Its special association by the English with the Britons was a product of their invasion of Britain. It contained a linguistic judgement, but it did not discriminate between the speakers of Latin and the speakers of British. But with the perishing of the spoken Latin of the island, and the concentration of English interests in Britain, walh and its derivatives became synonymous with Brett and brittisc, and in the event replaced them. [18]

In the same way the use of wealh for slave is also due solely to the situation in Britain. But again the gloss 'slave' is probably misleading. Though the word slave itself shows that a national name can become generalized in this sense, I doubt if this was true of wealh. The Old English word for 'slave' in general remained theow, which was used of slaves in other countries or of other origin. The use of wealh, apart from the legal status to which surviving elements of the conquered population were no doubt often reduced, must always have implied recognition of British origin. Such elements, though incorporated in the domain of an English or Saxon lord, must long have remained 'not English', and with this difference preservation in a measure of their British speech may have endured longer than is supposed.

This is a controversial point, and I do not deal with the question of place-names, such as Walton, Walcot, and Walworth, that may be supposed to contain this old word walh, [19] But the incorporation in the domains conquered by the English-speaking invaders of relatively large numbers of the previous inhabitants is not denied; and their linguistic absorption must have steadily proceeded, except in special circumstances.

What effect would that have, did that have, on English? It had none that is visible for a long time. Not that we should expect it. The records of Old English are mainly learned or aristocratic; we have no transcripts of village-talk. For any glimpse of what was going on beneath the cultivated surface we must wait until the Old English period of letters is over.

Unheeded language without pride or sense of ancestry may change quickly in new circumstances. But the English did not know that they were 'barbarians', and the language that they brought with them had an ancient cultivation, at any rate in its tradition of verse. It is thus to the appearance of linguistic class distinctions that we should look for evidence of the effects of conquest and the linguistic absorption of people of other language, largely into the lowest social strata.

I know of only one passage that seems to hint at something of the kind. It refers to a surprisingly early date, a.d. 679. In that year the Battle of the Trent was fought between the Mercians and Northumbrians. Bede relates how a Northumbrian noble called Imma was captured by the Mercians and pretended to be a man of poor or servile class. But he was eventually recognized as a noble by his captors, as Bede reports, not only by his bearing but by his speech.

The question of the survival in 'England' of British population and still more of British forms of speech is, of course, a matter of debate, differing in the evidence and the terms of the debate from region to region. For instance, Devonshire, in spite of its British name, has been said on the evidence collected by the Place-names Survey to appear as one of the most English of the counties (onomastically). But William of Malmesbury in his Gesta Regum says that Exeter was divided between the English and the Welsh as late as the reign of Athelstan.

Well known, and much used in debate and in the dating of sound-changes, are the Welsh place-names given in Asser's Life of Alfred: such as Guilou and Uisc for the rivers Wiley and Exe, or Cairuuis for Exeter. Since Asser was a native of South Wales (as we should now call it), Welsh was presumably his native language, though he may eventually have learned as much English, shall we say, as his friend the king learned of Latin. These names in Asser have been used (e.g. by Stevenson) as evidence for the survival of Welsh speech even as far east as Wiltshire as late as the end of the ninth century.

With the mention of Asser I will return, before I close, to the point that I mentioned when I began: the interests and uses of Welsh and its philology to students of English. I do not enter into the controversy concerning the genuineness of Asser's Life of Alfred, whether it is a document belonging approximately to a.d. 900, as it purports to be, or is in fact a composition of a much later date. But it is clear that in this debate we have a prime example of the contact of the two schools of learning: Welsh historical and philological scholarship and English. Arguments for and against the genuineness of this document arc based on the forms of the Welsh names in it, and an estimate of their cogency requires at least some acquaintance with the problems attending the history of Welsh. Yet the document is a life of one of the most remarkable and interesting Englishmen, and no English scholar can be indifferent to the debate.

To many, perhaps to most people outside the small company of the great scholars, past and present, 'Celtic' of any sort is, nonetheless, a magic bag, into which anything may be put, and out of which almost anything may come. Thus I read recently a review of a book by Sir Gavin de Beer, and, in what appeared to be a citation from the original,[20] I noted the following opinion on the river-name Arar (Livy) and Araros (Polybius): 'Now Arar derives from the Celtic root meaning running water which occurs also in many English river-names like Avon/ It is a strange world in which Avon and Araros can have the same 'root' (a vegetable analogy still much loved by the non-philological when being wise about words). Catching the lunatic infection, one's mind runs on to the River Arrow, and even to arrowroot, to Ararat, and the descent into Avernus. Anything is possible in the fabulous Celtic twilight, which is not so much a twilight of the gods as of the reason.

That was perhaps, in this time and place, an unnecessary aside. I am addressing those of rational mind and philological learning; but especially those who in spite of these qualifications have not yet for themselves discovered the interests and the uses of Welsh and its philology.

I have already glanced at the interest of this study to Romance philology, or the later history of spoken Latin, and of the special importance that it has for Anglo-Saxon. But the student of English as a Germanic tongue will find many things that throw new light on his familiar material; and some curious similarities interesting to note, even if they are dismissed as parallels produced by chance.

It would not be my place to treat them extensively, even if I had the time. I will only refer to two points in illustration. A traveller should at least produce some samples.

As an example of a curious parallelism I will mention a peculiar feature of the Old English substantive verb, the modern be'. This had two distinct forms of the 'present': A, used only of the actual present, and B, used only as a future or consuetudinal. The В functions were expressed by forms beginning with b-, which did not appear in the true present: thus, bīo, bist, bið pl. bīoð. The meaning of bið was 'is (naturally, always, or habitually)' or 'will be'.

Now this system is peculiar to Old English. It is not found in any other Germanic language, not even in those mot closely related to English. The association with the b-forms of two different functions that have no necessary logical connexion is also notable. But I mention this feature of Old English morphology here only because the same distinction of functions is associated with similar phonetic forms in Welsh.