Randy L. Brackett v. Howard Peters and Roland W. Burris
11 F.3d 78
7th Cir.
1993
Check Treatment
POSNER, Chief Judge.

Randy Brackett, an Illinois state prisoner who had been convicted after a bench trial of felony murder and given a long prison term (see People v. Brackett, 117 Ill.2d 170, 109 Ill.Dec. 809, 510 N.E.2d 877 (1987)), applied for federal habeas corpus, arguing that no rational finder of fact could have found that he had cаused the death of his victim. The district judge, after reviewing the record of Brackett’s trial, denied his application, precipitating this appeal, which raises interesting questions concerning the requirement of causation in criminal law.

Brackett, age 21 at the timе, had raped and severely beaten an 85-year-old widow, Mrs. Winslow, for whom he had previously done yard work. She was admitted to the hospital with a broken arm, a broken rib, and extensive bruises. During her stay in the hospital, which lasted several weeks, she— described as “feisty” before the rape and beating — became depressed, resisted efforts to feed her, and became progressively weaker. Transferred to a nursing home, she continued to deteriorate, even though her physical injuries were healing. Her appetite was very poor. Her doctor ordered a nasal gastric feeding tube for her but the tube could not be inserted, in part because facial injuries inflicted by Brackett made insertion of the tube too painful. About ten days after her admission to the nursing home, she died while a nurse was feeding her some pureed food through a feeder syringe. An autopsy revealed that a large quantity of food, some six or seven ounces, had become lodged in Mrs. Winslow’s trachea, asphyxiating her.

The question is whether Brackett’s assault on Mrs. Winslow could be found to be a cause of her death. If so, Brackett is guilty of felony murder; if not, not. So far as bears on this case, an act is a cause of an event if two conditions are satisfied: the event would not have occurred without the act; the act made the event more likely. The first condition is necessary to distinguish the attempted from the completed crime, the second to rule out cases in which, while the event in question would not have occurred but for the act, the act did not create the kind of dangerous condition that would make such events more likely to occur. Suppose, for example, that Mrs. Winslow had been killed by a fire at the nursing home. She would not have been in the nursing home (in all likelihood), so would not have been killed, but for Brackett’s assault. But as there would have been no greater danger of fire in a nursing home than in her own home, in our hypothetical case the assault would not have placed her in a situation of danger and therefore would not be considered a cause of her death. United States v. 1990 Toyota 4Runner, 9 F.3d 651, 652-53 (7th Cir.1993); DeShaney v. Winnebago County Dept. of Social Services, 812 F.2d 298, 302 (7th Cir.1987), aff'd, 489 U.S. 189, 109 S.Ct. 998, 103 L.Ed.2d 249 (1989).

Even with this qualification, which excludes from the concept of legally relevant causation certain purely adventitious “causes,” every event has multiple causes. Mrs. Winslow’s age was undoubtedly a cause of her death; a younger woman would have been much less likely to experience so rapid and complete a deterioration as a result of the assault. The autopsy revealed some *80 signs of senility, and senility is a common cause of depression, loss of appetite, and general weakening — all additional causes of Mrs. Winslow’s death, in the dual sense, which we have explained is the relevant sense, that had any of these conditions been absent she probably would not have died from the rape and beating and that each of the conditions made her death from thе assault more likely. None of them was related to the death merely fortuitously, as in our example of the nursing-home fire, which would be deemed “a supervening act disconnected from any act of the defendant,” People v. Meyers, 392 Ill. 355, 64 N.E.2d 531, 533 (1945); People v. Dordies, 60 Ill.App.3d 621, 18 Ill.Dec. 92, 96-97, 377 N.E.2d 245, 249-50 (1978), so that the defendant ‍​‌‌​​​​​​​​​​‌‌​‌​​‌​‌​‌​​​‌‌‌​‌​​‌‌​​‌​‌‌​‌​‌‌​‍would not be liable for thе death.

The immediate cause of Mrs. Win-slow’s death was the action of the nurse in depositing food into Mrs. Winslow’s trachea. Brackett’s lawyer argues that the nurse was grossly negligent, but this is far from plain— Mrs. Winslow appears to have exhibited no signs of distress until she keeled over dead— and even if it is true all that it would mean is that the nurse’s negligence was still another cause of Mrs. Winslow’s death. Hall v. State, 199 Ind. 592, 159 N.E. 420, 425-26 (1928); H.L.A. Hart & Tony Honoré, Causation in the Law 353-57 (2d ed. 1985). An event is, as we have emphasized, typically the consequence of multiple causes. But a murderer does not avoid conviction by pointing out thаt his act was only one of many causes that concurred to bring about his victim’s death. It is enough if his act was one of the causes — enough therefore if Brackett’s assault made Mrs. Winslow’s death more likely and if, but for the assault, she would not have died as soon as she did. People v. Reader, 26 Ill.2d 210, 186 N.E.2d 298, 300 (1962); People v. Fuller, 141 Ill.App.3d 737, 95 Ill.Dec. 885, 894, 490 N.E.2d 977, 986 (1986).

