If you have issues viewing or accessing this file contact us at NCJRS.gov.

CHAPTER 6

The Psychologica~ Autopsy

Edwin S. Shneidman, Ph.D.

It is probably best to begin by defining a does not automatically communicate whether psychological autopsy and its purposes, then the decedent struggled and drowned (acci­ to discuss some related theoretical background dent), entered the pool with the intention of and ways of actually performing psychological drowning himself (suicide), or was held under autopsies. The words psychological autopsy water until he was drowned (homicide). themselves tell us that the procedure has to It so happens that a considerable number do with clarifying the nature of a death and of deaths-the estimate is between 5 and 20 that it focuses on the psychological aspects of percent of all deaths which need to be certi­ the death. Two ideas, important to under­ fied-are not clear as to the correct or appro­ standing the psychological autopsy, need to priate mode. These unclear or uncertain be discussed. The first is what I have called the deaths are called equivocal deaths. The NASH classification of deaths; the second is ambiguity is usually between the modes of the idea of equivocal deaths. suicide or accident, although uncertainty can From the beginning of this century (and exist between any two or more of the four with roots that can be clearly traced back to modes. Elizabethan times), the certification and The main function of the psychological recordkeeping relating to deaths have implied autopsy is to clarify an equivocal death and that there are four modes of death. It needs to arrive at the "conect" or accurate mode of to be said right away that the four modes of that death. In essence, the psychological death have to be distinguished from the many autopsy is nothing less than a thorough retro­ causes of death listed in the current Inter­ spective investigation of the intention of the national Classification of Diseases and Causes , decedent-that is, the decedent's intention of Death (World Health Organization 1957; relating to his being dead-where the informa­ National Center for Health Statistics 1967). tion is obtained by interviewing individuals The four modes of death are natural, accident, who knew the decedent's actions, behavior, suicide, and homicide; the initial letters of and character well enough to report on them. each make up the acronym NASH. Thus, to Drug-related deaths can be anl0ng the most speak of the NASH classification of death is equivocal as to the mode of death. Propel' to refer to these four traditional modes in certification often necessitates knowledge of which death is cunently reported. Con­ the victim over and beyond standard toxic­ temporary death certificates have a category ological information, including such questions which reads "Accident, suicide, homicide, or as what dosage was taken (related to the exact undetermined"; if none of these is checked, time of death and the time at which autopsy then a "natural" mode of death, as occurs in blood and tissue samples were taken); the most cases, is hnplied.1 decedent's weight and build; the decedent's It should be apparent that the cause of long-term drug habits and known tolerances; death stated on the certificate does not neces­ the possible synergistics of other ingested sarily carry with it information as to the materials, notably alcohol or the effects of specific mode of death. For example, asphyxi­ certain combinations like hydromorphone ation due to drowning in a swimming pool (Dilaudid) and Methedrine; and the role of lethal action of drug overdoses, e.g., morphine lSee U.S. Standard Certificate of Death, appen­ sulfate (morphine) as opposed to the quicker dix F. acting diacetyl morphine (heroin). 42 THE PSYCHOLOGICAL AUTOPSY 43

HISTORICAL BACKGROUND a variety of publications. the bills became the earliest systematized American death certifi­ In 1662 John Graunt, a London tradesman, cates. pUblished a small book of "observations" on Meanwhile, back in England, "the London the bills of mortality thet was to have great Bills of Mortality remained among the eternal social and medical significance. By this time verities for Englishmen" (Cassedy 1969). the weekly bills were consolidated at the end Londoners were sure to find the bills on sale of each year, and a general bill for the year in each parish each week. Bills with lists of was published. Graunt separated the various cases of various diseases printed on the back bits of information contained in these annual brought twice the price of the regular bills. bills and organized them into tables. When the The American colonists had relatives or available data on deaths were believed accu­ fliends send copies to them: "Just as John rate, Graunt then focused on individual cuuses Graunt had found the English doing, the early of death. He next turned to the subject of Americans used the bills as grist for conversa­ popUlation estimation. Finally he constructed tion if for nothing else" (Cassedy 1969). Few; a mortality table, the first attempt to organize in the colonies even knew of Graunt's statis­ data in this manner. Of greatest significance tical applications to the bills of mortality for was his success in demonstrating the regulari­ years after his book was published. Not until ties that can be found in medical and social William Douglass commented upon statistical phenomena when one is dealing with large method in his history of the British colonies numbers. Thus John Graunt demonstrated in North America, published in 1'{51, did the how the bills of mortality could be used to colonists begin to make use ,)f Graunt's the advantage of both the physician and methods. government (Kargon 1963). The fact that no colonial bills were pub­ In 1741, the science of statistics came into lished during the 17th century may be attrib­ existence with the work of a Prussian clergy­ uted to printing priorities and especially to man, Johann SussmiIch, who made a systema­ lack of legislative requirements. Though a tic attempt to correlate "political arithmetic," printing press had begun operation in Cam­ or what we now call vital statistics. From this bridge, Massachusetts, in 1638, the next half study came what was subsequently termed century saw the introduction of only four the "laws of large numbers," which permitted more. These few presses were generally kept extended use of the bills of mortality to running at capacity turning out government supply important data in Europe as well as in documents. Though the government of each the American colonies. Cassedy (1969) says colony required the registration of all vital that Sussmilch's "exhaustive analysis of vital statistics, none of them made any provision data from church registers . . . became the for the information to be published. Thus the ultimate scientific demonstration of the publication of anything resembling the regularity of God's demographic laws." London bills was left to private enterprise. For many years no provisions existed any­ Despite the spread of colonial newspapers and where in the American colonies for anything the existence of a few church bills of mortal­ comparable to the London bills of mortality ity, British pUblications remained the colo­ (Cassedy 1969). At first there was no neces­ nists' best sources of vital statistical informa­ sity for detailed reeords since communities tion (Cassedy 1969). were small, but something better than hearsay The use of such nonstand':;trdized death or fading memories was needed as towns grew records continued into the 19th century. Im·ger. The birth of newspapers in the British Recognition of the need for informed medico­ sections of America, about the year 1700, legal investigation in England led to a series of provided a means of remedying this situation. refonns aimed at improving the quality of Some editors went to the trouble of obtaining death registration. In 1836 Parliament enacted infortI1ation from local church and town a bill requiring the recording of all deaths. records. They accumulated long lists to which Under the terms of this act, notification of summruies were added, which together were the coroner was not required unless the cause loosely called "bills of mortality." Issued in of death was one included in a special cate- 44 SHNEIDMAN gory. This system of death certification and be prescribed, and that in giving a certifi­ .! ;:egistration was to ensure that those special cate a medical practitioner should be re­ deaths that came under the coroner's jurisdic­ quired to use