Chapter IV · THE RELATION OF SOUL AND BODY. (continued 3)

Now for a right appreciation of this passage it must be borne in mind that Plato introduces his old tradition à propos of unintentional homicide. The actual penalties therefore are of a milder nature than those with which we have hitherto been concerned. Indeed it is not the difference in the penalties which should cause any surprise, but rather that an unintentional act should be punished at all; and it would seem perhaps that in citing this doctrine Plato sought to justify himself in retaining a provision of Attic law which at first sight appeared unjust. In Athens[1124], we know, the involuntary homicide was required not only to undergo purification but to withdraw for a whole year from the country of the man whom he had slain. The hardship of this was manifest, and yet Plato acquiesced in the righteousness of it for the reason apparently that the year’s retirement[1125] was not a penalty imposed by the state, but a satisfaction which, according to religious tradition, the dead man demanded and might even himself enforce.

Plato in fact recognises no less frankly than others the personal activity of the slain man. He differs indeed in limiting the duration of that activity, when he says that the dead man’s anger is hot against the slayer only while his death is still recent, and when by the provisions of his law he implies that the victim’s desire for vengeance is fully satisfied by the slayer’s withdrawal for the space of one year. But this difference is completely explained by the fact that Plato introduces the tradition in connexion with unintentional homicide, whereas previously we have had it treated in relation to wilful murder. Reasonably enough the man who has been accidentally slain is represented as angry only for a time, while the victim of deliberate murder nourishes a wrath implacable. The one drives the author of his misfortune into exile for a year and then repents him of the evil; the other dogs his enemy with vengeance not only for a year but throughout his life and even after death; and indeed Plato himself, when he passes from the subject of involuntary homicide to that of deliberate murder, proves his recognition of this difference by his enactments; for, at any rate in the most heinous case, namely the murder of a near kinsman, he expressly states[1126] that the old principle ‘as a man hath done, so must he suffer’ admits of no abatement; the guilty man must die, and his body be left unburied.

But I must not yet enter upon a discussion of the actual punishments inflicted. Here I am only concerned to point out how completely Plato’s ‘old doctrine’ harmonises with that which we have learnt from other sources concerning the personal activity of the dead man. First we read that the dead man terrifies and confounds the slayer to the utmost of his power, with the aid of the slayer’s own conscience; and then again that his next of kin is under an obligation to obtain satisfaction for him, and is punished by him if he neglects that duty. Clearly the slayer’s own conscience is no more than an instrument—a somewhat ineffective instrument, one might think, in a case of unintentional homicide—and the next of kin is no more than a minister, both of them employed and directed by the dead man himself. He it is who exacts his own vengeance.

The other literary method of mitigating the crude popular belief in a bodily revenant hunting down his enemy was to treat the murderer’s punishment as the result of a curse. Such a curse was denoted usually by the word μήνιμα, which may perhaps be more exactly rendered by the phrase ‘a manifestation of wrath (μῆνις)’ on the part of some supernatural being[1127], whether a god or the departed spirit of a man; when once provoked by deadly sin such as the murder of a kinsman or refusal of burial, this curse was held to cleave to the tainted family from generation to generation.

In the case of blood-guilt, which we are at present considering, the curse, as was said above, was held either to work spontaneously or to be executed by some powers of the nether world. The former view is more rarely adopted, but is clearly enough indicated in one or two passages of ancient literature. Plato in the Phaedrus speaks of most grievous sicknesses and sufferings being produced in certain families as the consequence of ancient curses (παλαιῶν ἐκ μηνιμάτων)[1128]; and from the reminiscences and verbal echoes of Euripides’ Orestes which appear in the passage[1129] it is abundantly clear that the particular family which Plato had in mind was the blood-guilty house of Atreus. Here then there is no mention of any gods, no suggestion that the curse was executed by them or in the first instance proceeded from them. And the negative evidence of Plato’s silence concerning the gods is turned to certainty by the positive statement of Aeschylus that, if a son neglect the task of vengeance, ‘betwixt him and the gods’ altars standeth the unseen barrier of his father’s wrath[1130]’; for if, in the case of the kinsman who by neglecting the duty of vengeance has made himself a partaker in the guilt and pollution of the murderer, the Wrath (μῆνις) by which he is punished both proceeds from the dead man and, far from needing the gods’ furtherance in order to take effect, stands as it were on guard to hold the polluted man aloof from their altars, then surely the Wrath which pursues the murderer himself must emanate from the same source and possess the same spontaneous efficacy. The dead man himself then both launches the curse and controls its course; and probably it was in deference to this doctrine that Plato formulated his own law, that, even in the case of a father being killed by his own son, the dying man might with his last breath remit the curse which such a deed incurred and exempt his son from all except the purifications and the temporary retirement imposed in cases of involuntary homicide[1131].

But more frequently the execution of the curse is conceived to be the work of certain powers of the nether world. These powers however do not act on their own initiative; they are instigated to the task of vengeance by the murdered man himself. Here, no less than in the other renderings of the old tradition, the sufferer himself is the supreme avenger of his own sufferings. The most famous example of this conception is furnished by the plot of the Eumenides. The Furies are represented as the servants of Clytemnestra, faithful witnesses to her wrongs, exactors of blood for blood on her behalf[1132]. When they slumber and allow Orestes to escape the while, her ghost approaches them in no suppliant manner for all their godhead, but chides them and urges them afresh, like hounds, upon the quarry’s trail[1133]. And, most significant of all, there is one passage in which they say of themselves that the name whereby they are known in their home beneath the earth is the name of Curses (Ἀραί)[1134]; they are in fact the personification of those curses which a murdered man himself directs against his murderer. Nor is this notion confined to drama. Xenophon is little prone to poetic imaginings; yet he can find an argument for the immortality of the soul in what he considers an established fact of human experience, namely, that the spirits of those who have been unjustly slain inspire terrors in their murderers’ hearts and ‘send against them’ certain ‘avengers of blood’ (παλαμναίους ἐπιπέμπουσι[1135]). And elsewhere again and again we hear of the same avengers under a variety of names—μιάστορες, ἀλάστορες, προστρόπαιοι—names which will receive consideration later and by their very meaning and usage will confirm once more my contention that, by whatever instrument or agency the murder is represented as being avenged, ancient literature only departed from the primitive belief in bodily revenants executing their own vengeance at the one point at which the grossness of popular superstition must have offended educated sensibilities, and followed the old tradition as faithfully as might be in conceding to the dead man, if not bodily, yet personal, activity.

The same popular beliefs, mutatis mutandis, probably attached also to another class of revenants, dead men whose bodies had not received due burial. The necessary modifications of the superstition would be two in number. First, the anger of the dead man would not endure for ever, unless his body had been so treated that burial was no longer possible, but would cease with the performance of that which he returned to demand; and secondly, he would not be represented as using for his agent his next of kin, who in most cases of the kind would be the very person responsible to him for the neglect of burial. Literature therefore had here no choice of versions; the bodily re-appearance of the dead man was reckoned too gross an idea; the employment of his nearest kinsman to act on his behalf became in this case impossible; a curse was the only expedient. And this is the expedient which we actually find adopted. In the Iliad Hector adjures Achilles not to fulfil his threat of throwing his dead body to the dogs and to the fowls of the air, but to give him burial, ‘lest,’ he says, ‘I become a cause of the gods’ wrath against thee’—μή τοί τι θεῶν μήνιμα γένωμαι[1136]—and the self-same phrase is put into the mouth of Elpenor’s spirit in the Odyssey when he craves due burial of Odysseus[1137]. The same idea occurs once more in Pindar’s reference to Phrixus, who bade go unto the halls of Aeetes (for there in a strange land he had died, and had not received the burial-rites of his own country) and bring his spirit to rest, and whose bidding Jason is besought by Pelias to fulfil, for that ‘already doth old age wait upon me; but with thee the blossom of youth is but burgeoning, and thou canst put away the wrath of powers beneath[1138].’ In each of these passages then the actual enforcement of the dead man’s will is by means of a curse or ‘manifestation of wrath’—for the same word μήνιμα (or μῆνις) is used; in each case also, as it happens, the curse does not operate automatically but is executed by gods—the method preferred also, as we saw, in cases of blood-guilt; but here also, as there, the personal activity of the dead man is frankly acknowledged; the phrase of Homer ‘lest I become ...’ and that of Pindar ‘Phrixus doth bid ...’ clearly suggest that the gods were instigated to intervene by the sufferer himself.

The case then stands thus. We learnt in the last chapter that the unburied dead no less than the murdered were popularly believed to become revenants. We have since learnt that the murdered, in the capacity of revenants, were popularly believed to avenge their own wrongs with their own hands, but that ancient literature commonly presents a modified version of that belief according to which the personal activity indeed of the dead man is recognised, but the instrument of his vengeance is a curse executed by demonic agents. We find now that literature assigns also to the unburied dead the same personal activity in punishing those whose neglect has caused their suffering, and by the same means. The reasonable inference is that here too we have a modified version of a popular belief that the unburied, like the murdered, not only became revenants, which we know, but, in the capacity of revenants, themselves punished those who refused or neglected to render them their due funeral rites.

