Of Woodchoppers, Respirators, and Flickering Flames
By Rabbi Aryeh Klapper
Dean of the Center for Modern Torah Leadership, member of the Boston Beth Din, and Instructor in Rabbinics at Gann Academy
“And G-d saw all that He had made, and behold it was very good” -
In the Torah of Rabbi Meir they found written: “And behold it was very good” – behold how good is death. (Bereshit Rabbah 9:5)
Human life is a centralbut not the supreme value in Judaism. This recognition challenges us to develop clear Jewish conceptions both of the value of life and of those values sometimes placed above it. Meeting this challenge is an urgent necessity in our time, when previously adequate halakhic formulations with regard to the value of human life have been overwhelmed by the astounding capacities of medical science. In particular:
a) a significant percentage of hospital deaths are now temporally volitional, that is to say medical intervention could delay them, however briefly, and
b) many patients can now be sustained for years in conditions that offer no hope of returning to any degree of consciousness, let alone physical autonomy.
These realities require us to rethink the goals of care.
A new set of halakhic formulations is necessary to meet these challenges. Discussions of end-of-life halakha should not presume – although they may conclude - that maximally extending every patient’s life is necessary or even desirable. We need to seriously consider the relative weight of quality and quantity of life.
Talmudic Evidence - Narratives
One Talmudic legend, taken at face value, suggests that life can legitimately become an unwelcome burden simply because of old age. Luz, the Talmud tells us, was a city of potential immortals, as the Angel of Death was not permitted to enter its walls. But eventually its residents would tire of old age – “their minds would come to disgust them” – and they would go out, or as some texts have it, they would be brought out, to meet the Angel voluntarily. Similarly, a midrash records that Rabbi Yose ben Halafta advised a very aged woman who sought death to end her life-sustaining practice of participating in communal prayer.
Two other stories seem to extend the rabbinic tolerance of thanatopsis to choosing death over agonizing pain. Rabbi Hanina ben Teradyon, while being burnt at the stake for teaching Torah in public, promises his executioner a share in the World to Come in exchange for hastening his death. The Maidservant of the House of Rebbe prays for his death when she sees the extent of his suffering from a diuretic disease, and her prayer overcomes the contrary prayer of a rabbinic assemblage. The latter narrative suggests that third parties as well can make the decision that life is no longer worth living.
But the road from narrative to law is long and winding, and not infrequently goes nowhere. Modern decisors have found more or less plausible ways of denying legal significance to each of these narratives. Thus the Luzites are Gentiles, and perhaps their behavior is not rabbinically condoned; Rebbe’s maidservant is not a halakhic authority; Rabbi Yose ben Halafta ended a miracle rather than naturally causing death; and Rabbi Hanina ben Teradyon acted on the basis of a case-specific suspension of the law (hora’at sha’ah).
Indeed, the case of Rabbi Hanina ben Teradyon illustrates well the near-impossibility of conclusively deriving law from Talmudic narrative. The story in and of itself makes a problematic distinction, as just before making his deal with the executioner, Rabbi Hanina turns down his students’ suggestion that he open his mouth, and thereby die faster, with the dramatic statement that “Better that it (my life) be taken by the One Who gave it”. Many explanations have been offered for why he nonetheless then permitted the executioner to increase the flame and remove the water-soaked wool packed around his heart, but none is compelling.
But the even more striking issue is the story’s treatment by recent decisors. Rav Eliezer Waldenberg suggested, as noted above, that Rabbi Hanina acted extralegally in permitting the executioner to hasten his death. In stark contrast, Rav Moshe Feinstein suggested that Rabbi Hanina went beyond the letter of the law in refusing to open his mouth and hasten his own death.
Taken together, the approaches of Rav Waldenberg and Rav Feinstein to this story show that a Talmudic story cannot necessarily be legally demonstrative. Narratives can be pushed to the margin of halakhic discourse once one concedes that actions and decisions valorized in narrative may have no normative implications. So far as I can tell, there are no principles internal to Halakha for determining which/when narrative actions are normative and which/when not.
At the same time, it seems clear to me that ad hoc legal distinctions, however reasonable they may be intellectually and literarily, and however consonant they may be with the norms of halakhic analysis, cannot obscure the overall sentiment of these narratives that life is not always a boon for the one living it. The question then is whether that sentiment generates halakhic outcomes, whether there are legally permitted ways of ending a life not worth living, or at least of allowing it to end without interference. In formal terms, the question is whether there exists a set of patients and methods such that the use of those methods to end those patients’ lives, or to allow those patients’ lives to end, does not violate the biblical prohibitions against manslaughter, murder, and standing idly by one’s fellow’s blood, or the positive injunction to preserve human life.
