A further suggestion is that the Homeric use of the chariot was due to the long shield[321]. The argument in this case is that such shields could not be carried on horseback, while their weight was too great to allow them to be borne for any distance on foot. This is an extremely dubious theory for two reasons. In the first place the Homeric use of the chariot is a problem which concerns not Greece alone but a considerable part of Europe, including countries where there is no evidence for the Mycenean shield[322]. Secondly, it may be regarded as a general rule that those peoples which use the long shield not only fight but also go to battle on foot. Moreover the evidence of the poems themselves does not really bear out the suggested connection. Thus there is no instance of the long shield more clear than that belonging to Aias the son of Telamon, and this hero is one of the few leading men who are never said to wear breast-plates. But he is also apparently one of the very few who do not possess chariots. Indeed a better case could probably be made out for connecting the chariot with the breast-plate.
The Homeric use of the chariot gives rise to another question which probably deserves more careful consideration. In the action itself driving appears to be universal; at most we have only one doubtful case of riding (Il. X 513). But there are at least two incidental references (Il. XV 679 ff., Od. V 371) which betray acquaintance with the latter art. It is quite possible therefore that the knowledge of riding was a comparatively recent accomplishment which the poet or poets knew to have been foreign to the Heroic Age. Unfortunately the history of equitation is a very obscure subject. In more eastern countries the use of chariots—which here were war-chariots in the true sense—continued until quite late times. In Cyprus we hear of them at the beginning of the fifth century; and in Italy they were introduced, or re-introduced, still later. But we do not know when riding began. In the wall-sculptures at Karnak[323] which commemorate the victories of Sety I and date from towards the close of the fourteenth century several Hittites are shown on horseback. The scene represents a battle, and it may have been the artist's intention to depict what was not a normal custom but the last resource of fugitives whose chariots had broken down. But even then the possibility remains that in emergencies or under special conditions riding may have been practised by the Greeks long before it was in regular use. The whole subject however requires further investigation, in particular the question whether there is reason for supposing the Hittites to have been more skilled in horsemanship than other nations.
Another series of inconsistencies has been pointed out in references to marriage customs, particularly in regard to the use of the word ἔεδνα (ἕδνα). In several passages this word evidently denotes the sum paid by the bridegroom to the bride's guardian; but in others it seems to mean presents given to the bride at marriage by her own relatives. We can scarcely doubt that a difference of custom is involved in these usages; but it does not necessarily follow that the second group of passages belong to a later period than the first. Account must be taken of local divergencies, for there is hardly any subject in which early Teutonic custom varied so much as in this, even from the time of our oldest records. Again, it has been suggested[324] with considerable probability that the inconsistencies of the Odyssey are to be attributed to a change of custom—not in the sense that the earlier parts of the poem reflect one form and the later parts another, but that the poem as a whole belongs to an age of transition when different forms of matrimonial arrangements were in vogue. This is a question with which we shall have to deal more fully in a later chapter.
We have yet to consider certain inconsistencies in regard to religious observances. Once only (Il. VI 303) is mention made of the figure—a seated figure—of a deity, and consequently it is held that the passage in question must be late. Recent discoveries however have tended to throw doubt on this view; for a number of statuettes, apparently representing deities, have come to light in deposits belonging both to the Mycenean and Geometrical periods, while primitive female figures, often in a sitting position, are quite common. The finding of a larger image would now scarcely call forth much surprise, though it cannot be supposed that such statues were of frequent occurrence.
In references to sanctuaries we sometimes hear of temples (νηοί), but more frequently only of shrines or sacred groves. The most certain examples[325] of the former are those of Apollo and Athene at Troy (Il. V 446, VI 88, VII 83, etc.) and another of Athene at Athens (Il. II 549, Od. VII 81). The sanctuary of Apollo at Chryse is also once described as a temple (Il. I 39), though the account of the sacrifice (ib. 440 ff.) suggests an open-air shrine. Besides these we have a general reference to temples among the Phaeacians (Od. VI 10) and a vow made by Odysseus' followers to construct a temple to Helios on their return home (ib. XII 346). Now it is held that all these passages belong to a late stage in the growth of the poems and that in their original form temples were unknown. Sanctuaries of the earlier period are described only by such terms as ἄλσος or τέμενος βωμός τε θυήεις. There can be little doubt that the grove and the open-air shrine represent more primitive types of sanctuary than the temple. But the advocates of this theory seem to have overlooked the fact that among many peoples the more primitive and the more developed forms of sanctuary are found existing side by side[326]. Thus beside the great temple at Upsala, the chief sanctuary of the North, there stood a grove which appears to have been regarded with still greater veneration, even down to the very end of heathen times. Close by was a spring in which human victims were sacrificed[327]. This case shows how entirely unjustifiable it is to assume the non-existence of a temple, when only a shrine or sacred grove happens to be mentioned. We may cite as an example the proposed sacrifice to Spercheios in Il. XXIII 145 ff.[328] In general however the Homeric evidence clearly suggests that temples—and the same may possibly be the case with images—were chiefly to be found in cities, while the more primitive forms of sanctuary remained in less populous places.