Thus the same principle governed the whole system of the punishment incurred by men who were guilty either of murder or of leaving the dead unburied—the principle that the dead man whom they had injured in either of these ways himself requited those injuries. Hence, when we proceed to examine the actual punishments inflicted, we need no longer concern ourselves with the fact that literature attributes the infliction now to the nearest kinsman of the dead man and anon to some divine avenger; but, whatsoever instrument or agency is employed, we know that the dead man himself was believed to control and direct it, and therefore that the punishment thus effected was conceived to be such as the dead man himself willed and, in popular belief, could with his own hands enforce. Thus in the Oresteia the punishment of Clytemnestra is actually effected by Orestes, and again the punishment of Orestes is entrusted to the Furies; but Orestes is only the minister of his dead father, carrying out the work of retribution under pain of incurring the same punishment himself if by inaction he should consent unto his mother’s crime; and the Furies in like manner are only the servants of the dead Clytemnestra, instigated by her to their pursuit. The slaying of Clytemnestra and the sufferings of Orestes are the punishments which the dead Agamemnon and the dead Clytemnestra, even in the literary version of the story, impose, and, in a more primitive and gross form of it, might themselves have inflicted.

But before examining the nature of those punishments in detail, it will be well to recall the fact that to the eyes of the ancient Greeks murder or homicide always presented itself in two distinct aspects[1139]. Regarded from one point of view, it was the gravest possible injury to the man who was slain. Viewed from the other, it was a source of ‘pollution’ (μίασμα, μύσος, ἅγος), an abomination to the gods and a peril to living men; for the taint of bloodshed was conceived as a contagious physical malady, which the polluted person by touch or even by speech[1140] might communicate to his fellow-men, and not to them only, but to places which he visited, the market, the harbours, the temples[1141]; nay, even the sanctity of the gods’ images was not proof against the contamination of his bloodstained hands[1142]. In brief, the two aspects of homicide were the moral and the religious aspects; and both moral and religious atonements were required. The wrong done to the dead man was requited by the sufferings which he in turn imposed; the pollution, being primarily a state of religious disability (for it involved, as Plato says[1143], the enmity of the gods), was removed by a religious ceremony of purification.

How clearly marked was this distinction in antiquity is evident from Plato’s laws on homicide, as a brief consideration of two or three special cases will show.

First, in the most venial case of homicide, where a man had killed his own slave, he incurred no punishment at all, but was bound none the less to get himself purified[1144].

Secondly, in cases of the utmost enormity, as where a man wilfully murdered his father or mother, religion provided no means of purification. Blood-guilt in general was ‘hard to cure’; but parricide belonged to the class of sins ‘incurable[1145].’ Such a murderer therefore must die, for, as Plato says, ‘there is no other kind of purification’ in this case than the paying of blood for blood. Religious purification in the ordinary sense of the word was refused, but the extreme punishment was demanded.

Thirdly, in the majority of cases of blood-guilt, where both purification and punishment were required, the two were clearly independent of each other. The purification of the involuntary homicide was to precede the year’s retirement[1146]. The religious ceremony cleansed the man from pollution, but could no more exempt him from making satisfaction to the dead man whom he had wronged, than absolution of sin pronounced in the Christian confessional can exempt from the legal consequences of crime. The Delphic priesthood itself, if we may trust the testimony of Aeschylus, claimed no more than the power to cleanse; for Apollo himself, holding Orestes guilty of manslaughter though not of murder, after granting him religious purification, does not intervene to save him from that exile which even the unintentional homicide was bidden by Attic law to undergo; nay, he even acquiesces in the necessity of Orestes’ flight, bids him not faint before his wanderings are done, and promises only to set a limit thereto and to free him from the pursuing Furies in the end[1147].

The distinction between the pollution and the injury, and between the purification and the punishment, being thus clearly recognised, it is necessary, in investigating the relations between the dead man and his murderer, to set the purely religious aspect of blood-guilt on one side, and to treat the punishments inflicted upon the murderer simply as the settling of an account between man and man. One point only as regards the pollution need be borne in mind, namely, that purification was granted to the homicide in the interests of gods and men whose abodes would otherwise be defiled by his presence, and that the dead man could not conceivably derive any satisfaction therefrom. On the contrary, his desire for vengeance would naturally lead him to interpose ‘the unseen barrier of his wrath’ betwixt the guilty man and those altars of the gods where alone purification could be won, and thus to keep his enemy still polluted; for his pollution, just because it was a peril to his fellowmen, carried with it the punishment of utter solitude until he was cleansed. When therefore, as will appear later, the murdered man is described not only as an avenger of his own wrongs, but as one who strives to keep alive the religious defilement of the murderer, there is no confusion of the moral and the religious aspects of murder, but rather the injured man is conceived as wreaking his vengeance by every possible means, not only directly by the sufferings which he can personally inflict, but also indirectly by the privation which the state of pollution necessarily involves.

The nature of the direct acts of vengeance, which are now to be examined, can best be learnt from that passage of the Choephori which depicts the horrible penalties awaiting Orestes if by inaction he should make himself a consenter to the crime of Clytemnestra. We have already learnt that in such a case the defaulting kinsman incurred precisely the same punishment as he should have assisted to inflict on the actual murderer. That therefore with which Orestes was threatened was that to which Clytemnestra was already condemned. The punishments named are those with which, according to popular superstition, a murdered man, risen in bodily substance from the grave, could requite his enemy. For no one, I suppose, would suggest that Aeschylus, who followed popular tradition so scrupulously in all that did not absolutely conflict with dramatic propriety, invented for himself the whole scheme of penalties here set forth. That he was bound to modify the means whereby the punishments were inflicted, in order to avoid the incongruity of a revenant upon the stage, we already know and shall see again; but how closely he adhered to the popularly accepted scheme of punishments, even when he was forced to find some new means of inflicting them, will incidentally be shown by that detailed examination to which his list of penalties must now be subjected.

The first penalty is the physical torment of leprous blains that consume the body and age the sufferer prematurely. At first we are inclined to wonder why leprosy is selected by the dead man as his means of retaliation against his enemy; but a little reflection will lead us to guess that in this particular act of vengeance Aeschylus could not actually reproduce the popular doctrine. The common-folk believed in the bodily activity of the dead; and, if they believed also that bodily sufferings were part of the punishment which the murderer incurred, the two beliefs must surely have been correlated; the physical sufferings of the murderer must have been conceived to be caused by the physical activity of the murdered; or, to put it more plainly, if we may elucidate ancient superstition by the aid of modern, the murdered man, in the form of a revenant bent on vengeance, was believed to leap upon his victim and rend him with his teeth and suck out his very life-blood. Clearly Aeschylus could not commit himself to so crude a presentation of a revenant; he could not conjure up before his audience the spectacle of the dead Agamemnon athirst for actual blood; but equally clearly he knew that popular superstition, and had it in his mind when he depicted the horrors of leprosy. For the bodily assault of a revenant he substituted a natural malady engendered by a dead man’s unseen wrath; but he described the operation of that malady in language suggested by the popular presentment of a personal avenger more reasonable indeed in his purpose but scarcely less ferocious in his acts than a Slavonic vampire—‘blains that leap upon the flesh and with savage jaws eat out its erstwhile vigour[1148].’ The means of inflicting the punishment is changed, but the actual punishment of the murderer is the same as if it were not leprosy but in very truth a vampire, which leapt upon him and gnawed his flesh and drained his life-blood. So faithful is Aeschylus to the crude popular idea of a retribution which required that he who had spilled another’s blood should have his own blood drunk by his victim.

The second penalty is the mental agony of one whom ‘madness and vain terror sprung of the darkness do shake and confound[1149].’ Here again the punishment is in strict accord with that law that a man must suffer as he has wrought. That old tradition recorded and revered by Plato, on which I have already touched, taught that every man who was slain by violence was himself filled thereby with quaking and terror and confusion of spirit, and accordingly sought his revenge in terrifying and confounding the slayer. No clearer commentary on the lines of Aeschylus could be desired. Plato explains how the terror and the confusion—for he employs the selfsame words as Aeschylus—by which the murderer is overwhelmed are the exact counterpart of the mental anguish which his violence brought upon his victim. Aeschylus then once again was following closely an old tradition of the popular religion. It matters not at all that in this case he names the Erinyes as the agents, just as previously he made leprosy the instrument, of the dead man’s vengeance. The actual sufferings which the murderer must undergo are in this case also identical in character with those which he caused to his victim.

