In the preceding chapters we have seen that the persons and events celebrated in the heroic poems apparently all belonged to the fourth, fifth or sixth centuries, and further that heroic poetry was flourishing among the Goths during the same period. For the existence of English, Scandinavian or German heroic poetry at this time we have no absolutely conclusive evidence. But the materials from which our poems are formed must largely be referred to the sixth century. This may be seen most clearly in cases where the poems of two or more nations not merely treat an identical theme but also agree in the motif or in comparatively small details, as in the stories of Ingeld, Waldhere and Svanhildr. A like age is probably to be attributed to resemblances in language, such as those shown by the hortatory addresses and the accounts of dragon-fights cited in the last chapter (p. 60 f.). The fact that these resemblances sometimes occur in stories relating to entirely different characters need not prevent us from believing that they spring ultimately from a common origin.

It cannot of course be proved that the materials from which the heroic poems are derived were themselves always in poetic (metrical) form. In principle we must admit the possibility that they were transmitted from one generation to another in a more or less stereotyped form of prose narrative, such as we find later in the sagas of Iceland and Ireland. But in point of fact we have no evidence whatever for the cultivation of such traditional prose narratives among any of the early Teutonic peoples, whereas there is good reason, as we shall see shortly, for believing that narrative poetry was both ancient and widely cultivated.

In the first place we may note that English and German poetry down to the ninth century shared a common system of metre and that the FornyrÐislag, which is used in most of the Edda poems, differed but little from this type, except of course that it was always arranged in strophes. The application of this common metre to narrative purposes can scarcely be regarded as a recent innovation, for English and German poems frequently exhibit verses and half-verses of very similar construction. Thus in the Hildebrandslied speeches are generally introduced with the formula: Hadubrant gimahalta, Hiltibrantes sunu, which is almost identical with a formula used in Beowulf: Wiglaf maÐelode, Weohstanes sunu. In the same German poem (v. 42) we find the verse: dat sagetun mi seolidante ('it has been told me by mariners'), with which we may compare Beow. 377: Þonne saegdon Þaet saeliÐende. Note should also be taken of such phrases as (v. 55) ibu dir din ellen taoc ('if thy prowess is sufficient') and definitely poetical expressions like (v. 43) inan wic furnam ('war carried him off'), as compared with Beow. 572 Ðonne his ellen deah and 1080 wig ealle fornam Finnes Þegnas. The number of such parallels might be greatly increased if we were to take into account passages from religious poems, especially the Old Saxon Heliand.

For a very much earlier period direct evidence is furnished by the Roman historian Tacitus, who says (Germ. 2) that the Germani possessed ancient poems or songs (carmina) even in his time and adds expressly that they had no other means of preserving a historical record[116]. That these poems were partly of what may be called a 'heroic' character is clear from another passage (Ann. II 88), where it is stated that Arminius was still a subject of poetry among barbarian nations (canitur ... adhuc barbaras apud gentes). In both cases the reference is in all probability to the peoples of western Germany rather than to the Goths.

On the whole then we need not doubt that the heroic poetry which we find in England and Germany during the seventh and eighth centuries had a long history behind it. Of course as to the form of the poetry current in the first century we are entirely without information. Many scholars hold that it was exclusively choric, not only in Tacitus' time but for four or five centuries later. This is one of the questions which we shall have to bear in mind in the course of our discussion.

The earliest historical reference to the cultivation of poetry, or rather perhaps minstrelsy, in England occurs in Bede's account of the poet Caedmon (Hist. Eccl. iv 24). In this story we are told that it was the custom that, when the villagers met together to drink and amuse themselves, everyone should take his turn in singing to the harp. Caedmon, who had never been able to learn a song, used to leave the festivities and make his way home as soon as he saw the harp coming in his direction[117]. No information is given as to the character of these songs. Probably they would as a rule be quite short—perhaps not much longer than the hymn learned by Caedmon from the angel, which contains only nine verses. It is not unlikely that they resembled some of the metrical riddles more than any other form of Anglo-Saxon poetry which has come down to us. Longer songs, of a narrative type, may of course have been known. But it is a question whether such songs would deal with heroic themes or with folk-tales. We may think of the Scandinavian story of Svipdagr and MenglÖÐ, which is preserved in a number of different versions, ranging in date probably from the tenth century to the seventeenth or later.

But we have already seen (p. 41 f.) that in the eighth century at least the recitation of heroic poetry was by no means unknown. Indeed we may infer from the language used by Alcuin (ridentium turuam in plateis) that it enjoyed a good deal of favour with the general public. Another of his letters[118] speaks to much the same effect. We may gather from the use of the word citharista that these recitations also were accompanied by the harp. But the language of the letters seems rather to suggest that the performers in such cases were persons who made minstrelsy more or less of a profession. A century earlier the existence of professional minstrels may be inferred from the well-known story of Aldhelm[119]—that he used to take up his position on a bridge, like a professional minstrel (quasi artem cantitandi professum), and sing to the people in order to call them back to church. We are not informed as to the character of the poems he recited, but clearly they were of a type calculated to attract the country people.

