist is entitled to make written submissions and to comment on the Office of Scientific Integrity's proposed findings and sanctions. If the Office concludes that there has been scientific misconduct, it can recommend to higher officials in the Department of Health and Human Services sanctions ranging from a letter of reprimand to the termination of a grant. The intermediate sanctions include requiring special approval for particular activities as a condition of a grant and prohibiting the scientist from serving for a specified period of time on peer review committees for Public Health Service grants. The Office of Scientific Integrity can impose some sanctions, including suspension of current grants or of eligibility for new grants, even before the investigation is completed.

The procedures have become somewhat more elaborate over the years. Notice of Policies and Procedures for Dealing with Possible Scientific Misconduct in Extramural Research, 56 Fed.Reg. 27384 (June 13, 1991). The scientist now has a right to an interview at which he can be assisted by counsel and a right to comment on the outside experts whom the Office of Scientific Integrity proposes to engage to assist in the investigation. If the office recommends terminating a grant or barring the scientist or his institution from future grants, further procedural safeguards click in. The new procedures are academic in this case because Abbs was offered an interview at which he could be assisted by counsel and an opportunity to comment on the outside experts and because the Office has not yet recommended any of the draconian sanctions that under the new procedures would entitle Abbs to still further safeguards.

The university became a coplaintiff with Abbs in this suit, pointing to its financial stake in the grants by the NIH that pay for Abbs's research and also defray some of the University's overhead expenses. Both sides moved for summary judgment. In December 1990 the district judge entered a purported judgment granting the plaintiffs' motion in part and the defendants' in part and denying the plaintiffs' motion for preliminary injunction as moot. 756 F. Supp. 1172 (W.D.Wis.1990). In an accompanying opinion the judge explained that while the procedures employed by the Office of Scientific Integrity were indeed invalid because adopted in violation of the Administrative Procedure Act (she rejected the government's argument that they are within various exemptions to the Act), they did not deny the plaintiffs due process of law--and anyway, she opined, Abbs has no liberty or property interest in continued funding by NIH, so even if the procedures were inadequate he had no constitutional claim. The plaintiffs moved the judge to vacate the portion of her opinion and judgment that pronounced the procedures constitutional. They argued that the constitutional issue had been rendered moot by the judge's ruling that the procedures violated the Administrative Procedure Act. The judge denied this motion, clearing the way for all parties--so they thought, anyway--to appeal.

Pending the decision of the appeal, the Office of Scientific Integrity has suspended its investigation of Dr. Abbs, although the district judge never entered an injunction against its conducting the investigation by the procedures that she had held invalid. The Public Health Service has not removed Dr. Abbs's name from the ALERT system; nor has either plaintiff attempted to get it to do so. Yet, so far at least, the accusation against Abbs, the posting of his name in the ALERT system, and the investigation now in limbo have not affected his funding by the NIH or any other agency. Indeed, so far as appears, it has not interfered with his activities or those of the university in the least.

We begin, and in fact end, with a hornet's nest of jurisdictional issues. The first concerns our jurisdiction of the appeal. There is no judgment in the district court--at least no judgment that conforms to Rule 58 of the civil rules. Entered on the usual form for Rule 58 judgments employed by the district court was the judge's ruling granting both sides' motions for summary judgment in part and denying the motion for a preliminary injunction. The latter order, the denial of the motion for a preliminary injunction, was appealable though interlocutory, 28 U.S.C. § 1292(a)(1), but has not been appealed. An order granting or denying summary judgment in whole or part is not itself a judgment; it is an interlocutory ruling, and is unappealable. Soo Line R.R. v. Escanaba & Lake Superior R.R., 840 F.2d 546, 549 (7th Cir.1988). The statement in the Wright and Miller treatise that "in the two-party, single-claim situation the granting of a summary judgment is a 'judgment' within the definition of Rule 54(a) and an appeal is proper," 10 Charles Alan Wright, Arthur R. Miller & Mary Kay Kane, Federal Practice and Procedure § 2715, at p. 626 (2d ed. 1983) (citations omitted), is misleading. Not only does it ignore Rule 58; in addition, a grant of summary judgment in the plaintiff's favor on liability alone in a damages suit does not end the case in the district court and is therefore unappealable wholly apart from Rule 58. Liberty Mutual Ins. Co. v. Wetzel, 424 U.S. 737, 96 S. Ct. 1202, 47 L. Ed. 2d 435 (1976); University Life Ins. Co. v. Unimarc Ltd., 699 F.2d 846, 849 (7th Cir.1983).

