Los filisteos

these marks that they are meant to express a modification of the phonetic value of the character, too slight to require a different letter to express it, but too marked to allow it to be neglected altogether. And obviously the most likely modification of the kind would be the elision of the vowel of a final open syllable. The mark would thus be exactly like the virāma of the Devanāgarī alphabet. 1 When we examine the text, we find that it is only in certain words that this mark occurs. It is found in βh, however declined, except when the suffixes w, ξ are present. It is found in the word nvhf, however declined, and appears in the two similar words µhtaσ and Mξtaσ. It is found in the personal name kq (in the formula pa Mξkq). There are only one or two of the eighteen examples of its use outside these groups, and probably if we had some more examples of the script, or a longer text, these would be found to fit likewise into series. This stroke would therefore be a device to express a final closed syllable. Thus, if it was desired to write the name of the god Dagon, it would be written on this theory, let us say, DA-GO-NA, with a stroke underneath the last symbol to elide its vowel. The consequences that may follow if this assumption should at any time be proved, and the culture which the objects represented by the various signs indicate, are subjects for discussion in later sections of this chapter. For further details of the analysis of the disk I must refer to my Royal Irish Academy paper above quoted: I have dwelt on it here, because if, as is most probable, the plumed head-dress shows that in this disk we have to deal with 'proto-Philistines', we must look to this document and others of the same kind, with which excavators of the future may be rewarded, to tell us something of the language of the people with whom we have to deal.

II. THEIR ORGANIZATION.

[figure]

A. Political.

From the time when the Philistines first appear in their Palestinian territory they are governed by Lords, ṣerānīm, each of whom has domination in one of the five chief cities, but who act in council together for the common good of the nation. They seem, indeed, to engage personally in duties which an Oriental monarch would certainly delegate to a messenger. They negotiate with Delilah. They convene the great triumph-feast to which Samson put so disastrous an

p. 88

end. There is a democratic instinct manifested by the men of Ashdod and Ekron, who peremptorily 'summoned' the council of lords to advise them what to do on the outbreak of plague: just as the merchants of the Zakkala obliged even a forceful ruler like Zakar-Baal to make an unsatisfactory compromise in the matter of Wen-Amon, and in much later times the people of Ekron deposed and imprisoned a ruler who persisted in the unpopular course of submission to Assyria. Achish makes arrangements with David, which his colleagues overrule. Of the methods of election of these officers we know absolutely nothing. From the Assyrian documents we hear of a series of rulers over Ashdod, father and son, but this does not necessarily prove that the hereditary principle was recognized. Such a political organization was quite unlike that of the nations round about: but the government of the Etruscans, who, as we have seen, were probably a related race, presents some analogy. There is a considerable similarity between the lucumones of Etruria and the Philistine ṣeranīm.

Nowhere do we read of a king of the Philistines. 1 To infer, as has actually been done, from 1 Kings iv. 21 ('Solomon ruled over all the kingdoms from the River unto the land of the Philistines') that their territory was organized as a kingdom, displays a sad lack of a sense of humour. When Hebrew writers speak of 'a king of Gath' (1 Sam. xxvii. 2), 'him that holdeth the sceptre from Ashkelon' (Amos i. 8), 'all the kings of the land of the Philistines' (Jer. xxv. 20), 'the king [perishing] from Gaza' (Zech. ix. 5), they obviously are merely offering a Hebrew word or periphrasis as a translation of the native Philistine title. The same is true of the analogous expressions in the Assyrian tablets. The case of the Etruscan 'kings' seems exactly similar, though there appears to have been an Achish-like king in Clusium.

In Gibeah, and probably in other towns as well, a resident officer, like a Turkish mudir, was maintained at the time of their greatest power.

It is possible that, if we had before us all the documents relating to the history of the Philistines, we might be able to divide them into clans, corresponding perhaps in some degree to the threefold division of the Egyptian monuments—Zakkala, Washasha, and Pulasati, i.e. as we have tried to show already, Cretans, Rhodians, and Carians. The continually recurring phrase 'Cherethites and Pelethites' suggests some twofold division. Ezekiel xxv. 16 ('Behold, I will stretch out my hand upon the Philistines, and I will cut off the Cherethites') may

p. 89

or may not imply a similar division. The report of the young Egyptian (1 Sam. xxx. 14) implies that the name 'Cherethites', if it had a specific meaning apart from 'Philistines', denoted the dwellers in the extreme south of Philistine territory: and we have already made passing note of the occurrence of the name Ziklag, a possible echo of the Zakkala, in that part of the country. The almost accidental allusion to Carians in the history of the kings must not be overlooked. But our data are so slender that very little can be built upon them. All we can say is that the origin of the Philistines makes it improbable that they were a single undivided tribe, and that the scanty hints which the history affords render it still more unlikely.

