The custom of fasting after a death has been ascribed to different causes by different writers. Mr. Spencer believes that it has resulted from the habit of making excessive provision for the dead.88 But although among some peoples the funeral offerings no doubt are so extensive as to reduce the survivors to poverty and starvation,89 I have met with no statement to the effect that they are anxious to give to the deceased all the eatables which they possess, or that the mourning fast is a matter of actual necessity. It is always restricted to some fixed period, often to a few days only, and it prevails among many peoples who have never been known to be profuse in their sacrifices to the dead. With reference to the Chinese, Dr. de Groot maintains that the mourners originally fasted with a view to being able to sacrifice so much the more at the tomb; and he bases this conclusion on the fact that the articles of food which were forbidden till the end of the deepest mourning were the very same as those which in ancient China played the principal part at every burial sacrifice.90 But this prohibition may also perhaps be due to a belief that the offering of certain victuals to the dead pollutes all food belonging to the same species.

Professor Wilken, again, suggests that the mourners abstain from food till they have given the dead his due, in order to show that they do not wish to keep him waiting longer than is necessary and thus make him kindly disposed towards them.91 This explanation presupposes that the fast is immediately followed by offerings or a feast for the dead. In some instances this is expressly said to be the case;92 the ancient Chinese, for instance, observed a special fast as an introductory rite to the sacrifices which they offered to the manes at regular periods after the demise and even after the close of the mourning.93 But generally there is no indication of the mourning fast being an essential preliminary to a sacrifice to the dead, and in an instance mentioned above the funeral feast regularly precedes it.94

It seems that Sir J. G. Frazer comes much nearer the truth when he observes that people originally fasted after a death “just in those circumstances in which they considered that they might possibly in eating devour a ghost.”95 Yet I think it would generally be more correct to say that they were afraid of swallowing, not the ghost, but food polluted with the contagion of death. The dead body is regarded as a seat of infection, which defiles anything in its immediate neighbourhood, and this infection is of course considered particularly dangerous if it is allowed to enter into the bowels. In certain cases the length of the mourning fast is obviously determined by the belief in the polluting presence of the ghost. The six days’ fast of the Paressí coincides with the period after which the dead is supposed to have arrived in heaven no longer to return; and they say that anybody who should fail to observe this fast would “eat the mouth of the dead” and die himself.96 Frequently the fasting lasts till the corpse is buried; and burial is a common safeguard against the return of the ghost.97 The custom of restricting the fast to the daytime probably springs from the idea that a ghost cannot see in the dark, and is consequently unable to come and pollute the food at night. That the object of the fast is to prevent pollution is also suggested by its resemblance to some other practices, which are evidently intended to serve this purpose. The Maoris were not allowed to eat on or near any spot where a dead body had been buried, or to take a meal in a canoe while passing opposite to such a place.98 In Samoa, while a dead body is in the house, no food is eaten under the same roof; hence the family have their meals outside, or in another house.99 The Todas, who fast on the day when a death has taken place, have on the following day their meals served in another hut.100 In one of the sacred books of India it is said that a Brâhmana “shall not eat in the house of a relation within six degrees where a person has died, before the ten days of impurity have elapsed”; in a house “where a lying-in woman has not yet come out of the lying-in chamber”; nor in a house where a corpse lies;101 and in connection with this last injunction we are told that, when a person who is not a relation has died, it is customary to place at the distance of “one hundred bows” a lamp and water-vessel, and to eat beyond that distance.102 In one of the Zoroastrian books Ormuzd is represented as saying, “In a house when a person shall die, until three nights are completed … nothing whatever of meat is to be eaten by his relations”;103 and the obvious reason for this rule was the belief that the soul of the dead was hovering about the body for the first three nights after death.104 Closely related to this custom is that of the modern Parsis, which forbids for three days all cooking under a roof where a death has occurred, but allows the inmates to obtain food from their neighbours and friends.105 Among the Agariya, a Dravidian tribe in the hilly parts of Mirzápur, no fire is lit and no cooking is done in the house of a dead person on the day when he is cremated, the food being cooked in the house of the brother-in-law of the deceased.106 In Mykonos, one of the Cyclades, it is considered wrong to cook in the house of mourning; hence friends and relatives come laden with food, and lay the “bitter table.”107 Among the Albanians there is no cooking in the house for three days after a death, and the family are fed by friends.108 So also the Maronites of Syria “dress no victuals for some time in the house of the deceased, but their relations and friends supply them.”109 When a Jew dies all the water in the same and adjoining houses is instantly thrown away;110 nobody may eat in the same room with the corpse, unless there is only one room in the house, in which case the inhabitants may take food in it if they interpose a screen, so that in eating they do not see the corpse; they must abstain from flesh and wine so long as the dead body is in the house;111 and on the evening of mourning the members of the family may not eat their own food, but are supplied with food by their friends.112 Among the Arabs of Morocco, if a person has died in the morning, no fire is made in the whole village until he is buried, and in some parts of the country the inmates of a house or tent where a death has occurred, abstain from making fire for two or three days. In Algeria “dès que quelqu’un est mort, on ne doit pas allumer de feu dans la maison pendant trois jours, et il est défendu de toucher à de la viande rôtie, grillée ou bouillie, à moins qu’elle ne vienne de quelqu’un de dehors.”113 In China, for seven days after a death “no food is cooked in the house, and friends and neighbours are trusted to supply the common necessaries of life.”114 There is no sufficient reason to assume that this practice of abstaining from cooking food after a death is a survival of a previous mourning fast, but the two customs seem partly to have a similar origin. The cooking may contaminate the food if done in a polluted house, or by a polluted individual. The relatives of the dead, or persons who have handled the corpse, are regarded as defiled; hence they have to abstain from cooking food, as they have to abstain from any kind of work,115 and from sexual intercourse.116 Hence, also, they are often prohibited from touching food; and this may in some cases have led to fasting, whilst in other instances they have to be fed by their neighbours.117