Sometimes, it is true, if one pieces together a ruling on summary judgment with the judge's opinion in which the ruling is explained, one may be able to discover a final judgment ("final decision" is the actual statutory term, 28 U.S.C. § 1291). That constructed or inferred judgment will not comply with Rule 58. The rule requires that the judgment order be separate from the opinion and thus that it be self-contained in the sense that the outcome of the case (albeit not the reasoning supporting the outcome) is stated in the judgment order, without any need to consult the opinion. Bethune Plaza, Inc. v. Lumpkin, 863 F.2d 525, 527 (7th Cir.1988). That emphatically is not the case here. But notwithstanding Fed.R.App.P. 4(a)(6), which says that a judgment is appealable when entered in conformity with Rule 58, compliance with Rule 58, although a great time saver, is not jurisdictional. If the district judge is through with the case there is a final judgment, even if not memorialized in conformity with Rule 58. Bankers Trust Co. v. Mallis, 435 U.S. 381, 98 S. Ct. 1117, 55 L. Ed. 2d 357 (1978) (per curiam); First National Bank v. Comptroller of the Currency, 956 F.2d 1360, 1363 (7th Cir.1992); Rosser v. Chrysler Corp., 864 F.2d 1299, 1305 (7th Cir.1988). That is the case here. The judge declared the government's investigative procedures for scientific misconduct invalid. That was a final declaratory judgment. Its finality was unaffected by the fact that the plaintiffs might later move for injunctive relief if the government refused to accept the declaration. 28 U.S.C. § 2201; Peterson v. Lindner, 765 F.2d 698, 703 (7th Cir.1985); cf. University Life Ins. Co. v. Unimarc Ltd., supra, 699 F.2d at 849; Disher v. Information Resources, Inc., 873 F.2d 136, 139 (7th Cir.1989); Adams v. Lever Bros. Co., 874 F.2d 393, 394 (7th Cir.1989). The government did accept it, subject to its right to appeal.

There is another question about appellate jurisdiction. Although the government lost in the district court, its appeal is styled as a cross-appeal. (When more than one appeal from a judgment is filed, the earliest filed is the appeal, and any later filed is a cross-appeal. 16 Charles Alan Wright, et al., Federal Practice and Procedure § 3950, at p. 367 n. 17 (1977).) The plaintiffs have appealed too. But what have they appealed? They won. Granted, they did not win as big as they would have liked. They knocked out the investigative procedures on the narrowest of procedural grounds--grounds that would allow the government, following the notice and comment period required by the Administrative Procedure Act, to repromulgate the identical procedures. The plaintiffs would much rather have won on constitutional grounds; then they would be entitled to better procedures, rather than to the same procedures following some additional paper shuffling. Their fear is hardly groundless, for in anticipation of possibly losing this appeal the government last summer gave notice of a new set of procedures for public comment--and the new procedures are much like the old (not surprisingly since the "old," as noted, include revisions also made last summer).