The third punishment of the blood-guilty man consists in wandering friendless and outcast; and this again is no arbitrary invention of Aeschylus, but was clearly prescribed by that old tradition which, in Plato’s reckoning, justified the legal imposition of a year’s retirement even upon those who had shed blood involuntarily. Where then is that correspondence, which our examination of the first two penalties has led us to expect, between this third punishment and the sufferings of the dead man who exacts it? Is there the same nicety of retribution? Clearly so. The dead man became in popular belief a revenant, a wanderer from out the grave, pitiable in his loneliness, cut off from all friendly intercourse with living men, not yet admitted to the fellowship of the departed, the sorriest of outcasts. Such was the misery to which the murderer by his act of violence had brought his victim; such therefore too the misery which the murderer himself must taste in his wanderings and loneliness here on earth, though it were but a foretaste of more consummate misery hereafter. Truly even in life the murderer was made to suffer as he had wrought.

And then comes the fourth penalty, death; for though Aeschylus, in the list of punishments which we have now before us, touches but lightly on this, the most obvious form of retribution, yet elsewhere he repeatedly affirms, and many another re-echoes, the doctrine that blood cries for blood[1150]. Perhaps in this passage he felt that by depicting the gnawing pangs of leprosy he had sufficiently proclaimed the sure approach of death; perhaps he passed it by as a slight thing in comparison with the horror that yet remained to be told. For death did not close the tale of punishments; the blood-guilty man, so chant the Furies, ‘though he be dead is none too free[1151].’

And so we pass to the last requirement of vengeance, that the outcast shall have no friend to honour his dead body with the due funeral-rites, whereby alone the desired dissolution could be secured, but is doomed to lie unburied, incorruptible. Such is my interpretation of the closing lines of the passage before us; and there is no need to repeat the defence of my contention that the word ταριχευθέντα must be understood in its literal and proper sense. But it will not be out of place to note here how, in the Eumenides, Aeschylus’ mind was still pervaded by the same popular belief. The word ταριχεύεσθαι means, in the literal sense in which I have taken it, to be withheld from corruption by some process of curing or drying; and, fantastic though it may seem, it is that process of ‘drying,’ if I may use the word, which the Furies are charged by Clytemnestra to carry out against her murderer. Let Aeschylus’ own words prove it. Hear first how Clytemnestra’s ghost with her last words spurs on the Furies to this special task:

And the Furies prove by their threats to Orestes that they are not unmindful of their charge. ‘Nay, in return for the blood thou hast shed, thou must give me to suck the red juices from thy living limbs. Thyself must be my meat, my horrid drink.’ ‘Yea, while thou livest, I will drain thee dry, ere I hale thee ’neath the earth[1153].’ And the same thought is emphasized yet again in that binding-spell which the Furies chant to draw him whom they already account their prey from his vain refuge at Athene’s altar:

Such is the wild, weird refrain of the Furies’ incantation; and in its closing phrase are re-echoed the closing words of Clytemnestra’s charge.

Will anyone then venture to say that Aeschylus had no special reason for thus repeating thrice within the compass of some two hundred lines the same threat? For the punishment threatened is substantially the same, though the means of inflicting it vary. Now it is the breath of the Furies which shall scorch up the victim’s very blood; now it is their lips that shall suck him dry; now a magic spell to parch and shrivel him; but ever the effect is the same; the bloodguilty man shall lie in death a sere and sapless carcase, already ‘damned to incorruption[1156] even in that doom which wastes all else.’ And the only reason which I can conceive for the poet’s insistence upon this thought is that here again, as in all the former punishments, he was reproducing a popular belief substantially the same then as it is in Maina now, namely, that the murdered man, having become a revenant, sucked his murderer’s blood and made him also in his turn a revenant.

Nor is Aeschylus the only ancient authority for the idea of some such retribution after death. Plato, in a passage of the Phaedrus already cited, contemplates the activity of a murdered man’s wrath (μήνιμα) not only in the present time but also hereafter[1157]; and in his Laws there is a provision, not assuredly of his own devising but dating from the very beginning of Greek legislation, which can only have been designed to insure the complete vengeance of the murdered man on his murderer even beyond death. A man convicted of the wilful murder of a near kinsman[1158] was punishable not only with death but with a further penalty: ‘the attendants of the jury and the magistrates having killed him shall cast out his corpse naked at an appointed cross-roads without the city, and all the magistrates, representing the whole city, shall take each a stone and cast it upon the head of the corpse and thereby free the whole city from guilt, and thereafter they shall carry the corpse to the borders of their land and cast it out, in accordance with the law, unburied[1159].’ Now the law, we know, in ordaining the penalty of death, ordained it as a satisfaction of the murdered man’s claims to vengeance. The State, so to speak, sided with the dead man and assisted him to exact blood for blood. Again the stoning of the dead body by representatives of the city was intended, we are expressly told, to free the whole city from guilt—from guilt, that is, in the eyes of the murdered man, who might otherwise visit his wrath upon the city as though it had consented to the crime or had too lightly punished it. Can it then be supposed that the State was actuated by any other motive in carrying out the rest of the penalty? It was surely still in deference to the murdered man’s desires that the murderer’s corpse was left unburied. To refuse burial was the surest means of condemning the man to resuscitation and thereby of satisfying his former victim’s uttermost demands.

Thus our detailed examination of the Aeschylean catalogue of penalties establishes beyond doubt that of which we had already had some evidence, namely, that all the punishments which were inflicted on the murderer—and, in popular belief, inflicted by the murdered man on his own behalf—were an exact reproduction of the sufferings which the murdered man himself had undeservingly endured, and culminated therefore, as they should, in the blood-guilty man becoming, like his victim, a revenant.

The main problem then of this section is now fully solved; but incidentally much light has been thrown upon the character ascribed by the Greek people in antiquity to those revenants who were not merely pitiable sufferers but were active in bringing a like doom upon those who had wronged them. And the character of these Avengers approximates very closely to that of the modern vrykolakes. True, there is one fundamental difference; the ancient Avenger directed his wrath solely against the author of his sufferings, or at the most extended it only to those who, owing to him the duty of furthering his vengeance, had proved lax and cowardly therein; the modern vrykolakas is unreasoning in his wrath and plagues indiscriminately all who fall in his way. But the actual sufferings which the vrykolakas inflicts are identical with those which furnished Aeschylus with his tale of threatened horrors. Modern stories there are in plenty, which tell how the vrykolakas springs upon his victim and rends him and drinks his blood; how sheer terror of his aspect has driven men mad; how, in order to escape him, whole families have been driven forth from their native island to wander in exile[1160]; how death has often been the issue of his assaults; and how those whom a vrykolakas has slain become themselves vrykolakes. Only his unreasoning and indiscriminate fury is necessarily of Slavonic origin; his acts are the acts of those ancient revenants whose own wrongs rightfully made them the Avengers of blood. Apart from the one Slavonic trait, the characters of the vrykolakas and the ancient Avenger are identical.

And perhaps this identity is most clearly seen in the one case in which the old Avenger punished not only the immediate author of his own wrongs, but a whole community which had subsequently given the guilty man an asylum. We have noticed how Antiphon ventured to threaten an Athenian jury with such punishment at the hands of the dead man if they wrongfully acquitted his murderer. In the same spirit Aeschylus makes the Furies, as the agents of the dead Clytemnestra, menace the whole land of Attica with a venomous curse that shall blast man and beast and herb in revenge for the wresting of Orestes from their grasp[1161]. And such too is the dread which in the Phoenissae of Euripides stirs Creon to make to the blood-guilty Oedipus this appeal: ‘Nay, remove thee hence: verily ’tis not in scorn that I say this, nor in enmity to thee, but because of thine Avengers, in fear lest the land suffer some hurt[1162].’ In such cases the punishments with which a whole community is threatened, although still a reasonable measure, approach most nearly to the indiscriminate violence of the modern vrykolakas.

For the fulfilment of such threats as these we must turn to the Supplices of Aeschylus, and there we shall find a description of just such a devastation as is said to have been suffered by the inhabitants of Santorini and many other places in the seventeenth century. The story of Aeschylus tells how ‘there came unto the Argive land, from the shore of Naupactus, Apis, son of Apollo, both healer and seer, and cleansed the land of monsters that destroyed mankind, even of those that Earth, tainted with the pollutions of blood shed of old, sent up in wrath to work havoc, fearsome as a dragon-brood to dwell among[1163].’ What then were these monsters? I will venture to say that any Greek peasant of to-day, could he but read and understand the Aeschylean description, would furnish a better commentary upon those lines than the most learned discourse thereon that any scholar has written; and his commentary would be summed up in the one word vrykolakes. For, vigorous as the description is, its vigour comes less of dramatic word-building than of fidelity to the horrors of popular superstition, and no other single passage could so fully establish the unity of ancient and modern belief. For while the actual language contains all the words[1164] which in antiquity were bound up with the superstition—the ‘pollution’ which comes of bloodshed, the ‘wrath’ which follows thereon and in which Earth herself is here made to share, and the ‘sending up’ by Earth of the Avengers—the thought of the passage is a faithful reflection of what the Greek peasants still believe, that a violent death is among the chief causes of resuscitation, that the earth sends up the dead man raging to deal destruction, and that with others of his kind he consorts and conspires in veritable dragon-bands; and men still tell of gifted seers and healers, such as Apis, summoned in hot haste to panic-stricken hamlets to allay the pest. The κνώδαλα βροτοφθόρα of Aeschylus, ‘the monsters that destroy mankind,’ are indeed but little removed from the modern vrykolakes.

