This class action on behalf of prisoners confined in the segregation unit at Stateville, the State of Illinois' maximum-security prison, charges that living conditions in the unit are so substandard as to be a form of cruel and unusual punishment within the meaning of the Eighth Amendment, made applicable to the states by interpretation of the Fourteenth Amendment. (The average number of prisoners confined in segregation at Stateville is 225, but this suit is limited to those confined there for more than ninety consecutive days.) A jury agreed, and awarded nominal damages to the class. The district judge set aside the award of damages on the ground that the defendants (Illinois correctional officials) were immune from liability for damages, but he entered an injunction which requires the defendants to provide "every Stateville segregation inmate who has been confined in segregation for 90 or more consecutive days the opportunity to shower at least three times each week and to exercise outside of his cell at least five hours each week." However, if the "inmate violates prison rules during his exercise or shower period, he may be denied the exercise or shower rights provided herein for a reasonable period of time pursuant to regulation approved by this court." 653 F. Supp. 649 (N.D.Ill.1987).

The defendants' challenge to the district court's decision is a narrow one. They do not challenge the jury's finding that the conditions of segregated confinement at Stateville violated the Eighth Amendment. They do not argue that the district judge should have paid no attention to the finding because ultimately he determined that the defendants were immune from liability for damages and set aside the jury's verdict. A public official's immunity in a suit for damages is, normally, immunity from trial, not just from the award of damages. Mitchell v. Forsyth, 472 U.S. 511, 526, 105 S. Ct. 2806, 2815-16, 86 L. Ed. 2d 411 (1985); Green v. Carlson, 826 F.2d 647, 651 (7th Cir.1987). If there are factual disputes that cannot be resolved without a trial, then trial there must be, id. at 652; but there was no trial on immunity here; after the trial the judge decided that the defendants were entitled to immunity after all, and set aside the jury's verdict. But the defendants do not argue that, by doing this, he deprived the jury's finding of liability of binding force in the injunction suit. They easily could so argue, as we shall see.

Nor do they argue that the ultimate question in an Eighth Amendment case is a question of law for the judge to decide rather than a question of fact for the jury. The question whether a rule of law has been violated--a question that requires applying the rule to the facts--is normally treated as a question of fact, see, e.g., Mucha v. King, 792 F.2d 602, 605 (7th Cir.1986), not because it is a question of fact (it isn't) but as a way of expressing a decision to leave the answer to the trial judge or jury to make, subject only to limited appellate review. The defendants miss this point in arguing, half-heartedly, that ultimate questions are not questions of fact. They base this argument on a case, no longer authoritative, which had held on the basis of a decision later expressly overruled by the Supreme Court in Pullman-Standard v. Swint, 456 U.S. 273, 285-86, 102 S. Ct. 1781, 1788-89, 72 L. Ed. 2d 66 (1982), that the question of discrimination in a Title VII case is one of law rather than of fact. See Stewart v. General Motors Corp., 542 F.2d 445, 449 (7th Cir.1976).

There are exceptions to the principle that ultimate questions are deemed questions of fact, an example being the question of actual malice in defamation cases; many of the exceptions are discussed in Miller v. Fenton, 474 U.S. 104, 106 S. Ct. 445, 88 L. Ed. 2d 405 (1985), and Bose Corp. v. Consumers Union of United States, Inc., 466 U.S. 485, 505-11, 104 S. Ct. 1949, 1962-65, 80 L. Ed. 2d 502 (1984). The defendants do not argue that this case falls within one of the exceptions, though in Rhodes v. Chapman, 452 U.S. 337, 101 S. Ct. 2392, 69 L. Ed. 2d 59 (1981), the Supreme Court had implicitly treated the question whether particular prison conditions amounted to cruel and unusual punishment as one of law, see id. at 347, 352, 101 S. Ct. at 2399-2400, 2402; none of the opinions in that case, however, discusses the standard of review. Several lower-court decisions treat the question as one of fact, but again without discussion. See, e.g., Shrader v. White, 761 F.2d 975, 980 (4th Cir.1985); Blake v. Hall, 668 F.2d 52, 54-55 (1st Cir.1981); cf. Joseph v. Brierton, 739 F.2d 1244, 1246 (7th Cir.1984). We need not try to resolve the issue, given the defendants' failure to raise it.