But a winner cannot appeal a judgment merely because there are passages in the court's opinion that displease him--that may indeed come back to haunt him in a future case. California v. Rooney, 483 U.S. 307, 107 S. Ct. 2852, 97 L. Ed. 2d 258 (1987) (per curiam); Electrical Fittings Corp. v. Thomas & Betts Co., 307 U.S. 241, 59 S. Ct. 860, 83 L. Ed. 1263 (1939). He can appeal only if the judgment gives him less relief than he considers himself entitled to. First National Bank v. Comptroller of the Currency, supra, 956 F.2d at 1364. Whether the judgment here does that would be easier to determine if there were a judgment order compliant with Rule 58, or an injunction compliant with Rule 65(d), which requires that an injunction be specific. Chicago & North Western Transportation Co. v. Railway Labor Executives' Ass'n, 908 F.2d 144, 149 (7th Cir.1990). (But instead of issuing an injunction the district judge issued a declaratory judgment, which was within her rights.) For example, if the judgment order had said that the government's investigative procedures for scientific fraud shall be suspended until repromulgated in conformity with the Administrative Procedure Act, rather than invalidated tout court, it would be plain that the plaintiffs had gotten less than they wanted, and they could appeal. But what we have called the constructed or inferred judgment contains no such qualification. It invalidates the procedures, period.

But we can look ahead, with the plaintiffs, and see that should the government repromulgate the same, or similar, rules, they would want to press their challenge to the rules' constitutionality; and they see the district judge's dictum as a roadblock. It is not. Dicta have no preclusive effect; even holdings do not; only judgments do. The judgment is that the rules are invalid. The fact that they are invalid for one reason and not another is no ground for appeal even if the reasons have different implications for future cases, which is to say different significance for the meaning of the decision as a precedent. A holding is a precedent, and it is constructed by the court asked to give it force as precedent from the facts recited in the earlier opinion and from what the opinion says about those facts. Normally the earlier decision can be read broadly or narrowly, and obviously what the decision says will influence, perhaps decisively, the interpretation placed on it by later decisions. But Rule 58 helps us see that a judgment and the reasons given in support of it are two different things. Anyway the unappealed holdings of district judges have no precedential weight--that is, no significance as authority--although their persuasiveness may influence a subsequent decision. Colby v. J.C. Penney Co., 811 F.2d 1119, 1124 (7th Cir.1987). And we are not talking about a holding but only about a dictum. "There is no known basis for an appeal from a dictum." Oxford Shipping Co. v. New Hampshire Trading Corp., 697 F.2d 1, 7 (1st Cir.1982).

It might seem that the plaintiffs got around this obstacle to appeal by moving the judge to vacate the constitutional part of her opinion, because her denial of that motion was a final, appealable judgment. But this is too facile a route to appeal. If as we believe a winning litigant cannot appeal from a phrase or sentence--or for that matter an extended discussionthat he doesn't like in the opinion announcing the judgment in his favor, he must not be permitted to lever his way into the appellate court by asking the judge to delete the offending phrase, or sentence, or whatever and then appealing from the judge's refusal to do so. Rules 59(e) and 60(b) speak of motions to amend or vacate or modify the court's judgment. Dicta are not judgments. A motion to vacate a dictum is outside the scope of these rules and an appeal from a denial of such a motion is frivolous and therefore does not engage our jurisdiction. Crowley Cutlery Co. v. United States, 849 F.2d 273 (7th Cir.1988).

An alternative route to this conclusion is found in the proposition that the denial of a Rule 59(e) motion is not "separately appealable" from the judgment unsuccessfully sought to be altered or amended. Cardoza v. Commodity Futures Trading Comm'n, 768 F.2d 1542, 1546-47 (7th Cir.1985). The premise of such a motion is that the judgment was wrong in some respect. This makes a challenge to the denial of the motion a challenge to the judgment itself, and is why the timely filing of the motion suspends the time for appealing from the judgment. If as in this case the "judgment" cannot be appealed--because it's not really a judgment--the party complaining of the "judgment" cannot be allowed to bootstrap his way to appeal by filing a post-judgment motion.

