Diglossia in Anglo-Saxon England, or what was spoken Old English like?

ABSTRACT

This paper suggests that diglossia in caste-like Anglo-Saxon societies consisted of O[E.sub.H] used by a very, small elite of largely Continental Germanic ancestry and O[E.sub.L] spoken by the bulk of the population. These shifted from British (Low) Latin and Late British to Old English (OE) after the Anglo-Saxon Conquest, in some areas over a period of about 300 years. It is hard to guess what the spoken language of the users of O[E.sub.H] was like and how great the gap between the spoken and written language was. The latter, through intensive networking, was kept remarkably constant over the entire OE period. Speakers of O[E.sub.L] are likely to have produced the attrition of inflections in the NP and the acquisition of aspectual distinctions in the VP. This surfaced in writing only after the whole-sale replacement of the Anglo-Saxon elite by the Normans. Cross-language contact linguistic research has shown that bottom-up language shift of large population groups is likely to produce grammatical (and phonological) contact phenomena in the target language rather than lexical ones. It is therefore claimed that Middle English started to be spoken as a low variety of English not after the Norman Conquest, but not long after the Anglo-Saxon Conquest.

**********

As we all know, the Anglo-Saxon period is the longest period in the history of the English language. By its external political demarcations (fifth century AD to the end of the eleventh century AD), Anglo-Saxon culture was present for around 600 years in the island of Britain. (1) This is a very long period, indeed. Within this period, Anglo-Saxon culture had a written presence for about 400 years. I would assume that Old English was enscripted in alphabetical characters around the beginning of the seventh century, with the impact of Christianization which started from both ends of the island, north and south. According to the Venerable Bede, Saint Augustin landed in Kent in 597 AD at the request of Pope Gregory the Great (ca. 540-604 AD) and Lindisfarne ("Holy Island") was founded by Saint Aidan in 634 AD. Aidan came from the island of Iona (Inner Hebrides) and was a representative of the Irish Church. The entire conversion process of the Anglo-Saxons lasted from 587-681 AD, coming to a close when Saint Wilfrid converted the South Saxons of the Isle of Wight to Christianity. And with Christianity came of course the acquisition of the cultural technology of writing. The Northern scriptoria exerted a lasting effect on the writing of Old English right up to the end of the Anglo-Saxon period, as the invariable use of the so-called "insular script" for the writing of Old English demonstrates, a script which is Irish in origin.

Interesting evidence of the process of the enscripting of Old English in the Northern cultural province is, for instance, provided by Bede's well-known story of Caedmon and the divine inspiration of the beginning of Old English literature (HE IV.23). Caedmon's poems were orally composed and later taken down in writing by the monks of Whitby. Bede says that Caedmon composed his songs in sua, id est Anglorum, lingua (Colgrave and Mynors 1969: 414), i.e. in English. But his name bears witness to his British ancestry.

A short side remark: it is interesting to note that the name of Caedmon is an Anglicization by oral loan of a Brittonic hero's or warrior's name, evidenced among the British princes in the seventh century. *Catu-mand-os in Brittonic meant 'war horse' or 'war pony'. (2) There is, for instance, the Cata-man-us inscribed stone in Anglesey, which has been securely dated to the first hall of the seventh century. (3) The Old English spelling (and pronunciation) of this Brittonic name is interesting, since both unstressed syllables of the compound *Catu-mand-os are dropped, by syncope of the composition vowel and by apocope of the inflectional ending (> Late British Cad[mu]ann, with lenition of the originally intervocalic */m/as*/v/in the environment of the voiced stop/d/) (cf. Coates (2002: 16)). This conforms to the historical development of Brittonic morphology during the four centuries the Britons lived under Roman rule. During that period, unstressed syllables were affected both by syncope and apocopy. Brittonic shed its inflectional endings in the NP, with the sole exception of number marking (singular and plural). (4) So Bede spelt and the monks and nuns of Whitby heard and pronounced the heroic Brittonic name of their cowherd as Caedmon. In today's Welsh, the name Cadfan is pronounced/'kadvan/. (5)

According to Bede's story (HE IV.24), Caedmon's religions poems were orally composed. These poems were later committed to parchment by the monks as were other poems which have been preserved in the Old Northumbrian dialect from about the eighth century (Caedmon's Hymn, Bede's Deathsong, Leyden Riddle, Ruthwell Cross). In his letter to his fellow teacher Cuthwin, Cuthbert diaconus writes that Bede was occupied by translating the Gospel of St. John and Isidor's book on De natura rerum into Old English (Colgrave and Mynors 1969: 582). Unfortunately, these translations have not survived. No doubt, Old Northumbrian was a fully-fledged literary language.

The written Old English language appears to have been or to have been kept remarkably constant over the entire period of Anglo-Saxon writing, in spite of the change of a few spelling conventions. These represented, among other things, both a further development and a Germanicization of the spelling because of the introduction of runic symbols into the Latin-derived alphabet (th > [thorn], uu > p). The dialectal variations are remarkably few. The oldest texts from Northumbria (seventh/eighth century) and the late West Saxon texts some three hundred years later (eleventh/twelfth century) show surprisingly little typological change of the grammatical structure of the language. This suggests that strong efforts were made to keep the written language unchanged. The late Anglo-Saxon efforts to this effect under the Benedictine Reform have been recently documented by Lucia Kornexl (2000) and Mechthild Gretsch (2001, 2003). It seems that the theocratic elite of late Anglo-Saxon England deliberately enforced the standardization of Old English as a means of political control, which was exposed to the threat of political disintegration at the hands of the Vikings. The Viking administration under King Knut and archbishop Wulfstan, however, followed suit in maintaining the West Saxon written standard. Written standard Old English only began to crumble during the reign of Henry I (1100-4135). The late annal entries in the Peterborough Chronicle, for instance, show that in Peterborough the OE written standard was only given up after 1121. The language of the First Continuation, covering the years 1122 to 1131, is still strongly influenced by the OE written standard, but already shows current features of the spoken language. But the annals of the Final Continuation, dealing with the years 1131-1154, are "incontrovertibly Middle English" (Clark 1970: lii) in lexis, morphology and spelling.

The earliest Middle English texts give evidence of a great typological change. With apparent suddenness appeared the drift away from syntheticity to analycity. All Germanic languages are subject to this drift, but here it appeared with particular strength. The general pattern of the accelerated typological drift in the early Middle English period was as follows: English was well ahead of the other Germanic languages and the North of England was well ahead of all other Middle English dialects in the spread of analyticity and of other linguistic innovations. Why was this so? Why this seemingly sudden und accelerated development?