Is it not then clear also on what sources Aeschylus drew for his picture of the Furies themselves? We have seen how, for dramatic purposes, they were substituted for a revenant wreaking his own vengeance. Clytemnestra herself in bodily form should have been the Avenger, if popular superstition had not been in this respect too gross; but the Erinyes take her place in the actual execution of vengeance, and she herself appears only as a ghost to instigate them to their work. But, when that substitution was effected, did not Aeschylus clearly transfer to the Erinyes the whole character and even the appearance popularly attributed to the human Avenger? They are black and loathly to look upon[1165]; their breath is deadly to approach[1166]; the smell of blood is a joy to them[1167]; they follow like hounds upon their victim’s trail[1168]; they torment him both body and soul[1169]; they fasten upon his living limbs and gorge themselves with his blood[1170]; and if any would harbour him from their pursuit, the venom of their wrath falls like a plague upon the land, and devastates it[1171]; they are monsters, κνώδαλα[1172]—and the recurrence of this word is significant—abhorrent alike to gods and to men[1173]. The description is surely not that which Aeschylus would himself have invented for beings who should come afterwards to be worshipped as ‘revered goddesses,’ σεμναὶ θεαί. The difficulty of that transition in the play itself cannot but arrest the attention of every reader; it is a difficulty which even the genius of Aeschylus could not remove. Why then did he draw so loathsome a portrait of the Erinyes in the earlier part of the play? Why did he create that difficulty? The reason, I suggest, was that he followed once more, and this time almost too faithfully, the popular traditions, and, while he would not represent a real revenant on the stage, transferred to those demonic agents, by whom the work of vengeance was vicariously performed, all the attributes popularly associated with the prototypes of the modern vrykolakas.

Thus then the history of the modern belief in vrykolakes has been fully traced. The ancients also believed that for certain causes—the same causes in the main as are still assigned—men were doomed to remain incorruptible after death and to rise again in bodily form from their graves, and that one class of these revenants, those namely who had wrongs of their own to avenge, inflicted upon their enemies (and upon any who shielded or harboured them) the same sufferings as are now generally believed to be inflicted in an unreasoning manner by all classes of vrykolakes alike upon mankind at large, with no justification, such as a natural desire for vengeance might afford, in the case of those whose resuscitation is not the outcome of any injury or neglect at the hands of other men, and with no discrimination between friend and foe on the part of those who have real wrongs to avenge. Remove the unreasoning element in the character of the vrykolakas, and the revenant in which the folk of ancient Greece believed remains.

But, if they believed in him, they must have called him by some name. Aeschylus’ phrase κνώδαλα βροτοφθόρα, ‘monsters that destroy mankind,’ is a description rather than a name. What were the reasonable vrykolakes of ancient Greece called? That is now the one question which must be answered in order to make our enquiry complete.

Briefly my answer is this, that the particular class of revenants with which the present section has mainly dealt, the Avengers of blood, were known by three several names, μιάστωρ, ἀλάστωρ, and προστρόπαιος, but that literature contains no word which could serve as a collective designation for all classes alike. I hope however to show that the Greek language was not originally defective in this respect, but that the term ἀλάστωρ, although regularly used from the fifth century onwards in the narrow sense of an Avenger, had originally a wider application and denoted simply a revenant.

Now the interpretation which I give to these three words is not that which is commonly accepted. Anyone who will turn to a lexicon will find that to each of the three is assigned a double signification in connexion with blood-guilt. All three are said to denote either a god who punishes the blood-guilty or the blood-guilty man who is punished. Thus a god, it is alleged, may be called μιάστωρ (literally a ‘polluter’) because he punishes the polluted—a somewhat obvious misnomer; or again ἀλάστωρ, because he ‘does not forget’ but punishes the sinner—a derivation which, as I shall show later, cannot be accepted; or thirdly προστρόπαιος, as the being who was ‘turned to’ by the murdered man and was besought to avenge his cause—a somewhat circuitous way for the word to arrive at its active sense of ‘Avenger.’ And, secondly, a man, it is said, was called μιάστωρ when, being himself polluted, he was liable to be ‘a polluter’ of other men with whom he came in contact—a view which is certainly defensible; ἀλάστωρ as one whose sin ‘could not be forgotten’—an interpretation almost beyond the pale of serious discussion; and προστρόπαιος because, being blood-guilty, he ‘turned towards’ some god for purification—an explanation which may be right—whence the word came to denote in general a polluted person who still needed purification.

Thus in my view, as I have indicated, the greater part of the information in the lexicons with regard to these three words is inaccurate; and my reasons for disputing the received interpretations will be set forth point by point as I offer my own interpretations in their stead.

In dealing with the first group of meanings assigned to the three words, by which they came, somehow or other, to be used with the common active signification of ‘Avenger,’ my main contention will be that, as regards their primary and strictest usage, all three words were applied not to gods but to men—men who, having been murdered, sought to requite their murderers—and were only secondarily extended to the agents, whether divine or human, to whom those dead men committed the task of vengeance; but I shall also endeavour to show, as regards the literal meaning of the three words severally, that the interpretation by means of which their final sense of ‘Avenger’ has generally been elicited from them is in each case wrong, and that, in the case of the word ἀλάστωρ in particular, a right understanding of its original meaning gives very important results.

And in dealing with the second group of meanings, by which the three words are said to denote three only slightly different aspects of one and the same person—a murderer who is μιάστωρ as polluted and spreading pollution, ἀλάστωρ as pursued by vengeance, and προστρόπαιος as still needing purification—I shall maintain that these alleged uses of the first two words do not exist, and, as regards the third, I will offer a suggestion, but a suggestion only, as to the means by which it acquired this signification which it unquestionably bore.

It will be convenient to deal first with μιάστωρ and ἀλάστωρ as being parallel in usage throughout, and to reserve προστρόπαιος for later consideration.

The clearest example of that which I take to be the original usage of μιάστωρ is furnished by Euripides. In that scene of mutual recrimination between Medea and Jason, after that in revenge for her husband’s faithlessness she has slain their children, there comes at last from her lips the brutal taunt, as she points to the dead, ‘They live no more: that truth at least will sting thee’; and Jason answers, ‘Nay, but they live, to wreak vengeance on thy head (σῷ κάρᾳ μιάστορες)[1174].’ No language could be more simple, more explicit. The very children who lay there murdered at Medea’s feet, they and none other should be the Miastores, the Avengers of their own foul deaths.

But of course the word has other applications also. When Aeschylus[1175] made the Erinyes threaten that even when Orestes should have fled beneath the earth, he should find another Avenger (μιάστορα) to plague him in their stead, the whole tenor of the passage compels us to understand that that other Avenger is some deity or demon of the nether world—a divine, not a human, Miastor, though at the same time one who will act, like the Erinyes themselves, on behalf of the murdered Clytemnestra.

And, yet again, the same term is applied to a living man, when, as next of kin to him who has been murdered, he is in duty bound to exact vengeance. This time Sophocles is our authority, and the person of whom the word is used is Orestes. ‘Oft,’ says Electra to Clytemnestra, ‘oft hast thou reproached me with saving him to take vengeance upon thee (σοὶ τρέφειν μιάστορα)[1176].’

These three passages then illustrate the threefold application of the name Miastor, and the question to be answered is which represents the primary usage of the word. To multiply instances of each or any would be of no avail; the question is not of the frequency of each usage; the commonest is not necessarily the earliest. How then is the question to be answered? It is, I think, already answered. We have seen that in popular belief the murdered man was the prime avenger of his own wrongs, and that even in literature, when the execution of vengeance is wholly transferred either to the nearest kinsman or to some demonic power, the murdered man is still recognised as the principal and the others are only his agents. It is this relation between them which settles the question. A principal does not act in the name of his agents, but the agents in the name of their principal. The name Miastor therefore belonged first to the dead man himself, and was only extended afterwards to those who wrought vengeance on his behalf.