However, an unclean individual may be supposed to pollute a piece of food not only by touching it with his hand, but in some cases by eating it; and, in accordance with the principle of pars pro toto, the pollution may then spread to all victuals belonging to the same species. Ideas of this kind are sometimes conspicuous in connection with the restrictions in diet after a death. Thus the Siciatl of British Columbia believe that a dead body, or anything connected with the dead, is inimical to the salmon, and therefore the relatives of a deceased person must abstain from eating salmon in the early stages of the run, as also from entering a creek where salmon are found.118 Among the Stlatlumh, a neighbouring people, not even elderly widowers, for whom the period of abstention is comparatively short, are allowed to eat fresh salmon till the first of the run is over and the fish have arrived in such numbers that there is no danger of their being driven away.119 It is not unlikely that if the motives for the restrictions in diet after a death were sufficiently known in each case, a similar fear lest the unclean mourner should pollute the whole species by polluting some individual member of it would be found to be a common cause of those rules which prohibit the eating of staple or favourite food.120 But it would seem that such rules also may spring from the idea that this kind of food is particularly sought for by the dead and therefore defiled.

Moreover, unclean individuals are not only a danger to others, but are themselves in danger. As Sir J. G. Frazer has shown, they are supposed to be in a delicate condition, which imposes upon them various precautions;121 and one of these may be restrictions in their diet. Among the Thlinkets and some peoples in British Columbia the relatives of the deceased not only fast till the body is buried, but have their faces blackened, cover their heads with ragged mats, and must speak but little, confining themselves to answering questions, as it is believed that they would else become chatterboxes.122 According to early ideas, mourners are in a state very similar to that of girls at puberty, who also, among various peoples, are obliged to fast or abstain from certain kinds of food on account of their uncleanness.123 Among the Stlatlumh, for instance, when a girl reaches puberty, she fasts for the first four days and abstains from fresh meats of any kind throughout the whole period of her seclusion. “There was a two-fold object in this abstention. First, the girl, it was thought, would be harmed by the fresh meat in her peculiar condition; and second, the game animals would take offence if she partook of their meat in these circumstances,” and would not permit her father to kill them.124

It should finally be noticed that, though the custom of fasting after a death in the main has a superstitious origin, there may at the same time be a physiological motive for it.125 Even the rudest savage feels afflicted at the death of a friend, and grief is accompanied by a loss of appetite. This natural disinclination to partake of food may, combined with superstitious fear, have given rise to prohibitory rules, nay, may even in the first instance have suggested the idea that there is danger in taking food. The mourning observances so commonly coincide with the natural expressions of sorrow, that we are almost bound to assume the existence of some connection between them, even though in their developed forms the superstitious motive be the most prominent.