Four possible scenarios have been proposed to explain this sudden shift from Old English to Middle English: three language internal scenarios and one external scenario, internal referring to systemic linguistic changes and external meaning "language contact."

One infernal answer to the question of sudden change in grammatical profile is, according to Robert M.W. Dixon (1997:67 ff.), the recourse to the punctuated equilibrium model. Is the sudden change from Old English to Middle English an example of a punctuated development? The punctuated equilibrium model claims that the rate at which languages change need not be constant. Languages may change or evolve very little over long periods of time and then, all of a sudden, they may be subject to radical typological changes.

equilibrium period [right arrow] sudden punctuation

Dixon writes: "I suggest that many types of change within a language are not gradual but rather happen fairly suddenly, often within the space of a generation or two.... If a new grammatical mechanism is innovated this is likely to happen all at once, rather than bit by bit" (Dixon 1997: 54). He also asserts that "overall, gradual change concerns matters of detail, while sudden changes relate to shift in grammatical profile" (Dixon 1997: 57). He explains that sudden changes in grammatical profile are "self-triggered" and produced "by the internal dynamics of the language" (1997: 57). (6) Dixon is a comparative linguist and a specialist of Australian aborigine languages. The examples he quotes in order to substantiate his thesis of punctuated historical developments are mostly taken from non-European languages. Could the punctuated equilibrium model also have applied to the relatively well-documented historical stages of the development of western

European languages? In particular, could it have applied to the dramatic typological shift of English from a predominantly synthetic to a predominantly analytic language first surfacing in twelfth century writing? And what internal dynamics triggered this sudden shift?

Another very traditional internal explanation of the typological shift from Old English to Middle English is that of the reduction of unstressed vowels under the effect of the strong stress accent which marked the first syllable of the stem of Old English lexemes. This stress accent was held responsible for having led to the reduction of unstressed vowels and, consequently, to the loss of inflectional endings (Campbell 1959: 30-37, 137-157). (7) Indications of the loss of the phonematic distinction between unstressed vowels of the inflectional morphemes only surface in the late Old English texts. This raises the important question of why German, for instance, whose lexemes are affected by the same type of stress accent, has not been subjected to the same type of attrition of inflectional morphology as English. Old Norse for that matter was as strongly inflected as Old English was and Icelandic still is. (8)

A third systems internal explanation suggested that many of the inherited Old English inflectional endings were redundant, i.e. that they were devoid of phonological distinctions. For instance, the inflectional paradigm of the OE noun lar fem. 'teaching' (cf. PDE folk-lore), which followed the so-called "strong" inflectional paradigm,

Sg    Nom      lar-[empty set]    Pl    lar-a
      Gen      lar-e                    lar-ea
      Dat      lar-e                    lar-um
      Acc      lar-e                    lar-a
      Instr    lar-e                    lar-um

only showed three distinctive endings {-e, -a, -um} and none at all for the nominative singular. These endings were polyfunctional and therefore not distinctive. In Herbert Pilch's structuralist grammar of Old English (1970: 104), the inflectional repertoire of the class of weak nouns is represented as follows: (9)

                Mask.             Fem./Neutr.

Nom.Sg.         -a                -e
Akk.Sg.
Gon.Sg.
Dat.Sg.                  -an
Nom./Akk.Pl.
Gen.Pl.                  -en-a
Dat.Pl.                  -um

The ending of the instrumental is already identical with the dative. Thus there were only five endings for altogether 10 morphological categories. For the so-called "strong" noun inflection, Pilch shows the following diagram with only six distinctive endings for 10 grammatical categories:

                 Sg.                         Pl.
        Mask.    Neutr.    Fem.    Neutr.    Fem.    Mask.

Gen.          -e-s
Nom.
                                             -a      -a-s
Akk.                               -u
               -e
Dat.                                       -u-m

Texts from the Anglian dialect area showed a further reduction of distinctivity, as they featured {-e} instead of {-a} in the nom/acc plural fem. In Late West Saxon texts, {-u} in fem. nouns came to be used in the oblique cases in the singular and {-a} in the nom./acc, neuter plural.

This type of increasing redundancy led scholars to the assumption that the inflectional endings had become largely "ornamental" in Old English and could therefore be dispensed with. The same question then arises as to why the corresponding Modern High German word has kept its unstressed syllables to this day:

Sg    Nom    Lehr-e    Pl    Lehr-en
      Gen    Lehr-e          Lehr-en
      Dat    Lehr-e          Lehr-en
      Acc    Lehr-e          Lehr-en

The German inflectional endings may very well be ornamental, but that does not seem to be a reason why they should be shed. (10)

It could be argued that the case distinctions in Old English within the NP were largely guaranteed by the inflected determiners preceding the noun, so that the inflectional redundancy was compensated for. The same holds true, of course, for Modern High German. But then in Middle English the definite article was indeclinable (as already the definite article in Old Welsh), while in German it is still fully inflected.

It seems to me that none of the system internal explanations so far discussed has any cogent explanatory force.

The one external answer given so far was that relating to the contact between Old English and Old Norse in the Danelaw areas. The view has been advanced that this contact may have led to a mild sort of creolization, whereby the speakers shed their endings due to the universal tendency of morphological simplification observed in creole situations (Poussa 1982, but cf. Gorlach 1986/1990; White 2003: 41). This would presuppose that, in the contact situation between the Northumbrians, Mercians and the Viking settlers, Old English and Old Norse were so different as to require first pidginization and then creolization in order to enable efficient communication. On the contrary, I would hold that, with a little effort, Northumbrians, Mercians and Scandinavians were very well able to communicate in their everyday dealings, much as I can communicate with speakers of vernacular Zurich or Berne German in Switzerland or even Carinthian German in Austria, although my native variety of German is Hanoverian Standard High German.

The creole hypothesis also does not explain why the earliest Northumbrian poems, which come exactly from that part of England which the Middle English innovations radiated from, already show first signs of inflectional attrition by the loss of final nasals, not the unstressed vowels, but the final nasals <n> and <m>. Thus the attrition of morphological distinctions had already started in the NP well before the Vikings arrived. The Vikings provided the necessary, but not the sufficient condition for the seemingly sudden Middle English innovations and their spread southwards across England over the centuries.

Let us now take a closer look at the innovating areas in Middle English, because not only the North strongly innovated compared to the written Old English Standard, but also the South and the South West, albeit on a different scale. The southern innovations were perhaps less obvious, because they related to the syntax of the VP and not to morphology of the NP. The most important South Western innovations were the rise of periphrastic aspect and DO-periphrasis. The East was slow to accept both the Northern and the South Western innovations and the South East was the most conservative area, resisting any innovation by dialect spread for the longest time. How can this be explained? Why was the North the innovator in the NP and the South and South West the innovator in the VP?