To almost every legal generalization there are exceptions; and the principle that the denial of a Rule 59(e) motion is not appealable apart from the judgment sought to be altered or amended is not an exception to that generalization. If the judgment was correct when rendered but afterward something happens that requires that the judgment be changed, the denial of a Rule 59(e) motion would be separately appealable from the judgment--the appellant's only quarrel with the district court being over its failure to take account of the post-judgment event. Mazzola v. Secretary of Health & Human Services, 795 F.2d 222, 223-24 (1st Cir.1986) (per curiam). But this is not such a case.

Although the plaintiffs had no right to appeal, they had every right to defend the judge's decision on the alternative ground that the challenged procedures violate the Constitution. But an appellee does not turn into an appellant by urging alternative grounds for affirmance that the district judge ignored or rejected. Rose Acre Farms, Inc. v. Madigan, 956 F.2d 670, 672 (7th Cir.1992).

The reader may wonder what difference it can possibly make if an appellee mistakenly describes himself as an appellant. It can make a big difference. A cross-appellant has 14 days after the appellant files his notice of appeal within which to file his own notice of appeal. Fed.R.App.P. 4(a)(3). Should the appellant file at the end of the permitted period, with the result that the cross-appellant's notice of appeal, though filed within 14 days, is outside the time within which the cross-appellant could have filed his appeal were it not a cross-appeal, and the first notice of appeal is later discovered to be invalid because the appellant had no right to appeal, the cross-appellant is out of luck. His appeal is untimely. First National Bank v. Comptroller of the Currency, 956 F.2d at 1364.

That is not a problem here. The government filed its improperly designated cross-appeal within the deadline for filing an original appeal. But by filing an invalid notice of appeal it could have lulled its adversaries into forgoing their right of appeal. We shouldn't like to see the Office of Scientific Integrity defraud its adversaries of their procedural rights.

So we have jurisdiction of these appeals, but did the district court have jurisdiction of the suit? Questions of jurisdiction to review the actions of administrative agencies usually are discussed under such murky rubrics as ripeness, prematurity, exhaustion, finality, and standing. However, we can frame the issue in this case (and perhaps not only in this case, but generally, though that remains to be seen) as a straightforward question of statutory interpretation. The statute that authorizes judicial review of federal agency orders not made reviewable by a specific statute confines that review jurisdiction to "final agency action for which there is no other adequate remedy in a court." Administrative Procedure Act, § 10(c), 5 U.S.C. § 704. Ordinarily that means a final order imposing some sort of sanction. It emphatically does not mean the issuance of the administrative complaint, kicking off the administrative proceeding. FTC v. Standard Oil Co., 449 U.S. 232, 101 S. Ct. 488, 66 L. Ed. 2d 416 (1980); R.R. Donnelley & Sons Co. v. FTC., 931 F.2d 430 (7th Cir.1991). Such a complaint "ha[s] no legal force or practical effect upon [the respondent's] daily business other than the disruptions that accompany any major litigation," 449 U.S. at 243, 101 S.Ct. at 495, and these disruptions are not enough to make the ordinary judicial remedy--review of the sanction when and if imposed--inadequate within the meaning of the statute. Otherwise a person could run into federal court the moment an agency filed an administrative complaint against him. Or before: the investigation of Dr. Abbs hasn't even gotten to the complaint stage yet.

It is true that the issuance of a rule is final agency action, and can be challenged by a suit such as this for a declaration of invalidity, provided that (as section 10(c) requires) there is no other adequate judicial remedy. Abbott Laboratories v. Gardner, 387 U.S. 136, 87 S. Ct. 1507, 18 L. Ed. 2d 681 (1967); Frozen Food Express v. United States, 351 U.S. 40, 76 S. Ct. 569, 100 L. Ed. 910 (1956). Abbs is challenging the set of rules governing the investigation; he is not trying to derail the investigation or the ensuing administrative proceeding (should the Office of Scientific Integrity conclude that Abbs committed misconduct, and recommend sanctions) by getting us to decide that he did not commit scientific misconduct. Though derailment has been the consequence of the suit and provides a pragmatic reason against allowing it, all he is explicitly asking us to do is to determine the ground rules by which the administrative investigation and any further administrative proceedings are to be conducted.

