The Origins of the British: The New Prehistory of Britain (46 page)

for English, although the trees are less informative, we know that there is considerable early borrowing from North Germanic … (Embleton … finds sixteen word loans in the Swadesh 200-word list from North Germanic to English). English does not quite move into North Germanic altogether …
⁷

What they seem to be saying is that, while English is structurally most similar to its neighbour Frisian, its vocabulary was more strongly influenced at an ‘early’ stage by Scandinavian languages. This effect was already apparent in Old English before the Vikings arrived on the scene to give the Anglo-Saxons some of their own invasion medicine. There are some pieces of textual evidence which are consistent with this idea and which I shall describe shortly, but the early Scandinavian loanwords are attested and stand by themselves.

The main problem this evidence raises for the Anglo-Saxon orthodoxy is that there is no specific record of Scandinavian invasions between the first arrival,
⁸
in the fifth century, of the Saxons, who were supposed to have introduced Old English as Anglo-Saxon, and the much later Viking raids, before which time the Old English texts were already showing strong Scandinavian influence on vocabulary. So, the Scandinavian influence on Old English should have been earlier than the Saxon invasions. But in that case there would have been no Anglo-Saxon to influence (and Latin and celtic had insufficient lexical effect on Old English to matter). So, what other language was around during the late Roman occupation to be influenced by Norse? Could it be that other Germanic languages, or even Norse, were already in residence in England before the fifth century
AD
? I shall come back to this possibility later.

An alternative explanation is that the language of Gildas’ invading ‘Saxons’ may not have been quite as ‘Saxon’ as he claimed. Were they, for instance, actually speaking the same language known from surviving Old Low German Continental texts such as the
Heliand
poem?
⁹
Could the ‘Saxon-invaders’ have already been influenced by Norse, or even be speaking it – or something else? Perhaps the oldest
possible
example of Norse words entering England (rather than Old English) comes from a single Germanic word Gildas himself interpolated when writing in Latin in the mid-sixth century
AD
. This is also the first extant written clue to the invaders’ language. The context is highly relevant since he applied it to the ‘Saxon’ invaders and their language in his famous line:

A pack of cubs burst forth from the lair of the barbarian lioness, coming in three ‘keels’ (
cyulas
), as they call war-ships in their language …
¹⁰

That these were high-prowed, sailing longboats, like those of the later Vikings, is explicit in the Latin of this and the next sentence. Forster argues from the
Oxford English Dictionary
(
OED
) etymology for ‘keel’ (
cyulas
) that ‘it is not clear that a Scandinavian origin for this word can be ruled out’. The relevant entry, among several in
OED
for ‘keel’, cites Old English
céol
(found in
Beowulf
) and Old Norse
kjl
.
¹¹
From my reading of the
OED
entry it could be either North Germanic or Old Low Saxon.
¹²
The Norse–English ship connection, however, pops up again in a rare Old English synonym for ship,
fær
(derived from the Norse word
faer
, meaning ‘journey’ – as in ‘seafarer’),
which is found in the sense of ‘ship’ only in two very early Old English poetic texts (
Beowulf
and Genesis A) and in Old Icelandic/Old Norse. Dennis Cronan, an expert on Old English poetry, points out that ‘Since Old Icelandic fær exhibits a similar semantic range, the meaning “ship, vessel” was probably inherited from Common Germanic’ (i.e. from the common ancestor of all Germanic languages).
¹³

The geneticist Peter Forster and colleagues have taken this speculation on Norse vocabulary’s influence on Old English rather further than many linguists would like. To do this, he has adapted the established genetic network methods I have used throughout this book and applied them to data on vocabulary (
Figure 8.1b
). He explains:

Firstly, the network diagram which our method generates (Bandelt et al. 1999) visualizes the changes of individual word use along the language branches, rather than providing only an abstract branch length or difference measure. This incidentally means that at its best, the network approach can reconstruct the lexicon, i.e. the specified word list, of a language which no longer exists. Moreover, our network method makes allowance for borrowing and convergent evolution by not forcing the data into a tree.
¹⁴

In other words, by comparing the origins of synonyms for common words, Forster is charting the ancestry of the use of individual words in different Germanic languages and how they got to their present locations. This helps to separate the history of Old English and later English vocabulary from the more conventional com parison of systematic structural and sound changes. By using a network without specifying a tree-root,
Forster avoids forcing English, ancient or modern, to associate with any particular ancestor, such as Frisian. This is important, because English is more profoundly affected by word borrowings than any other Germanic language, thus violating the assumptions of the simple tree-model of language evolution used to make conventional Germanic language trees (as in
Figure 8.1a
). Forster goes on:

Secondly, we have added four extinct lexical [vocabulary] sources from the first millennium
AD
(
Beowulf
, Ælfred [three Latin texts personally translated by King Ælfred],
Heliand
, and the Gothic Bible) into the analysis along with modern Germanic languages to investigate whether their inclusion would change the deep branching of English reported by Dyen
et al
. (1992), Gray and Atkinson (2003) and McMahon and McMahon (2003).
¹⁵

The results are revealing, but to see why, first we have to orient ourselves. To help comparison with the conventional Germanic tree, I have simulated an equivalent tree from Forster’s network with his help (
Figure 8.1c
). The network (
Figure 8.1b
) groups itself largely into two major poles, corresponding exclusively to North Germanic (i.e. Scandinavian) languages in the top right-hand corner, and exclusively to Continental West Germanic languages in the bottom left-hand corner, while some of the older extinct languages are scattered in between. Gothic, in an intermediate position, represents the extinct third or East Germanic branch.