According to David White (2002, 2003), Anglo-Saxon England consisted of four geographical zones which developed grammatical patterns reflecting the historical types of ethnic cum linguistic contact which speakers of Old English were subjected to during the many centuries of their existence in the island of Britain.

The languages spoken in sub-Roman Britain were British Latin in the British Lowlands, Brittonic in the Uplands, Wales and Cornwall, and Pictish north of the Clyde--Firth of Forth line. British Latin is likely to have been strong in the Lowlands, particularly in the east and south east (Schrijver 2002).

ZONE 1: In some areas of the Midlands and Northern zone, speakers of the post-conquest Anglian and Mercian dialects ruled the native population of the Britons as their slaves. These continued to speak Brittonic, their native language, for perhaps as many as six or seven generations (see the evidence of the Caedmon story for the North), before they shifted to Old English. Gelling (1993: 55) allows for more than four hundred years of the shift from Brittonic to Old English to have been completed and suggests that the process was only complete around 900 AD. During the process of shift, the Britons modified their target language by grammatical transfer from their native language (starting by imperfect adult learning). The main transfer from their source language consisted in the attrition of the inflectional endings in the NP, as Brittonic had already shed these before the advent of the Anglo-Saxons. The speakers of Brittonic compensated the loss of their inflectional morphology by means of rigid word order constraints and by the grammaticalized use of prepositions or other particles which lent the NP a very obvious analytical character. These shifters and now speakers of modified Old English or "Brittonic English" (as David White calls it) were later subjected to intimate linguistic contact with speakers of Old Norse. This contact led to reinforcing the already existing analytizing tendencies of Brittonic English and also led to adding new grammatical transfer features from Norse through superstratal and adstratal contact. (11) This twofold linguistic contact of English in the Anglian areas with substratal Brittonic and superstratal/ adstratal Norse helps to explain the sweepingly innovative character of the North.

ZONE 2: In the English south west, Old English experienced only one linguistic contact situation and not two as in the Upland Zone further north. The West Saxons only entered into contact with the language of their subjected Britons. The Treaty of Wedmore between the King Alfred and Guthrum (878 AD) defined the areas of Viking influences ("Danelaw"). Wessex was largely untouched by direct Viking influence, although the treaty also regulated trade relationships between the West Saxons and the Scandinavians. The treaty was formally observed only until the late 880, so that the boundary between Wessex and the areas under Scandinavian control will not have been sharply defined. But Scandinavian influence will have been considerably less in Wessex than anywhere else in the island of Britain, except for the south-east. The contact of the speakers of West Saxon with their Britons did, however, not have the same impact as that of the North, probably for two reasons: first, because the ratio of immigrant Anglo-Saxons settlers versus the subjected Britons was higher than in the North; and second, because the unsubjected Britons in Cornwall and Wales were considered to be fierce enemies and contact with them seems to have been constantly hostile. (12) The absence of Norse influence may perhaps explain why Brittonic substratal features characteristic of the VP, such as the rise of Verbal Aspect (imperfective/progressive) and of the DO-periphrasis (causative, emphatic, habitual) were able to develop by transfer in the South West and not elsewhere. In the north, Norse superstratal/adstratal influence would have precluded the use of aspectual and periphrastic DO-forms.

ZONE 3: This zone is that of the East, especially East Anglia, where the Anglo-Saxon prescrite was early and strong from the beginning (13) and may perhaps have induced the native population, either the Brittonic or British Latin speaking people, to a much more rapid acculturation than in the North and South West. A certain amount of violence between the newcomers and the native population can not, of course, be excluded. The East later formed part of the Danelaw, so that the relatively uncontaminated immigrant Old English of this zone experienced intensive language contact only with Old Norse. Since both contact languages were fairly close cognates and still fully inflected, superstratal/adstratal transfer from Norse to Old English would have been mostly lexical.

ZONE 4: The South East was the zone least subjected to linguistic contacts. The Anglo-Saxon presence was strong from the beginning (14) and the acculturation of the speakers of British Latin presumably was a fast one, taking only perhaps two generations or even less. This area, moreover, did not tome under the influence of the Vikings in the Danelaw. We know from psycholinguistic research on various types of shift of bilingual European language learners that fast shifts produce bilingual child language learners, who acquire the target language in native-like fashion. Truly bilingual children learn two native grammars. When they or their children stop using the less prestigious language, they behave as native speakers of the target language. Therefore, it comes as no surprise that the Kentish dialect area was the slowest to accept innovations (before, of course, it succumbed to the linguistic pressure of the greater London metropolitan area and finally disappeared).

What was the social stratification of Anglo-Saxon England like and which ethnicities were involved? The character of the social pyramid of Anglo-Saxon society is very unlikely to have been uniform all over the conquered regions. It is also unlikely that it remained unchanged during the length of the six centuries in its development from migration/conquest and ethnically divided society to the formation of the integrated Anglo-Saxon kingdoms. This is, for instance, made clear by the different provisions of the Anglo-Saxon laws concerning the payment of wergild. There seems to have been, however, a basically tripartite division of society. The total number of the population may have ranged between 1m and 2,5m inhabitants, and, after the dramatic drop of the size of the population in the late period of Roman Britain and after the withdrawal of the Roman troops, the size of the population showed no further drop during the conquest period and after (Harke 2002). The Anglo-Saxon elite was formed by a small number of aristocrats. For example, William the Conqueror replaced 4 to 5 thousand Anglo-Saxon aristocrats by 144 Norman barons within 20 years. Below the Anglo-Saxon aristocracy, there was the class of the freemen or "churls", which included the craftsmen. The majority of the population was, however, formed by the unfree people or "slaves", who had no right to own property. (15) Dorothy Whitelock said that

   [t]he unfree class consisted of persons of different origins. Some
   were the descendants of the British population, as the word for
   'Briton' to mean simply 'slave' testifies. The menial tasks
   described in some Anglo-Saxon riddles are performed by 'Britons'.
   (Whitelock 1952:111) (16)

If is to be noted that the riddles in the Exeter Book were codified, if not composed in the tenth century. Were the monastic and aristocratic audiences of the riddles able to relate to slaves who were still recognizably Britons, whatever this actually meant in a tenth century Wessex context? The barriers between the three classes of Anglo-Saxon society seem to have been more caste-like than class-like, as social risers were scarce. (17) Once born a slave of whatever extraction, nearly always a slave.