All that the defendants are arguing is that the most the Constitution guarantees the segregated inmates of Stateville under current conditions in the segregation unit--a vital qualification, as will shortly appear--is one hour of exercise outside the cell, and one shower, per week, and hence that the injunction is too severe. In so arguing the defendants may seem to have overlooked the possibility that the injunction might be intended to be more severe than the Constitution, in order to rectify more effectively the constitutional violations that the jury found and that the defendants no longer contest. See, e.g., Hutto v. Finney, 437 U.S. 678, 687, 98 S. Ct. 2565, 2571-72, 57 L. Ed. 2d 522 (1978). But we do not understand the injunction to reflect such a design; it seems rather to reflect the district court's view of the constitutional minimum in the circumstances.

The defendants' challenge to the terms of the injunction may seem inconsistent with their failure to challenge the jury's finding of a constitutional violation. But as they point out, much of the evidence on which that finding was based concerned conditions in F-House, the building that housed the segregation unit until shortly before the trial. Conditions in F-House, an old unit, were very bad. Not only were the cells small (4'9"' by 10'6"'), but they were dirty and roach-infested, toilets were frequently backed up, and food was served to the inmates cold. By the time of trial all the inmates had been moved to a new segregation unit, I-House, and not only is I-House newly built and modern but its cells measure 7'2"' by 13'2"', giving them almost twice the square footage as the cells in F-House (94 square feet versus 50 square feet). There is no evidence of unsanitary conditions or unheated food in I-House; and although the doors to the cells have only narrow slits rather than the barred windows of the doors to F-House's cells, each cell in I-House has an outside window at the back. The windows in the doors of F-House's cells, combined with the circular shape of F-House, made F-House a more sociable habitat for the inmates than I-House, but also and by the same token made the inmates more rowdy and by doing so may have been indirectly responsible for much of the filthy condition of the cells. Rowdy inmates are given to throwing food out of their cells, stopping up toilets, and preventing maintenance staff from entering.

There was testimony, both lay and medical, including testimony by the medical director of the Illinois Department of Corrections, Dr. Shansky, that four to seven hours of exercise outside the cell, and three showers, are the weekly minimum necessary to prevent serious adverse effects on the physical and mental health of inmates confined, as these inmates are, in what realistically is a form of solitary confinement. It is true that each cell contains a sink in which the inmate can wash himself as often as he wants and that the cells in I-House are large enough for the inmate to engage in various forms of exercise, including push-ups, sit-ups, step-ups, and running in place. It is also true that the inmates are allowed to leave their cells for a variety of reasons, such as to use the law library, see visitors, consult with lawyers and other counselors, and visit the medical unit. But these excursions consume in the aggregate only two to three hours per week on average, and the degree of constraint is considerable. A visit to the law library, for example, means being escorted in handcuffs to a caged carrel in the library--not browsing in stacks or working at a library table in a reading room.

Confinement in segregation is usually for a much shorter time than the prisoner's full term. Whereas most Stateville prisoners are violent criminals serving long terms (one-third are serving terms for murder and three-fourths have been convicted of murder at some point in their criminal careers), most prisoners confined to segregation serve only a few months there (at least consecutively). But the plaintiff class is, as noted earlier, limited to prisoners who served more than ninety consecutive days in segregation, and several of the class members spent more than a year there and one spent more than two and a half years there.

Confinement in Stateville's segregation unit involves considerable isolation, sometimes for protracted periods; and the record shows, what anyway seems pretty obvious, that isolating a human being from other human beings year after year or even month after month can cause substantial psychological damage, even if the isolation is not total. Of course, it is highly probable that the experience of being imprisoned inflicts psychological damage whether or not the prisoner is isolated, so it is only the marginal psychological damage from segregation that is relevant. And the infliction of disutility, to borrow a convenient economic term, is one of the objectives of criminal punishment; only if the only objective of punishment were incapacitation could it be argued that living conditions should be as comfortable in prison as outside. The Eighth Amendment has, however, been interpreted to place limitations on the harshness of prison conditions, and the defendants as we have said do not quarrel with the propriety of the jury's having determined in this case that conditions in the segregation unit at Stateville exceeded those limitations. The principal--though not, as it seems to us, an insuperable--difficulty in using this proposition to ground the injunction is that most of the testimony at trial was directed toward the totality of conditions in segregation at Stateville, including the small size and filthy conditions of the cells in F-House. The totality is of limited significance in appraising the injunction. The injunction looks to the future; and the future is I-House, not F-House.