Nor can we necessarily infer that the peculiar government by a council of the lords of five cities implies that they were divided into five tribes. For though there seems to have been an actual division of the territory into districts, each of them under the hegemony of one of these cities, the limits are rather indefinite; and to judge from the scanty materials at our disposal, seem to have varied from time to time. The recurrence of the phrase '[such a city] and the border thereof' 1 seems to indicate a definite division of the country into provinces governed each by one of the cities; and this is confirmed by David's speech to Achish, 2 'Give me a place in one of the cities in the country (‏באחת ערי השדה‎), for why should thy servant dwell in the royal city (‏בעיר הממלכה‎) with thee?' A similar polity is traceable in Etruria.

Of the division of the minor cities of the Philistine territory among the Pentapolis—perhaps Pentarchy would be a more correct term to use—we know very little. In the time of David's exile Ziklag was under the control of the king of Gath. Sargon, according to one interpretation of his inscription, supposes Gath itself to belong to Ashdod. We may compare 'Gazara that bordereth on Azotus' (1 Macc. xiv. 34), though they are about sixteen miles apart, and each only just visible on the other's horizon. Rather curiously, Joppa and the neighbouring villages depended, according to Sennacherib, on Ashkelon.

Besides these towns we hear of certain unwalled villages (1 Sam. vi. 18) which are not specified by name.

B. Military.

Certain functionaries called sārīm meet us from time to time in the history (1 Sam. xviii. 30, xxix. 3, 9). It is the sārīm whose protest

p. 90

prevents David from joining in the battle of Gilboa. The word is, of course, a commonplace Semitic term, and is applied in Deborah's Song to the princes of Issachar, and by Zephaniah to those of Jerusalem. Among the Philistines the officials denoted by this word were no doubt military captains.

It is obvious throughout the whole history, from the days of the Medinet Habu sculptures onwards, that the military forces of the Philistines were well organized. In 1 Samuel xiii. 5 we read of 30,000 chariots and 6,000 horsemen, which, even if the numbers are not to be taken literally, indicates a considerable wealth in war equipment. Elsewhere (ib. xxix. 2) we hear of 'hundreds and thousands', which may indicate a system of division into centuries and regiments. Of their methods of fighting we have no certain information: Judges i. 19 emphasizes their corps of war-chariots: in the account of the battle of Gilboa the archers are specially alluded to. The Medinet Habu sculptures and the description of the equipment of the champions are analysed in the following section.

C. Domestic.

On the subject of family life among the Philistines nothing is known. The high-minded sense of propriety attributed to Abimelech in the patriarchal narratives has already been touched upon. Samson's relations with his Timnathite wife can hardly be made to bear undue stress: a Semitic marriage of the ṣadīḳa type is pictured by the storyteller. The wife remains in her father's house and is visited by her husband from time to time. Men and women apparently mingle freely in the temple of Dagon at Gaza. No further information is vouchsafed us.

III. Their Religion.

[figure]

Of the religion of the Philistines we know just enough to whet a curiosity that for the present seeks satisfaction in vain. The only hints given us in the Old Testament history are as follows:

(1) The closing scene of Samson's career took place in a temple of Dagon at Gaza, which must have been a large structure, as different as possible from the native High Places of Palestine.

(2) In this temple sacrifices were offered at festivals conducted by the 'Lords' of the Philistines (Judg. xvi. 23). It is not unreasonable to suppose that Samson was destined to be offered in sacrifice at the great feast of rejoicing there described. This was probably an annual festival, occurring at a fixed time of the year, and not a special celebration of the capture of Samson: because an interval of some months, during which Samson's shorn hair grew again, must have taken place

p. 91

between the two events. We are reminded of the Athenian Θαργήλια, with Samson in the rôle of the φαρμακός. Human sacrifices were offered in the temple of Marna at Gaza down to the fourth century A. D., as we learn from a passage presently to be quoted from Marcus the Deacon.