The so-called "New Debate" in recent British archaeology and historiography (since the 1980s) has been concerned with the ethnogenesis of Anglo-Saxon England and the questions of the creation of English identity The origin of this new debate was also couched in the framework of the theoretical call of "Processual Archaeology" for the study of long-term cultural, social and economic processes underlying cultural change (Harke 2003: 15 f., Hines 2004: 17). Scholars such as Lloyd Laing, Nick Higham and Heinrich Harke have shown that the nineteenth century "Anglo-Saxonist" ideology of the Germanic racial "purity" of the Anglo-Saxon society cannot be maintained in the light of recent archaeological research. This means that the nineteenth century preconceptions of English historians such as Richard Green and Edward A. Freeman, that most of the island of Britain was ethnically cleansed was wrong and reflects the 19th century imperialist views (German 2000). These nineteenth century scholars advocated the double-X theory ("extermination and expulsion of the native Britons"). Today the double X-theory, i.e. the population replacement theory, is no longer upheld uncontroversially by historians and archaeologists. The current discourse advocates the theory of an elite take-over of the ca. two-million Romano-British population by Anglo-Saxon "fringe barbarians" (Harke 2002: 167). This means that the great mass of the Brittonic or British Latin speaking menial population of the island of Britain suffered an Anglo-Saxon take-over from the top. Perhaps they even welcomed the Anglo-Saxon take-over, since taxation appears to have been initially lower under the Anglo-Saxons than under the Romans or, after the withdrawal of the Roman troops, their sub-Roman masters.

The pattern of acculturation of the Britons was of course not uniform. The rapid acculturation of the Britons in the South East and East and the slower acculturation in the other parts of the island can be documented by a number of archaeological research results, the most important of which are:

1. Palaeobotany: the forests did not reclaim agriculturally exploited land after the adventus and the continuation of agricultural cultivation is very obvious; "in most areas the landscape was kept as open as in Roman times" (Harke 1989: 106, 2002: 159 f., 2003: 16)

2. New types of settlement: the Romano-British towns and country estates (villae) cum dispersed farmsteads were successively replaced by Germanic types of village settlements (Hooke 1997); this may explain why most English village names in the Lowland Zone are Germanic. (18)

3. Change of the building material of the houses from stone (Romano-British) to timber (Germanic) and of the ground plan of the houses; but there are also mixed types of houses which imply a mixed population (Harke 2003:18 f.)

4. Burial patterns: shift from native post-Roman inhumations without grave goods to inhumations with grave goods; in most parts the differences in body size of the male skeletons disappear at the end of the 7th century (Harke 2003: 19 ff.)

5. Mixed grave goods: Brittonic and Germanic artefacts, jewellery, etc.; cf. the Sutton Hoo hanging bowl (perhaps owned by king Raedwald of East Anglia) (Harke 2003: 17)

Recent DNA analyses have suggested that, since the Anglo-Saxons, (19) the Vikings (20) and the Normans (21) only came in small numbers and ruled a native population much larger in size, the DNA of the British population remained largely unchanged, in spite of the tact that they adopted the language and culture of the Anglo-Saxons and absorbed the languages of the Scandinavians and the Normans. In other words, the Britons became English by assuming Anglo-Saxon, Viking and Norman culture under the respective domination of their elites. This is, however, a very controversial matter, indeed, and speculations of this sort lie outside the domain of the sociology of language change.

The question now arises as to why the Britons did not adopt the Latin language during the almost 400 years of Roman domination, like their cousins, the Gauls, Belgians and Aquitanians had done in Gallia. The answer that has been suggested pertains to the different power structure of the Roman Empire, particularly in the peripheral areas, and that of the Anglo-Saxons. While the Empire seems to have been territorily defined, the rule of the Anglo-Saxons appear to have been ethnically defined. The ethnogenesis of the Germanic tribes needs to be viewed in connection with the Teutonic Migrations. The ethnogenesis both of the Continental Saxons and of the Anglo-Saxons is an interesting topic, in that it has been suggested that both had a very strong awareness of their ethnic lies and enforced these on their own people as well as on their colonized subjects ("assertive monoculturalism"). (22)

If we assume that the native Britons in the South West, the Midlands and the North slowly shifted to Old English in the course of two to four centuries (fifth to ninth), the type of linguistic contact and of language acquisition would have had to be that of the adult learner type. (23) Cross-linguistic research in contact linguistics has shown that adult second language acquisition in asymmetrical power situations leads to transfer of phonological and grammatical [L.sub.1] features to the target language. (24) Little or no lexical transfer is to be expected, because high status languages rarely borrow common lexis from dominated low status languages. (25) The shift from Late British to Old English exactly follows this pattern.

Why does the written documentation of the Old English dialects hot show grammatical transfer features, if the bulk of the population was culturally (and perhaps ethnically) British-derived? (26) The answer is complex and needs to be sought in at least two factors. The first is social and the second concerns the medium of communication, i.e. oral and written communication. The Anglo-Saxon kingdoms seem to have cultivated their "Englishness", i.e. their "Otherness" vis a vis their subjects, the Britons. "Englishness" was first affirmatively defined in opposition to the wealas, their unfree subjects working the land and the free Britons in Cumbria, Wales and Cornwall. In the later Anglo-Saxon period, "Englishness" was affirmatively defined in terms of religious difference regarding the Scandinavians. The self-definition of the Christian Anglo-Saxons seems to have been based on their understanding of themselves in salvational terms. The title of Bede's Historia ecclesiastica gentis anglorum reflects a telling ideological construct, welding the different elite groups of Anglo-Saxon society together in terms of their ethnic extraction.

The second important factor is that of the written language. At the beginning of this paper, I mentioned that Old English was enscripted in the seventh century AD and that the written standard was kept remarkably constant, typologically speaking, across the long centuries of the continued cultivation and preservation of Anglo-Saxon culture. The written language was, of course, that of a small powerful elite, ethnically the Anglo-Saxons. At the end of the seventh century, I would assume that the gap between the spoken and written language of the elite was comparatively small, I would assume. (27) In the later period, there must have been a tripartite division of the types of OE: a) the written language of the elite the norms of which were carefully maintained (O[E.sub.W]), (28) b) the spoken vernacular of the elite (O[E.sub.H]), and here we can only guess how large the gap between the written and the spoken language was, and c) the vernacular of the bulk of the population, which was largely of British and in the Danelaw areas also of Scandinavian extraction (O[E.sub.L]). The situation certainly differed significantly by area and gave rise to the differences of dialect that later emerged. The diglossia of spoken Anglo-Saxon, whether the high or the low prestige variety, never surfaced in writing during the entire Anglo-Saxon period. (29) Of the two types of spoken Old English, only the low variety surfaced after the replacement of the Anglo-Saxon elite by William the Conqueror, which took him some 20 years, as already mentioned.