The difficulty is compounded by the statement in the district judge's opinion that he "should accept the jury's findings as predicate for equitable relief." 653 F. Supp. at 663. If the judge meant by this that he was required to issue the injunction that would have been appropriate if I-House had never been built--appropriate if the members of the plaintiff class were still confined in F-House--he was wrong. It is true that if the jury's finding on liability had been valid, he could not have refused to issue an injunction on the ground that he disagreed with the jury's finding. Hunter v. Allis-Chalmers Corp., 797 F.2d 1417, 1421 (7th Cir.1986); Hussein v. Oshkosh Motor Truck Co., 816 F.2d 348, 355 (7th Cir.1987). The jury's verdict would bind the judge under the doctrine of collateral estoppel (thus demonstrating that the doctrine does not require a final judgment in the conventional sense). See Restatement, Second, Judgments Sec. 13 and comment g and illustration 3; Lummus Co. v. Commonwealth Oil Refining Co., 297 F.2d 80, 89 (2d Cir.1961) (Friendly, J.). But he set aside the jury's verdict. It is true that he did this not because he disagreed with its findings but because he thought that the defendants had immunity from damages liability. But by setting aside the verdict on a ground not challenged by the plaintiffs he deprived the findings of the collateral estoppel effect that they would otherwise have had. The plaintiffs had no right to a jury verdict, so how can they have a right to benefit from it? In any event, and more fundamentally, the judge was required to tailor the equitable relief to the actual circumstances before him, which included the circumstance that the prisoners are now housed in I-House.

Did he try to do this? A positive answer is strongly implied by his discussion of conditions in other prisons. See 653 F. Supp. at 663. A negative answer is weakly implied by his citation (also id.) to the page of Shelby County Jail Inmates v. Westlake, 798 F.2d 1085, 1092 (7th Cir.1986), at which this court, while indicating that the district judge must make an independent judgment on equitable issues insofar as they are not identical to the legal issues determined by the jury, found that the legal and equitable issues in that case were identical. Maybe the district judge in this case thought the same thing, in which event he was wrong, because the jury did not purport to determine the equitable relief necessary to prevent the perpetuation or recurrence of the defendants' unlawful acts. But although the district judge's opinion could have been clearer on this important point, we conclude that he did make an independent judgment on the equitable relief necessary to secure the plaintiffs' rights in I-House, which is all that is relevant since no one suggests that the defendants might decide to move the segregation unit back to F-House.

We are not disposed to quarrel with the district judge's conclusion that to allow even I-House inmates only one hour of exercise out of the cell per week (the defendants' position) would fall below the minimum decencies that the Eighth Amendment requires of incarceration, especially since the defendants' concession requires us to treat this as a question of fact and therefore to defer broadly to the answer given by the district judge. There was extensive evidence, much of it conflicting, and we cannot say that in sorting through it the judge committed clear error.

Of course, as the defendants argue, an inmate can preserve his cardiovascular fitness and overall muscle tone by running in place or by engaging in other forms of exercise that are feasible without leaving the cell; but this misses the point. The opportunity to exercise or even just to stand around, outside, in the company of other prisoners, is the inmate's principal break from what otherwise is solitary confinement except for his limited opportunities to visit other parts of the prison on specific and tightly supervised missions. With only one hour of exercise outside the cell a week and only two other hours outside the cell, an inmate spends 165 out of 168 hours in a 90-square-foot cage. That is too little, or so at least the district judge could find without committing a clear error.

This conclusion depends, obviously, on the proposition that the requirements of the Eighth Amendment have changed over time; for it was not so long ago that solitary confinement--real solitary confinement--was considered neither unusual nor especially cruel. But "the Eighth Amendment has not been regarded as a static concept." Gregg v. Georgia, 428 U.S. 153, 172-73, 96 S. Ct. 2909, 2925, 49 L. Ed. 2d 859 (1976) (plurality opinion); see also Estelle v. Gamble, 429 U.S. 97, 102, 97 S. Ct. 285, 290 50 L. Ed. 2d 251 (1976); Hutto v. Finney, 437 U.S. 678, 685, 98 S. Ct. 2565, 2570-71, 57 L. Ed. 2d 522 (1978); Rhodes v. Chapman, supra, 452 U.S. at 346, 101 S. Ct. at 2399 ("No static 'test' can exist by which courts determine whether conditions of confinement are cruel and unusual, for the Eighth Amendment 'must draw its meaning from the evolving standards of decency that mark the progress of a maturing society' "). It would not be an arid literalism--it would be a bold interpolation--to read the amendment as if the words "as so regarded on the date of this enactment" followed the words "cruel and unusual punishments." The term "cruel and unusual punishments" was taken verbatim from the English Declaration of Rights of 1689, see Granucci, "Nor Cruel and Unusual Punishments Inflicted:" The Original Meaning, 57 Calif.L.Rev. 839, 852-53 (1969), so if a historical view is taken perhaps the proper reference date is a century before the enactment of the Eighth Amendment. Like much of the rest of the Bill of Rights, the Eighth Amendment would be obsolete if its meaning were thought bounded by the mental horizons of its eighteenth-century framers, broad as those horizons were.