such a form. tion were in fact reported. The new certifica­ tion system was designed to utilize the data The United States Congress, trailing behind on causes of death for statistical purposes as the British Parliament, enacted no standard well as to prevent criminal practices. The registration act until 1903. Prior to that date, medical explanation of the death was the any attempts at standardization were left to essential information required for statistical the individual States. determination. A curious aspect of the law, In Massachusetts deaths were reported on however, was that doctors were specifically sheets of paper measuring 18 by 24 inches, enjoined to include no information on the with approximately 40 records to a page. mode of death on the death certificate. Information asked for (but not always pro­ The English 1836 registration act was vided) included date of d~ath, date of record, amended in 1874 to require that personal name of deceased, sex, marital status, age, information on the death be submitted to the residence, disease or other cause of death, registrar of the district within 5 days of its place of death, occupation, place of birth, and occurrence by the nearest relative of the names and birthplaces of parents. These deceased who was present at the death or in questions are generally to be found on all attendance during the last illness. A fine was records of that period. The California form to be levied for noncompliance. The new law added only a line for the signature of the also required a registered medical practitioner attending physician or coroner. New Jersey present during the last illness to complete a requested the same information but asl{ed for certificate stating the cause of death to the a more detailed exphmation of the cause of best of his knowledge and belief. death. This was optional, however, and the In 1893 a "Select Committee on Death space provided for it was usually left blank. Certification" of the House of Commons The New York form, the most comprehensive attempted to correct the shortcomings of the of its time, also required data concerning the previous legislation, particularly "the careless­ burial. ness and ignorance of the persons certifying, These early death registrations lacked many the absence of medical attendants during the significant data: cause of death and place of last illness and the indefinite character of the death were often omitted, and no questions disease itself" (Abbott 1901). That committee were asked regarding an autopsy report or the made a series of 10 recommendations, the time and manner of the death, relevant infor­ most important of which were the following: mation needed for statistical purposes and criminal investigation. When the United States Census Bureau was 1. That. in no case should a death be registered making its preparations for the 1880 census, without production of a certificate of the it decided to rely upon registration records cause of dea eh signed by a registered medi­ instead of mortality enumerations, wherever cal practitioner or by a coroner after possible, and made a study of State and local inquest. forms to determine where these registration 2. That in each sanitary district a registered records were adequate to its purpose. Wide medical practitioner should be appointed as variation was noted in the ways in which public medical certifier of the cause of items were worded and data recorded. The death in cases in which a certificate from a study revealed the inherent disadvantages of medical practitioner in attendance is not allowing each State to enact its own registra­ forthcoming. tion system without guidelines provided by 3. That a medical practitioner in attendance some central authority (Colby 1965). There­ should be required personally to inspect fore, with the aid of the American Public the body before giving a certificate of Health Association, the Census Bureau devel­ death. oped what may be termed a model death 4. That a form of a certificate of death should certificate and prescribed its use by the States, THE PSYCHOLOGICAL AUTOPSY 45 but the use of a standard certificate for the About 3.3 percent of the total suicide registration of deaths was not approved by rate for 1958 as compared with that for 1957 resulted from the transfer of a num­ Congress until 1903. ber of deaths from accident to suicide. In 1958 a change was made in the interpreta­ tion of injuries where there was some doubt as'to whether they were accidentally THE DEATH CERTIFICATE inflicted or inflicted with <;uicidal intent. Beginning with the Seventh Revision for '1'he impact of the death certificate is con­ data year 1958, "self-inflicted" injuries with no specification as to whether or not siderable. It holds a mirror to our mores; it they were inflicted with suicidal intent and reflects some of the deepest taboos; it can deaths from injuries, whether or not self­ directly affect the fate and fortune of a inflicted, with an indication that it is not fan1ily, touching both its affluence and its known whether they were inflicted acci­ mental health; it can enhance or degrade the dentally or with suicidal intent, are :::lassi­ fied as suicides. The change was made on reputation of the decedent and set its stamp the assumption that the majority of such on his postself career. But if the impact of the deaths are properly classified as suicide death certificate is great, its limitations ar(' of because of the reluctance of the certifier equal magnitude. In its present form the to designate a death as suicide unless evi­ death certificate is a badly flawed document. dence indicates suicidal intent beyond the shadow of a doubt. The magnitude of the Today most States follow the format of the comparability ratios for suicide varied U.S. Standard Certificate of Death. Most rele­ considerably with means of injury, from vant to our present interests is the item which 1.02 for suicide by firearms and explosives reads: "Accident, suicide, homicide, or un­ to 1.55 for suicide by jumping from high determined (specify)." When none of these is places. specified, a natural mode of death is, of course, implied. Only two States, Delaware It would seem that this redefinition led to and Virginia, have made all four modes of an apparent 55 percent increase from one death explicit on the death certificate (and year to the next in suicides by jumping from have included a "pending" category as well). high places. Even more interesting is the offi­ Curiously enough, Indiana included four cial observation that the death certifier would modes of death on the death certificate form be reluctant to indicate suicide "unless evi­ from 1955 to 1968, but then revised the form dence indicates suicidal intent beyond the in 1968 and now provides no item for mode shadow of a doubt." Clearly the certifier of death; nor, surprisingly, does the current plays an important role in the process of Massachusetts death certificate contain an generating mortality data. It is he who makes accident-suicide-homicide item. the subjective judgment of what constitutes In addition to the U.S. Standard Certificate, conclusive evidence of the decedent's intent. the International Classification of Diseases The Eighth Revision (National Center for and Causes of Death plays a major role in Health Statistics 1967), which made the cate­ determining the way a specific death may be gory "Und~termined" available, introduced counted-and thus in the apparent change in still further problems, apparently shifting causes of death statistics from decade to many suicidal deaths to the Undetermined decade. For example, the definitions of sui­ category. cides and accidents were changed in the What is urgently needed is an exploration Seventh Revision (1955) and Eighth Revision and description of the current practices of (1965) of the International Classification, and certifying deaths, especi.ally deaths by suicide. the numbers of suicides and accidents changed We need a uniform system that would elimi­ along with the definitions. When the Seventh nate such inconsistencies (or confirm the Revision was put into effect for the data year differential unequivocally), as for example, 1958, the death rate for suicides increBsed 10.9 deaths by suicide per 100,000 popula­ markedly over 1957. In part, one can find the tion for Idaho versus 20.2 for Wyoming. What explanation in this paragraph (National Center is required is a "correctional quotient" for for Health Statistics 1965a): each reporting unit-county, State, and 46 SHNEIDMAN Nation. Until such information