We may thus get ‘false archaisms’ – which are really natural parallels, if sometimes more violent and arbitrary, to the shifts of meaning of living words in ordinary linguistic tradition, owing to the misleading suggestion of familiar contexts. We may also get, as in the language of everyday, the fossilizing of words, embedded in a phrase or expression from which they can no longer be removed. The learners of a traditional poetry are, when confronted by rare words or by hapax legomena, in much the same position as are later students, though they rely on memory, having no printed Sprachschatz. Imagine them confronted with medostigge mæt mægþa hose (Beowulf l. 924). This line contains two hapax legomena.[6]

  Now medostig gives us little difficulty. We can analyze it. Although we may feel it a little bold in its ellipsis, we are sufficiently familiar with the ‘mead-hall’, and the poetic way of putting things oddly, to feel certain that ‘measured the mead-path’ means ‘paced the path towards the mead’ (sc. mead hall). The Abingdon road is after all the road to Abingdon. And so if our fancy is tickled by this, we can repeat it with variations. We know all about the elements: medo and stig are both familiar words we can place in any required context, or replace with equivalents. But mægþa hose is different. The construction is plain enough. We know mægþa. But hose we do not. Our only clue is that it must, in such a context and construction, mean vaguely ‘a company’. But this is not enough for free use. Imagine ourselves unequipped with any philology or knowledge of other Germanic tongues. Reference to Gothic or Old High German hansa was not available to the Old English poet. We do not then know (a) whether it is a general company word, or one suitable only to maidens, (b) what its form should be in any context when it is not used like this – in modern terms we do not know its gender or declension, or what its form should be in any case other than the (comitative) dative. In all probability then, if we did not simply forget it because of its rarity, we should only dare to use it in precisely similar contexts, and in the fossilized expression mægþa hose. Old English and Old Norse verse offer many examples of this close repetition.

  The case is similar with archaic forms. Some are general and their function, and equivalence with current forms, widely understood. Some are preserved only in certain phrases. We need not delay longer over this point; since language in general, and the language of our day as of any other, offers plenty of parallels. All know the use of thou and the corresponding verbal forms in est, and also art, wert, etc. They are preserved in the language of liturgy, and verse. We do not use them naturally, and may make mistakes in them, or find some now so odd that we avoid them (as resistedst), but they are available and preserved by ‘tradition’. But we are not obliged to use them. Tradition, when not reinforced by writing and by formally taught grammar, may preserve archaic forms, but it cannot dictate their use. Our feeling that if one uses thou art in a poem, one should not later (without some special reason) relapse into you are, is a sophisticated one. Under Old English conditions one was rather always free to fill the prescribed metrical scheme with current material which according to contemporary use and pronunciation obeyed the rules; one could call on certain traditional forms to assist in metrical obedience or to give the flavour of ‘poetry’, according to convenience and taste. Of abstract lexicographical teaching, there is in Old English no record. It is doubtful whether learners ever heard statements in schoolmaster terms such as: ‘hæleð means “man” and if convenient you can use hæleð also as plural instead of hæleðas; compare monaþ and monþas.’ Such abstracted teachings with regard to the element of poetic diction are not, at any rate, a necessary part of verse tradition. We do not really even now for all our dictionaries, grammars, and graduses, normally or necessarily acquire our poetic diction in this way. A way which marks the birth of philology – though that has naturally always been connected, especially in origin, with the study of traditional or archaic verse, and particularly with metre, whether in bardic lore or more recent philological investigations of Homeric and alliterative verse. Mnemonic lists of verse-words for horse, or helm or man and so forth may have existed. They existed in Old Norse, though to a certain extent their compilation is rather evidence of the decay of Skaldic craft than of its flourishing, due to antiquarian piety rather than practical poetical education.

  Let us now return to the metre and the sounds. If you believe still that metre exists, or existed, only in particular examples of supposedly metrical composition, there is no difficulty. For then all such traditional stuff is simply metrical, and that is all you can say: you can neither approve nor criticize its ‘metre’ at particular points; for that requires some standard independent of the example. And even the attempt while listening to hear verse metrically – to find the relation to the norm of the pattern of lines that may seem aberrant – implies the belief in the existence of a pattern in the abstract. So if this is your belief all you can do in listening is to perceive vaguely a rhythm (as you might in engines, or the beat of hoofs, or birds’ singing). And all you can do in emulation is to pour out words in an easy uncritical flow, and you can call it a rhythm if you wish. But that was certainly not the way Old English verse was written.