The term "cruel and unusual punishments" is relative rather than absolute. The conditions in which prisoners are housed, like the poverty line, is a function of a society's standard of living. As that standard rises, the standard of minimum decency of prison conditions, like the poverty line, rises too.

A knowledgeable witness testified that he knows of no other prison in the United States, including the federal penitentiary at Marion, Illinois--the nation's highest-security prison--that allows inmates of its segregation unit so little time for out-of-cell exercise. The defendants could have presented a witness who would testify that there were other such prisons--or, if not, still that Stateville is at or above the constitutionally permissible minimum. It cannot be right that the least comfortable prison conditions in the United States automatically violate the Constitution, cf. Rummel v. Estelle, 445 U.S. 263, 281-82, 100 S. Ct. 1133, 1142-43, 63 L. Ed. 2d 382 (1980); for then, by the inevitable progression of successive cases, all conditions other than the most comfortable would be found to violate it. The defendants did not present such a witness, however, and the district judge was entitled to draw the natural inference from their failure to do so. Nor do the defendants argue that the statement in Rhodes that opinions of expert witnesses do not establish constitutional minima, see 452 U.S. at 348 n. 13, 101 S. Ct. at 2400 n. 13, precludes the taking of expert testimony on questions (such as psychological harm, or comparative prison practices) logically related to the constitutional judgment. From the evidence before him the judge could reasonably conclude that the defendants must provide each inmate in I-House with at least five hours of exercise time per week in order to comply with the Eighth Amendment. The exception for fractious inmates protects the defendants' legitimate interest in security, well illustrated by United States v. Fountain, 768 F.2d 790 (7th Cir.1985); it is unfortunate but unavoidable that the exception may itself engender litigation by inmates claiming that the defendants are misapplying it.

Our conclusion that the district judge was authorized to establish a five-hour minimum for exercise is bolstered by other judicial decisions on prisoners' exercise rights. The decisions are fact-specific; not surprisingly, the federal courts have not attempted to specify detailed guidelines for the administration of state prisons. Nevertheless we are impressed by the number of decisions that hold or suggest that a failure to provide inmates (confined for more than a very short period--a qualification that distinguishes such cases as Harris v. Fleming, 839 F.2d 1232, 1236-37 (7th Cir.1988), and Shelby County Jail Inmates v. Westlake, supra, 798 F.2d at 1089) with the opportunity for at least five hours a week of exercise outside the cell raises serious constitutional questions. See, e.g., French v. Owens, 777 F.2d 1250, 1255-56 (7th Cir.1985); Ruiz v. Estelle, 679 F.2d 1115, 1151-52 (5th Cir.1982); Campbell v. Cauthron, 623 F.2d 503, 507 (8th Cir.1980) (pretrial detainees); Spain v. Procunier, 600 F.2d 189, 199-200 (9th Cir.1979); Sweet v. South Carolina Dept. of Corrections, 529 F.2d 854, 865-66 (4th Cir.1975); Frazier v. Ward, 426 F. Supp. 1354, 1367-69 (N.D.N.Y.1977); cf. Caldwell v. Miller, 790 F.2d 589, 601 (7th Cir.1986). We do not suggest that this is always and everywhere the constitutional minimum; much less may suffice when the period of incarceration is brief. But a judge who, on an adequate record, requires this level of exercise opportunity has not exceeded the outer bounds of permissible interpretation of the Eighth Amendment.