So much for the usage of the word. Next, how did it acquire the meaning of ‘Avenger,’ which it undoubtedly possessed? This can be only a matter of opinion. But since it appears to me unscholarly and illogical to suppose that a word, which on the grounds of formation must have first meant ‘one who causes pollution,’ could have come to mean ‘one who punishes pollution,’ I may at least offer an alternative suggestion. The murdered man, I admit, can hardly be said to have ‘caused’ the pollution of his murderer, or at any rate he could only have caused it involuntarily. But he might well be regarded as active in debarring the murderer from the means of purification and in keeping the pollution, as it were, fresh and virulent, with intent to isolate his enemy and to ban him from the abodes of his fellow-men. And some indication of such an activity is afforded by the Erinyes—acting, as always, on Clytemnestra’s behalf; they refuse to acknowledge the purification granted by Apollo to Orestes, and they say moreover that their task is to ‘keep dark and fresh the stain of blood[1177].’ The murdered man may therefore have been believed, if not actually to cause and to create, yet at least to promote and to re-create, the pollution of his foe, and, by keeping the stains of blood as it were from fading or being cleansed away, to wreak some part of his vengeance. In this way the transition from the sense of ‘polluter’ to that of ‘avenger’ is at least, I submit, intelligible. This however is only a side-issue. The important point is that the word Miastor, however it may have come to mean ‘Avenger,’ was primarily applied to the revenant himself, and only secondarily to any god.

The next name to be considered, ἀλάστωρ, is commonly accounted a synonym of μιάστωρ, denoting in actual usage a ‘god of vengeance,’ and meaning literally ‘one who does not forget’ blood-guiltiness. I too hold it to be a synonym of Miastor, but to denote therefore primarily not a god but a human revenant seeking vengeance, and only afterwards, by a transference of usage, a god or living man acting in the name of the dead; while, as for the supposed derivation, I count it absolutely untenable.

And first as regards the application of the word; after what has been, I hope, a fairly exhaustive study of the passages of classical literature in which it occurs, I am bound to confess that, though the instances of its use are far more numerous than those of Miastor, I am still unable to select three passages and to say ‘Here are my proofs of the triple application of the word.’ Indeed all that I can prove by the evidence of any single passage taken alone is curiously enough the existence of what I take to have been the rarest of the three usages—the application of the name Alastor to the kinsman of the dead man, as being the agent of his vengeance. Just as Sophocles speaks of Orestes being preserved as a Miastor to take vengeance on Clytemnestra for his father’s death, so does Aeschylus make the same Orestes name himself an Alastor on the score of the vengeance which he has taken. ‘Queen Athene,’ he prays, ‘at Loxias’ bidding am I come; receive thou me graciously, avenger as I am, no murderer, nor of defiled hand ... ἀλάστορα, οὐ προστρόπαιον, οὐδ’ ἀφοίβαντον χέρα[1178].’ Such, I am convinced, is the right rendering of the passage. The lexicons indeed cite the line as an example of the alleged passive meaning of ἀλάστωρ—one who suffers from divine vengeance, an accursed wretch[1179]; and I acknowledge that such a meaning would make passable sense of the passage; for Orestes was indeed suffering from the vengeance of the Erinyes. But I hold, and I shall endeavour to prove later, that ἀλάστωρ never possessed a passive meaning, and I claim moreover that the active meaning of ‘Avenger,’ which I attribute to the word here as elsewhere, is immensely preferable in itself. For Orestes throughout pleads justification[1180]; he has avenged murder, not committed it; he has discharged a duty to his dead sire, not perpetrated a wanton crime against his mother; he slew her indeed, but his motive was pious, and the ordaining of his act divine. On the grounds therefore, first, of the word’s own active meaning, secondly, of the whole trend of Orestes’ defence of his conduct, and last, but by no means least, of the exact parallel furnished by Sophocles’ use of the word Miastor, I am confident that Alastor as applied by Orestes to himself means an ‘Avenger.’

That the word however was not primarily applied to the kinsman acting on behalf of the murdered man will be universally conceded; in the vast majority of passages some supernatural being is clearly intended. But it has been too hastily assumed that the supernatural avengers were always gods or demons; that they were often so conceived I do not doubt; but, as a matter of fact, I have discovered no single passage of classical literature which can be said finally and absolutely in itself to demand that interpretation. In many instances the probabilities are in favour of the Alastores being regarded as a class of avenging demons; in many others it is equally good or even better to suppose that they are the dead men themselves in person.

What then are the foundations upon which the received notion, that the Alastores were always gods, is based? It might perhaps be urged that the word Alastor found a place among the many epithets and titles conferred by worshippers upon Zeus[1181] in order to indicate the particular exercise of his all-reaching power which their hearts desired. It might also be urged that Clement of Alexandria names the Alastores among those classes of gods whom the pagan Greeks had evolved from the naughtiness of their own imagination as types and personifications of the baser human passions[1182]. But neither of these facts can serve to substantiate the contention that the Alastores were primarily and necessarily gods. The occasional use of a word as an epithet of Zeus cannot be held to prove the general appropriation of that word to a class of lesser gods; while the statement of Clement is the statement of a man designedly vilifying the whole Greek religion, neither appreciating nor desirous to appreciate its refinements, but willing rather to overwhelm it utterly, its better and its worse elements alike, with the torrent of his invective and reprobation. To him the Alastores appeared as supernatural beings instinct with the pagan passion of revenge, false gods therefore or devils, fit objects whereon to pour out the vials of righteous wrath and Christian scorn. He was not concerned to be wholly just or wholly accurate. Indeed the very sources from which he drew the idea that the Alastores were gods are still open to us; it is the Greek Tragedians whom he holds guilty of this naughty invention; it is the Greek Tragedians who remain for us the fountain-head of information concerning these Avengers, and who will on examination make it clear that they were not primarily or necessarily gods.

The single passage in Greek Tragedy which has been often regarded as evidence in favour of Clement’s classification of Alastores among gods is on fuller enquiry rather a refutation of that view. In the Persae of Aeschylus the messenger, who reports to the queen the disaster which has befallen the Persian fleet, sets it down to supernatural agency:

This has generally been taken to mean that the beginning of the disaster was due to the sudden appearance of ‘some vengeful or malicious deity.’ But elsewhere in Tragedy ἀλάστωρ is treated not as adjective but as substantive; and, since there is no compulsion to suppose other than the ordinary use of the word here, it appears better to translate the phrase ‘some Avenger or some malicious god.’ In other words the real, if unemphatic, contrast implied in the phrase is not between ἀλάστωρ and κακός—no contrast is possible there[1184]—but between ἀλάστωρ and δαίμων. The inference therefore is rather that the Alastor in this passage was not conceived as a deity.

There are other passages of Greek Tragedy also in which the balance of probability seems to me to incline towards interpreting the name Alastor in the sense of a revenant and not of a god. Two such occur in the Medea of Euripides—the same play, be it noted, which contains that perfectly plain statement that the dead children of Medea are themselves the Miastores who will punish her. The first is in the scene in which Medea works herself up to the perpetration of her crime. Passionate love of her children, passionate jealousy and fury against their father, alternate in tragic turmoil, until the tense agony of spirit is let loose in that fierce oath,

Strong and terrible would be the oath even if the Alastores, whose wrath Medea thus defies, were gods or spirits; but the force and the horror are doubled, if the Alastores here are of the same order as those whom Jason names Miastores but a little later in the same drama, and if therefore among those Avengers, in whose name the murderous oath was sworn, were soon to be numbered those very children whom Medea loved best and yet bound herself to slay most foully.

The second passage occurs in Jason’s outburst of fury against Medea when he first learns her crime. ‘’Tis thine Avenger whom the gods have let light on me; for truly thou didst slay thine own brother at his own hearth, or ever thou didst set foot in Argo’s shapely hull[1186].’ Surely we are meant to understand that the dead Absyrtus is himself the Alastor—for one Alastor only is named this time, and that too as distinct from the gods (θεοί)—and that Jason diverted to himself a portion of the dead man’s wrath by wedding the blood-guilty woman. Again then the interpretation of Alastor in the same sense in which, only a little later in the same scene, Miastor is undoubtedly employed is, if not necessary, yet vastly preferable.

To review here all the passages of Greek Tragedy in which the word may advantageously be so understood, when at the same time no single one of them constitutes a final proof of my view, would be to encumber this enquiry to no purpose; but I may perhaps be permitted to select one instance from a story of blood-guilt other than that of which Medea is the centre.