I need to emphasize here the problems inherent to saying that we require a set of patients and methods, rather than saying that we require either a set of patients or else a set of methods. The halakhic givens here are that human beings are not only prohibited from killing one another, under the rubrics of shefikhut damim (bloodshedding) and retzihah (murder), but are also obligated to save one another, under the rubrics of vahai bahem (“and live by them”) and lo ta’amod al dam rei’ekha (do not stand idly by your peer’s blood). Allowing action or inaction that results in death requires the halakhic determination either that the obligations to save (and perhaps even the prohibitions against murdering) do not apply to the patient in question, or that they have not been violated. If the former, we have to explain why some actions remain prohibited (especially if those actions cannot be considered shefikhut damim or retzihah). If the latter, we have to explain why these actions are not permitted for all patients.
Nonetheless, it seems clear to me that no responsible halakhist has ever considered simply permitting a particular means of causing death, and leaving it to the conscience of individuals to decide whose death to licitly cause. Halakhists were historically also very uncomfortable declaring that medical conditions could vacate all the halakhot that ordinarily prevent us from causing human death, although we shall discuss the possibility that this has changed recently.
History of the Halakhic Discussion
Talmudic law records two categories of people whose physical condition might render them, if not legally dead, at least incompletely alive. These categories seem the most likely candidates to produce the result of vacating prohibitions against causing death. The first is the tereifah, a person with a vital organ that has been physically damaged in a manner that will probably kill him within a year. Killing a tereifah is not a violation of retzihah, but is still considered shefikhut damim, and to my knowledge no halakhist has suggested that there is no obligation to save a tereifah under ordinary circumstances. The category of tereifah accordingly has minimal practical implications for our study.
The second category is the goses. The etymology of the word is not entirely clear, and in any case halakhic literature uses it as a term of art. At first glance, it seems even less significant for our purposes, as the Talmudic discussion of the goses emphasizes repeatedly that he is fully alive, and that causing the death of a goses is at the least shefikhut damim and probably retzihah. Indeed, there is no Talmudic evidence whatever that the physical condition of a patient has any effect whatever on the halakhic obligation to save human life.
B. Sefer Hasidim
However, Sefer Hasidim suggests that some methods for causing the death of a goses are not only permitted but mandated. This suggestion is first cited in a halakhic context by Shiltei Gibborim, then codified by Rabbi Moshe Isserles (RAMA) in the Mapah, and to my knowledge its halakhic authority has not been directly challenged since. As we shall see however, its meaning may have changed dramatically even as its authority remained constant.
Sefer Hasidim 723-724 writes as follows:
One must not cause a person not to die rapidly.
For example, if someone was goses, and there was someone near to the house chopping wood, so that the neshamah was unable to leave, we remove the woodchopper from there.
One must not place salt on his tongue so that he won’t die,
If he is goses and says that he cannot die until they put him in a different place, they must not move him from where he is.
Even though they said that a person who is goses may not be moved from his place, if there is a fire he must not be left in the house, rather he must be taken out.
Sefer Hasidim offers the first halakhic formulation of a permission, even an obligation, to enable death under certain circumstances. He also clearly states that this obligation does not justify all means of enabling death. However, the distinction between permitted and forbidden means is unclear.
Sefer Hasidim mandates removing a woodchopper so as to enable death, and forbids placing salt on the tongue lest death be delayed, but forbids moving to a different room on request for the sake of death, and mandates removal from a house to save from fire. What principle or principles do these example instantiate? The obligation to remove the goses from a burning house seems to indicate that there is still an obligation to save the goses’ life, but if that is so, why is it permitted to cause his death, however indirectly, by removing the woodchopper?
Perhaps the obligation to remove the goses from the house is to save him from the pain of burning rather than from death. On that assumption, we could conclude that there is no obligation to save a goses, but that there remains a prohibition against causing his death directly. The last line of Sefer Hasidim would then make the statement that one may risk causing the goses’ immediate death to save him pain, or in halakhic terms, to risk hayyei sha’ah (short-term life) to avoid tza’ar. It would be difficult to generalize the case, however, as moving the goses also carries the chance that he will survive longer than he would have in the flames.
Another way to explain Sefer Hasidim is to say that he permits ‘removal of a prevention’, and a fortiori permits failure to place a prevention, but forbids even indirect action. This again requires the interpretation that one is obligated to remove the goses from a burning building to save him pain rather than to extend his life.
A third explanation of Sefer Hasidim is that he forbids causing death by means that require physical contact with the goses, but permits all means that don’t require physical contact. On the surface this explanation seems absurd – what halakhic significance does contact have? – but it is rooted in a longstanding understanding of the Talmudic evidence that deserves full presentation.
We begin with Mishnah Shabbat 23:5:
They perform all actions necessary for the dead (on Shabbat).
They anoint and rinse him
so long as they move none of his limbs.
They remove the cushion from underneath him
And place him on the sand so that he will delay (putrefying).
They tie the cheek
not so that it should rise but so that it should not further (sink),
and similarly, a beam that was broken (on Shabbat), they may support it with a bench or with bedstaves, not that it should rise but so that it should not further sink.
They do not close (the eyes) of the dead on Shabbat
nor during the week together with the departure of the nefesh,
and one who closes (the eyes) together with the departure of the nefesh is considered a bloodshedder.