is obtained, In the Western world death is given its available suicide statistics are highly suspect. administrative dimensions by the death cer­ At the turn of the century, an early refer­ tificate. It is the format and content of this ence book of the medical science (Abbott document that determine and reflect the cate­ 1901) urged reliable death registration book­ gories in terms of which death is concep­ keeping and listed the following as "objects tualized and death statistics reported. The secured by a well-devised system": ways in which deaths were described and categorized in John Graunt's day and earlier 1. Questions relating to property rights are set deep precedents for ways of thinking often settled by a single reference to a about death, and they govern our thoughts record of death. and gut reactions to death to this day. Deaths 2. The official certificate of a death is usually were then assumed to fall into one of two required in each case of claim for life insur­ categories: There were those that were truly ance. adventitious-accidents, visitations of fate or 3. Death certificates settle many disputed fortune (called natural and accidental)-and questions in regard to pensions. there were those that were caused by a cul­ 4. They are of great value in searching for prit who needed to be sought out and punished records of genealogy. (called suicidal and homicidal deaths). In the 5. A death certificate frequently furnishes case of suicide, the victim and the assailant valuable aid in the detection of crime. were combined in the sanle person, and the 6. Each individual certificate is a contribution offense was designated as a crime against one­ causa scientiae. Taken collectively, they are self, a felo-de-se. England did not cease to of great importance to physicians, and classify suicide as a crime until 1961, and in especially to health officers, in the study the United States it remains a crime in nine of disease, since they furnish valuable infor­ States. mation in regard to its causes, its prev­ The historical importance of certifying the alence, and its geographical distribution. mode of death-i.e., of the coroner's function- At least three more functions for the can now be seen: It not only set a stamp of death certificate might be added to the innocence or stigma upon the death, but also half-dozen listed in 1901: determined whether the decedent's estate could be claimed by his legal heirs (natural or 7. The death certificate should reflect the accidental death) or by the crown or local type of death that is certified-brain death lord (a suicide or a murder). That was cer­ (a flat electroencephalographic record), tainly one important practical effect of the somatic death (no respiration, heartbeat, death certificate. The NASH categories of reflexes), or whatever type is implied. death were implied as early as the 16th cen- 8. The death certificate should include space tury in English certification, and this sub- for the specification of death by legal exe­ manifest administrative taxonomy of death cution, death in war or military incursions, has beguiled most men into thinking that death by police action, and others of that there really are four kinds of death, which, sort. of course, is not necessarily so at all. 9. Perhaps most importantly, the death cer­ Although it may be platitudinous to say tificate should abandon the anachronistic that in each life the inevitability of death is an Cartesian view of man as a passive biological inexorable fact, there is nothing at all inexora­ vessel on which the fates work their will. ble about our ways of dimensionalizing death. Instead, it should reflect the ccntemporal'Y Conceptualizations of death are man-made view of man as a psycho-socia-biological and mutable; what man can make he can also I organism that can, and in many cases does, clarify and change. Indeed, changes in our I playa significant role in hastening its own conceptualizations of death are constantly iI demise. This means that the death certifi­ occurring, notwithstanding the NASH notions I cate should contain at least one item on the of death have held on for centuries after they ·1

decedent's intention vis-a-vis his own death. became anachronistic. Each generation be- I . _J THE PSYCHOLOGICAL AUTOPSY 47 comes accustomed to its own notions and Prevention Center, to assist him in a joint thinks that these are universal and ubiquitous. study of these equivocal cases, and it was this From the time of John Graunt and his effort-a multidisciplinary approach involving mortuary tables in the 17th century, through behavioral scientists-which led to my coin­ the work of Cullen in the 18th century and ing the term "psychological autopsy.,,2 Willian1 Farr in the 19th century, the adoption (Curphey 1961, 1967; Litman et al. 1963; of the Bertillon International List of Causes Shneidman and Farberow 1961; Shneidman of Death in 1893, and the International Con­ 1969, 1973). ference for the Eighth Revision of the Inter­ In the last few years, especially with the national Classification of Diseases as recently interesting and valuable work of Litman et al. as 1965 (National Center for Health Statistics (1963), Weisman and Kastenbaum (1968), 1965b), the classification of causes of death and Weisman (1974), the term "psychological has constantly been broadening in scope. The autopsy" has come to have other, slightly changes are characterized primarily by at­ different meanings. At present there are at tempts to reflect additions to knowledge, least three distinct questions that the psycho­ particularly those contributed by new pro­ logical autopsy can help to answer: fessions as they have developed-anesthesi­ 1. Why did the individual do it? When the ology, pathology, bacteriology. immunology, mode of death is, by all reasonable measures, and advances in obstetrics, surgery, and most clear and unequivocal-suicide, for example­ recently, the behavioral sciences. the psychological autopsy can serve to account for the reasons for the act or to dis­ cover what led to it. Why did Ernest Heming­ PURPOSES OF THE PSYCHOLOGICAL way "have to" shoot himself (Hotchner AUTOPSY 1966)? Why did former Secretary of Defense James Forrestal kill himself (Rogow 1963)? As long as deaths are classified solely in We can read a widow's explicit account of tenns of the four NASH categories, it is how she helped her husband, dying of cancer, immediately apparent that some deaths will, cut open his veins in Lael Tucker Werten­ so to speak, fall between the cracks, and our baker's Death of a Man (1957). Some people familiar problem of equivocal death will con­ can understand such an act; others cannot. tinue to place obstacles in our path to under­ But even those who believe they understand standing human beings and their dying. Many cannot know whether their reasons are the of these obstacles can be cleared away by same as those of the cancer victim or his wife. reconstructing, primarily through interviews What were their reasons? In this type of with the survivors, the role that the deceased psychological autopsy, as in the following played in hastening or effecting his own type, the mode of death is clear, but the death. This procedure is called "psychological reasons for the manner of dying remain autopsy," and initially its main pUl·pone was puzzling, even mysterious. The psychological to clarify situations in which the mode of autopsy is no less than a reconstruction of the death was not immediately clear. motivations, philosophy, psychodynamics, The origin of the psychological autopsy and existential crises of the decedent. grew out of the frustration of the Los Angeles 2. How did the individual die, and whell­ County Chief Medical Examiner-Coroner, that is, why at that particular time? When a Theodore J. Curphey, M.D., at the time of the death, usually a natural death, is protracted, reorganization of that office in 1958. Despite his efforts, which were combined with those of toxicologists and nonmedical investigators, 20n the lighter side: From the viewpoint of his he was faced with a number of drug deaths colleages, Dr. Curphey became suspect as to his pro­ for which he was unable to certify the mode fessional ancestry by virtue of his extending the autopsy procedure to include behavioral scientists. He on the basis of collected evidence. As a result has told me that around that time he was asked he invited Norman Farberow, Ph.D., and me, whether he was a psychiatric pathologist or a patllo· then Co-Directors of the Los Angeles Suicide logical psychiatrist.