To an extent this broad pattern of lexical divergence, or changes in actual word values, with the two surviving major branches of Germanic languages on the periphery of the network,
and the more archaic ones rooting in the middle, is consistent with the conventional evolutionary linguistic tree shown in
Figure 8.1a.
But, as may be apparent already, there the similarities between the two depictions stop. For instance, English forms a fourth deep branch of its own. This is consistent with, and would be predicted from, the studies on Dyen’s cognate set cited above, which all show English rooting deeply. In contrast, the other modern West Germanic languages on the Continent form a tightly knit group at the bottom of the network (
Figure 8.1b
) with a very similar choice of words in their vocabularies, while the High–Low German split and other structural differences apparently count for little. This separation of the English branch on the other side of the network from its theoretical colleagues would not be expected at all from the Anglo-Saxon orthodoxy and cannot simply be explained by later borrowings from Vikings and Normans and new words introduced by Latin scholars.

It may help to add a little perspective to the non-English part of the network. Based on the conventional tree of Germanic languages shown in
Figure 8.1a
, the centre or root of the network should be at the three-way division between North, West and East Germanic. On the network this is logically somewhere between Gothic and
Heliand
and the Old English roots. From the perspective of the languages in the modern West German pole, this root makes sense because the language of the
Heliand
poem (
AD
825) was Old Low German, and should be very close to their theoretical common ancestor. From the perspective of this inferred root, however, the vocabulary of one contemporary Old English text,
Beowulf
, ends up leaning a little towards the Scandinavian pole.

While this might explain why the tale of
Beowulf
is set entirely in a Scandinavian context (see
Chapter 10
), it does not explain why two ‘dialects’ of Old English,
Beowulf
(Anglian) and Ælfred (Wessex) were already diverging so much from each other in England so soon after the invasion. Not only that but, within several centuries of the Anglo-Saxon invasion they are already each lexically further from their supposedly close relative in
Heliand
than the latter is to Gothic (
Figure 8.1b
). As Forster says, ‘The lexical diversity which formerly existed in Britain [more than a thousand years ago] was comparable to the present diversity within mainland Scandinavia, or within the present Dutch/German language area.’ This complex picture is hardly consistent with the aftermath of a sudden Dark Age founding event by languages all closely related to Old Low Saxon. It is not as if they had picked up much in the way of celtic or Latin words on arrival – a known reason for rapid change. An alternative, deeper timescale and relationship needs to be sought in Old English dialects.

The position of modern English on the network is several word-pegs closer to Scandinavian than to
Beowulf
but it also forms a deep branch of its own, taking off from a different part of the network slightly nearer to the Scandinavian than to the Old Saxon part (
Figure 8.1b
). In this context, Forster suggests that ‘the Scan dinavian relationship may date back at least as far as
Beowulf
if the word for “sleep” is taken as an indicator.’ As far as the marked divergence of the English branch is concerned, much of this happened surprisingly early. Again, Forster claims:

Thus, in a time span of around 1,200 years, English seems to have changed about ten words within the shortened 56-word list used for the network … Moreover, according to the
Oxford English Dictionary
(1989), which documents the earliest appearances for each of these words, all 10 changes originated before
AD
1300. This compares with only five changes in the 1,200 years between Heliand (approximately
AD
825) and modern Low German, and two changes in around 1,000 years between Icelandic and ‘Old Norse’.

Given that the roots of English vocabulary appear, like the two ‘Old English’ texts in the network, to have diverged so early, it is not clear whether either of the latter are directly ancestral to modern English. This might mean that the connection between Old English and the ancestral Common Germanic root predated the arrival of the Angles and Saxons. English would then be neither directly descended from Old Saxon nor from Scandinavian (i.e. Old Norse). In Forster’s words:

The network analysis reveals a Scandinavian influence on English and apparently a pre-Scandinavian archaic component in Old English. All Germanic lexica spoken today appear to converge in the network on an ancestral Common Germanic lexicon spoken at an unknown time, but constrained to before
AD
350 and probably after 3600
BC
. The results presented in this study are based on only fifty-six words and must be considered preliminary pending more detailed networks using longer word lists.
¹⁶

In other words, Old English, rather than just borrowing a few Scandinavian words at the start, may have been around for some time before the Anglo-Saxon invasion
and
these effects are
reflected in modern English. I have included Forster’s note of caution in the quote, because this is a very short word list from which to draw such profound historical conclusions. I understand that he and his colleagues are in the process of extending the word list to 200, which might strengthen their conclusions. Their date range is also hedged by extreme academic qualification, but starts to look as if it could include some of the period of the Later Neolithic and Bronze Age, when, as we have seen, there do appear to be genetic and cultural influences coming into eastern England from southern Scandinavia and north-west Europe.