Mishnah Shabbat 23:5 presents the law of the goses as a tangent. It begins with the general rule that one may do everything necessary to prepare a corpse for burial on Shabbat. One may, for example, wash the corpse or anoint it with oil. However, one must be careful not to move its limbs. The reason for this is that the corpse is muktzah, a category that inter alia bans Jews from moving objects that are not plausibly useful on Shabbat. One may nonetheless remove a cushion from underneath the corpse so as to slide it onto sand, if that will enable the corpse to last longer without rotting, as indirect movement is not forbidden. One may tie the cheekbones to preserve facial appearance but not to improve it (the same principal applies to supporting a splintered beam). One may furthermore not close the eyes of the corpse on Shabbat.
At this point the Mishnah has discussed only the already-dead, and how to treat them on Shabbat. It now shifts course and mentions that one may not close the eyes even during the week “with the departure of the nefesh”, and that anyone who does so “is committing bloodshed”. The Mishnah offers no explanation for this startling claim. The Babylonian Talmud (Shabbat 151a), however, cites a beraita that offers an explanatory analogy.
“One who closes the eyes with the departure of the nefesh is committing bloodshed.
This can be compared to a candle that is going out. When a person puts his finger on it, it goes out immediately”.
The straightforward intent of the analogy is that closing the eyes hastens the death of the dying in the same way that snuffing out a flickering candle hastens its extinction. Apparently there was a custom of closing the eyes of the dead as soon after death as possible; very likely this custom reflected a view of the eyes as the “window of the soul”, and there was a fear that the soul would reenter if the window were not closed, causing long and agonizing death throes. People therefore tried their best to time the closing perfectly, and accordingly ran the risk of closing the eyes just slightly early. The Mishnah informs them that this is prohibited, that shortening any life even infinitesimally is tantamount to manslaughter.
A parallel text found in Masekhet Semahot (also known as Evel Rabati) takes a different route to the same end.
The goses, behold he is like one alive for all matters . . .
They must not tie his cheeks,
and they must not plug his orifices,
and they must not place a metal utensil or anything that cools on his abdomen,
until the time he dies.
They must not move him,
and they must not rinse him,
and they must not place him neither on sand nor on salt.
until the time he dies.
They must not close his eyes.
One who touches him and moves him – behold he is a bloodshedder.
Rabbi Meir would analogize him to a candle that is flickering – once a person touches it, he immediately extinguishes it.
This beraita’s opening general statement is that a goses is considered “like the living” for all purposes. Accordingly, one is forbidden to perform on the goses roughly those activities which the Mishnah in Shabbat permits one to perform on a corpse, including moving it, nor may one make funeral arrangements in the goses’ presence. Finally, one is forbidden to close the goses’s eyes “until the nefesh leaves”.
Here, however, the flickering flame in the analogy is not snuffed out, but rather simply touched. The associated prohibition accordingly is not to close the goses’ eyes, but rather to “touch him and move him” – whether this requires a touch that moves, or prohibits either touching without moving or moving without touching, is unclear. Closing the eyes here is only an example of touching/moving.
This alternate understanding of the analogy makes it superficially plausible to read Sefer Hasidim as distinguishing between modes of causing death that involve physical contact with the goses, i.e. touching, and those that don’t. Revisiting the issue, however, we recognize that the prohibition against touching is based entirely on the presumption that touching the goses kills him directly and actively, and would apply equally to every other direct and active means of causing death. There is therefore no basis for using such contact as the basis for halakhically distinguishing among modes of causing death.
There are furthermore three good reasons to reject the idea that it is forbidden to touch a goses. First of all, no contemporary medical condition corresponds to a state in which touching causes death, and thus maintaining the prohibition mythicizes the status of goses for no reason – goses in the Talmud seems a perfectly straightforward condition readily recognized by laypeople. Second, asserting such a prohibition potentially deprives dying patients of the comfort of the touch of their loved ones, who may be legitimately concerned lest their comforting touch be considered murderous by Halakha. Third, it seems likely to me that this prohibition stems from an erroneous text. As mentioned above, the analogy as cited in Tractate Shabbat refers exclusively to the act of closing the eyes, not to mere touch, or even of touch which moves the goses. That prohibition applies only on Shabbat, on the basis of muktzah. I suggest that the close parallels between the Mishnah regarding what may be done for a goses on Shabbat, and the beraita regarding what may be done for him during the week, eventually led to the transfer of the prohibition against even minor motion from one context to the other, and generated a misunderstanding of the analogy to the flickering candle. The version of the beraita preserved in “Beraitot Me’evel Rabbati in Masekhet Semakhot” preserves the analogy in what I believe to be its original form, and contains no prohibition against touching.
Such historical speculation has generally had limited influence on halakhic decision-making. In this case, however, studying the history of the texts yields a surprising result. Many decisors and commentators, beginning with Behag, simply cite both versions of the text, with the analogy relating to closing the eyes in one, and touching in the other, without comment. The only rishon I can find whose published edition clearly preserves what I believe to be the original text of the beraita is Rabbi Yitzhak ibn Gayyat. Nonetheless, both Tur and Shulhan Arukh cite the beraita in that form, and make no mention whatever of a prohibition to touch a goses. I see this absence, combined with the physical implausibility and pastoral inhumanity of banning touch, as sufficient ground for ruling that no such halakhic prohibition exists. 