This shall be from that scene in the Hercules Furens in which the hero, sane now and overwhelmed with horror at the ghastly slaughter of his own children which in a moment of sudden madness he had wrought, receives from Theseus some measure of consolation and advice. Early in that colloquy, ere yet Theseus has had time to soothe the sufferings or to guide the course of his stricken friend, Heracles cries to him in bitterness of soul,

and Theseus answers with gentle simplicity,

And then follow the lines:

It is the connexion and significance of the last two lines which I wish briefly to discuss. Theseus has used the word ‘taint’ (μιαίνεις), and Heracles at once seizes on it, emphasizes it, and warns his friend to begone lest he be contaminated; and then Theseus answers (to give a literal rendering) ‘No Avenger of blood proceeds from them that love against them that love.’ What does this mean? The line is often translated as if Theseus meant, ‘No, I will stay, for though an Avenger of blood may probably pursue you, Heracles, I have no fear that he will touch me who love you as a friend[1188].’ A generous and sympathetic utterance indeed! And how consistent with that fine burst of feeling with which he had but a moment before refused to be warned away:

Is this the man whose words, spoken but a moment later, shall be interpreted to mean, ‘I will not run away, because the danger that threatens my friend cannot hurt me’? The thought is deeper, more generous, than that. Theseus is thinking not of himself, but of his friend. It is the word ‘pollution,’ used first by himself and caught up by Heracles, which arrests his attention. Was his friend ‘polluted’ by a deed of blood, wrought in madness, expiated in tears? Polluted? Yes, in the sense that religious purification was required[1190]. He cannot deny the pollution. But could the deed also be punished as the murder of close kinsfolk was wont to be punished? Could the children, albeit slain by their own father’s hand, desire revenge upon him who loved them and was loved of them? ‘No,’ he answers boldly, ‘pollution (μίασμα) there is, but no Alastor, no Avenger of blood, can come from them that love against them that love.’ How then does Theseus picture the Alastor who, but for the bond of love between the father and his dead children, would seek vengeance for their death? The phrase which he uses is ambiguous—perhaps deliberately ambiguous—οὐδεὶς ... ἐκ τῶν φίλων. It may mean equally well ‘no one of those who love’ or ‘no one coming from those who love.’ But when the close correspondence of μίασμα, ‘pollution,’ and ἀλάστωρ ‘avenger,’ is noted in this passage, and when it is also remembered that the dead children of Medea are elsewhere plainly named Miastores, it is hard to suppose that an audience familiar with the belief that the dead themselves avenged their own wrongs would not have interpreted the ambiguous phrase to mean ‘none of these children shall rise up from the grave as an Alastor, for love is stronger than vengeance.’

But such doubt as still remains is set at rest when we turn from the usage of the word Alastor to its origin and enquire how it obtained the sense of ‘Avenger.’ What is its derivation?

Two conjectures seem to have been made by the ancients and are recorded by early commentators and lexicographers[1191]. The one connects the word with the root of λανθάνω, ‘I escape notice,’ and extracts a meaning in a variety of ways, leaving it open to choice, for example, whether it shall mean a god whose notice nothing escapes or a man who commits acts which cannot escape some god’s notice. The other conjecture refers the word to the root of ἀλάομαι, ‘I wander.’ It is between these two proposed derivations that our choice lies; nor can we obtain much help from the greatest modern authorities. Curtius[1192] unhesitatingly adopts the latter, Brugmann[1193] the former, nor does either of them so much as mention the possibility of the alternative. I must therefore discuss the question without reference to these authorities, knowing that, if I run counter to the one, I have the countenance of the other.

Is then ἀλάστωρ, in the sense of a ‘non-forgetter,’ a possible formation from the root of λανθάνω? My own answer to that question is a decided negative, and my reasons are as follows. Substantives denoting the agent and formed with the suffix -τωρ (-τορ-) can only be so formed direct from a verb-stem, as ῥήτωρ from ϝρε or ϝερ appearing in ἐρῶ etc., μήστωρ from the stem of μήδομαι, ἀφήτωρ answering to the verb ἀφίημι, ἐπιβήτωρ to ἐπιβαίνω. It is among these and other such examples that Brugmann places the anomalous ἀλάστωρ, to be connected with ἄλαστος, λήθω. But evidently, in order that ἀλάστωρ may be parallel, let us say, to ἀφήτωρ, we must postulate the existence of an impossible verb ἀ-λήθω or ἀ-λανθάνομαι, ‘I non-forget.’ Nor would it mend matters to suppose, first, the formation, direct from λήθω, of a nomen agentis of the form λάστωρ, a ‘forgetter’; for the privative ἀ- appears only in adjectives and adverbs and in such verbs and substantives as are formed directly from them, as ἀμνημονεῖν from ἀμνήμων etc., and cannot be prefixed at pleasure to a substantive or verb not so formed; ἀλάστωρ could no more be formed from an hypothetical substantive λάστωρ[1194], than could an hypothetical verb ἀ-λανθάνεσθαι be formed from λανθάνεσθαι. Etymologically then the derivation of ἀλάστωρ from ἀ- privative and the root of λήθω is impossible, and its sense of ‘Avenger’ was not developed from the meaning ‘one who does not forget.’

On the other hand, to the connexion of ἀλάστωρ with the verb ἀλᾶσθαι, ‘to wander,’ no exception can be taken. Not only is the formation simple, but an exact parallel is forthcoming. As the substantive μιάστωρ stands to the verb μιαίνω, so does the substantive ἀλάστωρ stand to a by-form of ἀλάομαι, which is fairly frequent in Tragedy, ἀλαίνω[1195]. It follows then that ἀλάστωρ meant originally a ‘wanderer.’

But, when once that primary meaning is discovered, there can be no further doubt as to the primary application of the term. Of the three possible exactors of vengeance—the revenant himself, some demonic agent, and the nearest kinsman—the first alone could be aptly described as a ‘wanderer’; moreover we know that the murdered man was actually so conceived, and that, among the punishments by which he sought to make his murderer suffer the same lot as he himself endured, one of the most conspicuous was the punishment of wandering and exile. The name Alastor therefore, like Miastor, denoted first of all the dead man himself, and was only secondarily extended to human or divine agents seeking vengeance on his behalf.

It remains only to enquire how the meaning ‘Avenger’ was evolved from the meaning ‘Wanderer,’ and so completely superseded it that the name Alastores was extended to those agents who were in no obvious sense ‘Wanderers’ but simply ‘Avengers.’

The first occurrence of the word is in the Iliad, as the proper name of a Greek warrior[1196]. This fact tends to show that the word had as yet acquired none of that ill-omened sense which it undoubtedly bears in Greek Tragedy. It was used rather, we may believe, in its original and literal sense of ‘wanderer,’ and the adoption of such a word as a proper name is entirely consistent with the principles of Homeric nomenclature. Hector, Nestor, Mēstor, are famous names of the same class.

Otherwise than as a proper name the word is not used in Homer, nor does it occur at all again, so far as I am aware, before the time of Aeschylus. It is during this interval then that the evolution of meaning must have taken place; for by the age of Aeschylus the idea of vengeance—and vengeance of a horrible kind—had become the ordinary signification of the word. My view then is that the intervening centuries had witnessed a gradual differentiation of the several words which alike originally meant a ‘wanderer,’ a differentiation such that ἀλήτης remained the ordinary and general term, while ἀλάστωρ was little by little restricted to the wanderer from the dead, the revenant; and that subsequently from meaning a revenant of any and every kind it became limited to that single class of revenants whose wanderings were guided by the desire for revenge—the class to whom the name Miastores had always belonged.

Some evidence for the first stage in this development of meaning is furnished by the Tragic usage of the verb from which the substantive is derived; for in both its forms, ἀλᾶσθαι and ἀλαίνειν, it continued to be applied to any of the restless dead, when the substantive ἀλάστωρ, as I conceive, had come to be appropriated to the Avenger only. Indeed it might almost be thought that both Aeschylus and Euripides had an inkling of the derivation and earlier meaning of the substantive; for while idiom debarred them from using ἀλάστωρ in the large sense of any revenant, they certainly used the corresponding verb in contexts which suggest that those who thus ‘wander’ were not imagined by them as vague impalpable ghosts, but possessed for them rather the real substance and physical traits of a revenant. Thus in the Eumenides, though Clytemnestra could not be permitted to play the part of a revenant and appears only as a ghost, yet the more gross and popular conception of her is clearly present to the poet’s mind. Though a ghost, she points to the wounds which her son’s hands inflicted[1197]; though a ghost, she is made to exhort the Erinyes to vengeance ‘on behalf of her very soul’ (τῆς ἐμῆς πέρι ψυχῆς)[1198]. Strange gestures and strange language indeed, if the so-called ghost had been conceived as a mere disembodied soul! But the popular conception of the revenant penetrated even here. And was it not the same conception which suggested the phrase αἰσχρῶς ἀλῶμαι, ‘I wander in dishonour[1199]’? In the popular belief, as we know, the murderer was bound to wander after death, suffering as he had wrought; and it is as a murderess[1200] that Clytemnestra avows herself condemned to shameful wanderings. ‘To wander,’ ἀλᾶσθαι, sums up the suffering which the murderer, like his victim, must incur after death. It is likely then that the name ἀλάστωρ too was originally applied to any ‘wanderer’—whether murderer or murdered—before it acquired the connotation of vindictiveness and so became appropriated to the latter only.