'±::::;::zaz:s 48 SHNEIDMAN the individual dying gradually over a period of WIDOW, 10,1, DIES IN COTTAGE time, the psychological autopsy helps to SHE ENTERED AS 1887 BRIDE illumine the sociopsychological reasons why Mrs. John Charles Dalrymple, 104 years he died at that time. This type of psycho­ old, died here [Randolph Township, N.J.] logical autopsy is illustrclted by the following yesterday in the cottage to which she came as a bride in 1887. blief case from Weisman and Kastenbaum Her husband brought her in a sleigh to (1968): the house, which she was to leave next week to make way for the new Morris An 85-year-old man had suffered with County Community College .... chronic bronchitis and emphysema for many years but was alert and active other­ wise. He had eagerly anticipated going to The main question here, as in Weisman and his son's home for Thanksgiving and when Kastenbaum's case of the old man who was the day arrived he was dressed and ready, left alone on Thanksgiving, is: Might even this but no one came for him. He became more person have lived at least a little longer had concerned as the hours went by. He asked the nurse about messages, but there were she not suffered the psychologically traumatic none, and he finally realized that he would threat of being dispossessed from the home have to spend the holiday at the hospital. where she had lived for 81 years? Or does the After this disappointment the patient kept question in this particular case tax one's com­ more and more to himself, offered little monsense credulity? and accepted only minimal care. Within a few weeks he was dead. 3. H/hat is tile most probable mode of death? This was the question to which the The implication here is that the patient's dis­ psychological autopsy was initially addressed. appointment ahd his resignation to it were When cause of death can be clearly estab­ not unrelated to his sudden downhill course lished but mode of death is equivocal, the and his death soon afterward, i.e., if his son purpose of the psychological autopsy is to had come to take him out for Thanksgiving, establish the mode of death with as great a the old man would have lived considerably degree of accuracy as possible. Here are three longer than he did. This man's death like simplified examples: some others-voodoo deaths, unexplained Cause of death: asphyxiation due to deaths under anesthesia, and "self-fulfilling drowning. A woman found in her swimming prophecy" deaths, for example- must be con­ pool. Question as to correct mode: Did she sidered subintentioned. There can be little "drown" (accident), or was it intentional doubt that often some connec~ion exists be­ (suicide)? tween the psychology of the individual and Cause of death: multiple crushing injuries. the time of his death (Shneidman 1963). A man found dead at the foot of a tall build­ There is, of course, a wide spectrum of ing. Question as to correct mode: Did he fall applicability of this concept. When a person (accident), jump (suicide), or was he pushed has been literally scared to death by his belief or thrown (homicide)? in the power of voodoo, the role of the vic­ Cause of death: barbiturate intoxication tim's psychological state seems fairly obvious; due to overdose. A woman found in her bed. and it is difficult to believe that there was no Question as to correct mode: Would she be psychological connection between the fatal surprised to know that she was dead (acci­ stroke of Mrs. Loree Bailey, owner of the dent), or is this what she had planned (sui­ Lorraine Motel in Memphis, and the assassina­ cide)? tion of Martin Luther King, Jr., at the motel The typical coroner's office, whether 3 hours earlier. But in many other cases any headed by a medical examiner or by a lay relationship between the individual's psycho­ coroner, is more likely to be accurate in its logical state and the time of his death seems certification of natural and accldental deaths difficult or impossible to establish. than of those deaths that might be suicides. As an example of the problems raised by Curphey says, "A major reason for this, of this concept, consider the following case, course, is that both the pathologist and the reported in the New York Times of June 26, lay investigator lack sufficient training in the 1968: field of human behavior to be able to esti- THE PSYCHOLOGICAL AUTOPSY 49 mate with any fair degree of accuracy the not be followed slavishly a..'1d that the investi­ mental processes of the victim likely to lead gator should be ever mindful that he may be to suicidal death. It is hem that the social asking questions that are very painful to scientists, with their special skills in human people in an obvious grief-laden situation. The behavior, can offer us much valuable assist­ person who conducts a psychological autopsy ance" (Curphey 1961). should participate, as far as he is genuinely The professional personnel who constitut~ able, in the anguish of the bereaved person a "death investigation team" obviously should and should always do his work with the hold no bnef for one particular mode of mental health of the survivors in mind. death over any other. In essence, the members Here, then, are some categories that might of the death investigation team interview be included in a psychological autopsy persons who knew the deceased-and attempt (Shneidman 1969): to reconstruct his lifestyle, focusing particu­ larly on the period just prior to his death. If 1. Information identifying victim (name, the infOl1nation they receive contains any age, address, marital status, religious prac­ clues pointing to suicide, their especially tices, occupation, and other details) attuned ears will recognize them. They listen 2. Details of the death (including the cause for any overt or covert communications that. or method and other pertinent details) might illumiaate the decedent's role (if any) 3. Brief outline of victim's history (siblings, in his own demise. They then make a rea­ marriage, medical illnesses, medical treat­ soned extrapolation of the victim's intention ment, psychotherapy, suicide attempts) and behavior over the days and minutes pre­ 4. Death history of victim's family (suicides, ceding his death, using all the information cancer, other htal illnesses, ages at death, they have obtained. and other details) 5. Description of the personality and life­ style of the victim CONDUCTING THE 6. Victim's typical patterns of reaction to PSYCHOLOGICAL AUTOPSY stress, emotional upsets, and periods of disequili brium How is a psychological autopsy performed? 7. Any recent-from last few days to last 12 It is done by talking to some key persons­ months-upsets, pressures, tensions, or spouse, lover, parent, grown child, friend, anticipations of trouble colleague, physician, supervisor, coworker­ 8. Role of alcohol or dnlgs in (a) overall life­ who knew the decedent. The "talking to" is style of victim, and (b) his death done gently, a mixture of conversation, inter­ 9. N~ture of victim's interpersonal relation­ view, emotional support, general que1:itions, ships (including those with physicians) and a good deal of listening. I always tele­ 10. Fantasies, dreams, thoughts, premonitions, phone and then go out to the home. After or fe~ of victim relating to death, acci­ rapport is established, a good general opening dent, or suicide question might be: "Please tell me, what was 11. Changes in the victim before death (of he (she) like?" Sometimes clothes and habits, hobbies, eating, sexual patterns, material possessions are looked at, photo­ and other life routines) graphs shown, and even diaries and corre­ spondence shared. (On one occasion, the 12. Information relating to the "life side" of victim (upswings, successes, plans) widow showed me her late husband's suicide 13. Assessment of intention, i.e., role of the note which she had hidden from the police!­ victim in his own demise rather changing the equivocal nature of the 14. Rating of lethality (described in the final death.) section of this chapter) In general, I do not have a fixed outline in 15. Reaction of informants to victim's death mind while conducting a psychological 16. Comments, special features, etc. autopsy, but, inasmuch as outlines have been requested from time to time, one is presented In conducting the interviews during a below with the dual cautions that it should psychological autopsy, it is often best to ask J