  If you believe that metre was to the ancient poets a recipe independent of the pudding, then you have to consider how they got to know of the recipe; and also what happened when traditional verse, which in course of generations had changed its pronunciation, seemed here and there not to obey the recipe. As to the first question: We do not know. But it is probable that it was not in kind different from the way in which poets now acquire ‘metres’, though it may have been less haphazard; for a knowledge of such matters was probably more general and more generally esteemed. In other words, with the memorizing and hearing of verse there went doubtless also a certain amount of actual abstract description of the rules. Not organized, perhaps, or possessing a whole vocabulary of technical terms – though these possibly existed, as they did in Icelandic – but still a laying down of certain laws by which composition was to be directed, and verse judged. And certainly there existed the tacit assumption that there were ‘rules’, that verse could be regular or irregular, good or bad metrically.

  It is unlikely that there was any definite ‘profession’ of minstrelsy, which one entered by apprenticeship and by which one after gained one’s sole livelihood. Certainly there was no rigid bardic organization; and equally certainly one could acquire the mysteries of verse and make and recite it with approval without belonging to any craft. So it was in Norway and Iceland. Widsith, the imagined minstrel of the compiler of the necklace of legendary kings and heroes in the Exeter Book, expresses his gratitude for praise and substantial reward, and whither so he wandered he brought out his songs hoping for both. But he was a king’s ambassador, and his reward was princely, and he gave it to his feudal lord Eadgils even as the warrior Beowulf gave the rich gifts of Hrothgar to Hygelac; for both received their inherited lands from their kings. Egill Skallagrímsson was no less eager for reward and praise; but he was a great warrior, of ancient house, a chieftain in Iceland. He was certainly a scop or a skáld, but he was not an initiated bard.[7]

  Where then did he learn the rudiments? In his own home. It was the custom in that house and the houses of his father’s friends that people should amuse themselves over their drinking by reciting and making verses.[8] And such also was the custom in ancient England. In Bede’s story Cædmon was very odd in being unable to make even a moderate contribution to the general play. It was his peculiar incompetence – so unusual that even a cowherd was covered with shame because of it – that threw his later miraculous inspiration into relief. A point rather missed now, when the making-up of verse (and even the reciting of it from memory) is so odd that it is rather versifying that covers men with blushes.

  In this fireside school both metre and diction could be learned. The earliest efforts of Egill are reported. There is no real ground for robbing him of their credit, even if we may believe the age at which he composed the difficult stanzas in ‘court metre’ – Kominn emk til arna (xxxi) and Síþögla gaf söglum (xxxi) – to have been exaggerated in earliness. But the stanza Þat mælti mín móðir (xl) attributed to him when about seven has the ring of truth. In Iceland at any rate a boy of talent could master the outlines of dróttkvætt, a much more complicated metre than the Old English epic verse, while still young and living in his father’s house. And the only later schooling in such matters he got or needed, beyond doubtless listening, and learning verse by heart, was the companionship and rivalry with other skálds in the Norwegian court.