I cannot but think that to any student of comparative religion the argument derived from the references to sanctuaries will appear entirely worthless. The argument against the antiquity of the round shield also can hardly be maintained. Of the rest all except the one based on the use of the metals contain a certain element of doubt. But even if we grant their validity in every case it cannot be said that either individually or collectively they necessitate the lapse of a long interval between the earlier and later portions of the poems. If the interval had amounted to anything like three centuries discrepancies of a far more striking character must have come to light.
As yet we have taken no account of linguistic inconsistencies. The poems as we have them present a medley of forms belonging to different ages and different dialects. But these inconsistencies appear everywhere; it is not the case that certain portions of the poems contain only early forms and others only late ones. The Odyssey is said to contain a certain number of apparently late usages, especially of prepositions and conjunctions, which are wanting or only occur rarely in the Iliad. It is held also that within the poems themselves earlier and later portions can be distinguished to a certain extent by similar differences of usage. But the evidence on the whole is slight and generally somewhat ambiguous[329]; and consequently linguistic criteria have as a rule played only a subordinate part in the attempts which have been made to determine the stratification of the poems. We must conclude then either that the later parts were composed in an artificial type of language, thoroughly permeated with archaisms, or that the earlier parts have undergone a very considerable amount of modernisation. Indeed the latter explanation must be admitted to some extent in any case; for relatively modern forms occur frequently in what are usually regarded as the very earliest parts of the poems. But if the preservation of the poems was dependent on oral tradition for any considerable length of time it is very difficult to set a limit to the operations of such a process. In Anglo-Saxon poems—even in those which were not entirely dependent on oral tradition—modernisation prevailed to such an extent that archaic forms disappeared practically everywhere, while substitutions of one word for another were very frequent. The more elaborate character of Greek metre doubtless acted as a check on this tendency to change, but at the same time I cannot help thinking that its importance in the history of Homeric poetry has been greatly underrated.
On the question of dialect something more must be said. It is true no doubt that the language of the poems as we have them must be regarded as Ionic—but only with certain reservations. In the first place we have to note the regular preservation of h-, which in strict Ionic—the language of the Asiatic coast—was lost in the seventh century, if not earlier. Again, we find a considerable number of forms which cannot be assigned to any Ionic dialect, e.g. such as contain -ā- (λαός, Ἀτρείδαο, αἰχμητάων). Special attention must be paid to forms which are definitely Aeolic, such as πίσυρες, ἄμμε, ἐρεβεννός. Whatever may be the explanation of the preservation of h-, it is universally agreed that the Aeolic element is deeply rooted in the history of Homeric poetry. Some scholars indeed hold that the poems were originally composed in Aeolic and that their present form is practically an Ionic translation. The more general view however is that Ionic was the language of epic composition from the beginning and that such forms as πίσυρες and ἄμμε are derived from early Aeolic lays on which the epics were for the most part based.
The legends as to Homer's birthplace are perhaps not without significance for this question. In ancient times, from Pindar[330] downwards, Smyrna's claim to this honour was the one most generally recognised, and the majority of modern scholars are inclined to the view that the birthplace of Homeric poetry is to be sought in or around that city. Now Smyrna was originally an Aeolic state, but was captured by the Ionians of Colophon[331], apparently towards the close of the eighth century, from which time onwards it appears as Ionic. The mixture of dialect found in the poems—an older stratum of Aeolic underlying a later stratum of Ionic—would therefore be perfectly in accord with what we know of the history of the city, although unfortunately no early inscriptions are extant[332].