It also seems to me that when Sefer Hasidim writes “even though they said ‘one may not move him from his place’”, this implies that he understood the beraita’s phrase “ein mezizin oto” as referring to moving the goses rather than to simply touching him. Thus Sefer Hasidim as well does not see touching/not touching as a relevant distinction between modes of enabling death. It follows that Sefer Hasidim can only be understood by assuming that the relevant distinction is either
a) between the obligation to save, which does not apply to a goses, and the prohibitions against causing death, which do, or
b) between removal of a prevention, which is not considered an action, and actions.
No adequate halakhic rationale, however, has ever been offered for this second distinction; surely it would be forbidden to remove something preventing the death of a healthy person. Sefer Hasidim must therefore be understood as arguing that the obligation to preserve life is waived in the case of a goses. This is consistent with its prohibition in Sefer Hasidim 234 against screaming at the point of death, lest the soul be returned to the body, which also assumes that Halakha acknowledges that restoring someone’s life is not always doing them a favor.
C. Shiltei Gibborim
While this seems to me the correct reading of Sefer Hasidim, this text has a complex history of interpretation in halakhic tradition. It is first introduced into the halakhic mainstream by the 14th century work Shiltei Gibborim in his commentary to Talmud Moed Kattan 16b
From here it seemed right to forbid that which some people practice, when the met is goses and the neshamah is unable to depart, that they push the cushion out from underneath him so that he will die rapidly, for they say that there are in the bed feathers of birds that cause the nefesh not to depart.
Many times I yelled like a rooster to remove this evil practice, but I was not successful,
and my teachers disagreed with me,
and the Rav Rabbi Natan of Igra of Blessed Memory wrote to permit this.
After several years, I found in Sefer Hasidim 723 support for my words, for it is written there: “If he is goses and unable to die until they put him in a different place – they must not move him from there.”
It is true that the words of Sefer Hasidim require investigation, because initially he wrote that
“If a man was goses, and there was someone close to that house who was chopping wood, and the neshamah is unable to depart, we remove the chopper from there”,
but we can answer as follows and say
that certainly to do something that will cause the goses not to die rapidly is forbidden,
for example to chop wood there so that the neshamah will be delayed from departing,
or to place salt on his tongue so that he will not die rapidly – all this is forbidden, as is implied there by his language,
and in all similar cases it is permitted to remove such a cause
but to do something that will cause his death to occur more rapidly and the departure of his nefesh is forbidden.
Therefore it is forbidden to move a goses from his place and leave him in a different place so that his neshamah can depart,
and therefore it is also forbidden to put the key of bh”kn under the head of a goses so that he die rapidly, because this too speeds the departure of his nefesh.
According to this,
If there is something there that causes his nefesh not to depart, it is permitted to remove that cause, and this is not at all significant, as he is not placing his finger on the candle, and he is not doing an action,
but to place something on the goses, or to carry him from place to place so that his neshamah will depart rapidly, appears definitely forbidden, as he is placing his finger on the candle.
The progression of opinions in Shiltei Gibborim’s text is very complicated. For example, it is not clear whether he opposed the custom of removing the pad from underneath a goses because he shared the popular belief that the feathers served to prevent death, or rather because he believed that removing the cushion might cause death directly by jolting the goses. There is even some controversy among aharonim as to whether his conclusion still opposes the custom of removing the pad! I read him as opposing the custom because the jolt might be fatal, and as opposing it even in his conclusion.
Shiltei Gibborim reads Sefer Hasidim as distinguishing between “hastening death” on the one hand, and “removing an obstacle to death” on the other. Thus he bans procedures that “hasten death” even if they involve no contact with the goses, such as leaving a key under his pillow, and presumes that all obligations to save and prohibitions against killing apply to a goses. Shiltei Gibborim offers no source other than Sefer Hasidim for the suggestion that “removing an obstacle to death” is legally permitted,, and offers no guide for distinguishing between methods that “hasten death” and those that “remove obstacles to death”. For example, he gives no explanation as to why providing the goses’ soul with a key is regarded as hastening his death rather than as removing the obstacle of a locked door, or as providing the goses with the means to remove that obstacle.
One might understand Shiltei Gibborim as positing that the obligations to save and not kill apply only when a person ought not yet have died, but when that point passes, they lapse as well. This logic would permit, for example, shooting the goses who has lived past his time in order to overcome the obstacle, for example the feathers. However, his emphasis that removal of an obstacle “is not an action” seems to falsify that understanding. This distinction between action and non-action is also not explained, however, and seems to fly in the face of his assertion, not found in Sefer Hasidim, that one may actually remove salt from the tongue of a goses to enable him to die. Perhaps he is trying to set up a set of patients and methods, but I don’t see the compelling argument for his particular combination.