  Now let us examine the second question and consider the effect of old poems upon people who believed in metre and knew it. Alliterative verse is specially useful in such an enquiry. Certain of its features are so readily seizable that it is impossible to imagine that they were not known to be ‘rules’ apart from poems – for example, at least it must be admitted that everyone knew that such verse must ‘alliterate’, that is have a minimum of two words a line beginning with the same consonant or with a vowel;[9] and that lines that did not have this alliteration would not seem ‘right’. On the other hand this important feature of alliteration is very liable to disturbance by phonetic change, and yet is at the same time naturally intolerant of inaccuracy – of anything in the least comparable to traditionally inaccurate end-rhymes, or to the preservation as eye rhymes of the changed rhymes of past centuries. It is naturally intolerant of such things because in Old English verse it is used economically, and primarily not as an adornment but as a fundamental structural feature, linking the hemistichs into a compact line and often defining the way in which they are to be metrically analyzed: without it a line is more unacceptable than a couplet that totally fails to rhyme in heroic verse. But this essential[10] alliteration depends in principle upon a correspondence, readily recognized by the ear, between two sounds, often brief and not sonorous (like k or p), widely and yet variably spaced. Plainly inaccuracy will be much less tolerated than in end-rhyme, consisting of the recurrence of a sequence of sounds (including a vowel), normally at regular intervals, and often repeated many times. It would be interesting then to discover if we could what people, while alliterative verse was still current and living, made of old verse that phonetic change had damaged. One thing seems plain and that is that they could not keep up archaic pronunciation so as to preserve the lines. We are not dealing with traditional pronunciation of sacred books, with Vedas or Bibles; but with a general oral tradition, not devoted as far as we know to the reverent repetition of poems regarded as ‘classics’, but to the practice of the ars poetica, to the rules of verse and its adornment. And even the Latin of Scripture and liturgy were not in the early west preserved from phonetic change by a clerkly tradition of pronunciation. Latin pronunciation was taught, and exercised the schoolmasters a good deal; but it was in doubt, for it had been affected by the varying fortunes of sounds in different vernacular areas.

  If a man who knows ‘metre’ is confronted with a poem in archaic language he is bound to come across lines that are no longer ‘right’. He can then assume error, or if he is an acuter critic perceive that change has occurred. In either case in dealing with an oral poem he may change the line or passage so as to make it ‘right’ according to contemporary usage. This must often have occurred. In purely oral tradition it must have normally occurred. Though the loss of final u and i in certain conditions is a change that probably occurred long after the English came to England – it is normally ascribed indeed to the seventh century; and though such poems as Beowulf, and with perhaps more credibility other pieces such as Widsith, are widely held to contain embedded in them very ancient matter – still the critics look in vain for any certain case where a defective line occurs that only requires the restoration of this u or i to make it scan. Which is all the more remarkable in contrast with the loss of medial h. This is still in place in the earliest written document of English; but as a full consonantal sound it must have disappeared as well during the latter part of the eighth century, though the contraction of the resulting hiatus into one long syllable (that was not metrically appreciable as two) as in giþiohan>geþio’an>geþeon, must have taken place somewhat later. Now two things are here to be noted: (1) that defective lines require the restoration of the older syllables as man geþeon in Beowulf are common in the older poetry but (2) their imitation in verse demonstrably written in a period after the contraction was an accomplished colloquial fact is nowhere to be seen.

  The deduction from this is, I think, that we are allowing our thought to be confused between oral and scribal preservation. Writing down – probably a somewhat exceptional occurrence – has preserved for our perusal some poems originally written in an early period. But this was no part of normal tradition. Before ever our West-Saxonized copies were made the poems were dead – as far as oral tradition goes. We have no idea what they would have become, if England had not been ravaged, and they had descended by living repetition until say 950 or 1000 or later – then to be collected, like the Elder Edda, by amateurs of the past, on the verge of the Middle Age, and the invasion of new modes.

  Oral tradition contributed to their past – in the case of such long formal poems as Andreas or Beowulf, provided them with material – but however much traditional matter and phraseology they used, their authors were scholars who knew the art of verse, and such things as these (Beowulf, or Andreas) were never in this form oral, but made and soon written down: their later history has been in the scriptorium. Thus the upper limit for the composition of Beowulf is after the loss of u and i; the lower before the completing of the contraction of þeo(h)an > þeon. And before that later date Beowulf must have been in writing, mummified, safe from the corrosion of oral tradition, but not from the worms of scribal editing. It is even likely on various grounds – e.g. direct errors in West Saxonization, due to misunderstanding of archaic spelling, present in our text – that its scribal tradition has not been continuous, but has had a leap: transcription in the late OE period from an MS made in the late eighth or early ninth. There is indeed no need to assume more than two copies between our text and the first writing down.

  Now this gives extraordinary interest to the question of alliteration on g in Old English verse – a question that has not been given the attention it deserves. For here we have another phonetic change (directly affecting alliteration not scansion this time) occurring later, right in the full tide of Old English and in the very century of most of our manuscripts, the tenth. Again I think we can show that certain poems were composed and their texts written before that event, and not rehandled by oral tradition after. But others were either made after that event or else entirely remodelled after it. All the older poetry treats g of all varieties as equivalent. This is still so in Judith. None of the Chronicle poems do – not even those in strict metre (937, 973, 975, 1065); and it can be shown that for the author of Maldon also the front and back g were distinct non-alliterating sounds.