Again Euripides uses the same verb of one whose body has not received burial. This time there is no connexion with blood-guilt at all, but the lines are simply the plaint of captive wife for husband slain in battle: ‘oh beloved, oh husband mine, dead art thou and wanderest unburied, unwatered with tears’—σὺ μὲν φθίμενος ἀλαίνεις, ἄθαπτος, ἄνυδρος[1201]. ‘To wander unburied’—could there be a simpler description of a revenant? Does not the whole misery of the unburied dead consist in this—that they must wander? It is almost inconceivable then that the name Alastor, ‘wanderer,’ should have been originally applied only to a single class of the wandering dead—to those whose wanderings were directed towards vengeance, and not also to those whose wanderings were more aimless, more pitiable, whose whole existence might have been summed up in that one word ‘wandering.’ At some time then between the age of Homer and that of Aeschylus Alastor, I hold, meant simply revenant.

How then shall we explain that caprice of language which, according to this Tragic usage, permitted all the unhappy dead to be said ‘to wander’ (ἀλᾶσθαι, ἀλαίνειν), but apparently forbade them to be collectively named ‘wanderers’ (ἀλάστορες)? How did Alastor acquire its sense of ‘Avenger’ and become restricted to one class of revenant only?

It might be sufficient answer to point out that those revenants who were bent on avenging their own wrongs are likely always to have occupied a prominent place in popular superstition simply because they inspired most terror in the popular mind; other revenants were harmless, and, as harmless, liable to be little regarded and seldom named; and the most conspicuous class might thus have appropriated to itself the name which properly belonged to all. But there is another influence which, if it did not cause, may at least have facilitated and quickened the change—the influence of the word ἄλαστος, ‘unforgotten,’ which, as I have noted above, was commonly and naturally, in an age when etymology was not science but guess-work, connected with ἀλάστωρ. Etymologically the two words have nothing in common; but that is no obstacle to the supposition that, in their usage, their casual but close similarity of form rendered the meaning of the one susceptible to the influence of the other. Nay more, the fact that the two words, it matters not how erroneously, were actually in early times referred to a common origin[1202] warrants the suggestion that such influence had been exercised. Now ἄλαστος always remained in meaning true to its derivation. Itself employed in the passive sense, ‘unforgotten,’ it seems to have made over the active meaning, ‘unforgetting,’ ‘vindictive’ (which, on the analogy of ἄπρακτος and a score of similar forms, it should naturally have possessed), to the apparently kindred word ἀλάστωρ. This adventitious meaning accorded well with the popular conception of the most conspicuous class of ‘wanderers’ from the grave—those whose wanderings had a vindictive aim; and thus, by the help of the accidental resemblance of two words, it seems to have come to pass that the term Alastores ceased to be applicable to all kinds of revenants and denoted only the ‘Avengers.’ At this point it became in fact synonymous with Miastores, and, like that word, enlarged its scope so as to denote not only the prime Avenger, the revenant himself, but also any divine or human agents employed by him as subsidiary Avengers.

So much then for the first meaning which the lexicons attach to the words Alastor and Miastor; the second interpretation of them, in relation to a blood-guilty man, may be more briefly treated. Alastor in this passive sense is alleged to mean a man who suffers from the vengeance of one who is an Alastor in the active sense; and Miastor to mean a man who is himself polluted and therefore pollutes those with whom he associates.

As regards Alastor, this explanation stands already condemned by the fact that it pre-supposes the derivation from λανθάνομαι, and even then it does fresh and incredible violence to language; a sane philologist may commit the error of deriving ἀλάστωρ from λανθάνομαι and making it mean ‘one who does not forget’; but only the maddest could dream of interpreting it as ‘one who does deeds which others do not forget.’ But, if in spite of this we trouble to turn up the references which the lexicons give under this heading, it is obvious at once that there is no more support for such a meaning in idiomatic usage than in etymological origin. Three references are cited. The first is to that passage of the Eumenides in which Orestes declares himself ἀλάστορα, οὐ προστρόπαιον[1203], a phrase which means, as I have already shown, ‘an avenger, not a murderer.’ This then should be classified as an example of the active, not of the hypothetical passive, meaning of Alastor. Of the other two passages, one is from the Ajax of Sophocles, where the hero in his anger and despair speaks of the guileful enemies who robbed him of his prize as Alastores[1204], and the other a passage from Demosthenes in which he criticizes Aeschines for applying the word as an opprobrious name to Philip of Macedon[1205]. But in what possible sense could either Ajax’ enemies or Philip of Macedon be described as ‘suffering from Avengers’? On the contrary, at the times when the word Alastor was applied to them, their success should surely have suggested that they were favoured by heaven, and their opponents rather were the sufferers. What then was the meaning of the word thus opprobriously employed? A meaning, I answer, very little removed from that of ‘Avenger’ and arising out of it. For how was the Avenger—be he the revenant himself or a demon acting on his behalf—constantly pictured? Was it not as a fiend tormenting with every torment the object of his wrath, plaguing him, maddening him, sucking his very blood? Little wonder then if the justice of that vengeance was sometimes obscured in men’s minds by their horror of it, and if the word Alastor suggested to them a fiend, a merciless tormentor. In that sense Ajax might well apply the name to his enemies, and Aeschines to Philip. Nor are other instances of it lacking. Demosthenes himself, for all his criticism of Aeschines’ vulgarity in calling Philip βάρβαρόν τε καὶ ἀλάστορα, ‘a foreign devil,’ used the same word of Aeschines and his friends[1206]; again, in Sophocles, the lion of Nemea for the loss and havoc that he inflicted is unique among beasts that perish in having merited the same sorry title—βουκόλων ἀλάστωρ, the ‘herdsmen’s Tormentor[1207]’; and indeed, apart from living men and animals, there are many instances in Tragedy[1208] in which the word Alastor, applied to some supernatural foe, revenant or demon, may be more appropriately rendered by ‘fiend’ or ‘tormentor’ than by ‘avenger.’

And the same thing is true, I hold, of the word Miastor. The theory of the lexicons, namely, that the word denotes a polluted and blood-guilty man because such an one is inevitably a ‘polluter’ of others, is certainly not intrinsically bad; for it recognises the primary meaning of the word, ‘polluter,’ and bases the secondary meaning ‘polluted’ upon a right understanding of the old belief that pollution was contagious. But at the same time it gives some occasion to wonder why the word should have been diverted from its most natural meaning in order to denote that which the cognate word μιαρός already expressed more simply. Moreover, when examination is made of those passages which are claimed as examples of such an usage, the theory becomes wholly unnecessary. The two most specious examples are two passages from Aeschylus[1209] and Euripides[1210], in both of which the persons called Miastores are Aegisthus and Clytemnestra. Now the authors of Agamemnon’s death were certainly polluted, and might with justice have been called μιαροί—that is admitted. But because they might have been called μιαροί and actually are called μιάστορες, it does not follow that, though the words have the same root, they also bear the same meaning. Obviously the word ‘fiends,’ if μιάστορες ever has that sense, would be an equally apt description of the murderous pair. The choice therefore between these two renderings here must be guided by more certain examples of usage elsewhere.

Two may be selected as eminently clear. In one Orestes calls Helen τὴν Ἑλλάδος μιάστορα[1211], where the word cannot mean a ‘polluted wretch,’ for the construction postulates an active meaning in Miastor; nor yet can the phrase be intelligibly rendered ‘the polluter of Greece,’ for there was no pollution involved in the warfare which Helen had caused; clearly Orestes means ‘the tormentor of Greece,’ the fiend who had proved the bane of ships and men and cities. In the other passage Peleus applies the word to Menelaus: ‘I look upon thee,’ he says, ‘as on the murderer—the fiend-like destroyer (μιάστορ’ ὥς τινα)—of Achilles[1212].’ Here again Miastor clearly bears an active sense, and at the same time cannot be rendered ‘polluter.’ Menelaus had brought upon Achilles not pollution but death, and the word Miastor explains the word ‘murderer’ (αὐθέντην) which precedes it—explains that the murder laid to Menelaus’ charge was not the open violence of a stronger foe, but resembled the death-dealing of some lurking fiend. In these two passages then the interpretation of Miastor in the sense of ‘fiend,’ ‘tormentor,’ ‘destroyer,’ is necessary and proven; and, this being known, common reason bids us read more ambiguous scriptures in the light thus obtained. There is therefore no call to suppose that μιάστωρ ever meant ‘polluted’; from the active meaning ‘Avenger’ it developed, like Alastor, the broader sense of ‘Tormentor’ or ‘Fiendish Destroyer’; and these meanings completely satisfy the conditions of Tragic and other usage of the words.

There remains the word προστρόπαιος, to which the lexicons, I admit, rightly ascribe a twofold meaning. It is clearly used both of the Avenger of blood and also of the blood-guilty person who is seeking purification. But as regards both the means by which the first signification was obtained, and the primary application of the word in that signification, I join issue. The second meaning is more satisfactorily explained, and my criticism of it will not go beyond an alternative suggestion.