On the Continent we find very similar evidence. A passage in the Annals of Quedlinburg, to which we have already alluded and which dates perhaps ultimately from the tenth century, states that the country people used to sing of Dietrich von Bern[120]. From a much earlier period we have an interesting reference to a Frisian minstrel named Bernlef, who became a disciple of St Liudger. He had been blind for three years, when he was brought to the missionary, but "he was greatly loved by his neighbours because of his geniality and his skill in reciting to the accompaniment of the harp stories of the deeds of the ancients and the wars of kings[121]." This incident appears to have taken place before 785. That minstrelsy was definitely recognised as a profession among the Frisians is shown by the last clause in the Lex Frisonum, which fixes a special compensation for injury to the hand of a harpist[122].

From all this we gather that in the eighth century there existed both in England and Germany a class of minstrels whose practice it was to play the harp and recite heroic poetry in the village-streets or on bridges or wherever they could gather an audience. Now if we turn to the poems themselves we find that they also contain references to professional minstrels; but these appear to have been quite a different class of persons from those with whom we have been dealing.

At the close of his elegy (v. 35 ff.) Deor gives the following account of himself: "With regard to myself I will say that formerly I was the bard (scop) of the Heodeningas and dear to my lord. My name is Deor. For many years I have had a good office and a gracious master. But now Heorrenda, a skilful poet, has received the domain which the king had before given to me." There may be some difference of opinion as to the precise meaning of the word londryht[123], but it is clear enough that the poet had been a court-minstrel and that he had been supplanted in the king's favour by a rival. In Beowulf also we find mention of a person who seems to hold a similar position. The word scop occurs three times in this poem (vv. 90, 496, 1066)—always perhaps with reference to the same man. In each case he is represented as singing or reciting, and twice mention is made of the harp. On the last occasion, when he recites the story of Finn and Hnaef at the banquet (cf. p. 2), he is called HroÐgares scop, which seems to imply a sort of official position.

The case of Widsith is somewhat different. The poet is a traveller who prides himself on the large number of nations he has visited. He states also that he served under various princes by whom he had been handsomely rewarded. The poem ends with some reflections on the life of wandering minstrels; but these verses may be a later addition, like the introduction. At all events in v. 94 the poet speaks of his return home, when he presented to his lord, Eadgils prince of the Myrgingas, a valuable 'ring' which had been given him by Eormenric. This present was a reward to Eadgils for his kindness in granting the poet the land formerly held by his father. It would seem then that the poet is represented as a man of good position. Whether we describe him as a wandering minstrel or not, his occupation is clearly to be regarded as court-minstrelsy and different therefore from that of Bernlef and the harpists mentioned by Alcuin.

Apart from these personal notices there can be little doubt that the heroic poems which have come down to us were of courtly and not of popular origin. In the first place we may note their strongly aristocratic tone. This may be appreciated from the fact that all the women mentioned in the Anglo-Saxon poems are of royal birth, while the men are either princes or persons, apparently of noble or knightly rank, attached to the retinues of princes. On the rare occasions when persons of humbler rank are referred to, their names are not mentioned. In Beowulf no name is given even to the court-minstrel. Again the poems frequently refer to details of court etiquette, with which they seem to be well acquainted. In the later German poems this feature must of course be attributed to the conditions of the twelfth and thirteenth centuries, when the poems were composed. But it is quite as marked even in Beowulf. We may note especially the long and detailed account of Beowulf's arrival at the Danish king's hall and the conversation which the chamberlain holds with the king on the one hand and the visitor on the other, before the latter is invited to enter. The chamberlain's exact position is remarked, when he approaches the king and it is added that "he knew the custom of knighthood[124]." Other members of the court also have their position or duties described (vv. 500 f., 1161 f., 1165 f., 1794 ff.); but most of all the poet loves to picture the movements of the king and queen (vv. 612 ff., 920 ff., 1162 f.).

Then again it should be observed that persons of royal rank are very seldom spoken of with disrespect. The rare exceptions to this rule probably all refer to persons of a remote past, Eormenric, Thrytho and Heremod, and in the last two cases the reprobation is qualified in a very marked way. Moreover the ground of censure is invariably violence, cruelty or treachery. Of immoral or unseemly conduct we have no mention. Indeed, except in the story of Weland—which stands by itself in many ways, as we shall see later—such subjects seem to be studiously avoided. More than this the Anglo-Saxon heroic poems are entirely free from coarseness of language, such as we frequently find in Saxo's history, and indeed from references of any kind which could offend even the most fastidious taste. In general the same remarks are true also of the German and Scandinavian heroic poems, though not in the same degree. But the gnomic and theological poems of the Edda show a wholly different tone, which at its worst (e.g. in Lokasenna) verges on bestiality.