In contrast, we do not think that the provision of the injunction requiring the defendants to allow the inmates to take three showers a week has adequate support in the record or in constitutional doctrine. No doubt Americans take the most showers per capita of any people in the history of the world, but many millions of Americans take fewer than three showers (or baths) a week without endangering their physical or mental health, and abroad people as civilized and healthy as Americans take many fewer showers on average, as every tourist knows. Any inmate of Stateville's segregation unit who wants to keep as clean as a free person can wash himself daily, or if he wants hourly, in the sink in his cell--a point ignored by Dr. Shansky, the medical expert for the plaintiff class on the issue of showers. There is evidence that inmates of the segregation unit consume on average more medical services than nonsegregated inmates, but there is no evidence connecting this datum to the number of showers for each group. The importance of the daily shower to the average American is cultural rather than hygienic; as Dr. Shuman, the defendants' expert, testified, "I know of nothing that says that a shower is necessary more than one day a week. I believe in this culture it is done that way but I don't know of any medical need for it." The judge was not obliged to believe the defendants' witness, but Dr. Shansky made the same point as Shuman when he said that being able to take only one shower a week makes inmates feel "less than human." While there is plenty of medical and psychological literature concerning the ill effects of solitary confinement (of which segregation is a variant), see, e.g., Grassian, Psychopathological Effects of Solitary Confinement, 140 Am.J.Psychiatry 1450 (1983), we have not found any corresponding literature on the effects of limiting the number of showers that prisoners may take. The deprivation merely of cultural amenities is not cruel and unusual punishment. Harris v. Fleming, supra, at 1235-37. Moreover, the evidence on this aspect of the case was heavily contaminated by evidence about conditions in F-House; for of course the filthier the conditions under which inmates are living, the greater their need for frequent showers.

There would be a special irony in requiring the State of Illinois to provide the security escorts and other services and facilities necessary to enable the most incorrigible of its prisoners to take three showers a week, when many law-abiding poor people in dilapidated public housing projects do not have working showers. The necessity to make painful tradeoffs between competing public projects in an era of governmental fiscal scarcity counsels caution in the formulation of equitable relief requiring public expenditures.

Preston v. Thompson, 589 F.2d 300, 303 (7th Cir.1978), upheld a district court's injunction that required a minimum entitlement of two showers a week, but the circumstances were different. The prison officials had imposed a prison-wide deadlock or lockdown, with drastic consequences for all prisoners. The district court thought the prison was too slow in returning to normalcy, and the two-shower provision was intended to nudge the defendants in that direction. The plaintiffs in our case rely heavily on a district court decision, Lightfoot v. Walker, 486 F. Supp. 504, 511 (S.D.Ill.1980), which finds that allowing inmates only one shower per week "promotes deterioration of inmates physically." But the district court's opinion in that case does not explain the basis of its finding or indicate whether the deterioration resulted solely from the infrequency of the showers or from that in combination with a failure to allow the inmates more than an hour of exercise per week. On the record in the present case one shower a week for inmates of Stateville's segregation unit is constitutionally sufficient and the district judge's contrary finding is clearly erroneous. The injunction is modified to delete the reference to showers and as modified is affirmed. No costs on appeal.

MODIFIED AND AFFIRMED.

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

After marching up the hill to uphold very ably and convincingly the five hour exercise prescription, the majority marches down the other side to somehow find three showers clearly erroneous. This is puzzling, because in the minds of most Americans there is some considerable linkage between occasions of vigorous exercise and the need for showers. In addition, I doubt that American tourists, whose observations the majority cites, would recommend importing at random foreign standards of balneation, even into maximum security prisons. As a matter of fact, the majority may be maligning foreigners since the Japanese, for example, have traditionally surpassed Westerners in their zeal for frequent bathing.

Whatever may be the dubious logic of the majority's stand on showers, a one-shower-a-week regimen takes us back in memory to the days of the Saturday night bath in a washtub with water heated on a wood stove. But that is history. Indoor plumbing has now been with us for some years--even in prisons. I have the impression the jury in bringing in its verdict understood some of these things better than the majority.

No less an authority than Dr. Shansky, the medical director of the Illinois Department of Corrections, testified that three showers are the weekly minimum (in connection with adequate exercise) necessary to prevent serious adverse effects on the physical and mental health of inmates in what amounts to solitary confinement. On that basis alone, I have great difficulty in seeing how Judge Hart's conclusion can be called clearly erroneous. I therefore respectfully dissent on the shower reversal.

On another point, the majority says that, "the infliction of disutility ... is one of the objectives of criminal punishment...." Supra p. 1313 (Emphasis supplied). This sounds a bit ominous to me--assuming that I understand it. On that shaky premise, I would also most respectfully disagree on this point.