The lexicons elucidate the first meaning as follows: he to whom one turns, especially with supplications, θεός or δαίμων προστρόπαιος the god to whom the murdered person turns for vengeance, hence an avenger, like ἀλάστωρ ... hence also of the manes of murdered persons, visiting with vengeance, implacable.

The objections to this explanation are obvious. It may well be questioned whether προστρόπαιος is at all likely to have had any passive meaning—as it were a person who ‘is turned to’—when the verb προστρέπω itself was, so far as I can ascertain, never so used; and further, if a god had really been called προστρόπαιος because the murdered man turned for vengeance to him, the extension of the term to the manes of murdered persons must imply a conception of the murdered man turning for vengeance towards—himself. This is not a little cumbrous; and for my part I deny the existence of any passive sense of προστρόπαιος.

I do however find two senses of the word, the one active, corresponding to the transitive use of the verb προστρέπειν or προστρέπεσθαι (for the middle as well as the active voice might be used transitively, as will shortly appear), the other middle, corresponding to the ordinary usage of the middle προστρέπεσθαι. Thus the active meaning of προστρόπαιος will be turning something towards or against someone; the middle meaning, turning oneself towards someone.

The active usage is best illustrated by a passage of Aeschines, in which he accuses Demosthenes of wilful perjury in calumniating him, and then appeals to the jury in these words—ἐάσετε οὖν τὸν τοιοῦτον αὑτοῦ προστρόπαιον (μὴ γὰρ δὴ τῆς πόλεως) ἐν ὑμῖν ἀναστρέφεσθαι[1213]; ‘Will you then allow this perjurer, who has turned upon his own head (for I pray that it be not on the city) the anger of the gods in whose name he swore, to continue in your midst?’ Here the very brevity of the Greek, which I am compelled to expand in translation, proves that Aeschines’ audience were perfectly familiar with an active meaning of προστρόπαιος with an evil connotation, ‘turning some misfortune or punishment or vengeance upon someone.’

The middle sense of προστρόπαιος is equally clearly exhibited by Aeschylus, who in telling the story of Thyestes says that after his banishment by his brother Atreus he came again προστρόπαιος ἑστίας[1214], ‘turning himself (as a suppliant) towards the hearth’ of his father’s home, so that his own life at least was spared out of respect for the place.

Thus the two meanings of the word are established, and it remains only to show how they were specially used in connexion with blood-guilt.

In the active sense προστρόπαιος was primarily applied, I hold, like Miastor and Alastor, to the murdered man himself, who ‘turned’ his wrath ‘against’ the murderer, or, if it so happened, against the next of kin who had failed in his duty of bringing the murderer to justice. It is precisely thus that Plato uses the verb προστρέπεσθαι in recording the old tradition in which he apparently reposed so much faith as to base his own laws upon it. ‘If the nearest of kin,’ so runs the passage, ‘do not seek vengeance for the deed, it is held that the pollution devolves upon him, and that the sufferer (i.e. the dead man) turns upon him the suffering (i.e. that which the homicide himself should have incurred), and anyone who will may bring a suit against him, etc.[1215]’ The words which I have italicised are in the Greek τοῦ παθόντος προστρεπομένου τὴν πάθην, where the middle presumably was preferred to the active because the sufferings which the dead man inflicts are, as we already know and as the language of the particular phrase itself suggests, exactly those which he himself suffers. This usage of the verb, though it is distinctly rare and probably a technicality of religion or law, is so perfectly clear in this one example[1216], that there should be no hesitation about understanding the cognate word προστρόπαιος in the same sense. And indeed one lexicographer, Photius, shows that he did so understand it; for he tells us that Zeus was sometimes invoked under this title, as turning against murderers the pollution (including perhaps the punishments) of their crime: Ζεὺς ... προστρόπαιος, ὁ προστρέπων τὸ ἄγος αὐτοῖς (sc. τοῖς παλαμναίοις)[1217]—such are his actual words, and this time of course the verb is rightly in the active, for Zeus is in no way personally concerned but acts only in the interests of the dead man. Clearly then it was in virtue of this active meaning that προστρόπαιος came to be practically a synonym of Miastor and Alastor in the sense of an Avenger of blood.

Once more then we return to the same question which has been propounded and answered with regard to those two other names—to whom was the term προστρόπαιος primarily applied?

I find the application of it more restricted than that of the other two words. It was used of the dead man himself, and it was used of demons avenging his cause; but it was never used[1218] of the next of kin in the character of avenger—and that for the very good reason that when the word was applied to a living man it bore an entirely different meaning, which has yet to be discussed, the meaning of ‘blood-guilty.’

A few examples of each usage must be given. Both Antiphon and Aeschylus apply the word to murdered men; Antiphon, in a speech in which the kinsman, who has, as in duty bound, undertaken the prosecution of the murderer, claims that, if the jury wrongfully acquit, the dead man will not become προστρόπαιος, an Avenger, against his kinsmen who have done their best in his service, but will visit his anger on the jury for condoning and thereby sharing the blood-guilt[1219]; Aeschylus, in that list of penalties which has been discussed, when he depicts the ‘madness and vain terror,’ which will befall Orestes if he fail in his task, as an arrow that flieth in darkness sped by powers of hell ‘at the behest of fallen kindred that turn their vengeance upon him’ (ἐκ προστροπαίων ἐν γένει πεπτωκότων[1220]). But equally clearly in other passages the Avenger indicated is not the murdered man, but some divine being. Antiphon again is an authority for this usage. Twice, in a context similar to that which has just been noticed, he speaks not of the murdered man himself becoming an Avenger, but of certain divine powers—whom he also calls ἀλιτήριοι, the powers that deal with sin—acting as Avengers (προστρόπαιοι) of the dead[1221]. And similarly in later time Pausanias also speaks of ‘the pollution (μίασμα) incurred by Pelops and of the Avenger (προστρόπαιος) of Myrtilus[1222].’

Since then there is no question but that the word προστρόπαιος was actually applied both to dead men and to gods, to which of the two did it refer primarily? We already know the answer. The dead man himself, as a revenant, was the prime and proper Avenger of his own wrongs; demons of vengeance acted only in his name, as his subordinates and agents. To him therefore the name primarily belonged. And even if we had not already learnt this from other sources, the passage of Aeschylus, to which I have just referred, might well guide us to the same conclusion. The arrow that flieth in darkness is sped indeed, he says, ‘by powers of hell’ (τῶν ἐνερτέρων)—the demonic agents of the dead—but ‘at the behest of fallen kindred.’ The activity both of the principal and of the agent is recognised in the same passage, and either might have been called προστρόπαιος: but, because the activity of both was plainly asserted, Aeschylus reserved the name for the one to whom it primarily belonged, the murdered man, who turns his wrath, who turns indeed those powers of hell who execute his wrath, against his enemies.

There now remains for consideration only the second meaning of προστρόπαιος; how could a word, which in reference to dead men or to deities meant ‘an Avenger of blood,’ bear, in relation to living men, the sense of ‘blood-guilty’? Very likely the dictionaries are right in accepting the explanation of this use which Hesychius[1223] and others give. We have seen one case[1224] in which the word clearly has a middle sense ‘turning oneself towards’ a place or a person in supplication; and there is no difficulty in supposing that the word was used technically in the same sense of a blood-guilty man who turned to some god or to some sanctuary in quest of purification. This, I say, is very probably the right explanation. But I may perhaps offer an alternative explanation which I do not count preferable but merely possible. The active meaning of προστρόπαιος, ‘turning something upon someone,’ might conceivably have produced this sense of ‘blood-guilty’ as well as the other sense ‘an Avenger of blood.’ As the dead man was held to turn something, namely his wrath, against his enemy, so might the murderer have been pictured as trying to turn something, namely the pollution which he had incurred, upon some object and so to cleanse himself therefrom. Now the chief feature in the Delphic ceremony of purification was the slaying of a sucking-pig[1225]. This may of course have been merely a propitiatory sacrifice; but it is possible also that the animal was really a surrogate victim for the murderer himself, that by laying his polluted hand on its head he transferred the religious uncleanness from himself to it, and that, by the subsequent slaughter of the now blood-guilty animal, he vicariously satisfied the old law that blood could only be washed out by blood. This is only a conjecture, and I leave others to judge of its probability; but, if the ceremony had followed the lines which I have suggested, it is easily intelligible that, in the technical language of religion, the murderer who sought to turn his own pollution upon the victim might have been called προστρόπαιος.

Thus then the problem of the ancient nomenclature of revenants is solved, and the results are briefly these: all revenants were originally called ἀλάστορες, ‘Wanderers’; but subsequently that name was restricted only to the vengeful class of revenants, to which the names μιάστορες and προστρόπαιοι had always belonged; and for the more harmless and purely pitiable revenants no name remained, but men said of such an one simply, ‘He wanders.’