Lastly, we must not overlook the fact that in dignity and polish of style the heroic poems far surpass any narrative works which the English language has to offer for many centuries later. It has been remarked that the composition of epic poetry requires a more or less professional training, and in the case of such poems as Beowulf this is doubtless true, not only on account of its length but also because a very large vocabulary is needed for the constant interchange of epithets which is one of its chief characteristics, while the allusions with which it abounds point to the possession of much historical or traditional lore[125]. At the time when it was produced the knowledge and leisure necessary for such composition is scarcely likely to have been found outside the entourages of kings.

On the other hand we have seen that minstrelsy of some kind was cultivated even by peasants in Caedmon's time. We can hardly doubt that such was the case to a higher degree in court circles. In Beow. 867 ff. we find a 'king's thegn' composing an account of the hero's adventure immediately after its occurrence, and utilising apparently by way of illustration the story of Sigemund. This person may be the court minstrel; but the identity of the two is scarcely certain. In a later passage (v. 2105 ff.), referring to the banquet after the fight with Grendel, we hear of the king himself taking his turn with the harp: "There we had poetry and music. The old Scylding (Hrothgar) related stories of old time out of his great store of information. Now the martial[126] hero would lay his hands on the joyous harp, the instrument that makes good cheer; now he would recite a poem, true but sad; now a story of marvel would be related in due course by the magnanimous king. Now again, bowed with age as he was, the old warrior would begin to lament that he had lost the martial vigour of youth. His heart surged within him as he called to mind the manifold experiences of a long life." It is held that the reference here is to lyrical effusions rather than to anything in the nature of epic narrative[127]; but I am inclined to doubt if we are justified in totally excluding the latter. I would rather favour the view that the cultivation of minstrelsy, including narrative as well as lyric poetry, was a general accomplishment in royal households, and that the office of court-minstrel was an honour given to that member of the court who had attained the greatest proficiency in his art.

The statements of the poems as to the prevalence of court-minstrelsy during the Heroic Age are fully confirmed by the testimony of contemporary Roman writers. Perhaps the most important reference is a passage in Priscus' account of his visit to Attila in the year 448. After describing the banquet given by the king to his guests he proceeds as follows[128]: "When evening came on torches were lighted and two barbarians stepped forth in front of Attila and recited poems which they had composed, recounting his victories and his valiant deeds in war. The banqueters fixed their eyes upon them, some being charmed with the poems, while others were roused in spirit, as the recollection of their wars came back to them. Others again burst into tears, because their bodies were enfeebled by age and their martial ardour had perforce to remain unsatisfied."

It will be noticed that this account bears a curious resemblance to the passage from Beowulf which we have just quoted. Nothing is said as to the language in which the poems were composed[129], but at all events we need hardly doubt that in this as in other respects Attila was following Gothic custom. For the duet we find an interesting parallel in a passage of Widsith (v. 103 ff.): "Then Scilling and I began to sing with clear voices before our victorious lord; loudly rang out our music as we played the harp. Then it was openly confessed by many brave-hearted and experienced men that they had never heard a better song."

Evidence for the cultivation of minstrelsy at Teutonic courts in Gaul is furnished by letters of Cassiodorus (Variarum II, 40 f.) and Sidonius Apollinaris (Ep. I, 2). The former document is an answer from the Ostrogothic king Theodric to a request from Clovis, king of the Franks (d. 511), who had asked him to send him a skilled minstrel. Sidonius' letter is a very full account of Theodric II, king of the Visigoths, who reigned from 453 to 466. He states that the king seldom admitted jesters when he was dining, and that he took no pleasure in music except when it encouraged manliness of spirit as well as pleased the ear. In neither of these cases however is it certain that the performers were Goths.

A clearer case of Teutonic minstrelsy, dating from the same period, occurs in one of Sidonius' poems (Carm. 12[130]), in which the author complains that he has to live among troops of longhaired and greedy Burgundians, listening with polite attention, in spite of his disgust, to their songs. More than a century later we hear of Frankish court-minstrelsy in a neighbouring district. In a poem addressed to Lupus, duke of Aquitaine, about the year 580, Venantius Fortunatus (Carm. VII 8. 61) says: "Let the Roman sound your praise with his lyre and the barbarian with his harp[131]"; and again: "Let us frame verses for you, while barbarian poets compose their lays; thus the hero will be greeted with like honour, though in diverse strains[132]." In the introduction to his poems the same author complains of the constant buzzing of the harp, as it resounds to the barbarian lays[133].