A rational finder of fact could find these conditions satisfied. Cf. Tucker v. Commonwealth, 303 Ky. 864, 199 S.W.2d 631 (1947). The proposition that raping and beating an 85-year-old woman creates a risk of death requires no discussion, but we must also consider the first part of the dual test of cause, and thus ask whether she would have died anyway when she did. That is unlikely. Death was the last link in a continuous series of events that began with the assault. She died a month later, never having returned home. Her condition deteriorated from the start of her hospitalization, and when she was transferred to the nursing home her dоctor already believed her to be near death. Of course she was very old and the autopsy revealed a condition of senile atrophy that must have existed before the assault. But judging from the description of her as “feisty” her senility could not have been so far advanced that death was imminent. It is more than unlikely that had she not been assaulted on October 21, 1981, she still would have entered the hospital the next day and died a month later. The assault appears to have precipitated her rapid declinе. Of course there are dangers in inferring consequence from sequence. But they are slight when as in this case the event not only follows the act closely in time but is the kind of event frequently produced by the kind of act, and no persuasive evidence of an alternative causal sequence is presented, as had been done in People v. Benson, 19 Ill.2d 50, 166 N.E.2d 80 (1960), an otherwise similar case. Illinois, incidentally, has abolished the archaic (see Comment, “Taming a Phoenix: The Year-and-a-Day Rule in Federal Prosecutions for Murder,” 59 U.Chi.L.Rev. 1337 (1992)) “year and a day” rule, People v. Carter, 168 Ill.App.3d 237, 118 Ill.Dec. 983, 989, 522 N.E.2d 653, 659 (1988), a rule designed to cut оff criminal liability when the inference of causation is attenuated by time.

Brackett’s lawyer fastens on the statement in the state trial judge’s otherwise uninformative opinion that the judge was rejecting the state’s theory of “psychological murder.” The theory had been advanced in a press conference called by the prosecutor. We are unable to determine exactly what the theory was or even whether it was pursued at the trial. Brackett’s lawyer describes it as foUows: the assault caused Mrs. Winslow to become clinically depressed, clinically depressed people (we know) are prone to suicide, Mrs. Winslow committed suicide by refusing to eat, and suicide is one of those *81 “supervening acts disconnected from any act of the defendant” which cut off liability for causing death. If this is “psychological murder,” we are puzzled by the trial judge’s rejection of it. We think — to take a hypothetical case somewhat clearer than this case — that if a person, desiring the death of another whom he knew to suffer from deрression, ‍​‌‌​​​​​​​​​​‌‌​‌​​‌​‌​‌​​​‌‌‌​‌​​‌‌​​‌​‌‌​‌​‌‌​‍stole his intended victim’s anti-depression medicine hoping to precipitate the victim’s suicide, and his hope was fulfilled, this would be murder. The victim’s depressive state would no more be a “supervening act” than any other vulnerability of the victim; in criminal law as in tort lаw, the injurer takes his victim as he finds him. Cunningham v. People, 195 Ill. 550, 63 N.E. 517, 525 (1902). It would be only a shade removed from a case of death by fright, Rollin M. Perkins & Ronald N. Boyce, Criminal Law 822-23 (3d ed. 1982), but well removed from a case in which the defendant, not desiring the victim’s death, made it more likely, for example by teaching him to play Russian roulette, as in Lewis v. State, 474 So.2d 766 (Ala.Cr.App.1985), as distinct from playing it with him. Commonwealth v. Atencio, 345 Mass. 627, 189 N.E.2d 223 (1963); cf. Jacobs v. State, 184 So.2d 711 (Fla.Apр.1966). Well removed too from a case of assisted suicide, where (as possibly in the Lewis case also) the victim desires his death. In our hypothetical case the victim presumably was taking anti-depression medicine because he did not want to commit suicide, a recognized risk of depression. The fact that a psychiatric condition, whether or not by precipitating suicide, is one of the causes of a victim’s death does not excuse his murderer. State v. Govan, 154 Ariz. 611, 616, 744 P.2d 712, 717 (App.1987); cf. Commonwealth v. Feinberg, 433 Pa. 558, 253 A.2d 636, 642 (1969). Otherwise it would be open season on sufferers from mental illness.