(3) There was also a temple of Dagon at Ashdod, which indicates that the deity was a universal god of the Philistines, not a local divinity like the innumerable Semitic Ba‘alim. Here there were priests, and here a rite of 'leaping on (or rather stepping over) the threshold' was observed. A sculptured image of the god stood in this temple.

(4) There was somewhere a temple of Ashtaroth (Samuel) or of Dagon (Chronicles) where the trophies of Saul were suspended. It is not expressly said that this temple was in Beth-shan, to the wall of which the body of Saul was fastened.

(5) The Philistines were struck with terror when the Ark of Yahweh was brought among them. Therefore they believed in (a) the existence and (b) the extra-territorial jurisdiction of the Hebrew deity. This suggests a wider conception of the limitations of divine power than was current among the contemporary Semites.

(6) Small portable images (‏עצבים‎) were worn by the Philistines and carried as amulets into battle (2 Sam. v. 21). 'This practice lasted till quite late (2 Macc. xii. 40).

(7) News of a victory was brought to the image-houses, probably because they were places of public resort, where they could be proclaimed (1 Sam. xxxi. 9).

(8) At Ekron there was an oracle of Baal-zebub, consulted by the Israelite king Ahaziah (2 Kings i. 2).

Let us clear the ground by first disposing of the last-named deity. This one reference is the only mention of him in the Old Testament, and indeed he is not alluded to elsewhere in Jewish literature. He must, however, have had a very prominent position in old Palestinian life, as otherwise the use of the name in the Gospels to denote the 'Prince of the Devils' (Matt. xii. 24, &c.) would be inexplicable. A hint in Isaiah ii. 6 shows us that the Philistines, like the Etruscans, were proverbial for skill in soothsaying, and it is not unlikely that the shrine of Baal-zebub should have been the site of their principal oracle. If so, we can be sure that Ahaziah was not the only Israelite who consulted this deity on occasion, and it is easy to understand that post-exilic reformers would develop and propagate the secondary application of his name in order to break the tradition of such illegitimate practices. It is, however, obvious that the Philistines who worked the oracle of Baal-zebub simply entered into an old

p. 92

Canaanite inheritance. This is clear from the Semitic etymology of the name. When they took over the town of Ekron and made it one of their chief cities, they naturally took over what was probably the most profitable source of emolument that the town contained. The local divinity had already established his lordship over the flies when the Philistines came on the scene.

This was no contemptible or insignificant lordship. A man who has passed a summer and autumn among the house-flies, sand-flies, gnats, mosquitoes, and all the other winged pests of the Shephelah will not feel any necessity to emend the text so as to give the Ba‘al of Ekron 'a lofty house' or 'the Planet Saturn' or anything else more worthy of divinity 1; or to subscribe to Winckler's arbitrary judgement: 'Natürlich nicht Fliegenba‘al, sondern Ba‘al von Zebub, worunter man sich eine Oertlichkeit in Ekron vorzustellen hat, etwa den Hügel auf dem der Tempel stand' (Geschichte Israels, p. 224). The Greek Version lends no countenance to such euhemerisms, for it simply reads τῷ Βάαλ μυῖαν. Josephus avoids the use of the word Ba‘al, and says 'he sent to the Fly' (Ant. ix. 2. 1). The evidence of a form with final l is, however, sufficiently strong to be taken seriously. Although the vocalization is a difficulty, the old explanation seems to me the best, namely, that the by-form is a wilful perversion, designed to suggest zebel, 'dung.' The Muslim argot which turns ḳiyámah (Anastasis = the Church of the Holy Sepulchre) into ḳumámah (dung-heap) is a modern example of the same kind of bitter wit.

The Lord of Flies is hardly a fly-averter, like the Ζεὺς ἀπόμυιος of Pliny and other writers, with whom he is frequently compared. In fact, what evidence there is would rather indicate that the original conception was a god in the bodily form of the vermin, the notion of an averter being a later development: that, for instance, Apollo Smintheus has succeeded to a primitive mouse-god, who very likely gave oracles through the movements of mice. That Baal-zebub gave oracles by his flies is at least probable. A passage of Iamblichus (apud Photius, ed. Bekker, p. 75) referring to Babylonian divinations has often been quoted in this connexion; but I think that probably mice rather than flies are there in question. Lenormant (La divination chez les Chaldéens, p. 93) refers to an omen-tablet from which auguries are drawn from the behaviour or peculiarities of flies, but unfortunately the tablet in question is too broken to give any continuous sense. 2

p. 93

A curious parallel may he cited from Scotland. In the account of the parish of Kirkmichael, Banffshire, is a description (Statistical Account of Scotland, vol. xii, p. 464) of the holy well of St. Michael, which was supposed to have healing properties:

'Many a patient have its waters restored to health and many more have attested the efficacies of their virtues. But as the presiding power is sometimes capricious and apt to desert his charge, it now [A. D. 1794] lies neglected, choked with weeds, unhonoured, and unfrequented. In better days it was not so; for the winged guardian, under the semblance of a fly, was never absent from his duty. If the sober matron wished to know the issue of her husband's ailment, or the love-sick nymph that of her languishing swain, they visited the well of St. Michael. Every movement of the sympathetic fly was regarded in silent awe; and as he appeared cheerful or dejected, the anxious votaries drew their presages; their breasts vibrated with correspondent emotions. Like the Dalai Lama of Thibet, or the King of Great Britain, whom a fiction of the English law supposes never to die, the guardian fly of the well of St. Michael was believed to be exempted from the laws of mortality. To the eye of ignorance he might sometimes appear dead, but, agreeably to the Druidic system, it was only a transmigration into a similar form, which made little alteration in the real identity.'

In a foot-note the writer of the foregoing account describes having heard an old man lamenting the neglect into which the well had fallen, and saying that if the infirmities of years permitted he would have cleared it out and 'as in the days of youth enjoyed the pleasure of seeing the guardian fly'. Let us suppose the old man to have been eighty years of age: this brings the practice of consulting the fly-oracle of Kirkmichael down to the twenties of the eighteenth century, and probably even later.

Leaving out Baal-zebub, therefore, we have a female deity, called Ashtaroth (Aštoreth) in the passage relating to the temple of Bethshan, and a male deity called Dagon, ascribed to the Philistines. We may incidentally recall what was said in the first chapter as to the possibility of the obscure name Beth-Car enshrining the name of an eponymous Carian deity: it seems at least as likely as the meaning of the name in Hebrew, 'house of a lamb.' Later we shall glance at the evidence which the Greek writers preserve as to the peculiar cults of the Philistine cities in post-Philistine times, which no doubt preserved reminiscences of the old worship. In the meanwhile let us concentrate our attention on the two deities named above.

I. Ashtoreth. At first sight we are tempted to suppose that the Philistines, who otherwise succeeded in preserving their originality, had from the first completely succumbed to Semitic influences in the

p. 94

province of religion. 'As immigrants', says Winckler in his Geschichte Israels, 'they naturally adopted the civilization of the land they seized, and with it the cultus also.' And certainly Ashtaroth or Ashtoreth was par excellence the characteristic Semitic deity, and worshippers of this goddess might well be said to have become completely semitized.

But there is evidence that makes it doubtful whether the assimilation had been more than partial. We begin by noting that Herodotus 1 specially mentions the temple of ἡ Οὐρανία Ἀφροδίτη as standing at Ashkelon, and he tells us that it was the oldest of all the temples dedicated to this divinity, older even than that in Cyprus, as the Cyprians themselves admitted: also that the Scythians plundered the temple and were in consequence afflicted by the goddess with a hereditary νοῦσος θήλεια. 2 The remarkable inscription found at Delos, in which one Damon of Ashkelon dedicates an altar to his tutelary divinities, brilliantly confirms the statement of Herodotus. It runs:

[figure]

'To Zeus, sender of fair winds, and Astarte of Palestine, and Aphrodite Urania, to the divinities that hearken, Damon son of Demetrios of Ashkelon, saved from pirates, makes this vow. It is not lawful to offer in sacrifice an animal of the goat or pig species, or a cow.' 3

p. 95

The Palestinian Astarte is here distinguished from the Aphrodite of Ashkelon; and though there obviously was much confusion between them, the distinction was real. From Lucian 1 we learn that there were two goddesses, whom he keeps carefully apart, and who indeed were distinguished by their bodily form. The goddess of Hierapolis, of whose worship he gives us such a lurid description, was in human form: the goddess of Phoenicia, whom he calls Derkĕto (a Greek corruption of the Semitic Atargatis, ‏עתר-עתה‎), had the tail of a fish, like a mermaid.