The OE written standard survived the Norman Conquest for something like two generations, since Anglo-Norman had not yet been subjected to enscripting (Clanchy 1989:58 ff.) (30) Latin had lost something of its prestige as the language of religion and administration in Late Anglo-Saxon England due to the affirmative OEH vernacular endeavours by the tenth century Benedictine reformers. When English re-emerged in writing in the first half of the 12th century, it was of course the spoken language of the formerly repressed low variety of Anglo-Saxon, which now rose to the status of a strongly regionalized middle class written language.

In what way does written Middle English reflect modelling on Brittonic? The most obvious grammatical innovations were:

In the Northern zone (NP):        * invariable case and gender
                                  inflection of nouns, pronouns and
                                  adjectives, as earlier in
                                  Brittonic (31)
                                  * invariable article, as earlier in
                                  Brittonic (32)
                                  * fixed word order, as earlier in
                                  Brittonic (33)
In the South Western zone (VP):   * rise of periphrastic aspect, (34)
                                  as in Middle Welsh (35)
                                  * DO-periphrasis in a variety of
                                  uses, as in Middle Welsh (36)

Quite a few other grammatical Brittonicisms only surfaced even much later in Middle English, such as for instance clefting. Clefting already occurs in the oldest documented Welsh texts, which probably date to the seventh or eighth century (by absolute chronology), although the dating is controversial. Clefting is most likely to be linked to the rise of a fixed word order after the loss of inflections. It serves various expressive purposes within the framework of a functional sentence perspective. Thus, in written documentation, clefting in Old Welsh occurs some 400 years or so earlier than in written English. (37)

My conclusion then is that I do not believe in the punctuation theory concerning the sudden rise of Middle English due to a language internal development, although I agree with Dixon that punctuation is possible theoretically. But I would like to know what the purely inernal conditions might have been. Nor do I believe in the internal typological restructuring of English due to the effect of the stem initial stress accent or in the deletion of merely ornamental inflections. I do believe, however, that there must have existed a very large gap between the high variety of spoken Old English (O[E.sub.H]) and the low variety (O[E.sub.L]). In terms of population numbers, we may perhaps assume that the high variety was spoken by some 4 or 5 thousand people and that the low variety was spoken by 1 to 2.5 million speakers of learner Old English. The percentage of speakers of the high variety must have been very low (0.2 to 0.4%), perhaps even lower than the percentage of today's speakers of RP as a community language (2%). (38) The low variety of Old English would already have featured most if not all the basic grammatical characteristics of Middle English, but it never entered into the realm of writing, because of the essential caste character of Anglo-Saxon society and because of the elite's exclusive control of the technology of writing. The well-known Middle English dialect zones reflect the former ethnic contact situations which the language of the Anglo-Saxon conquerors in Britain experienced over the centuries, i.e. Anglo-Saxon with British Latin, Brittonic and Old Norse.

Thus the assumption of a substantial diglossia in Anglo-Saxon England helps to explain why, after the removal of the Anglo-Saxon elite, Middle English dialect writing appears to feature such "sudden" innovations emanating or radiating from the two focal centres in the North and in the South West.

CODA 

Angelika Lutz (2002a) has made an interesting case against the tripartite division of the history of the English language by Henry Sweet. He divided its history into Old English, Middle English and Modern English on the basis of the degree of morphological synthesis. Lutz points out that the widely differing views held by various scholars of the beginning of Middle English suggest that the Norman Conquest and its social restructuring of England has nothing to do with the analyticization of English. Instead, she suggests a bipartite division of the history of the English language on the basis of the influx of Romance vocabulary (end of thirteenth to fifteenth century). This changed the language to such an extent that Renaissance scholars did not consider the earlier period to be "English", but "Saxon". (39) According to Lutz, the (partial) relexification of English at the end of the Middle Ages was a more important change of the communicative system than the much earlier loss of inflections and the grammatical changes this entailed (Lutz 2002: 161).

In my opinion, Sweets' (and his followers') morphological and Lutz' lexical criteria resorted to for the sake of the periodization of the long history of the English language do not really exclude each other. They look at language change from different linguistic angles, i.e. the system internal change of morphological structure and the change of the communicative function by (partial) relexification. Because of the views advanced in this paper, I would, however, endorse Lutz's view concerning an unbroken continuation of English across the divide of the Norman Conquest. There was a political divide of ethnic rulership, but there was no linguistic divide as far as the spoken language of the bulk of the population was concerned. The attrition of the inflections and the restructuring of the syntax started with the adult learners of Old English whose native language was Late British and who shifted to O[E.sub.L]. In the Danelaw areas, attrition was reinforced by the contact of the former language shifters with the Norse speaking settlers. To put it bluntly, outside the East and the South East spoken Middle English, as far as grammar is concerned, began in the sixth and seventh centuries AD. Like in France, where the result of the contact of Gaulish Latin, as a spoken substratum, with superstratum Frankish only surfaced in writing in the ninth century after the Carolingian Reform, it needed a strong external impetus to adjust the written language to the spoken practice. The Carolingian reformers attempted to restore the "proper" use of Latin in the Empire and thereby allowed the spoken languages of Old French and Old High German to appear in writing (Strassburg Oaths 842 AD, Eulalia 878 AD). Spoken low status Old English, however, could only surface in writing after the demise of Anglo-Saxon culture. Unfortunately, the historical grammars of English never gave credit to the speakers of the "real" language of Anglo-Saxon England.

(1) I am very grateful to Prof. Ilse Wischer (Potsdam), Dr Hartmut Burmeister (Potsdam/Kiel), Dr Susanne Kries (Potsdam), Dr Elvira Veselinovic (Potsdam) and Dr David L. White (Austin TX) for commenting on earlier drafts of this paper and for suggesting improvements. I also received linguistic help, for which I am very grateful, from Gary Lovan (Potsdam). I am, of course, solely responsible for any errors or infelicities.

(2) Cf. Forster (1921: 179, 1942: 664), Evans (1967: 222), Jackson (1953: 120, 244, 272, passim), Koch and Carey (1995: 33 f.), Sims-Williams (2003: 294, 364, 378).

(3) The full inscription reads: Catamanus rex sapientisimus opinatisimus omnium regum (Jackson 1953:120, Sims-Williams 2003: 378).

(4) The morphonologically conditioned marking of the constituents of the NP and VP by what is called the "mutations" in Celtic Studies, which in a way compensate for the loss of inflections, may have been linked to the basically syllable-timed character of the insular Celtic languages in their pre-historical stages and does not seem to have exerted a transfer effect on English.

(5) On the palatalization of the short Brittonic stem vowel /a/as/ae/in Old English names see Jackson (1953:271 f.).