Courts worry, naturаlly, about problems of evidence and inference in Gaslight-type settings where cruel, deceptive, or even simply inconsiderate behavior might be claimed with more or less plausibility to have driven a susceptible person to suicide, or, as here, to loss of the “will to live.” E.g., Tucker v. Commonwealth, supra, 199 S.W.2d at 634; Annot., “Homicide by Fright or Shock,” 47 A.L.R.2d 1072, 1074-76 (1956). But we do not think ‍​‌‌​​​​​​​​​​‌‌​‌​​‌​‌​‌​​​‌‌‌​‌​​‌‌​​‌​‌‌​‌​‌‌​‍that there is or should be a categorical bar to the imposition of criminal liability in such eases. Note, “Criminal Liability for Assisting Suicide,” 86 Colum.L.Rev. 348, 364-66 (1986). It has long been the rule in tort law (the “thin-skull” or “eggshell-skull” rule) not only that the tortfeasor takes his victim as he finds him, but also that psychological vulnerability is on the same footing ■with physical. If for example the victim is predisposed to schizophrenia, and the tort-feasor inflicts a minor injury which precipitates the schizophrenia, he is liable for the entire consequences even though they were both highly unlikеly and unforeseen. Lancaster v. Norfolk & Western Ry., 773 F.2d 807, 822 (7th Cir.1985); Steinhauser v. Hertz Corp., 421 F.2d 1169, 1172-73 (2d Cir.1970). In superficial tension with the eggshell-skull cases is the famous (or infamous) “taunting” case, Yania v. Bigan, 397 Pa. 316, 155 A.2d 343 (1959), where one business rival taunted another into jumping into a deep trench filled with water, where he drowned, the taunter refusing to assist him — yet was held not liable for his death. Thе court emphasized, however, that there was no contention that the victim had not been in full possession of his mental faculties at the time of the incident. The analogy in criminal law would be to the pure assisted suicide, where the person who commits suicide is in full possession of his faculties. An intermediate case is State v. Lassiter, 197 N.J.Super. 2, 484 A.2d 13 (1984), where the victim had been severely beaten by the defendant and when she threatened to jump out the window he said “go ahead and jump,” she did, and his conviction of murder was affirmed. Other cases (reviewed in Note, supra, 86 Colum.L.Rev. at 364-65) hold that the defendant cannot be guilty of murder in such circumstances, but they are not cases in which, as in our hypothetical case, the defendant deliberately induces a psychotic reaction in his victim. The fact that a normal person would not havе had such a reaction does not excuse the defendant.

The eggshell-skull principle does not quite fit a case of intentional murder, for the murderer must intend his victim’s death *82 and ordinarily this will presuppose some awareness of the likely consequences of his act. It is not murder to kill a person by a slight blow harmless to an ordinary person ‍​‌‌​​​​​​​​​​‌‌​‌​​‌​‌​‌​​​‌‌‌​‌​​‌‌​​‌​‌‌​‌​‌‌​‍if you do not know the person is unusually vulnerable; there is even a presumption in Illinois that one who beats another with his bare fists does not intend to kill him. People v. Brackett, supra, 109 Ill.Dec. at 814, 510 N.E.2d at 882. But felony murder is different, and it is time to remind the reader thаt this a felony-murder case. No intent to kill, as in our hypothetical “psychological murder” ease, is required; and though the Illinois cases do require that death be a “foreseeable” consequence of the felony, e.g., People v. Hickman, 59 Ill.2d 89, 319 N.E.2d 511, 513 (1974); People v. Tillman, 70 Ill.App.3d 922, 27 Ill.Dec. 204, 207, 388 N.E.2d 1253, 1256 (1979), all they mean is that the death must be сaused by the felony; for remember that “cause” in law means not just but-for cause but also an enhancement of the likelihood (what in law is often called “foreseeability”) that the class of events would occur. O.W. Holmes, Jr., The Common Law 58 (1881); Perkins & Boyce, supra, at 817. (Some courts do not require proof of such enhancement in a felony-murder case. Note, “Felony Murder: A Tort Law Reconceptualization,” 99 Harv.L.Rev. 1918, 1926-28 (1986). But the Illinois courts do.) In Hickman the defendant’s burglary increased the probability that one police officer would shoot another by accident. Here the only issue is whether а rational finder of fact could conclude that the felonies which Brackett committed caused the death of his victim; the answer, we have seen, is yes.

We have emphasized analogies from tort law because the doctrine of causation is more developed there than in criminal law. See, e.g., Note, supra, 99 Harv.L.Rev. at 1926-28. The reason is that tort law, which has compensatory as well as deterrent functions, focuses on injury, while criminal law, which emphasizes deterrence and incapacitation, focuses on thе dangerousness of the defendant’s conduct. There is no tort liability without proof of injury, but there are plenty of crimes that are punishable though no injury resulted — many attempts and conspiracies, for example. A victim’s eggshell skull may require a refined adjustment in damagеs to reflect the likelihood that the victim would because of his vulnerability have been injured sooner or later nontor-tiously. Lancaster v. Norfolk & Western Ry., supra, 773 F.2d at 822; Steinhauser v. Hertz Corp., supra, 421 F.2d at 1173-74. But a criminal assailant is punishable as a first-degree murderer “no matter how feeble the spark” of life that his blow extinguished. Cunningham v. People, supra, 63 N.E. at 525. Uncompleted crimes are often punished severely; and when injury or death ensues from deliberate wrongdoing, even if (as in the case of felony murder) it is ‍​‌‌​​​​​​​​​​‌‌​‌​​‌​‌​‌​​​‌‌‌​‌​​‌‌​​‌​‌‌​‌​‌‌​‍not an intended consequence, the criminal law comes down heavily on the defendant without worrying overmuch about the precise amount of harm inflicted.

AFFIRMED.