The name of this goddess, as written in Sidonian inscriptions, was long ago explained as a compound of ‏עתר‎ and ‏עתה‎, ‘Atar and ‘Ate. These are two well-established divine names; the former is a variant of ‘Ashtart, but the latter is more obscure: it is possibly of Lydian origin. 2 In Syriac and Talmudic writings the compound name appears as Tar‘atha.

The fish-tailed goddess was already antiquated when Lucian wrote. He saw a representation of her in Phoenicia (op. cit. § 14), which seemed to him unwonted. No doubt he was correct in keeping the two apart; but it is also clear that they had become inextricably entangled with one another by his time. The figure of the goddess of Hierapolis was adorned with a cestus or girdle, an ornament peculiar to Urania (§ 32), who, as we learn from Herodotus, was regarded as the goddess of Ashkelon. There was another point of contact between the two goddesses—sacred fish were kept at their shrines. The fish-pond of Hierapolis is described by Lucian (§§ 45, 46) as being very deep, with an altar in the middle to which people swam out daily, and with many fishes in it, some of large size—one of these being decorated with a golden ornament on its fin.

To account for the mermaid shape of the Ashkelonite goddess a story was told of which the fullest version is preserved for us by Diodorus Siculus (ii. 4). 'In Syria is a city called Ashkelon, and not far from it is a great deep lake full of fishes; and beside it is a shrine of a famous goddess whom the Syrians called Derketo: and she has the face of a woman, and otherwise the entire body of a fish, for some reason such as this: the natives most skilful in legend fable that Aphrodite being offended by the aforesaid goddess inspired

p. 96

her with furious love for a certain youth among those sacrificing: and that Derketo, uniting with the Syrian, bore a daughter, and being ashamed at the fault, caused the youth to disappear and exposed the child in certain desert and stony places: and cast herself in shame and grief into the lake. The form of her body was changed into a fish: wherefore the Syrians even yet abstain from eating this creature, and honour fishes as gods.' The legend is told to the same effect by Pausanias (II. xxx. 3).

This legend is of great importance, for it helps us to detect the Philistine element in the Ashkelonite Atargatis. An essentially identical legend was told in Crete, the heroine being Britomartis or Dictynna. According to Callimachus’ Hymn to Artemis Britomartis was a nymph of Gortyna beloved of Artemis, whom Minos, inflamed with love, chased over the mountains of Crete. The nymph now hid herself in the forests, now in the low-lying meadows; till at last, when for nine months she had been chased over crags, and Minos was on the point of seizing her, she leaped into the sea from the high rocks of the Dictaean mountain. But she sprang into fishers’ nets (δίκτυα) which saved her; and hence the Cydonians called the nymph Dictynna, and the mountain from which she had leaped called they Dictaean; and they set up altars to her and perform sacrifices.

The myth of the Atargatis of Ashkelon fits very badly on to the Syrian deity. She was the very last being to be troubled with shame at the events recorded by Diodorus Siculus: she had no special connexion with the sea, except in so far as fishes, on account of their extreme fertility, might be taken as typical of the departments of life over which she presided. There can surely be little question that the coyness of the Cretan nymph, her leap into the sea, and her deliverance by means of something relating to fishes, has been transferred to the Ashkelonite divinity by the immigrants. The Atargatis myth is more primitive than that of Britomartis: the union from which Britomartis was fleeing has actually taken place, and the metamorphosis into a fish is of the crudest kind; the ruder Carians of the mainland might well have preserved an earlier phase of the myth which the cultured Cretans had in a measure refined.

The cult of Britomartis was evidently very ancient. Her temple was said to have been built by Daedalus. The name is alleged to mean uirgo dulcis 1; and as Hesychius and the Etymologicon Magnum give us respectively γλυκύ and ἀγαθόν as meanings of βριτύ or βρίτον,

p. 97

the explanation is very likely correct. The name of the barley drink, βρύτος or βρύτον, may possibly have some connexion with this word. See also the end of the quotation from Stephanus of Byzantium, ante p. 15.