50 SHNElDMAN open-ended questions that permit the re­ That is the general way that one would I spondent to associate to relevant details with­ inquire, if it were relevant, into, say, drug out being made painfully aware of the specific patterns of behavior. Where suicide or homi­ interests of the questioner. As an example: I cide is a possible mode of death, it is rather might be very interested in knowing whether important to know whether or not the dece­ or not there was a change (specifically, a dent was "into" drugs, an habitual user, or a recent sharp decline) in the decedent's eating dealer, on what terms he was with his dealer, habits. Rather than ask directly, "Did his etc. appetite drop recently?" a question almost calculated to elicit a defensive response, I have asked a more general question such as, FUNCTIONS OF THE "Did he have any favorite foods?" Obviously, PSYCHOLOGICAL AUTOPSY my interest is not to learn what foods he pre­ ferred. Not atypically, the respondent will tell The questions should be as detailed (and me what the decedent's favorite foods were lines of inquiry pursued) only as they bear on and then go on to talk about recent changes clarifying the mode of death. All else would in his eating habits-"Nothing I fixed for him seem to be extraneous. And to do this de­ seemed to please him "-and even proceed to pends, of course, on having established rapport relate other recent changes, such as changing with the respondent. patterns in social or sexual or recreational The results of these interviewing procedures habits, changes which diagnostically would are then discussed with the chief medical ex­ seem to be related to a dysphoric person, not aminer or coronel'. Because it is his responsi­ inconsistent with a suicidal or subintentioned bility to indicate (or ameud) the mode of death. death, all available psychological information In relation to a barbiturate death3 where should be included in the total data at his dis­ the mode of death is equivocal (between sui­ posal. Since a sizable percentage of deaths are cide and accident), it might be callous to ask equivocal as to mode precisely because these the next of kin, "Did your husband (wife) psychological factors are unknown, medical have a history of taking barbiturates?" A examiners and coroners throughout the coun­ more respectful and productive question try are robbing themselves of important infor­ might be, "Did he (she) take occasional medi­ mation when they fail to employ the special cation to help him (her) sleep at night?" If skills of the behavioral scientists in cases of the response to this question is in the affirma­ equivocal deaths. The skills of behavioral tive, one might then ask if the respondent scientists should be employed in the same knows the name of the medication or even way as the skills of biochemists, toxicologists, the shape and color of the medication. If one histologists, microscopists, and other physical determines that the deceased in fact had a scientists. The time has long since passed history of taking sleeping medication, one when we could enjoy the luxury of disregard­ might then ask if the decedent was accus­ ing the basic teachings of 20th century tomed to having some occasional alcoholic psychodynamic psychology and psychiatry. beverages prior to going to sleep. If these facts Certification procedures (and the death cer­ can be brought into the open, it may well be tificates on which they are recorded) should that one can then establish the quantity of reflect the role of the decedent in his own the medication and alcohol content that the demise, and in equivocal cases this cannot be decedent was taking immediately prior to his done without a psychological autopsy, death. The general method of questioning' is The retrospective analysis of deaths not one of "successive approaches," wherein the only serves to increase the accuracy of certifi­ respondent's willingness to answer one ques­ cation (which is in the best interests of the tion gives a permission to ask the next one. overall mental health concerns of the com­ munity), but also has the heuristic function of 31 am grateful to Dr. Michael S. Backenheimer of providing the serious investigator with clues the National Institute of Drug Abuse for the sugges­ that he may then use to assess lethal intent in tions contained in this paragraph. living persons. THE PSYCHOLOGICAL AUTOPSY 51 And there is still another function that the these meetings "should be small enough to psychological autopsy serves: In working with exclude those interested primarily in the the bereaved survivors to elicit data relative to sensational aspects of the event in question, appropriate certification, a skillful and em­ and should h1c1ude only those whose presence pathic inVE'stigator is able to conduct the would serve one of the above purposes." interviews in such a way that they are of Following, reprinted with permission, are actual therapeutic value to the survivors. A verbatim repol'ts (except for a few minor psychological autopsy should never be con­ changes to disguise identity) of two psychia­ ducted so that any aspect of it is iatrogenic. tric inquests from that university hospital Commenting on this important mental health setting. function of the psychological autopsy, Cur­ Case 1. A 32-year-old male graduate stu­ phey (1961) has stated: dent took his life by drug overdose. He had first been hospitalized in 1974 for treat­ The members of the death investigation ment of strychnine poisoning and was dis­ team, because of their special skills, are charged eight days later with a diagnosis of alert in their interviews with survivors '.0 cyclothymic personality. He was readmitted evidences of extreme guilt, serious depres­ late the same year with depression and sion, and the need for special help in paranoid delusions, both of which cleared formulating plans for solving specific rapidly on Triavil (a combination of the problems such as caring for children major tranquilizer-antipsychotic agent per­ whose parents committed suicide. Since phenazine-and the tricyclic antidepressant we noted this phenomenon, the coroner's amitriptyline), was discharged after a two­ office has, in some few cases, referred dis­ week hospitalization and was followed in h'aught survivors of suicide victims to mem­ the outpatient clinic by a resident who had bers of the team specifically for supportive undertaken his care only one week before interviews even when the suicidal mode of the discharge date. Gradual improvement death was not in doubt. was repol'ted, although the absence of the supervisor (who was aware of the suicide This therapeutic work with the survivor­ risk and would have questioned termination victims of a dire event is called postvention of treatment) leaves the reported improve­ and has been presented in some detail else­ ment open to question. Two months later where (Shneidman 1967, 1971, 1973). the patient said that he had discontinued A large university hospital in the east his medications because they slowed him down and that he was confident about his (which has asked not to be identified, but to work, feeling well, and would not try sui­ which I am appropriately beholden) conducts cide again. He was discharged from out­ what they call "psychiatric inquests" on