(6) For a critical review of Dixon's evolutionary approach see Robert A. Orr (1999). Dr Burmeister (Potsdam/Kiel, pc. December 2003) pointed out to me that there are surprising parallels for sudden punctuations both in [L.sub.1] and [L.sub.2] acquisition ("spurts").

(7) For the successive reduction of OE unstressed syllables see Pilch (1970:74 f.)

(8) Lexical stress does not necessarily imply the reduction of unstressed syllables in polysyllabic lexemes. This is shown for instance by Finnish. The first syllables of Finnish lexemes are invariably stressed, but the following syllables retain their full quality and quantity (pe. Juhani Klemola, August 2001). Estonian, on the other hand, closely related to Finnish, lost some of its final syllables.

(9) The weak class inflection of OE nouns is derived from the IE n-stems (cf. Lat. hom-o, hom-inis etc.).

(10) In his lecture on the attrition of inflectional morphology in the South Slavic languages ("Zur Geschichte des Flexionsverlustes im Balkanslavischen", Potsdam, 10.01.04), Dr Andrzej Sobolev (Marburg) pointed out that the ornamental character of the inflectional endings led to their attrition. His thesis is that the analytizing tendency of the respective languages are reinforced by their contact with strongly analytical languages, provided that their own phonological structure undergoes a process of restructuring. See also Hinrichs and Buttner (2004).

(11) For transfer features from Norse, see Miller (2004).

(12) The famous entry for the year 755 AD in the Anglo-Saxon Chronicle ("Cynewulf and Cyneheard") needs to be interpreted in this light. Cumbra, the West Saxon aldorman, who was unjustly killed by the deposed West Saxon king Sigebryht, seems to have been an acculturated Briton, as his name suggests. The British hostages who are mentioned in this entry seem to suggest hostile encounters with the unsubjected Britons. The presence of Asser at King Alfred's court and his role as his biographer is extraordinary in this connection. He came from St. David's in South Wales and was invited by Alfred to his court.

(13) Cf. the map of the progress of the Anglo-Saxon occupation in Jackson (1953: 208-209).

(14) See fn. 13.

(15) Cf. Stenton (1971: 277-318), Harke (1997: 141-148).

(16) The OE term wealh seems to have shifted in meaning from originally 'foreigner (Latin or Late British speaking)' to 'Briton' and finally to 'slave' (cf. Faull 1976). But the social, status oriented and the ethnic denotations of the term seem to have been ambiguous. There are, for instance, six references in the seventh century Laws o fine (promulgated between AD 688-694), referring both to the wergild values of British slaves and free Britons. On the other hand, the wealas are mentioned some 44 times in the Anglo-Saxon Chronicle (until the first quarter of the twelfth century). In the later entries of the Chronicle, the term mostly refers to the inhabitants of Cornwall and Wales. In the Anglo-Saxon riddles on the other hand, the meaning of the term seems to be 'slave' rather than 'Briton', as clear ethnic connotations are missing. Cf. Tolkien (1963: 26) and Mielmel Cichon, "Indigenous 'foreigners': legal, poetic and historical sources for Old English wealh", (paper given at the 12th International Congress of Celtic Studies, University of Wales, Aberystwyth, 26 August 2003, fc.). For slavery in Anglo-Saxon England, see Pelteret (1981, 1995).

(17) N.R. Ker's Catalogue of Manuscripts containing Anglo-Saxon (1957) lists a number of records consisting of "manumissions". See also: <http://www.trin.cam.ac.uk/chartwww/manumit.html> (accessed 17.30.04). Conversely, freemen could also go into slavery for various crimes.

(18) On the other hand, Gelling (1993: 55) suggests that there was a general tendency to replace Brittonic place names by English ones, which was completed by the mid tenth-century. (18)

(19) Minimalist view: 10.000 (Higham 1995) and 25.000 (Laing and Laing 1990), maximalist view: 150.000 to 250.000 (Harke 2002). (19)

(20) 10.000 to 40.000 Vikings (Harke 2002: 158).

(21) 10.000 to 200.000 settlers (Normans, Bretons, Flemish and other) (Harke 2002: 162).

(22) Cf. Geary (1983), Pohl (1991, 1997), Shennan (1991), Higham (1995) and Hines (1990: 31, 1994, 1996, 2000: 78). See also Banham (1994).

(23) The same type of process also occurred in the shift of the Gaelic speaking population in Ireland and Scotland from the 18th to the 20th century.

(24) The type and degree of adult transfer depends on two major factors: a) the typological closeness or difference of the involved languages and b) the linguistic levels concerned (phonology, syntax, lexicon, inflectional morphology--in this order of transfer likelihood). I owe this observation to Dr. Hartmut Burmeister (Postdam/Kiel, pc. December 2003).

(25) With the exception, of course, of place names or names of cultural artefacts which are very different from the culture in power and are useful to be known by their native names (cf. Leonard Bloomfield's (1933: 461) terminology: intimate borrowing).

(26) A possible exception is the present tense beon : wesan distinction which may perhaps have been calqued on the distinction between the copula and the verbum substantivum in Brittonic as well as in the other Celtic languages.

(27) I am not saying that the oldest OE texts represent the spoken language of the elite, as they are all highly stylized literary texts. Early spoken elite Anglo-Saxon will, however, have been much closer to the written texts than in the later times of the OE period.

(28) Cf. the linguistic and the stylistic artificiality of Beowulf, the action of which is set in the sixth century in a continental environment (today's Denmark and Sweden). The names of the protagonists, however, are transparent English heroic names and would have sounded very differently in their pre-English and primarily oral contexts.

(29) Malone (1930) and Magoun (1937) were, unfortunately, not accessible to me.

(30) Since in 1066 Norman French was still considered to be a low status vernacular, William the Conqueror "at first issued his written instructions in English just like his Anglo-Saxon predecessors. But in the 1070s, after the numerous rebellions had caused William to rely more on foreigners ..., English ceased to be the written language of the government, although a few royal charters for Canterbury continue to be recorded bilingually (in Latin and in English) until Henry II's reign in 1155" (Clanchy 1989: 58).

(31) On the precedence of the loss of inflections in the oldest Welsh written texts see Tristram (2002a; see also 2002b).

(32) On the precedence of the invariable article in Old Welsh, see Tristram (2002a).

(33) On the precedence of fixed word order as a result of the loss of inflections, see Tristram (2002a).

(34) Cf. Prof. Ilse Wischer's (Potsdam) ongoing research into the use and distribution of periphrastic aspect forms in the OE Orosius (paper given at the Potsdam "Linguistisches Forschungskolloquium", winter semester 2003/4).

(35) For a comparison of aspect in Welsh and English, see Tristram (1995, 1999: 255-262).