Note that we also have the option of ruling against the author of Shiltei Gibborim and in favor of his teachers and Rabbi Nathan of Igra that removing the cushion is permitted. Their position certainly does not distinguish between action and non-action, and on the surface contradicts the Mishnah’s prohibition against moving the goses. It is possible that they, like Sefer Hasidim, read the Mishnah only as banning moving the goses “from his place”. In that case they could be distinguishing between actions that cause death directly and those that cause it indirectly, although this again would be a novel halakhic conclusion. Alternatively, they may agree in principle with Shiltei Gibborim but hold that removing the cushion should be classified as ‘removing a prevention’, as they do not see the direct effect on the goses, but rather the removal of the feathers, as fatal.
Rav Yosef Caro in Shulhan Arukh makes no mention of Sefer Hasidim or Shiltei Gibbotim, and presumably was aware of neither. Shulkhan Arukh writes merely that
The goses, behold he is like one alive for all his matters . . . and they may not close (his eyes) until his nefesh departs. And anyone who closes (his eyes) together with the departure of the nefesh, behold he is a bloodshedder
Shiltei Gibborim is cited by RAMA in Darkei Mosheh and subsequently codified in the Mapah as follows:
But some say that one must not dig a grave for him, even though it is not with him in the house, until after he dies.
It is forbidden to dig any grave that will be open even though it is not with him in the house, until after he dies.
It is forbidden to dig any grave to be open until the next day if they will not bury the dead person that day, and there is danger in this matter.
And it is also forbidden to cause the dead to die rapidly,
as for example one who was goses for a long time and was unable to separate, it is forbidden to shove the cushion or bolster from underneath him, on the ground that they say that there are feathers from some birds that cause this.
and they must also not move him from his place,
and it is also forbidden to place the keys of the synagogue under his head, so that he separate.
But if there is something there that causes a delay in the departure of the nefesh,
as for example if there is close to that house a pounding noise,
such as a woodchopper,
or if there is salt on his tongue,
and these prevent the departure of the nefesh,
it is permitted to remove it from there,
as this involves no action at all, rather he is removing the prevention.
RAMA’s citation in Darkei Moshe leaves out the issue of action/non-action, but his codification in Mappah restores it with the added claim that removing an obstruction is not an action “at all” because it merely “removes the prevention”. RAMA also rules explicitly against the practice of removing the cushion. The difference between removing the salt on a goses’ tongue on the one hand, and removing the cushion on the other, is explained by the assertion that the first merely removes a prevention.
It is important to emphasize that RAMA would certainly forbid removing necessary medicine from a patient’s tongue, and so his declaration that removing the salt is “not an action at all” cannot be an independently sufficient ground for permitting it. Rather, he must believe that a patient kept alive only by salt on the tongue is not, from a halakhic perspective, fully alive, or more technically, that there is no obligation to save the life of such a patient. His reading is therefore congruent with what we have suggested is the formal halakhic reasoning of Sefer Hasidim.
But RAMA follows Shiltei Gibborim’s practical conclusion in applying Sefer Hasidim’s reasoning to allow removal of salt from the patient’s tongue. I suggest that this is not a necessary analogy to removing the woodchopper, and would have been opposed by Sefer Hasidim. Removal of the woodchopper is not an action; rather, it is the cessation of the woodchopper’s action, and can therefore be defined as “failing to act to save from death” rather than as “acting to cause death”, and is therefore easily compatible with the position that one has no obligation to save a goses. Removing the salt, however, seems clearly to be, in ordinary language, causing death; what defines it halakhically as “removal of a prevention” therefore is not the nature of one’s action, but rather the nature of the life it ends.
RAMA’s formulation therefore generates the conclusion that causing death for some patients is by definition merely the removal of an obstacle. It should therefore be permitted to directly kill those patients as well. In RAMA’s circumstance, however, this practical result was insignificant, as the only way one could legally know that patients had reached the condition in which one could legally kill them was by performing otherwise innocuous actions and seeing if they resulted in death.
Both RAMA and Sefer Hasidim conclude that there exists a set of patients who may not be killed but nonetheless need not be saved. In Sefer Hasidim, that set may be limited to the set of those who are goses. As imminent and inevitable dying seems to be at least part of the definition of goses, Sefer Hasidim therefore has limited if any direct application to long-term vegetative cases and their ilk.
However, our reading of Sefer Hasidim also suggested removing a goses from a burning building is required not to save the goses’ life, but rather to spare him pain, and that this may stand for the proposition that one may risk short-term life for the sake of palliation, especially where there is a possibility that one may end up extending life as a result. This has direct implications for the very common question of whether one can give terminal patients morphine shots, as the morphine is addictive and requires increasing doses that will eventually repress respiration to the point of death, although in some cases the morphine at initial doses may extend life by easing respiration.