An important survival of the mourning fast is the Lent fast. It originally lasted for forty hours only, that is, the time when Christ lay in the grave.126 Irenaeus speaks of the fast of forty hours before Easter,127 and Tertullian, when a Montanist disputing against the Catholics, says that the only legitimate days for Christian fasting were those in which the Bridegroom was taken away.128 Subsequently, however, the forty hours were extended to forty days, in imitation of the forty days’ fasts of Moses, Elijah, and Christ.129

Not only on a death, but on certain other occasions, food is supposed to pollute or injure him who partakes of it, and is therefore to be avoided. In Pfalz the people maintain that no food should be taken at an eclipse of the sun;130 and all over Germany there is a popular belief that anybody who eats during a thunderstorm will be struck by the lightning.131 When the Todas know that there is going to be an eclipse of the sun or the moon, they abstain from food.132 Among the Hindus, while an eclipse is going on, “drinking water, eating food, and all household business, as well as the worship of the gods, are all prohibited”; high-caste Hindus do not even eat food which has remained in the house during an eclipse, but give it away, and all earthen vessels in use in their houses at the time must be broken.133 Among the rules laid down for Snâtakas, that is, Brâhmanas who have completed their studentship, there is one which forbids them to eat, travel, and sleep during the twilight;134 and in one of the Zoroastrian Pahlavi texts it is said that “in the dark it is not allowable to eat food, for the demons and fiends seize upon one-third of the wisdom and glory of him who eats food in the dark.”135 Many Hindus who revere the sun do not break their fast in the morning till they catch a clear view of it, and do not eat at all on days when it is obscured by clouds136—a custom to which there is a parallel among some North American sun-worshippers, the Snanaimuq Indians belonging to the Coast Salish, who must not partake of any food until the sun is well up in the sky.137 Brahmins fast at the equinoxes, solstices, conjunctions of planets, and on the days of the new and full moon.138 The Buddhist Sabbath, or Uposatha, which, as we have noticed above, occurs on the day of full moon, on the day when there is no moon, and on the two days which are eighth from the full and new moon, is not only a day of rest, but has also from ancient times been a fast-day. He who keeps the Sabbath rigorously abstains from all food between sunrise and sunset, and, as no cooking must be done during the Uposatha, he prepares his evening meal in the early morning before the rise of the sun.139

Among the Jews there are many who abstain from food on the day of an eclipse of the moon, which they regard as an evil omen.140 We have also reason to believe that the Jews were once in the habit of observing the new moons and Sabbaths not only as days of rest, but as fast-days; and the Hebrew Sabbath, as we have seen, in all probability owes its origin to superstitious fear of the changes in the moon.141 Or how shall we explain the curious rule which forbids fasting on a new moon and on the seventh day,142 if not as a protest against a fast once in vogue among the Jews on these occasions, but afterwards regarded as an illegitimate rite?143 This theory is not new, for Hooker in his ‘Ecclesiastical Polity’ observes that “it may be a question, whether in some sort they did not always fast on the Sabbath.” He refers to a statement of Josephus, according to which the sixth hour “was wont on the Sabbath always to call them home unto meat,” and to certain pagan writers who upbraided them with fasting on that day.144 In Nehemiah there is an indication that it was a custom to fast on the first day of the seventh month,145 which is “holy unto the Lord”;146 and on the tenth day of the same month there was the great fast of atonement, combined with abstinence from every kind of work.147 I venture to think that all these fasts may be ultimately traced to a belief that the changes in the moon not only are unfavourable for work, but also make it dangerous to partake of food. The fact of the seventh day being a day of rest established the number seven as a sabbatical number. In the seventh month there are several days, besides Saturdays, which are to be observed as days of rest,148 and in the seventh year there shall be “a sabbath of rest unto the land.”149 In these Sabbatarian regulations the day of atonement plays a particularly prominent part. The severest punishment is prescribed for him who does not rest and fast on that day “from even unto even”;150 and it is on the same day that, after the lapse of seven times seven years, the trumpet of the jubilee shall be caused to sound throughout the land.151 Most of the rules concerning the day of atonement are undoubtedly post-exilic. But the fact that no other regular days of fasting but those mentioned by Zechariah are referred to by the prophets or in earlier books, hardly justifies the conclusion drawn by many scholars that no such fast existed. It is extremely probable that the fast of the tenth day of the seventh month as a fast of atonement is of a comparatively modern date; but it is perhaps not too bold to suggest that the idea of atonement is a later interpretation of a previously existing fast, which was originally observed for fear of the dangerous quality attributed to the number seven. Why this fast was enjoined on the tenth day of the seventh month remains obscure; but it seems that the order of the month was considered more important than that of the day. Nehemiah speaks of a fast which was kept on the twenty-fourth day of the seventh month.152