(36) On the DO-periphrasis in Middle Welsh, Cornish and Breton, see Tristram (1997).

(37) On clefting in Old Irish and in Old Welsh, see Tristram (2002b: 265-267, German 2004). The earliest English example dates from the West Saxon Gospels (German 2004), but clefting only began to appear more commonly in thirteenth century documents.

(38) Cf. Trudgill (20004: 19, 2001).

(39) Cf. Lutz (2002b) for the difference between English and German as two Germanic languages as to their import of French Man words and patterns of word formation.

REFERENCES

Bammesberger, Alfred--Alfred Wollmann (eds.)

1990 Britain 400-600: Language and history. Heidelberg: Winter.

Banham, Debby

1994 "Anglo-Saxon attitudes: In search of the origins of English racism", European Review of History 2: 143-156.

Bloomfield, Leonard

1933 Language. London: Allen & Unwin.

Boehmer, Heinrich (ed.)

1921 Texte und Forschungen zur englischen Kulturgeschichte. Festgabe fur Felix Liebermann. Halle/Saale: Niemeyer.

Campbell, Alistair

1959 Old English grammar. Oxford: Clarendon Press.

Clanchy, M.T.

1989 England and its rulers 1066-1272. London: Fontana.

Clark, Cecile

1957 The Peterborough Chronicle 1070-1154. Oxford: Clarendon Press. [1970]

Coates, Ian

2002 "The significance of Celtic place-names in England", in: Markku Filppula et al. (eds.), 47-85.

Colgrave, Bertram--R.A.B. Mynors (eds.)

1969 Bede's Ecclesiastical history of the English people. Oxford: Clarendon Press.

Dixon, R.M.W.

1997 The rise and fall of languages. Cambridge: Cambridge University Press.

Evans, D. Ellis

1967 Gaulish personal names. Oxford: Clarendon Press.

Fanego, Teresa--Elena Seoane (eds.)

2002 Sounds, words, texts and change. Selected papers from 11 ICEHL, Santiago de Compostela, 7-11 September 2000. Amsterdam: Benjamins.

Faull, Margaret Lindsay

1976 "The semantic development of Old English wealh", Leeds Studies in English 9: 20-44.

Filmer-Sankey, William--David Griffiths (eds.)

1994 Anglo-Saxon studies in archaeology and history. Vol. 7. Oxford: Oxford University Committee for Archaeology,

Filppula, Markku et al. (eds.)

2002 The Celtic roots of English. (Studies in Language 37). Joensuu: University of Joensuu Press.

Forster, Max

1921 "Keltisches Wortgut im Englischen", in: Heinrich Boehmer (ed.), 119-242.

1942 Der Flussname Themse und seine Sippe, Studien zur Anglisierung keltischer Eigennamen und zur Lautchronologie des Altbritischen. Munich: Verlag der Bayer.

Geary, P.

1983 Ethnic identity as a situational construct in the Early Middle Ages. (Mitteilungen der Anthropologischen Gesellschaft in Wien 113). Wien: Berger.

German, Gary

2000 "Britons, Anglo-Saxons and scholars. 19th-century attitudes towards the survival of Britons in Anglo-Saxon England", in: Hildegard L.C. Tristram (ext.), 347-374.

2004 On the rise of it-clefting and other related structures in English: Contact or indigenous development? (Typescript).

Gorlach, Manfred

1986 "Middle English--a creole?", in: Dieter Kastovsky--Alexander Szwedek, (eds.), 329-344.

[1990] [Reprinted in: Manfred Gorlach (ed.), 65-78].

Gorlach, Manfred (ed.)

1990 Studies in the history of the English language. (Anglistische Forschungen 210). Heidelberg: Winter.

Graves-Brown, Paul--Sian Jones--Clive Gamble (eds.)

1996 Cultural identity and archaeology. London and New York: Routledge.

Gretsch, Mechthild

2001 "Winchester vocabulary and standard Old English: The vernacular in late Anglo-Saxon England", The T. Northcote Taller Memorial Lecture 2000 (Bulletin of the John Rylands University Library of Manchester 83), 41-87.

2003 "In search of standard Old English", in: Lucia Kornexl--U. Lenker (eds.), 33-67.

Habermann, Mechthild--Peter O. Muller--Horst Haider Munske (eds.)

2002 Historische Wortbildung des Deutschen. Tubingen: Niemeyer.

Hatfield, James Taft (ed.)

1930 Curme Volume of Linguistic Studios, edited on the occasion of his seventieth birthday. (Language Monographs/ Linguistic Society of America 7). Baltimore: Waverly Press.

Harke, Heinrich

1995 "Finding Britons in Anglo-Saxon graves", British Archaeology Dec 1995: 7.

1997 "Early Anglo-Saxon Social Structure", in: John Hines (ed.), 125-170.

1998 "Briten und Angelsachsen im nachromischen England: Zum Nachweis der einheimischen Bevolkerung in den angelsachsischen Landnahmegebieten", Studien zur Sachsenforschung 11, 89-119.

1999 "Sachsische Ethnizitat und archaologische Deutung im fruhmittelalterlichen England", Studien zur Sachsenforschung 12: 109-122.

2002 "Kings and warriors: Population and landscape from post-Roman to Norman Britain", in: Paul Slack--Ryk Ward (eds.), 145-175.

2003 "Population replacement or acculturation? An archaeological perspective on population and migration in post-Roman Britain", in: Hildegard L.C. Tristram (ed.), 13-28.

Higham, Nicholas

1995 An English empire: Bede and the early Anglo-Saxon kings. Manchester and New York: Manchester University Press.

Hines, John

1990 "Philology, archaeology and the adventus Saxonum vol Anglorum", in: Alfred Bammesberger--Alfred Wollmann (eds.), 18-36.

1994 "The becoming of the English: Identity, material culture and language in Early Anglo-Saxon England', in: William Filmer-Sankey--David Griffiths (eds.), 49-59.

1996 "Britain after Rome: Between multiculturalism and monoculturalism", in: Paul Graves-Brown--Sian Jones--Clive Gamble (eds.), 256-270.

2000 "Attitude problems? The Old Saxon and Old English Genesis poems", in: Magnus Ljung (ed.), 69-90.

2004 Voices in the past. English literature and archaeology. Cambridge: D.S. Brewer.

Hines, John (ed.)

1997 The Anglo-Saxons from the migration period to the eighth century. Woodbridge: The Boydell Press.

Hinrichs, Uwe--Uwe Buttner (eds,)

2004 Die europaischen Sprachen auf dem Weg zum analytischen Sprachtyp. Wiesbaden: Harrassowitz.

Higham, Nicholas

1992 Rome, Britain and the Anglo-Saxons. London: Seaby.

1995 An English empire: Bede and the early Anglo-Saxon kings. Manchester and New York: Manchester University Press.

2002 "The Anglo-Saxon/British interface: History and ideology", in: Markku Filppula et al. (eds.), 29-46.