Athenaeus (viii. 37) gives us an amusing piece of etymology on the authority of Antipater of Tarsus, to the effect that one Gatis was a queen of Syria who was so fond of fish that she allowed no one to eat fish without inviting her to the feast—in fact, that no one could eat ἄτερ Γάτιδος: and that the common people thought her name was 'Atergatis' on account of this formula, and so abstained from fish altogether. He further quotes from the History of Asia by Mnaseus to the effect that Atargatis was originally a tyrannous queen who forbade the use of fish to her subjects, because she herself was so extravagantly fond of this article of diet that she wanted it all for herself; and therefore a custom still prevails to offer gold or silver fish, or real fish, well cooked, which the priests of the goddess eat. Another tale is told by Xanthus and repeated by Athenaeus in the same place, that Atargatis was taken prisoner by Mopsus king of Lydia, and with her son Ἰχθύς ('fish') cast into the lake near Ashkelon (ιν τῇ περὶ Ἀσκάλωνα λίμνῃ) because of her pride, and was eaten by fishes.

Indeed, the Syrian avoidance of fish as an article of food is a commonplace of classical writers. A collection of passages on the subject will be found in Selden, De Diis Syris, II. iii.

Lucian further tells us (§ 4) that the temple at Sidon was said to be a temple of Astarte; but that one of the priests had informed him that it was really dedicated to Europa, sister of Cadmus. This daughter of King Agenor the Phoenicians honoured with a temple 'when she had vanished' (ἐπειδή τε ἀφανὴς ἐγεγόνεε), and related the legend about her that Zeus, enamoured of her, chased her, in the form of a bull, to Crete.

Here then we have distinctly a legend to the effect that a certain temple of the Syrian goddess was really dedicated to a deity who had fled from an unwelcome lover, and who was directly connected with Crete. In fact, we have here a confused version of the Britomartis legend on the Syrian coast. And when we turn to the Metamorphoses of Antoninus Liberalis, ch. 30, we find a version of the Britomartis story that is closely akin to the tale told by the Sidonian priest to Lucian. We read there that 'of Cassiepeia and Phoenix son of Agenor was born Carmē: and that Zeus uniting with the latter begat Britomartis. She, fleeing from the converse of men, wished to be a perpetual virgin. And first she came to Argos from Phoenicia, with Buzē, and Melitē, and Maera, and Anchiroē, daughters of

p. 98

[paragraph continues] Erasinos; and thereafter she went up to Cephallenia from Argos; and the Cephallenians call her Laphria; and they erected a temple to her as to a deity. Thereafter she went to Crete, and Minos seeing her and being enamoured of her, pursued her; but she took refuge among fishermen, and they caused her to hide in the nets, and from this the Cretans call her Dictynna, and offer sacrifices to her. And fleeing from Minos, Britomartis reached Aegina in a ship, with a fisherman Andromēdes, and he laid hands on her, being desirous to unite with her; but Britomartis, having stepped from the ship, fled to a grove where there is now her temple, and there she vanished (ἐγένετο ἀφανής); and they called her Aphaea, and in the temple of Artemis the Aeginetans called the place where Britomartis vanished Aphaē, and offered sacrifices as to a deity.' The relationship to Agenor, the love-chase, and the curious reference to 'vanishing' can scarcely be a mere coincidence. Lucian, though careless of detail and no doubt writing from memory, from the report of a priest who being a Syrian was not improbably inaccurate, has yet preserved enough of the Britomartis legend as told in Sidon to enable us to identify it under the guise of the story of Europa.

To the same Cretan-Carian family of legends probably belongs the sea-monster group of tales which centre in Joppa and its neighbourhood. The chief among them is the story of Perseus the Lycian hero and Andromeda; and a passage in Pliny seems to couple this legend with that of Derketo. 1 Some such story as this may have suggested to the author of the Book of Jonah the machinery of his sublime allegory; and no doubt underlies the mediaeval legends of St. George and the Dragon, localized in the neighbouring town of Lydd. We can scarcely avoid seeing in these tales literary parallels to the beautiful designs which the Cretan artists evolved from the curling tentacles of the octopus.

We are now, I think, in a position to detect a process of evolution in these tangled tales. We begin with a community dwelling somewhere on the sea-coast, probably at the low cultural level of the tribes who heaped the piles of midden refuse on the coasts of Eastern Denmark. These evolved, from the porpoises and other sea-monsters that came under their observation, the conception of a mermaid sea-goddess who sent them their food; and no doubt prayers and charms and magical formulae were uttered in her name to ensure that the creeks should he filled with fish. The sacredness of fish to the goddess would

p. 99

follow as a matter of course, and would be most naturally expressed by a prohibition against eating certain specified kinds. 1 And aetiological myths would of course be developed to account for her fish-tail shape. The Dictynna legend, with a Volksetymologie connecting the name of the nymph with a fishing-net, is one version; the legend afterwards attached to Atargatis is another.