There is, though, a critical difference between challenging a rule of conduct that carries sanctions for its violation and a rule of procedure (broadly conceived--for we don't want to get into the question whether "Policies and Procedures for Dealing with Possible Misconduct in Science" fits the APA's exemption for procedural rules, 5 U.S.C. § 553(b)(A)) that specifies what sanctions, and how, are applicable to (alleged) misconduct already committed. If the rule of conduct with sanctions attached cannot be challenged in advance of violating it, the people subject to it are placed in a dilemma: comply with a rule that harms them and that they believe to be invalid or violate the rule at the risk of incurring a heavy penalty (criminal, in both Abbott Laboratories and Frozen Food Express ) if they've guessed wrong and the rule is upheld in the penalty proceeding. 4 Kenneth Culp Davis, Administrative Law Treatise § 25:6, at p. 369 (2d ed. 1983). In such a case, the judicial remedy that comes at the end of the penalty proceeding is inadequate in the familiar equity sense that permits, for example, the issuance of a preliminary injunction in a damages case upon a showing of irreparable harm. Roland Machinery Co. v. Dresser Industries, Inc., 749 F.2d 380, 386 (7th Cir.1984). So the rule can be challenged directly, in a separate proceeding such as the one here. National Automatic Laundry & Cleaning Council v. Shultz, 443 F.2d 689, 696 (D.C.Cir.1971).

Dr. Abbs, however, is not challenging a rule of conduct. It is not as if the NIH had told him that to keep his grant he must stop x-raying human beings. Or as if he were claiming a right to trace old graphs. The rule of conduct has (or has not) been violated, and the question now is only what procedures to employ. So he faces no dilemma, runs no risks. If as he argues the methods that the Office of Scientific Integrity is employing in its investigation violate due process or the APA, then if and when a sanction of any sort is meted out to him he can challenge it on the ground that it violates his procedural rights, and if the challenge succeeds he will be home free. If the challenge fails, he will be no worse off than he would have been had he never brought the present suit, assuming (for symmetry) that this suit terminated adversely to him on the merits. A challenge to administrative action, whether the action is denominated a rule or a complaint, falls outside the grant of jurisdiction in section 10(c) of the Administrative Procedure Act when the only harm the challenger seeks to avert is the inconvenience of having to go through the administrative process before obtaining a definitive declaration of his legal rights. That is the teaching of the Standard Oil decision and finds an echo in many lower-court cases as well, not only Donnelley but also General Finance Corp. v. FTC, 700 F.2d 366 (7th Cir.1983); Bethlehem Steel Corp. v. EPA, 536 F.2d 156 (7th Cir.1976); Solar Turbines, Inc. v. Seif, 879 F.2d 1073, 1081-82 (3d Cir.1989), and Association of National Advertisers, Inc. v. FTC, 617 F.2d 611, 621-22 (D.C.Cir.1979).

It might seem, however, that insofar as Dr. Abbs challenges the ALERT system as a form of blacklist, even full exoneration in the administrative proceeding or some subsequent judicial review proceeding would not eliminate the injury of which he complains, because the "blacklist" (if that is what it is) comes in before the formal investigation, let alone the formal determination (and judicial review thereof), of the alleged fraud. The district judge said that it was "not mere speculation to think that the placement of [Abbs's] name in [the ALERT] system will lead to delays in approval of funding requests for research on which he is the principal investigator; indeed, it would be difficult to argue that the system is working properly if it did not lead to such delays." 756 F. Supp. at 1180-81. A number of cases, most famously Joint Anti-Fascist Refugee Comm. v. McGrath, 341 U.S. 123, 71 S. Ct. 624, 95 L. Ed. 817 (1951), hold that a blacklist can inflict the sort of harm that entitles people to maintain a suit in federal court. In that case, as a result of being labeled communists, the plaintiffs lost tax exemptions and licenses, not to mention supporters, members, and places where they could hold meetings. Id. at 158-60, 71 S.Ct. at 641-42 (Frankfurter, J., concurring); see also Bantam Books v. Sullivan, 372 U.S. 58, 68-69, 83 S. Ct. 631, 638, 9 L. Ed. 2d 584 (1963). The issue was not the adequacy of the plaintiffs' alternative remedies--they had none--but whether they had suffered the kind and amount of harm necessary to satisfy the requirements of Article III. If Abbs has suffered a like harm, he can maintain this suit because the harm would not be removed by ultimate vindication years after. The harm would therefore be irreparable, the alternative judicial remedy therefore inadequate, and the requirements of section 10(c) therefore satisfied.