Jackson, Kenneth

1953 Language and history in Early Britain. Edinburgh: Edinburgh University Press.

[2000] [Reprinted with a new introduction by Willie Gillies. Dublin: Four Courts Press].

Kastovsky, Dieter--Alexander Szwedek, (eds.)

1986 Linguistics across historical and geographical boundaries. In honour of Jacek Fisiak. Berlin: Mouton de Gruyter.

Ker, Neil Ripley

1957 Catalogue of manuscripts containing Anglo-Saxon. Oxford: Clarendon Press.

[1990] [Reissued].

Koch, John T.--John Carey (eds.)

1995 The Celtic heroic age. Literary sources for ancient Celtic Europe and early Ireland and Wales. Malden MA: Celtic Studies Publications.

Kornexl, Lucia

2000 "Concordes equali consuetudinis usu--Monastische Normierungsbestrebungen und sprachliche Standardisierung in spataltenglischer Zeit", in: D. Ruhe--Karl-Heinz Spiess (eds.), 237-273.

Kornexl, Lucia--U. Lenker (eds.)

2003 Bookmarks from the past. Studies in English language and literature in honour of Helmut Gneuss. Frankfurt/Main: Peter Lang.

Krygier, Marcin--Liliana Sikorska (eds.)

2004 For the love of the Inglis lede. (Medieval English Mirror 1). Frankfurt/Main: Peter Lang.

Laing, Lloyd--Jennifer Laing

1990 Celtic Britain and Ireland. The myth of the Dark Ages. Dublin: Irish Academic Press.

Lenz, Katja--Ruth Mohlig (eds.)

2002 "Of dyuersitie & change of langage". Essays presented to Manfred Gorlach on the occasion of his 65th birthday. Heidelberg: Winter.

Ljung, Magnus (ed.)

2000 Language structure and variation. A Festschrift for Gunnel Melchers (Acta Universitatis Stockholmiensis: Stockholm Studies in English, XCII). Stockholm: Almqvist & Wiksell International.

Lutz, Angelika

2002a "When did English begin", in: Teresa Fanego--Elena Seoane (eds.), 145-171.

2002b "Sprachmischung in der deutschen and englischen Wortbildung", in: Mechthild Habermann--Peter O. Muller--Horst Haider Munske (eds.), 407-437.

Magoun, F.P.

1937 "Colloquial Old and Middle English", Harvard Studies and Notes in Philology and Literature 14: 167-173.

Malone, Kemp

1930 "When did Middle English begin?", in: James Taft Hatfield (ed.), 110-117.

Miller, Gary

2004 "The morphosyntactic legacy of Scandinavian-English contact", in: Martin Krygier--Liliana Sikorska (eds.), 9-39.

Orr, Robert A.

1999 "Evolutionary biology and historical linguistics", Diachronica 16: 123-157.

Pilch, Herbert

1970 Altenglische Grammatik. Munich: Hueber.

Pelteret, David

1981 "Slave raiding and slave trading in early England", Anglo-Saxon England 9: 99-114.

1995 Slavery in Early Medieval England. Woodbridge: Boydell.

Pohl, Walter

1991 "Conceptions of ethnicity in Early Medieval studies", Archaeologia Polona 29: 39-49.

1997 "Ethnic names and identifies in the British Isles: A comparative perspective", in: John Hines (ed.), 7-40.

Poussa, Patricia

1982 "The evolution of Early Standard English: The creolization hypothesis", Studia Anglia Posnaniensia 14: 69-85.

Ramisch, Heinrich--Kenneth Wynne (eds.)

1997 Language in time and space. Festschrift fur Wolfgang Viereck. Stuttgart: Franz Steiner.

Riehle, W. (ed.)

1995 Anglistentag 1994 Graz. Tubingen: Niemeyer.

Ruhe, D.--Karl-Heinz Spiess (eds.)

2000 Prozesse der Normbildung und Normveranderung im mittelalterlichen Europa. Stuttgart: Franz Steiner Verlag.

Shennan, Stephen J.

1991 "Some current issues in the archaeological identification of past peoples", Archaeologia Polona 29: 29-37.

Sims-Williams, Patrick

2003 The Celtic inscriptions of Britain: Phonology and chronology, c. 400-1200 (Publications of the Philological Society 37). Oxford: Blackwell.

Slack, Paul--Ryk Ward (eds.)

2002 The peopling of Britain: The shaping of a human landscape. (Linacre Lectures 1999). Oxford: Oxford University Press.

Stenton, F. M.

1971 Anglo-Saxon England. (3rd edition). Oxford: Clarendon Press.

Tristram, Hildegard L. C.

1995 "Aspect in Contact", in: W. Riehle (ed.), 269-294.

1997 "DO in Contact?", in: Heinrich Ramisch--Kenneth Wynne (eds.), 401-417.

1999 "The Celtic Englishes--Zwei grammatische Beispiele zum Problem des Sprachkontaktes zwischen dem Englischen und den keltischen Sprachen", in: Stefan Zimmer--Rolf Kodderitzsch--Arndt Wigger (eds.), 254-276.

2002a "Attrition of inflections in English and Welsh", in: Markku Filppula et al. (eds.), 111-149.

2002b "The politics of language: Links between Modern Welsh and English," in: Katja Lenz--Ruth Mohlig (eds.), 257-275.

Tristram, Hildegard L. C. (ed.)

2000 The Celtic Englishes II. Heidelberg: Winter.

2003 The Celtic Englishes III. Heidelberg: C. Winter Universitatsverlag.

Trudgill, Peter

2000 Sociolinguistics: An introduction to language and society. (4th edition). London: Penguin.

2001 "Received pronunciation: Sociolinguistic aspects", Studia Anglica Posnaniensia 36: 3-13.

White, David

2002 "Explaining the innovations of Middle English: What, where, and why?", in: Markku Filppula et al. (eds.), 153-174.

2003 "Brittonic influence in the reductions of Middle English nominal morphology", in: Hildegard L.C. Tristram (ed.), 29-45.

Whitelock, Dorothy

1952 The beginnings of English society. (Pelican History of England 2). Harmondsworth: Penguin.

Zimmer, Stefan--Rolf Kodderitzsch--Arndt Wigger (eds.)

1999 Akten des zweiten deutschen Keltologensymposiums (Bonn, 2.-4. April 1997). Tubingen: Niemeyer.

HILDEGARD L.C. TRISTRAM

University of Potsdam