Rabbi Shlomo Zalman Auerbach (Responsa Minhat Shlomoh 2: (2-3) 86) permits this practice as follows:
“since suffering is very difficult for a person and he cannot endure it, as our Sages said ‘Had they tortured Hananiah etc.’, it is a reasonable presumption that one needs to be merciful toward the sick and to lighten his burden and to somewhat quiet his pains, especially as it is also possible that strong pains weaken and damage a person more than the medications, but if the patient is aware it seems that they need to inform him of this if he otherwise knows his condition, but even if he does not know it we nonetheless find on Talmud Sanhedrin 84b and the Rashi thereupon that “You must love your neighbor as you love yourself – Jews were only cautioned from doing to their peers things they do not wish to do to themselves”. In the case before us, every sick person seeks and pleads for pain relief even by damaging means, and since this is the standard case, the court can constructively testify that he would prefer it. It is understood that all this is exclusively to relieve pain, so that the likelihood that this will bring his death nearer is like an inevitable but unwanted consequence (my reference to that legal construct is only illustrative). We have also found in the Talmud in several places about many things that are very nearly dangerous that nonetheless, since the general practice is to do them, we say “G-d guards the simple”, so since the way of all patients is to receive such pain relief, it is plausible that it is also relevant to apply that principle even in our case, and so one indeed needs to relieve the pain, and Hashem should have mercy.”
Rabbi Auerbach’s rationales are obscure and to some degree frightening, as he seems in principle to permit any generally practiced form of medical euthanasia so long as the practice has a formal palliative intent, even if it were inevitably fatal. With all due respect to his awesome scholarship and judgment, and knowing my relative insignificance, I would be deeply uncomfortable ruling on the basis of such a rationale. However, it seems to me that Sefer Hasidim provides a possible precedent with narrower implications, as it would permit only a risk rather than an inevitable death, and might require that there be a serious chance that life actually be extended.
In contrast to Sefer Hasidim, RAMA does not necessarily limit the set of patients whom one need not save to those who are goses. Rather, RAMA creates a new category of “patients whose death is being prevented, rather than whose life is being continued”. This has both a narrowing and expanding effect on the set. It narrows in that not all those who are goses necessarily meet this requirement, and indeed RAMA’s formulation contains the novel phrase “One who has been goses for a long time”, although he gives no clue as to how long “a long time” is. But his formulation also raises the possibility that some patients who are not goses, and indeed whose death might be prevented for a very long time, nonetheless need not have their deaths prevented.
This possibility is seized upon by two great recent decisors, Rabbi Eliezer Waldenberg and Rabbi Hayyim David HaLevy, to address the question of withdrawing care from patients who are permanently unconscious (the precise degree of brain damage necessary to enable the halakhic determination that their unconsciousness is permanent is not specified) and who would die soon after being removed from artificial respiration.
Rabbi HaLevy argues in a remarkable theological vein that G-d no longer desires these patients to be alive, and concludes that removing such patients from respirators is not only permitted but mandatory.
“Rather, it further seems to my impoverished intellect that even if the doctors wish to continue and keep them alive via the respirator, they are not permitted to do so, as indeed it has already been explained that it is forbidden to extend the life of a goses by artificial means, for example by placing salt on his tongue and chopping wood, when there is no longer hope for his life. While the Halakha discusses a goses who is alive owing to his own powers and who therefore has great agony, whereas in our case he feels no pain or suffering, nonetheless it seems to my impoverished intellect that not only is it permitted to detach him from the respirator, but even that there is an obligation to do so, for indeed the nefesh of a human being, which is the property of The Holy One Who is Blessed, has already been taken by The Holy One Who is Blessed from this human being, as behold he will die immediately once the machine has been removed. Quite the reverse – via this artificial respiration we are leaving the nefesh within him and causing it (the nefesh, not the goses) suffering, for it is not able to separate and return to its place of rest.
Therefore it seems to my impoverished intellect that is is permitted to you, when you have reached a clear conviction with no room for doubt or challenge that there is no longer hope for this human being to be healed, to detach him from the respirator, and you can do this without any pangs of conscience”.
Rabbi HaLevy provides no explanation beyond citing RAMA for distinguishing between removal from a respirator and directly causing death, and his notions of purely spiritual pain, and of souls that G-d has taken remaining within the body, are highly creative.
Rabbi Waldenberg uses very strong rhetoric, and a very extreme case, to argue that doctors must withdraw care in some cases so as “to allow him to die a ‘mitah yafah’ – a death of dignity, and not to degrade him through his transformation, without purpose, into a figure with many tubes that destroys the tzelem Elokim within him.” This intuitive claim is then legally formalized as follows: “The foundational point is that when one removes a cause that it was forbidden to introduce, since it delays death, that status is a sign that this does not bring independent life to the goses, and therefore one can remove it, as its removal is not an action of interrupting his life, but rather the removal of an external cause that prevents the departure of his nefesh . .
It emerges from the above that even according to RAMA it would be permitted in our case to do something, even actively, so as to remove the respirator, since thereby one would just be interrupting the life force that comes to him from the outside, and they are doing “no deed at all” that involves interrupting independent life, as it has become clear that this is already interrupted.