When the Carian-Cretan league, after their repulse from Egypt, settled on the Palestine coast, they of course brought their legends with them. In their new home they found a Bona Dea all powerful, to whom inter alia fish were sacred, and with her they confused their own Virgo Dulcis, patroness of fishermen. They built her temples—a thing unheard-of before in Palestine—and told of her the same tales that in their old home they had told of Britomartis. They transferred the scene of the tragedy from the eastern headland of Crete to the λίμνη of Ashkelon, and they fashioned the legend into the form in which it ultimately reached the ears of Diodorus Siculus.

To the legend of Atargatis Diodorus adds that the exposed child was tended and fed by doves till it was a year old, when it was found by one Simma, who being childless adopted it, and named it Semiramis, a name derived from the word for 'dove' in the Syrian language. In after years she became the famous Babylonian queen: and the Syrians all honour doves as divine in consequence. The etymology is of the same order as Justin's derivation of 'Sidon' from 'a Phoenician word meaning "fish"': the tale was no doubt told primarily to account for the sacredness of doves to the Syrian goddess. The goddess of Ashkelon was likewise patroness of doves, and this bird frequently figures on coins of the city.

II. Dagon was evidently the head of the pantheon of the Philistines, after their settlement in Palestine. We hear of his temple at Gaza, Ashdod, and, possibly, according to one version of the story of the death of Saul, at Beth-Shan. 2 Jerome in commenting on 'Bel boweth down, Nebo stoopeth', in Isaiah xlvi. 1 (where some versions of the Greek have Dagon for Nebo), says Dagon is the idol of Ashkelon, Gaza, and the other cities of the Philistines. 3 The important temple

p. 100

of Gaza is mirrored for us in the graphic story of the death of Samson, as we shall see in the following section.

In the temple of Ashdod there was an image of the god—a thing probably unknown in the rude early Canaanite shrines. Josephus (Wars, v. 9. 4) calls it a ξόανον, which possibly preserves a true tradition that the figure was of wood. Some interesting though obscure particulars are given us regarding it in 1 Samuel v. 1–5. The Ark, captured at Aphek, was laid up two nights in the temple. The first night the image of Dagon fell on its face before the Ark, and was replaced by 'the priests of Dagon'; the only reference we have to specifically religious functionaries among the Philistines. The second night he was fallen again, and the head of the figure and the palms of its hands were broken off and lay on the threshold.

The account of the abasement of Dagon is of considerable importance with regard to the question of the form under which he was represented. The current idea is that he was of merman form, the upper half man, the lower half fish. This theory is by modern writers derived from the mediaeval Jewish commentators: Rabbi Levi, in the third century, said that Dagon was in the figure of a man: the first statement of his half-fish form, so far as extant authorities go, is made by David Ḳimḥi, who writes, They say that Dagon had the shape of a fish front his navel downwards, because he is called Dagon [‏דג‎ = fish] and upwards from his navel the form of a man, as it is said "both the palms of his bands were cut off on the threshold".' Abarbanel appears to make the god even more monstrous by supposing that it was the upper end which was the fishy part. But the idea must have been considerably older than Ḳimḥi. As we shall see presently, it underlies one of the readings of the Greek translation: and the attempts at etymology in the Onomastica 1 show clearly that the idea arose out of the accident that ‏דג‎ means 'a fish', while the story in 1 Samuel v requires us to picture the god with hands; coupled with vague recollections of the bodily form of the Atargatis of Ashkelon.

If we examine the passage, we note, first, that he had a head and hands, so that he must have been at least partly human. Next we observe that exactly the same phrase is used in describing both falls of the idol. The first time it was unbroken, and the priests could

p. 101

put it in its place again. The second time it was fallen again, but the projecting parts of it were broken off. In other words, the first fall of the statue was just as bad as the second, except that it was not broken: there is no statement made that on the second occasion the image, whatever its form, snapped across in the middle. In both cases it fell as a whole, being smashed the second time, just as might happen to a china vase; this would imply that what was left standing and intact was not so much any part of the statue itself, as the pedestal or some other accessory.