But in striking contrast to the Joint Anti-Fascist Refugee case there is no indication that Abbs has suffered, or will suffer, any harm whatsoever as a consequence of being placed in the ALERT system. We do know that he has lost no grants as yet. His funding is undiminished. He, or his lawyer, is sufficiently insouciant about the whole business to have made no effort to remove Abbs's name from the ALERT system even after the invalidating of the investigative procedures by the district judge removed the ground on which the government might have resisted such a move. We know that the scientific grants business is a competitive one and we can understand the district judge's assuming that the ALERT system could have no purpose other than to induce granting agencies within the Public Health Service to look upon grant applications with a fishier eye. We emphatically disassociate ourselves from the government's argument (offered for a different purpose--that of showing that Abbs has no constitutional claim) that placing Abbs in the ALERT system, and even finding him guilty of scientific misconduct, could not possibly invade a legally protected interest of his because even if he loses all his grants--even if he becomes a pariah to grantmakers, public and private--he can always go back to teaching neurology in the classroom. To those unfamiliar with modern education, "teaching" is one profession, wherever or however performed. In fact it is a highly stratified profession. Even "higher" education is highly stratified. A college teacher of neurology is not, in a realistic sense, in the same occupation as the director of a world-renowned research laboratory, who though employed by a university does little classroom teaching because his primary value to the university and to his academic specialty--perhaps also to society--is in the conduct and administration of advanced scientific research. Put this director back in the classroom, having stripped away his scientific standing and access to research funds by condemning him for scientific fraud, and you exclude him from his occupation.

That is putting things in their worst light. It would be quite enough, so far as Abbs's right to challenge the "blacklist" is concerned, if by reason of placement in the ALERT system Abbs were in serious danger of having his grant money reduced. But it has been five years since he was placed in the system and nothing has happened to him and so far as appears nothing will happen to him unless and until the Office of Scientific Integrity recommends sanctions. And we are far from that.

The complaint is appropriately cautious. It says that Abbs "may be subject to a wide range of sanctions ...; his reputation and good standing may be destroyed; his opportunity and ability to secure governmental and private research grants may be destroyed" (emphasis added); and so on in this vein. And these iffy consequences are predicted to arise either from a determination of misconduct, which has not been made, or from sanctions, which have not been imposed. There is no suggestion that the mere listing of Abbs's name in the ALERT system could have any of these dire consequences. Nor does the complaint seek, as part of the relief asked for, that his name be removed from the system. Have he and his lawyer simply blundered, by overlooking a palpable source of harm to Abbs? Probably not, because an uncontradicted affidavit states that "an entry in the ALERT system is confidential," that it is available to "other PHS [Public Health Service] components ... on a need to know basis," and that "the individuals who are reviewing Dr. Abbs' grant applications for scientific merit have not been informed of the OSI investigation of Dr. Abbs, which is confidential." The affidavit is consistent with the administrative regulations. Notification of Altered Privacy Act System of Records, 52 Fed.Reg. 19929 (May 28, 1987). No longer is it so mysterious how Dr. Abbs could have remained in the ALERT system for five years without any adverse effect on his career. When and if he is exonerated his name will be removed from the system, and it is entirely speculative that he will suffer any harm in the interim. His complaint does not so much as allege any such harm. We explained earlier why his challenge to the investigative procedures is outside of section 10(c); we now see that the pre-investigative "blacklist" adds nothing to the case for immediate judicial review. Kukatush Mining Corp. v. SEC, 309 F.2d 647, 649-51 (D.C.Cir.1962).