A curious case of royal minstrelsy is recorded by Procopius (Vand. II 6) in his account of the siege of Mount Pappua (A.D. 534). Gelimer, king of the Vandals, wrote a letter to Faras, the Herulian chief who commanded the besieging army, begging him to send him a harp, a loaf and a sponge. The explanation given by the messenger was that the king had composed a song upon his misfortunes and as he was a good minstrel he was anxious to accompany it with a mournful tune on the harp as he bewailed his fate[134]. We are again reminded of Hrothgar, but perhaps still more of Hrethel's dirge over his son (Beow. 2460 f.).

These examples are sufficient to show that minstrelsy was widely cultivated in the courts of the Teutonic princes who had established themselves in Roman territory. Taken together with the references in Anglo-Saxon poetry, which deals of course with the more northern peoples, they leave no doubt that the prevalence of court-minstrelsy was one of the characteristics of the Heroic Age. For the existence of professional minstrels the Roman evidence is not so clear, though we may regard in this light the two 'barbarians' mentioned by Priscus (p. 84). On the whole the impression which we gain from our authorities is that the cultivation of the art was more or less general as in the north.

But it must certainly not be assumed that the poems of the period were always of an ephemeral character. Jordanes (cap. 5), in a passage to which we have already referred, says that the Goths "used to sing to the strains of the harp ancient poetry dealing with the deeds of their ancestors, Eterparmara, Hanala, Fridigernus, Vidigoia and others who are very famous in this nation[135]." Again, when he mentions the migration of the Goths to the Black Sea (cap. 4), he refers to "their ancient poems," where this event is commonly related in almost historical style[136]. From such expressions we must conclude that Jordanes knew, directly or indirectly, of Gothic poems which he believed to have been in existence for a considerable time.

Jordanes' language may be compared with Einhard's reference to the 'barbarous and very ancient poems' collected by Charlemagne (cf. p. 62). But in reality there appears to be an essential difference between the minstrelsy of the Heroic Age and that of Charlemagne's time. In the latter period we hear only of 'very ancient poems' or of poems dealing with 'the deeds of the ancients,' as in the story of Bernlef. The two expressions may really be more or less equivalent, for there is nothing to show that Bernlef was an original composer. For the existence in his time of court-minstrelsy, or indeed of any poetry dealing with contemporary persons and events, we have no evidence which can be called satisfactory[137].

In the Heroic Age on the other hand we have references not only to 'ancient poems' but also to original compositions, dealing with the praise or fortunes of living men. The exploits of Attila, one of the leading figures of that age, were sung in his own presence, as we know from an eye-witness. We need scarcely doubt therefore that Beowulf truly reflects the spirit of the times when it makes one of the Danish king's thegns compose a poem on the hero's exploit immediately after the event. It is to such compositions that heroic poetry—indeed in a sense we may say the Heroic Age itself—owes its origin.

The beginning of the process may be seen from a few passages in Widsith. At the end of the poem we are told that "he who wins praise (lof) shall have his glory (dom) established on high beneath the sky." The meaning of the first expression is explained by another passage (v. 70 ff.): "I have been in Italy with Alboin. No human being, as far as my knowledge goes, had a readier hand than had Audoin's son for the winning of praise, nor a heart more ungrudging in the distribution of rings and shining bracelets." We may compare also what the poet says of his patroness, the princess Ealhhild (v. 99 ff.): "Her praise spread through many lands, when I set myself to declare in song where under heaven it was that I knew of a gold-adorned queen who lavished presents in the noblest fashion." The chief object which the characters of the Heroic Age set before themselves is to 'win glory'—to have their fame celebrated for all time. Thus in Beow. 1387 ff. the hero says: "Let him, who can, win for himself glory before he dies; that is the best thing which can come to a knight in after times, when he is no more." Such glory may be won by brave deeds, as when Beowulf says to Hrothgar before his second adventure (ib. 1490 f.): "I will win for myself glory with Hrunting (a sword), or death shall take me," or again when the queen says to Beowulf (ib. 1221 ff.): "Thou hast brought it about that men shall esteem thee far and near and for all eternity, wheresoever the sea encircleth its wind-swept barriers." But the same object can be attained by generosity, which will ensure one's praise being sounded from court to court.

Now perhaps we are in a better position to understand why the Heroic Age ends when it does. The latest person mentioned in the heroic poems is Alboin who died about 572. The last Roman author who mentions Teutonic court-minstrelsy is Venantius Fortunatus, who wrote apparently about ten years later. Is there a connection between these two facts? It should be remembered that we have felt some hesitation in including Alboin among the characters of the Heroic Age, for though his praises were sung among the Saxons and Bavarians, as well as in England, he does not figure in any widely known story. We may reasonably expect that such stories would as a rule—not necessarily—require a certain time in which to be elaborated. Is it possible that in Alboin's time the conditions favourable to such elaboration were no longer in existence—that court-minstrelsy was dying out or had lost its creative power?

