supplanted by other synonyms in prose and in common speech; nor is this different from what we find in the heroic poetry of (for example) Virgil and Milton. It is also true that modern dictionaries still include after definitions of many uncolloquial words the quasiapologetic explanation ‘obsolete except literary’ or ‘archaic and poetic’. But such dictionary descriptions are relics of a rule of English that has been (temporarily at least) suspended. The reaction against ‘rhetoric’ has been so strong that present-day readers of poetry will not now readily accept as ‘poetic’ anything that is ‘archaic’, as ‘literary’ anything that is ‘obsolete’. And here is the crux of the modern translator’s problem, which Tolkien does not estimate at its most formidable. There is no use being faithful to the poetic archaisms of the original if the result cannot be couched in terms acceptable to one’s poetic co-readers and co-writers. If it is a case of losing an archaism or losing the poetry, the archaism must go. Whatever the tradition of the original poetry may have been, the translator’s duty is as much to speak to his own age as it is to represent the voice of a past age: these are, indeed, equal tasks. If the poet of Beowulf was, like Virgil, sometimes amantissimus vetustatis, there is no fear of the flavour of an antiquity, a weight and a grandeur, and a melancholy of history, being lost in a version that uses current English; for the background cannot be modernized – the social system, the customs, the entertainments, the supernatural beings, the battles and the weapons . . . and in the inescapable bedrock vocabulary of king, lord, and retinue, gold-giving and mead-drinking, coat of mail and dragon and burial-mound, the twentieth-century reader will find enough that is remote from his own experience without any superadded linguistic crinkum-crankum and mock-epopeanism.
The metrical problem has been as much discussed as that of diction has been neglected. In the main, this problem has been seen as one of carrying over a poetic tradition based on accent or stress and syllabically flexible into a basically syllabic tradition. (The distinction between ‘long’ and ‘short’ syllables, important in the old metric, holds little or no aesthetic meaning for the modern reader’s ear, and is usually not taken into account; even if experimentally reproduced in modern verse, it adds no pleasure to it; its significance was lost when the Old English inflections were lost.) Solutions of the problem may be distributed into three principal groups: blank verse, rhyming syllabic metres with heroic or semi-heroic associa-
xvi