those patient treatment at his request, with the (rare) occasions when a patient commits sui­ assurance that he ('ould call back if he cide or makes a serious suicide attempt. A needed further help. He did not contact th(~ clinic again, but three days before his death staff psychiatrist, emphasizing the therapeutic called his faculty advisor, made accusations aspects of the psychological autopsy procedure, against a fellow graduate student and re­ stated that "the inquest is a kind of postven­ quested a departmental inquiry. He was tion, designed primarily for the benefit of a advised to go on a vacation but went instead shocked and grieving staff." He stated further to another city, where it was reported that that, for them, there are essentially three he spent two days praying in a chapel, re­ turned to his boarding house, ate supper main purposes of such a procedure: "(1) To and took a fatal drug overdose. Those com­ review with those responsible for the patient, menting on the case stressed the question­ the status of the pati.ent prior to the act, and able aspects of taking at face value the to determine what course of clinical manage­ statements of the patient which led to his ment would more likely have led to its antici­ discharge from the clinic. Furthermore, the patient sent out danger signals in his last pation and prevention; (2) To facilitate ex­ call to his faculty advisor. pression of feeling appropriate to the event on the part of staff members; and (3) To detel" Case 2. An 18-year-old male patient spent three months on the Child Psychiatry Serv­ mine whether dissemination of the results of ice two years prior to committing suicide. the inquest would serve an educational pur­ He had been evaluated for a mild aortic pose, and arrange for this (e.g., Grand Rounds) stenosis about which he and his family when approp'iate." Further, he states that were greatly concerned. During his first 52 SHNEIDMAN

admission, he showed fragmentation, loose reported as accidental death in anotl1('~ The associations, grimacing, and bizarre move­ factors that determine decisions of co;:~mers ments. Within three months of treatment and medical examiners must be made clearly with haloperidol (a major tranquilizer with antipsychotic properties), milieu therapy visible as attempts are made to develop cri­ and psychotherapy he improved enough for teria for gathering vital baseline data in tl:e discharge to outpatient treatment. Halo­ area of suicide. peridol was discontinued three weel{s prior Tlwre is an urgen" need to explore and de­ to his discharge from the hospital. He was scribe present practices of reporting suicides carried as an outpatient for six months. About a month prior to readmission loose­ and the degree of consistency or inconsist­ ness of associations and paranoid ideation ency of such reporting in the Unit"d State:::. recurred, 3lld he was reported by his family Until such information is obtained, it wlll be to have wandered nude out of his house. impossible to interpret the available st.atil;ti;s. His school performance declined, and he said he had strange thoughts and could not The coroners and medical examiners are the trust anyone. He refused to take the halo­ keys to the meaningful reporting of statistics peridol which was again prescribed f01 him. on suicide. Upon readmission to the hospital he It is belieVed that it is of the highest pri­ appeared disorganized, suspicious, and ority that an investigation be focused around regressed. Two months later he had im­ proved sufficiently to wanan t a reducti on the following questions directly related to this of the haloperidol which he had been given problem: since admission, but he again regressed and 1. What percentage of all deaths are autop­ was placed on higher doses of haloperidol sied? by a staff supervisor. His bizarre behavior, 2. Who, at prpsent, are the certifying offi­ open sexual advances to staff and patients, and age combined to necessitate a transfer cials, officers, or agencies? Are these medical to a ward where he could be more appro­ examiners, physicians in the community, priately managed. There it was reported sheriffs, coroners? How are they selected? that he accepted seclusion when required, how trained? seemed to respond to medications, but was 3. What are the present official criteria in general withdrawn and regressed. He was angry at his therapist, who had informed given to certifying officials in various jurisdic­ him of his departure some months hence, tions to guide them in reporting a death as and on the day of his suicide made known suicide? a desire to cut off his penis. Later that day 4. What are the present actual practices of he struck an attendant, spent some time in cC'rtifying officials in reporting suicidal seclusion, and shortly after his release from seclusion left the ward undetected and deaths? To what extent rue these practices went to the tenth floor where, finding an consistent with or different from the official unguarded window, he jumped to his critel'ia? death. After the event it was revealed in a 5. By what actual processes do the certlfy­ patient meeting that he had informed ing officials arrive at the decision to list a another patient of his intention and had death as suicide? given away his radio. The discussion pointed out the strong suicide potential of 6. How often ar(~ autopsies performed? young disturbed males, the risks of multi­ Who determines when an autopsy is to be ple therapists and multiple absences, the performed? Are the services of a toxicologist hazards of disagreement about diagnosis, and biochemist available? prognosis, and treatment method. 7. What percen tage of deaths are seen as equivocal, or undetermined, or as a com bina­ tion of two or more modes (for example, SUICIDE STATISTICS: SOME QUESTIONS accident-suicide, undetermined)? 8. What are the criteria for special proce­ In relation to suicide statistics in the United dures in an equivocal death? States, we know that accurate figures do not 9. How much of the total investigation of now exist. There is widespread confusion and a death is dependent upon the police reports? comiderable difference of interpretation as to What is the relationship of the coroner's how to classify deaths. For example, what is investigation to the local police department? considered suicide in one locality is often 10. When, if ever, are behavioral or social THE PSYCHOLOmCAL AUTOPSY 53 scientists involved in the total investigatory In o1'd('1' to avoid tlw inadequacies of this procedure of a dC'ath? conc('ptual confusion, it has been proposed 11. What p(~rcentage of certifying officials that all human deaths 1)(~ elassified among in the United States are medically trained? three types: intentioned, subintentioned, and Does medieal training significantly influence unintentioned (Shneidman 1963, 1973). tlw way in which deaths Urt' reported? An illtelltiolled death is any death in which From data dealing with these questions, the decedent plays a dirf.'ct, conscious role in based on appropriate sampling from regions effecting his own demise. On the other hand, and taking into account rural-urban diffcr­ an IIllilltelltiol1cd death is any death, whatever ences, size of municipalities9 ete., aPPl'opriatn its determiIwd cause 01' apparent NASH agencies could then address themselves to a mode, in which the decedent plays no effec­ numbC'r of important genC'ral questions, in­ tive role in effecting his own demise-wherp eluding the following: dl'ath is dup entirt'ly to independent physical trauma from without, or to nonpsycho­ • What local, State, regional, or other differ­ logically ladpn biolof:,>1cal failure from within. ences em(>rgl' in th(' practices of reporting But most importantly-and, in a fashion I the various modes of death? helieVI' to be chal'aci,E'l'istic of a sizable p(>1'­ • What