We have also learned from the above that if they removed the respirator in order to clarify whether he is alive, or if it was interrupted for whatever reason, and it becomes clear that he no longer has independent life either from the brain or from the heart, that not only is it permitted to not reconnect him, but rather there is a prohibition to do so, since it is already clear that this will not bring him independent life but rather will only cause extension of the delay of the departure of the nefesh, and if they transgressed and connected him there is an obligation to disconnect him”.
The distinction between “independent life” and “externally provided life” that Rabbi Waldenberg makes here is to my mind unprecedented, and he provides no criteria for applying it in future cases that are not purely intuitive. Furthermore, as he explicitly allows actively causing death, it is entirely possible that he permits any painless form of euthanasia for such patients.
As may be clear from the above, I find the distinctions made respectively by Rabbi HaLevy and Rabbi Waldenberg deeply troubling. Rabbi HaLevy’s claim to know that G-d desires certain patients to be dead seems to assume facts that cannot possibly be in evidence, and Halakha is not generally decided on the basis of claims to direct knowledge of the Divine Will. Rabbi Waldenberg’s assertion that a body can lose its Tzelem Elokim because of medical intervention has a certain intuitive power but ultimately both reduces human beings to bodies and gives those bodies too much significance. The great paraplegic physicist Stephen Hawking retains his tzelem Elokim no matter how many machines are required to sustain his powerful intellect. Furthermore, Rabbi Waldenberg’s distinction between “externally sustained” and “independent life” seems entirely arbitrary – are scuba divers, for instance, artificially alive? Are patients undergoing surgery that requires them to be ventilated? Would it be perfectly legitimate halakhically to cut off divers’ and surgical patients’ respirators? Note further that Rabbi Waldenberg explicitly rejects the active/passive distinction, and implicitly the killing/lifesaving distinction; such patients have a life that it is a mitzvah to end. It seems to me that he would in fact endorse any means of painlessly euthanizing such patients.
But while the specific arguments each great decisor makes are, laaniyut daati, unconvincing and carry implications very difficult to countenance, they reflect an apparent intuitive consensus among great poskim that Halakha cannot in fact demand that society expend great sums and effort to sustain the lives of patients who will never experience any quality of life. I suggest that this indicates that a viable halakhic formulation that permits withdrawal of care in such cases must be discoverable, although it has not yet been discovered. How halakhic decisions should be made in the interim is to my mind an extraordinarily complicated question.
Finally, I should note that a formulation along the lines of those of Rabbi Waldenberg and Rabbi HaLevy might have another interesting consequence. The controversy about whether “brain-stem death” is halakhic death is generated by the reality that effectively transplanting some organs currently requires them to be removed from a body with a beating heart, and Halakha views the removal of vital organs as murder. Under their formulations, however, it seems to me possible that such transplants might nonetheless be legitimate, as the prohibitions against taking life simply do not apply to patients such as these, and thus transplants might be permitted halakhically even from patients who do not meet the “Harvard criteria”. I leave it to others to decide whether this prospect argues for or against adopting a formulation along these lines.
Contemporary medical science presents Halakha with challenges that cannot be easily met simply by extending the rays of precedent. The capacity to indefinitely keep patients alive who have no quality of life by any available standard requires sober assessment of the purposes and effects of medical intervention. I hope to have shown that contemporary poskim have utilized the resources of the masoret with great humanity and creativity to fashion immediate responses to some of these challenges. At the same time, I hope to have shown that accepting their formulations might bring us into uncharted and dangerous halakhic waters and that accordingly, much work remains to be done in the task of being developing a Halakhic medical end-of-life ethic that is both rigorous and relevant.
 Lifesaving mandates breaking the Sabbath and all laws other than those against idolatry and adultery, and if enough lives are involved, adultery too may be permitted. The Biblical injunction “Choose life!” resonates through the history of Jewish thought and practice, and gratitude for life is at the core of the Jewish sense of obligation to G-d.
 Jews may not commit idolatry or adultery to save life, nor take one life to save two others, and the Tosafists permitted Jews to choose death rather than commit any sin. Some authorities take literally the Talmudic statement that “Better for a person to throw himself into a burning furnace than to publicly humiliate someone else”, and the language of that statement valorizes suicide rather than merely allowing oneself to be killed. Jewish law also mandated some wars and legitimates others despite the inherent risk to life involved in war.
 I acknowledge here my indebtedness to a lecture by Rabbi Saul Berman many years ago that first made me aware of the weaknesses of the standard halakhic presentations in this area.
 Sotah 46b
 Yalkut Shim’oni Proverbs 943
 Avodah Zarah 17b
 Ketubot 104a
 Responsa Tzitz Eliezer 18:48
 Responsa Igrot Mosheh Hoshen Mishpat 2:73
 I except here narratives that are explicitly cited for the purpose of legal precedent, such as the legal maaseh, although the relationship of even that genre to law is at the least complicated.