It will perhaps be urged that the absence of reference to court-minstrelsy after Venantius' time may be due to mere accident. But a short consideration of the political position will show that there is good reason for thinking otherwise. During the period which had elapsed since the time of Priscus and Sidonius—we may say roughly about a century and a quarter—the Teutonic world had undergone great changes. Many kingdoms had disappeared, among them those of the Huns, Rugii, Heruli, Alamanni, Thuringians, Burgundians, Vandals and Ostrogoths, and probably also the Warni and Gepidae. Of the nations which survived the Visigoths and Langobardi, and to a large extent the Franks also, were settled among alien peoples and thus exposed to denationalising influences. This is partly true also of the Bavarians, and they moreover had become subject to the Franks before the middle of the sixth century. Probably the only other Teutonic kingdoms which remained on the Continent were those of the Frisians and the Danes, for there is no evidence that the Old Saxons were under kingly government at this time[138]. Moreover the Danes were almost cut off from the western peoples by the irruption of the Slavs who now occupied the greater part of ancient Germany. Hence we may conclude that even if court-minstrelsy survived in a few places the poems had now no longer any chance of obtaining a wide international circulation.

The change of faith is of course another consideration which must be taken into account. One of its effects was to cut off the Christian kingdoms from those of the Frisians and Danes. Probably also it had an adverse influence on the cultivation of court-minstrelsy, for there can be little doubt that this was originally permeated by heathen ideas. At all events we find in later times surprisingly few traces of heroic poetry in the territories of the Franks, Visigoths and Langobardi.

Of the purely Teutonic kingdoms, excluding Denmark, that of the Frisians was the last to retain both its independence and its religion. It can hardly be due to accident therefore that some of the most important of the heroic poems, such as KÛdrÛn (cf. p. 34) and probably also the Norse version of the story of SigurÐr (cf. p. 59), appear to be derived from Frisian sources, though this region was not their original home. Further, we have noticed that in Frisian law a special compensation was fixed for injury done to the hand of a harpist. Still more significant is the fact that in the passage quoted above (p. 80) from the Vita Liudgeri, describing Bernlef's skill in reciting heroic poetry, one text adds the words more gentis suae. It would seem then that minstrelsy of this type was regarded as a distinctive characteristic of the Frisians and that heroic poetry retained its hold upon them at a time when it was little known elsewhere.

In England the conditions appear to have been quite different from those with which we have been dealing, for at the end of the sixth century this country probably contained more Teutonic kingdoms than did the whole of western and central Europe. We have seen reason for believing that Beowulf was composed within about half a century of Venantius' time and that the other heroic poems may date from the same period. From the evidence which we have discussed above we should naturally conclude that court-minstrelsy lasted somewhat longer in England than elsewhere, although it dealt entirely with stories derived from abroad. It is true that there is no external evidence for such minstrelsy; but that is fully explained by the fact that we have practically no literature of any kind before the last decades of the seventh century. Most probably its extinction was due to that wave of religious fervour which was started by the Kentish king Erconberht and which in the course of the following half century seems to have succeeded in enforcing conformity to the new faith throughout the whole country.

It will be convenient now to consider briefly the court-poetry of the Viking Age. The history of heroic poetry in the North unfortunately cannot be traced in its entire course. We have seen that there is a long gap, extending over some two centuries and a half, in Danish tradition, and also that the poems which have come down to us are probably all of Norse (Norwegian-Icelandic) origin. Yet the social conditions of the Viking Age were very different from those which prevailed on the Continent during the same period and unquestionably nearer than the latter to those of the Heroic Age. It is not unreasonable therefore to expect that the court-poetry of the Viking Age may throw light on the earlier period.

We saw in an earlier chapter (p. 15 f.) that, apart from the Edda, Old Norse literature is rich in narrative poems of the ninth and tenth centuries. These are usually the work of known authors and deal for the most part with contemporary persons and events, though they contain frequent references to characters of the Heroic Age, as well as to the ancestors of reigning princes. Many of the authors, such as ThiÓÐolfr of HvÍn, ThÓrbiÖrn Hornklofi and GoÐÐormr Sindri, were what we may call court-minstrels—or rather court-poets, for the harp seems not to have been used by such persons, at least in the latter part of the Viking Age. But they can scarcely be regarded as professionals in any strict sense of the term. As a rule they appear to have been men of good family. ThiÓÐolfr was a familiar friend of Harold the Fair-haired, who entrusted him with the education of one of his sons. GoÐÐormr Sindri, who composed poems for the same king, refused to receive any reward and had sufficient influence with Harold to insist on his being reconciled with his son Halfdan Svarti. Another poet of the same period, Einarr[139], commonly called Torf-Einarr, was earl of Orkney and practically an independent prince. Eyvindr Skaldaspillir, who was attached to the service of Haakon I and Haakon, earl of Lade, was himself a descendant of Harold the Fair-haired.