arc the gpneral implications from til(' cl'ntag(> of all dea1.hs--subiutellfiull£'d deaths data for the accuracy of present death sta­ are dl'aths in whieh the decedc:'l1t plays some tistics, eSfwcially tIlt' statistics for each partial, covert, or unconscious role in hasten­ separate mode? ing his own demise. The objectiv(' evidences • What suggestions can be made for improve­ of the presence of 1.h£>so roles lie in such ment in conceptualization, praetice, and behavioral manifestations as, for example, training which point toward more aecuratE' poor judgment, excessive risk-taking, abuse of and meaningful reporting? alcohol, misuse of drugs, neglect of self, sclf­ destructive style of life, disregard of pre­ seribed life-saving mE'dical regimen, and so on, SUGGESTIONS FOR THE CONCEPTUAL where tIl(> individual fosters, facilitates, ex­ IMPROVEMENT OF THE DEATH acerbates, 01' hastens the process of his dying. CERTIFICATE That individuals may play an unconscious role in their own failures and act inimically to The current NASH classification of death their own best welfare and even hasten their grew out of a 17th centmy way of thinking own deaths seems to be well documented in about man (as a biological vessel who was sub­ the psychoanalytic and general clinical prac­ ject. to whims of fate) and tended to leaVE' tice. This concept of subintentioned death is man himself out of his own death. Twentieth similar, in some ways, to Karl Menningt>r's century psychology and psychiatry have concepts of chronic, focal, and organic suicide at.temptl~d to put man-conscious and un­ (1938). Menninger's idea') relate to self­ conscious-back into his own life, including defeating ways of continuing to live, whereas the way in which he dies. Tlw NASH classifi­ the notion of subintentioned cessation is a cation of modes of death is not only apsycho­ deseription of a way of stopping the process logical but it tends to emphasize relatively of living. Included in this subintention cate­ unimportant dE'tails. For example, it is essen­ gory would be many patterns of mismanage­ tially a matter of indifference to a human ment and brinl<-of-death living which result in being whether a light fixtul'(l above him falls death. In terms of the traditional elassifica­ and he is invaded by a lethal chandelier tion of modes of death (natural, accident, (accidental mode), or someone about him homicide, and suieide), some instances of all coughs and he i') invaded by a lethal virus foUl' types can be subsumed under this cate­ (natural mode), or someone shoots a gun at gory, depending on the particular details of him and he is invaded by a lethal bullet each case. (homicidal mode), if the fact IS that he does Confusion also discolors and obfuscates our not wish (intention) any of these events to thinking in the field of suicide. CU1'l.'ently occur. there is much overattention paid to the cate- 54 SHNEIDMAN gOl'ies of attempted, threatened, and commit­ imprudence, poor jUdgment, provoking others, ted suicide. These categories are confusing disregard of pre&cribed life-saving medical because they do not tell us with what inten­ regimen, active resignation to death, mis­ sity the impulse was felt or the deed was management of drugs, abuse of alcohol, done. One can attempt to attempt, attempt to "tempting fate," "asking for trouble," etc., commit, or attempt to feign, and so on. One where the decedent himself seemed to have can threaten or attempt suicide at any level of fostered, facilitated, or hastened the process intensity. What is needed is a dimension of his dying, or the date of his death. which cuts across these labels and permits us Low lethality. The decedent played some to evaluate the individual's drive to self­ small but not insignificant role in effecting imposed death. We propose a dimension or hastening his own demise. The same as called lethality, defined as the probability of medium above, but to a much less deg1·ee. a specific individual's kUling himself (Le., end­ Absellt lethality. The decedent played no ing up dead) in the immediate future (today, role in effecting his own death. The death tomorrow, the next day-not next month). A was due entirely to assault from outside the measure of the lethality of any individual can body (in a situation where the decedent be made at any given time. When we say that played no role in causing this to happen), 01" individual is "suicidal" we mean to convey death was due entirely to failure within the the idea that he is experiencing an acute ex­ body (in a decedent who wished to continue acerbation (or heightening) of his lethality. to live). All suicide attempts, suicide threats, and The item on the certificate might look like committed suicides should be rated for their this: lethality. Thfl rule of thumb would be that IMPUTED LETHALITY: (Check One) beyond' a certain point one must be wary of High Medium Low Absent the danger of explosion into overt behavior. (See Instructions) What is suggested is that, in addition to the present NASH classification, each death cer­ The reasons for advocating the suggestion tificate contain a new supplementary item are as follows: First, this classifIcation permits which reflects the individual's lethality intent. reflection of the role that the dead individual This item might be labeled Imputed Lethality played in his own dying, in hastening his own (recognizing its inferential character) and death; the ways in which he might have par­ would consist of four terms, one of which ticipated in his own death, etc. Next, it is would then be checked. The terms are: High, more fair. At present, individuals of higher llJedill11l, L01V, Absent, and would be defined social status who commit suicide are more as follows. likely to be assigned the mode of accidental High lethality. The decedent definitely or natural death than are individuals of lower wanted to die; the decedent played a direct social status who no more evidently commit conscious role in his own death; the death suicide. It the term is to have any meaning at was due primarily to the decedent's openly all, it should be fairly used across the board, conscious wish or desire to be dead, or to hiR . measured by the individual's int.ention. (her) actions in carrying out that wish (e.g., Finally, the lethality intention item provides jumping rather than falling or being pushed an unexampled source of information by from a high place; he shot himself to death; means of which biostatisticians, public health he deliberately interrupted or refused life­ officials, and social scientists could assess the saving procedures or medical regimen). mental health of any community. It is ob­ Medium lethality. The decedent played an vious that the number of deaths that are important role in effecting or hastening his caused, hoped for, or hastened by the dece­ own death. Death was due in some part to dents themselves is a measure of the prev­ actions of the decedent in which he played alence of psychological disorder and social some partial, covert, or unconscious tole in stress. At present we do not have this measure, hastening his own demise. The evidences for and we need it. this lie in the decedent's behaviors, such as It might be protested, inasmuch as the his carelessness, foolhardiness, neglect of self, assessments of these intention states involve THE PSYCHOLOGICAL AUTOPSY 55 the appraisal of unconscious factors, that none; medium, 20 (54%); low, 9 (24%); some workers (especially lay coroners) cannot absent, 8 (22%). legitimately be expected to make the kinds of Four deaths were of unknown origin. psychological judgments required for this The first thing we notke is that some type of classification. To this, one answer natural, accidental, and homicidal deaths were would be that medical examiners and cor­ classified as having some degree of lethal in­ oners