The difficulty lies in the words which follow the account of the fracture of the statue—‏רק דגון נשאר עליו‎. In the English version these are rendered 'only [the stump of] Dagon was left'. The words in brackets, for which the Hebrew gives no warrant, are inserted as a makeshift to make some kind of sense of the passage. Wellhausen ingeniously suggested omission of the ‏ן‎ at the end of ‏דגון‎, supposing that it had been inserted by dittography before the initial ‏נ‎ of the following word. This would make the word mean 'only his fish was left'. But this assumes the thesis to be proved.

When we turn to the Greek Version we find that it represents a much fuller text. It reads thus: …. The passage in brackets has no equivalent in the Hebrew text: it suggests that a line has been lost from the archetype of the extant Hebrew Version. 1 If with some MSS. we omit the first χειρῶν (which makes no satisfactory sense with ἴχνη), this lost line would imply that Dagon's feet were also fallen on the threshold (ἀμαφέθ = Hebrew ‏המפתתן‎). This does not accord with the 'fish-tail' hypothesis. But, on the other hand, it shows that the fishtail conception is considerably older than Kimhi, for χειρῶν must in the first instance have been inserted by a glossator obsessed with it.

And what are we to make of πλὴν ἡ ῥάχις ὐπελείφθη? 'The backbone of Dagon was left' is as meaningless as the traditional Hebrew, if not worse. But when we look back at the Hebrew we begin to wonder whether we may not here be on the track of another Philistine word—the technical term for, let us say, the pedestal or console on which the image stood; or, it may be, some symbol associated with it. Wellhausen (Text d. Buch. Sam. p. 59) has

p. 102

put forward the suggestion that ῥάχις really depends on ‏רק‎ 'only'. But the translators would presumably have understood this simple word—they have indeed rendered it correctly, by πλήν. We need a second ‏רק‎ to account for ῥάχις, and such, I submit, must have stood in the Hebrew text. Some word like (let us say) ‏רקד‎, especially if unintelligible to a late Hebrew copyist, would certainly drop out sooner or later from the collocation ‏רק רקד דגון‎. It would be very natural for the original author to use such a word, for the sake of the paronomasia; and it would fully account for ῥάχις, which in this case is not the Greek word at all, but a transliteration of an unknown word in the Hebrew original. The word ἀμαφέθ, immediately before, which has given much trouble to the copyists of the Greek text (see the numerous variants in Holmes and Parsons), is an example of an even easier word in the Hebrew being transferred to the Greek untranslated.

Further we are told that the priests and those who entered the house of Dagon—an indication that the temple was open to ordinary worshippers—did not tread on the threshold of the temple in Ashdod, in consequence, it was said, of this catastrophe; but, as the Greek translators add 'overstepping they overstepped it' (ὑπερβαίνοντες). That the explanation was fitted to a much more ancient rite we need not doubt: the various rites and observances relating to thresholds are widespread and this prohibition is no isolated phenomenon. 1 It is not certain whether the threshold of the Ashdod temple only was thus reverently regarded, or whether the other Dagon temples had similar observances: the latter is probable, though evidently the writer of Samuel supposed that the former was the case. The possible connexion between the Ashdod prohibition and the 'leaping on (preferably over) the threshold' of Zephaniah i. 9, has already been noted.

We must, however, face the fact that Dagon cannot be considered as exclusively a Philistine deity, even though the Semitic etymologies which have been sought for his name are open to question. There are ‏דג‎ 'fish', as already mentioned, and ‏דגן‎ 'corn'. Philo Byblios favoured the second of these. The inscription of Eshmunazar, king of Sidon, is well known to refer to Joppa and Dor as ‏ארין דגן‎, which seems at first sight to mean 'the land of Dagon'. But more probably this is simply a reference to that fertile region as 'the land of corn'. However we have, through Philo, references associating Dagon with the Phoenicians. In the Sanchuniathon cosmogony reported in the

p. 103

fragments of Philo we have an account of his birth from Ouranos and Ge, 1 with his brethren Ēlos and Kronos and Baetylos; he is equated to Σίτων 'corn', which is apparently personified; and by virtue of this equation he is identified with a Ζεὺς Ἀρότριος. All this is very nebulous: and not more definite is the curious note respecting the gods Taautos, Kronos, Dagon and the rest being symbolized by sacred letters. 2 If these passages mean anything at all, they imply that the people who taught the Phoenicians the use of letters (and possibly also of baetylic stones) also imparted to them the knowledge of the god Dagon. But stories which ostensibly reach us at third hand afford a rather unsafe apparatus criticus.