It has been mentioned that from the middle of the tenth century onwards most of the poems quoted in the sagas are of Icelandic authorship. A considerable number of them may be regarded as court-poems, since they were composed in honour of princes whom the authors were visiting at the time. As an example we may take a verse quoted by Gunnlaugs Saga Ormstungu (cap. 7) from the poem composed by the hero, when he visited London in 1001: "The whole nation reveres England's generous ruler as a god; all ranks, warrior prince and people alike, bow down to Aethelred." Such poems were often handsomely rewarded. Aethelred presented Gunnlaugr with a scarlet cloak, lined with fur and embroidered with lace, while Sigtryggr, king of Dublin, gave him a fur-lined cloak, a lace-embroidered tunic and a gold ring which weighed a mark. It happened very frequently that men like Gunnlaugr would enter a king's retinue and remain with him for months and even years. But they would seldom consent to recognise any lordship permanently, since as a rule they had lands of their own in Iceland, to which they eventually returned.

It is hardly probable that any class of persons exactly corresponding to this existed in the Heroic Age itself, for with the somewhat doubtful exception of the Old Saxons we have no evidence for independent commonwealths during that period. In Widsith, it is true, we have the case of a minstrel who claims to have wandered far and wide and to have visited many princes by whom, like Gunnlaugr, he was handsomely rewarded. But Widsith had a lord at home to whom he subsequently returned. Indeed the introduction, if we may use it as an authority, seems to make him set out at first on a definite commission from that prince. The permanent lordless state was probably altogether foreign to the conditions of the Heroic Age. The lordless man in the poems is either one who has lost his lord, as in the Wanderer, or one who has been dismissed from his lord's service, like Deor. Until he finds another lord he has neither home nor security, and his condition is pitiable in the extreme.

But it is by no means so clear how the court-minstrels of the Heroic Age differed from the Norwegian poets of the Viking Age. In KÛdrÛn the minstrel HÔrant (Heorrenda) seems to hold a position quite comparable with that of ThiÓÐolfr or GoÐÐormr Sindri; indeed he is even described as a relative (mÂc) of the king[140]. Again, we may take the case of StarkaÐr. He is commonly regarded as the type of a roving poet-warrior of the Viking Age. But in reality he seems to belong to the Heroic Age; for however late the poems attributed to him may be in their final form, they had their counterparts in England probably as early as the sixth century. There is no conclusive evidence for denying either that he was the foster-father of Ingialdr (Ingeld) or that he entered the service of a number of different kings.

The story of StarkaÐr is bound up with the history of heroic poetry in the Viking Age—a difficult problem, to which we shall have to refer again shortly. We may note here however that the poems of ThiÓÐolfr and Hornklofi must have been preserved for some three centuries by oral tradition before they were committed to writing. There seems to be no reason therefore for denying in principle the possibility that poems may have survived even from the Heroic Age. It is to be observed that court-poets were expected to be able to recite old poems, as well as works of their own composition. Thus on the morning of the battle of Stiklestad (A.D. 1030) St Olaf ordered the Icelander ThÓrmÓÐr KolbrÚnarskald to recite the old BiarkamÁl. This story is interesting as it shows that the love of the heroic poems was strong enough to assert itself in an hour of supreme danger and under a most religious king.

In the course of this chapter we have seen that one of the characteristics of the Heroic Age was the prevalence of court-minstrelsy of a certain type, namely the recitation of metrical speeches accompanied by the harp. The cultivation of such minstrelsy seems to have been more or less general, and it is certain that princes had their praises and exploits celebrated in poems of this kind during their lifetime and even in their presence. But with the close of the Heroic Age the evidence for minstrelsy of this type apparently ceases altogether. In the eighth century we hear only of wandering minstrels who are invited into houses or perform in the streets. The minstrelsy of this period seems not to have been creative. At all events it deals only with characters belonging to former times, Ingeld or Alboin or ancient kings in general. Between these two periods we have to set the composition of the English heroic poems and probably also those German poems which were regarded as ancient in Charlemagne's time. Lastly, we find in Germany a series of poems dating from the twelfth and thirteenth centuries in which the old stories are treated again but in a new form.

It appears then that the history of heroic poetry falls naturally into four stages. To Stage I belong the court-poems of the Heroic Age itself; to Stage II the epic and narrative poems based on these; to Stage III the popular poetry of the eighth and following centuries; to Stage IV the German poems of the twelfth and following centuries, composed at a time when heroic subjects had again come into favour with the higher classes.

To Stage I we may assign not only laudatory poems dealing with the victories and valour of living princes, but also such compositions as Gelimer's dirge and choric songs like the funeral chant over Attila[141]. From this stage probably nothing has come down to us—though it would be difficult to point out any essential difference between Gelimer's dirge and the Elegy of Deor. We can form an idea however of these earliest poems from the poetry of the Viking Age, which seems to have been composed under very similar conditions. As instances we may cite Gunnlaugr's poem on Aethelred II (cf. p. 91 f.) and Eyvindr's poem on the death of Haakon I, which we shall have to discuss later.