  I will take another point. If secondary ‘stresses’, certainly present in the archaic period, seem in the freer verse of Maldon sometimes neglected, sometimes observed, this is to be attributed rather to a general looseness of metrical structure, and not (as I used to be taught) to the sporadic use of archaistic half-lines preserved by tradition. Thus the repetition of old poems, or parts of old poems, might preserve, say, ham siðian ‘to journey home’ in memory, and also reveal that this had once been used to fill a hemistich. But if it did not still fill a hemistich according to current pronunciation, it would not be used by people who knew the rules (and were trying to obey them). Either the phrase must be altered or added to so as to conform to the rules – very easy to do – or the rules must be altered to take in the example. The latter procedure is in Old English unlikely and in any case is nowhere seen clearly to have happened.[11] And still less likely in Anglo-Saxon conditions is it that ‘tradition’ would not only preserve ham siðian but would preserve in bardic lore knowledge that such cases were to be rectified by a pronunciation: hám síðìan – if and when such a pronunciation had in fact ceased to be used. Since tenth century poets eschewed the alliteration of front and back g – although their non-equivalence would render 100 lines or more of Beowulf defective – because they did not in fact any longer in the current tongue alliterate, while in every point of scansion and diction their verse was encrustedly traditional, it seems plain that on this point, bygone sounds, a tradition had little to tell.

  Now in Maldon we meet Nu mæg cúnnìan 215, a mæg gnórnìan 315 – so to be scanned according to the probable metrical interpretation of the lines in which they stand; and we have: hám síðìe 251 which must be so interpreted. Compare hám síðìan in Genesis 2161. But this seems to conflict with such examples as siðian mote 177, which might appear to be a normal A-hemistich: sídĭăn móte. But this cannot be explained by arguing that hám síðìe is ‘traditional’ and archaic, while sídĭăn móte is a contemporary, and ready-made fresh A. You cannot call in the rules to explain the setting aside of traditional pronunciations in favour of the current, at the same time as you are calling on tradition to explain the non-current pronunciation in other cases. If tradition was strong enough to make ham siðie scan, it was strong enough to make siðian mote over-long. The reason is rather that hám síðìe was still so pronounced, and still a sufficient (if minimum) line-filling; whereas síðìan móte was a maximum line filling: Sievers A2,[12] seldom indeed found in the second hemistich in Beowulf, and is normally lightened in the first by double alliteration. The frequent use of these heavier groups in places where they are usually avoided in say, Beowulf, is one of the prosodic marks of this kind of freer verse. Other examples are stiðlice clypode 25b, 234b, 265b, and with anacrusis as well 72b. Related to this is the not infrequent occurrence of anacrusis in the second hemistich before A and E types: as 72 already cited, also A mid gafole forgyldan 32b, 242b, etc.; E and ne forhtedon na 21b, 49b. (Similarly A: 11, 55, 66, 68, 96 146, 189, 231, 240). Many of these could be emended by the omission of small unnecessary words, such as his, to, mid, or slight alteration as ealle gemanodest (for hafast ealle gemanode) 231; but not so 21, 49, 96, 189, 240, 242. Whatever be the reason of these divergences[13] from strict verse (and others not yet dealt with) it is plain that it cannot be mere incompetence, or ignorance of the author. The great mass of the poem is in verse that, judged by Beowulf even, is regular, though it is less compact, and the style freer and less studied. It is going beyond the evidence therefore to say as Sievers does that the author of Maldon (grouped with those of Solomon and Saturn and the metrical Psalms!) is one who ‘had only an imperfect command of the old rules’. This can only mean either he did not quite know what they were, or else he knew and could not conform in all cases. The latter alternative is absurd in the case of Maldon. The former almost equally so. Why he should have been unable to get knowledge of rules which others before and after him knew (and which Icelandic children acquired in the farmhouse) is difficult to say. And still more difficult is it to explain the enormous preponderance of lines that do conform – this could only be due to his poem being in the main a string of pure reminiscences or quotations of ancient verse: and that is in Maldon clearly not the case.

 

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16