 Genesis 9:6
 Exodus 21:14; Number 35:16-18
 Leviticus 18:5
 Leviticus 19:16
 Maimonides, Laws of Murder 2:8
 See Encyclopedia Talmudit (Hebrew) 5:393-4
 Sefer Hasidim, traditionally ascribed to Rabbi Judah the Hasid of Regensburg (d.1217) but not entirely from his hand, is an eclectic combination of stories, halakhic statements, folklore, and other genres. It had significant influence on Ashkenazi Halakhah.
 a standard commentary on the Talmudic digest of Rabbi Isaac Alfasi, authored by the 16th Century Italian Rabbi Yehoshua Boaz ben Shimon Barukh.
 Rabbi Isserles lived in 16th Century Poland. His Mapah, or Tablecloth, added Ashkenazic halakhic traditions as glosses to Rabbi Yosef Caro’s Shulhan Arukh, and the combined work has been the standard codification of Jewish law ever since.
 medieval halakhic authority
 11th century Spain
 It is worth mentioning a possibly significant implication of such a ruling. Some criteria for determining death have been deemed functionally irrelevant under Halakha because testing for them requires touching the dying person. If the argument above is correct such testing would of course be permitted.
 Nonetheless, it must be conceded that the mainstream of Halakha has followed and continues to follow this position. For example, Shaarei Tzedek Medical Center in Jerusalem, in accordance with the advice of many great halakhic decisors, places its respirators on timers that turn them off, absent human intervention, at regular intervals. If a decision is made not to continue the life of a patient, then, the medical staff need merely fail to restart the machine, rather than turning it off. For the reasons set forth throughout this essay, I find this a very unsatisfying result.
 post-medieval halakhic authorities
 The text of the printed Shiltei Giborim refers to mafteiach BhKn,= Beit haKnesset, or synagogue key. No explanation is given as to why a synagogue key would aid death by magical or other means. However, as will be discussed more extensively below, Darkei Moshe has only B”h, and I suggest that the original intent was Beit haKvarot, or graveyard, in which case the mechanism is obvious.
 as he reports their positions – we have no other source for their views
 Note that one version of the beraita but not the other explicitly forbids removal of cushions from underneath the goses.
 Shulhan Arukh and Beit Yosef seem unaware of Sefer Hasidim and Shiltei Gibborim. One might therefore conclude, although one cannot demonstrate, that they reject any distinctions and forbid contributing in any way to the death of a goses.
 In the printed text of Shiltei Gibborim the abbreviation here is BhK”n, or Beit HaKnesset. However, Mapah and Darkei Mosheh both have B”H, which could stand for Beit haKvarot, or cemetery, which seems a better fit for context. But I have no outside evidence for this emendation, and no record of an alternate edition of Shiltei Gibborim.
 “It is written in Hagahot Alfasi:
From that which we learned in a Mishnah: “One who closes the eyes with the departure of the nefesh is considered like a bloodshedder”, it seems correct to forbid that which some people practice, that when someone is goses and the neshomoh is unable to leave, that they shove the cushion from underneath him so that he will die rapidly, as they say that the bed contains feathers of birds that cause the nefesh not to depart.
But Rabbi Nathan of Igra wrote to permit this.
Afterward I found written in Sefer Hasidim a support for our words, for he wrote:
If someone was goses and unable to die until placed in another place, they must not move him.
And he wrote:
“If someone was goses, and there was someone near that house chopping wood, so that the neshomoh is unable to depart, we remove the woodchopper from there.
But his words do not contradict one another,
for certainly to do something that will cause him not to die rapidly is forbidden, as for example to chop wood there so the neshomoh will be prevented from leaving, or to place salt on his tongue so that he will not die rapidly, is certainly forbidden,
and everything similar it is permitted to remove,
but to do something that will cause his rapid death and the departure of his nefesh is forbidden,
and therefore it forbidden to move the goses from his place and leave him in a different place.
It is also forbidden to place the keys of a B”h under the head of a goses so that he will die rapidly, because this also hastens the departure of the nefesh,
but it is permitted to remove something that delays the departure of the nefesh.”
 This formulation carries the implication that one would also have no obligation to save such a patient from other threats to his life, an implication that has no support in RAMO’s language, but I have no other way of translating his language into formal halakhic terms. I’ll return to this point at greater length below.
 Some recent research indicates that the concern that respiration will accidentally be fatally repressed does not apply to conscious patients.
 Dr. Ernest Mandel notes that, if one reads Minchat Shlomoh’s reference to pesik reisha (inevitable consequence) as pure exaggeration, he may also agree that palliation should be permitted only if it risks short-term life and not if it inevitably shortens life.
 Tehumin 2:297-305
 Responsa Tzitz Eliezer 13:89
 For the argument about costs see the end of Tzitz Eliezer 13:89.
 This capacity also raises challenges as to the appropriate allocation of societal resources, a topic beyond the scope of this paper. On that topic generally see Rabbi M.D. Tendler in Kavod HaRav, ed. Moshe Sherman, pp. 167-169 (Student Organization of Yeshiva, 1992).