Stage II is represented by the Anglo-Saxon poems, which are clearly products of court-life, as we have seen (p. 81 ff.). From its general resemblance to these it seems probable that the Hildebrandslied belongs to the same class. Some writers draw a distinction between Beowulf and Waldhere on the one hand and Finn and the Hildebrandslied on the other, classifying the former as epics and the latter as lays (Lieder). It may be granted that the style of the two latter poems appears to be more rapid and less diffuse than that of the others. Still I should prefer to speak of short and long epics, or rather perhaps of short and long narrative poems. Very probably the earliest narrative poems were comparatively short. It may be that poems on the scale of Beowulf were first composed in England—though this can hardly be proved. But the difference between the two classes seems to me to be one of degree and not of kind. At all events no one will suggest that the Hildebrandslied is even approximately contemporaneous with the events which it professes to describe. One would expect it to be at least as remote from them as is the case with Beowulf.

Stage III is directly represented only by certain ballads such as the Seyfridslied, which in their present form date from a time considerably later than the poems belonging to Stage IV. Much indirect evidence however can be obtained from various sources of earlier date, e.g. from ThiÐreks Saga af Bern, which is largely based on the popular heroic poetry of northern Germany, and from parts of Saxo's History which seem to be derived from Danish ballads. So far as we can judge from our authorities the popular poems seem to have differed in many ways from those which we have been discussing. They tended to simplify complex stories by the loss of minor characters and to amalgamate stories which were originally quite unconnected. Again, they appear to have had a preference for biographical sketches, whereas the court poems are usually occupied with accounts of adventures which lasted only a few days. We may add also the absence of any detailed acquaintance with court-life and a general approximation to the characteristics of folk-tales, e.g. in the introduction of nameless characters and persons of humble station.

It must be remembered of course that our authorities knew the popular poems only as they existed in the twelfth century[142]. We cannot say with any confidence that Bernlef's poems possessed the same characteristics. There is nothing to show that the Hildebrandslied was written down before his time, and it may be to persons of his type that we owe its preservation. The Anglo-Saxon poems may not have been committed to writing till a still later period. All that we can say is that they show no obvious signs of popular corruption and that their diction is much more archaic than that of poems which were composed in the eighth century. We have seen that in Iceland poems of a rather elaborate type could be preserved by oral tradition for over three centuries. This however was in a community which was largely given up to the cultivation of poetry. A knowledge of the old poems would be a necessary part of the training of those who hoped to win rewards for their art in foreign courts. Such favourable conditions can hardly have existed either in England or Germany. The process of disintegration which the poems underwent in the latter country points to their being preserved only by village minstrels, who as time went on became less and less expert in their profession.

Stage IV is represented by the Middle High German epic poems, which both in form and spirit show all the characteristics of the age in which they were composed. In England this stage was never reached. There may have been a revival of interest in heroic poetry during the ninth and tenth centuries, but we have no evidence for the composition of new poems on these subjects.

There can be no doubt that the poems of Stage IV are derived from those of Stage III. But the question may be raised whether the latter were necessarily derived from poems of Stages I and II—whether some heroic poems may not have been entirely of popular origin. It may be freely granted that the poetry of Stage II was constantly exposed to popular influence, especially in the form of folk-tales. Most scholars indeed hold that some of the best known heroic stories, such as those of SigurÐr-Siegfried and Weland, are derived from popular mythology. With this problem we shall have to deal later. On the whole however I am inclined to doubt whether we possess a single heroic story which has not been treated in court-poetry at an earlier stage in its career.

Again, the relationship between Stage I and Stage II is not so simple as the bare statement given above might seem to imply. In the first place it is only as a class that poems of Stage I can be regarded as the earlier. Individual poems may very well be later than others which belong to Stage II. Thus it is extremely probable that Gothic princes were listening to laudatory poems about themselves at a time when other Gothic poems, of a definitely narrative type, were coming to be regarded as ancient. Then again, poems of a more or less narrative type may have been composed quite soon after the events which they treat. Hornklofi's HrafnsmÁl and the poem on the battle of Brunanburh must be regarded as analogous to poems of Stage I, but it is only a short step from such works to purely narrative pieces. Indeed, in a sense, we may almost class among narrative poems Eyvindr's HÁkonarmÁl, which describes the fall of Haakon I at the battle of Fitje. Yet the author was himself present at the battle. It is by no means clear then that poems belonging to Stage II necessarily presuppose the existence of earlier poems or indeed of materials of any kind beyond the author's personal knowledge and imagination.

But, more than this, we are scarcely justified in denying the possibility that even epics may have been composed upon quite recent events. Few will deny that the poem on the battle of Maldon has a good claim to that title, whatever its original length may have been. The extant portion contains nine speeches, by seven different persons. Twenty-two warriors in the English army are mentioned by name, and in about a dozen cases the names of their fathers or other relatives are also given. The poem differs from the heroic type in the fact that it does not record the name of a single person among the enemy; but that need not prevent us from regarding it as an epic. Yet there can be no reasonable doubt that it was composed within a few years, possibly even months, of the battle[143].