throughout the country are ma~ing tention. If the medium-intention and low­ judgments of precisely this nature every day intention categories are combined, then over of the week. In the situation of evaluating a one-fourth (26%) of all natural, accidental, possible suicide, the coroner often acts (some­ and homicidal deaths (216 in 843) were times without realizing it) as psychiatrist and deemed to be subintentioned. If one then psychologist, and as both judge and jury in a adds the suicidal deaths, in which the dece­ quasijudicial way. This is because certification dent has obviously played a role in his own of death as suicide does, willy-nilly, imply death, then only 64 percent of all deaths (625 some judgments or reconstruction of the vic­ in 978) were deemed to have been totally tim's motivation or intention. Making these adventitious; conversely, 36 percent were judgments-perhaps more coroners ought to deemed to have had some psychological com­ use the category of Undetermined-is a part ponents. of a coroner's function. But it might be far Also of special interest in these Marin better if these psychological dimensions were County data is the finding that coroners can, explicit, and an attempt, albeit crude, made with apparently no more difficulty than they to usc them, than to have these psychological experience in assigning deaths to the NASH dimensions employed in an implicit and un­ categories, simultaneously (and by essentially verbalized (yet operating) manner. The the same process of inference and induction) dilemma is between the polarities of the pres­ assign deaths to intentional categories as well. ently used oversimplified classification, on It is an important pioneer effort and deserves the one hand, and a somewhat more complex, widespread emulation. but more meaningful classification, on the other. The goal should be to try to combine In summary, the following points may be greatest usefulness with maximum meaning­ emphasized. fulness. Causes. The classification of causes of In Marin County, California, the coroner's death has been rather well worked out and is office is currently assessing each death proc­ consistent with contemporary knowledge. essed by that office in terms of both the There is currently an accepted international traditional NASI-! classification of mode of classification which has wide acceptance. death and the lethality intention of the dece­ Modes. The modes of death have not been dent. For a 2-year period, 1971-1972 (978 stated explicitly and have not been too well cases), the breakdown was as follows:,e1 understood. In general, four currently implied Natural deaths (630): high lethality intent, modes of death-natural, accidental, suicidal, none; medium lethality, 33 (5%); low, 37 and homicidal-suffer from the important. (6%); absent, 560 (89%). deficiency of viewing man as a vessel of the Accidental deaths (176): high lethality intent, fates and omitting entirely his role in his own 2 (1%); medium, 77 (44%); low, 40 (22%); demise. absent, 57 (33%). Intent. The addition of the dimension of Suicidal deaths (131): high lethality intent, lethal intention serves to modernize the death 131 (100%). certificate, just as in the past advances have Homicidal deaths (37): high lethality intent, been made from the teachings of bacteriology , surgery, anesthesiology, immunology, etc. The time is now long overdue for the intro­ duction of the psychodynamics of death into 41 am especially grateful to Keith C. Craig, coro­ the death certificate. The addition of a single ner's deputy. Marin County, for his interest and help item on imputed lethal intent (High, Medium, in supplying these data. Low, Absent) would provide an appropriate 56 SHNEIDMAN reflection of the psychological state of the ference for the Eighth Revision of the Inter­ subject and begin, at last, to reflect the teach­ national Classification of Diseases, Geneva, Swit­ ings of 20th century psychology. In this way zerland, July 6-12. Vilal and Health Statistics. PHS Pub. No. 1000-Series 4-No. 6. Public Health Serv­ we might again permit the certification of ice. Washington: U.S. Government Printing Office. death to reflect accurately our best current understanding of man. National Center for Health Statistics. 1967. Rightlz Revision International Classification of Diseases, Adapted f01' Use in the United States. PHS Pub. No. 1693. Public Health Service, Washington: REFERENCES U.S. Government Printing Office. Rogow, Arnold A. 1963. James FOl'l'estal: A Study of Abbott, Samuel W. 1901. Death certification. In Personality, Politics and Policy. Los Angeles: Albert H. Buck, ed., Reference Handboolz of the Boulevard Bookshop. Medical Sciences. New York: William Wood, Shneidman, Edwin S. 1963. Orientations toward Cassedy, James H. 1969. Demography in Early death: A vital aspect of the study of lives. In America. Cambridge: Harvard University Press. Robert W. White, ed., The Study of Lives. New Colby, Marian. 1965. The significance, evolution, and York: Atherton. implementation of standard certificates. Am. J. Shneidman, Edwin S. 1967. Sleep and self-destruc­ Public Health. 55 :596-99. tion: A phenomenological study. In Edwin S. Curphey, Theodore J. 1961. The role of the social Shneidman, ed., Essays in. Self-Destruction, New scientist in the medicolegal certification of death York: Science House. from suicide. In Norman L, Faberow and Edwin Shneidman, Edwin S. 1968. Orientations toward S. Shneidman, eds., The Cry for Help. New York: cessation: A re-examiIlation of current modes of McGraw-HilI. death. J. Forensic Sci. 13:33-45, Curphey, Theodore J. 1967. The forensic pathologist Shneidman, Edwin S. 1969. Suicide, lethality, and and the multidisciplinary approach to death. In the psychological autopsy. In Edwin S. Shneid­ Edwin S. Shneidman, ed., Essays in Self-Destruc­ man and Magno Ortega, eds., Aspects of Depres­ tion. New York: Science House. sion. Boston: Little, Brown. Hotchner, A.E. 1966. Papa Hemingway: A Personal Shneidman, Edwin S. 1971. Prevention, intervention, Memoir. New York: Random House. and postvention of suicide. Ann. Intel'll. Med. Kargon, Robert. 1963. John Graunt, Francis B.ac?n, 75:453-58. and the Royal Society: The reception of statistICs. Shneidman, Edwin S. 1973, 1974. Deaths of Man. J. Hist. Med. 18:337-48. New York: Quadrangle Books; Penguin Books. Litman, Robert E. 1967. Sigmund Freud on suicide. Shneidman, Edwin S., and Norman L. Farberow. In Edwin S. Shneidman, ed., Essays in S,d{­ 1961. Sample investigations of equivocal deaths. Destruction. New York: Science House. In The Cry for Help. New York: McGraw-Hill. Litman, Robert E., T.J. Curphey, E.S. Shneidman, Weisman, Avery D. 1974. The Realization of'Death: N.L. Faberow, and N.D. Tabachnick. 1963. A Guide for the Psychological Autopsy. New Investigations of equivocal suicides. J.A..M.A. York: Behaviol'al Publications, 184 :924-29. Weisman, Avery D., and Robert Kastenbaum. 1968. McGee, Richard K. 1974. Crisis Intervention in the The psychological autopsy. Community Ment, Community. Baltimore: University Park Press. Health J. Monograph No.4. New York: Be­ Menninger, Karl A. 1938. Man Against Himself. New havioral Publications. York: Harcourt, Brace. Wertenbaker, Lael Tucker. 1957. Death of a Man. National Center for Health Statistics. 1965a. Mortal­ New York: Random House .. ity trends in the United States, 1954-1963, Vital World Health Organization. 1957. Manual of the . and Health Statistics, PHS Pub. No. 1000-Series International Statistical Classification of Diseases, 20-No. 2. Public Health Service. Washington: U.S. Injuries, and Causes of Death, Based on the Government Printing Office. Recommendations of the Seventh Reviqion Con­ National Center for Health Statistics. 1965b. Report ference, 1955. Geneva: World Health Organiza­ of the U.S. delegation to I;he International Con- tion.