Of course no one likes to be accused of misconduct. The district judge remarked that "the fact that Abbs might ultimately be cleared of the accusations against him ... is as unconvincing as arguing that persons charged with felonies have no cognizable interest in the trial procedures afforded them because they might be acquitted." 756 F. Supp. at 1181. But a criminal defendant cannot bring a suit to enjoin the procedures under which he is to be tried. No more can Dr. Abbs.

We have not forgotten that there are two plaintiffs. But the university has not demonstrated any irreparable harm to itself either. Neither plaintiff can maintain this suit consistently with section 10(c) of the Administrative Procedure Act. The judgment of the district court is therefore vacated and the suit is dismissed.

CUDAHY, Circuit Judge, concurring in part and dissenting in part.

I agree fully with the majority that the procedural rule here in question is not judicially reviewable unless or until sanctions have been imposed under it. The reasons for this conclusion are persuasively set forth in the majority opinion. I have strong doubts, however, that this result should be a matter of jurisdiction under section 10(c) of the Administrative Procedure Act.

I find the Supreme Court's decision in Toilet Goods Ass'n v. Gardner, 387 U.S. 158, 87 S. Ct. 1520, 18 L. Ed. 2d 697 (1967), issued the same day as Abbott Laboratories, particularly telling in this regard. In Toilet Goods, the petitioners brought a preenforcement declaratory judgment action challenging the following regulation promulgated by the Commissioner of the Food and Drug Administration:

"(a) When it appears to the Commissioner that a person has:

....

(4) refused to permit duly authorized employees of the Food and Drug Administration free access to all manufacturing facilities, processes, and formulae involved in the manufacture of color additives and intermediates from which such color additives are derived;

he may immediately suspend certification service to such person and may continue such suspension until adequate corrective action has been taken."

387 U.S. at 161, 87 S. Ct. at 1523 (quoting 21 C.F.R. § 8.28). The Court explicitly held that there was "no question that this regulation--promulgated in a formal manner after notice and evaluation of submitted comments--is a 'final agency action' under § 10 of the Administrative Procedure Act." Id. at 162, 87 S.Ct. at 1523. Nevertheless, the Court found that the question of the regulation's validity was not appropriate for judicial resolution--that is, that the issue was not "ripe." In so doing, the Court observed that "[t]his is not a situation in which primary conduct is affected.... This regulation merely states that the Commissioner may authorize inspectors to examine certain processes or formulae; no advance action is required of cosmetics manufacturers...." Id. at 164, 87 S.Ct. at 1524.

Thus, the Toilet Goods Court makes precisely the same distinction as that on which the majority relies here--the difference in preenforcement impact between a rule of procedure and a rule of conduct. And the end result is the same--no review. But the Toilet Goods Court based its holding on the doctrine of ripeness, not on a conclusion that there was no "final agency action" and therefore no "jurisdiction." (I do not think the fact that the Toilet Goods regulation had been promulgated by notice and comment rulemaking really distinguishes the cases; that distinction would make no sense where the issue is whether the agency should have engaged in such rulemaking. Nor does the majority suggest any basis for such a distinction.)

Whether a particular case is decided on grounds of ripeness or on lack of final agency action may not make any difference, since as a practical matter the result is the same. But the latter approach may have greater potential to constrain judicial review of agency action, given the relative inflexibility of "jurisdictional" requirements.

For these reasons, although pragmatically there may be little difference in the outcome, I would find the agency rule here unripe for judicial review rather than finding the court without jurisdiction.