It is likely enough that the author of this poem was well acquainted with heroic poetry and that his treatment of the subject was affected thereby. But is that any objection to supposing that poems of this type may have been composed within the Heroic Age itself? We have no reason whatever for denying that this age was capable of such compositions. Indeed there is one piece of evidence which points very much to the contrary.

This is a passage which occurs in Procopius' History of the Gothic War (IV 20). After describing the embassy sent to Justinian by the Utiguri in the year 551, the author goes on to state that 'about this time' hostilities broke out between the nation of the Warni and the Angli (???????) who inhabit the island of Britain (???tt?a). Not long before the Warni had been ruled by a king named Hermegisklos who, being anxious to establish his throne on a secure foundation, sought and obtained in marriage a sister of Theodberht, king of the Franks. By a previous wife he had an only son named Radiger, who at this time was betrothed to a sister of the king of the Angli and had paid her a large sum of money in furtherance of his suit[144]. One day, when the king was riding in a certain place with the chief men of his nation, he saw a bird sitting on a tree and croaking loudly. Now, whether it was that he really understood what the bird said, or whether he had some other source of information but pretended that the bird was uttering a prophecy which he understood—at all events he declared on the spot to his companions that he would die forty days later, for this had been clearly foretold to him by the bird[145]. Thereupon he gives advice as to what should be done after his death, namely that Radiger should marry his widow, in accordance with national custom, and dissolve his engagement with the English princess. On the fortieth day the king died, and Radiger proceeds to carry out his injunctions. But the princess, infuriated at his conduct, gathers together a fleet of four hundred vessels and invades the land of the Warni with 100,000 men. Radiger's army is completely defeated, and he takes refuge in a dense forest. The princess insists on his being taken alive at all costs, and eventually he is captured and brought before her. He expects to be put to death, but after explaining the cause of his action, he is pardoned on condition that he returns to his former engagement.

This story contains a number of features, such as the prophecy of the bird, the payment of the 'bride-fee' and Radiger's marriage with his stepmother, which show clearly that it was derived from someone who was well acquainted with the peoples of northern Europe. There is no ground for disputing that it has a historical basis; but at the same time it is obviously much more than a mere record of facts. Apart from the incident of the bird and the gross exaggeration apparent in certain details[146], the pictorial character of the narrative, especially in its earlier part, indicates a close affinity with heroic poetry. Un-heroic features are not wanting—the introduction of nameless persons and the political reflections of Irmingisl—but they may with probability be attributed to Procopius himself. I do not mean to suggest that his source of information was an epic poem, but I do think that the intellectual atmosphere which could produce such a story must have been exactly of the kind most favourable to the growth of such compositions.

Procopius, as we have seen, places this war in or about the year 551, and we can certainly understand the course of events more easily if it took place after Theodberht's death (A.D. 548). Therefore, since Procopius' work appears to have been written within the next seven or eight years, the story had had little time to develop before it came to his ears. On the other hand we must remember that it refers to a distant region—a fact I think not without significance for the history of contemporary narrative poetry. To persons who had themselves taken part in the events poems like that on the battle of Brunanburh would appeal much more strongly than purely narrative pieces. It is for persons who were either ignorant of the events or knew them only by hearsay that the latter would seem to be primarily intended.

In this discussion we have taken no account of the heroic poems of the Edda. There can be little doubt that these poems must be assigned to Stage IV of our scheme; for though some of them are probably three or four centuries older than the German poems of this class, they bear fairly obvious marks of the disintegrating process which seems to characterise popular poetry[147]. It may be asked how such an opinion is compatible with the view that the story of SigurÐr was derived from the Continent during the seventh century (cf. p. 59). The true explanation is, I believe, that heroic poetry passed through Stage III after it became known in Norway.

We noted in an earlier chapter (pp. 30, 33) that Norway plays no part in stories of the Heroic Age. On the other hand it has what we may call a Heroic Age of its own—namely the Viking Age. We have no poems—and few stories of any kind—dealing with persons of Norwegian nationality who lived before that period. The remoteness, poverty and mountainous nature of the country doubtless retarded its development, not only politically but also in the cultivation of poetry. The strophic character of all Norse poetry is generally held to point to a choric origin, and it may very well be that this primitive type of poetry was the only one used in Norway when the heroic stories first became known there. Possibly it was to Frisian and Danish minstrels that the change was due[148]. But the conditions suitable for a flourishing court-poetry, like that of the Heroic Age, scarcely existed before the last years of the eighth century, when Norwegian princes had begun to enrich themselves with maritime enterprises.

[116] celebrant carminibus antiquis, quod unum apud illos memoriae et annalium genus est, etc.