Worth(y).

To an earthbound empiricist Roger Lass is an annoying fellow. He is so full of bright ideas racily expressed. He adduces plenty of real data, and tells mediaevalists home truths, e.g. on difficulties in practice of maintaining a spelling-system with patterns that cut across those of speech (1992: 105). Sometimes his very titles evoke scenes from Old English poetry (1978). (1) Yet he goes by complicated routes even to simple conclusions, e.g. multivalency of monophthongization products (1988); and he will take off stratospherically into heights of theoretical abstraction long before resources of empirical testing are exhausted, and let that skew or even reverse the argument implied in the material he has so helpfully presented (1993: 28-41). Well, in my stamping-ground of Anglo-Saxon charter boundaries fricative voicing must stay unvisited: they offer practically no direct data. For the longer-running Lass theme (1988, 1989, 1992, etc.) of front rounded vowels they do however yield plenty, namely on questions of relation between eo and y in late Old English. Some have been exhibited (Kitson 1993: 14-20). A few more constitute this offering of a humble empiricist to the virtuoso theoretician to do with what he will.

The grammarians (Campbell 1959: [section][section]320-324, Sievers-Brunner 1965: [section][section]113, 118, Luick 1914-1940: [section]286, Hogg 1992: [section]5.183-184) have long recognized a falling together of eo and y in late West Saxon in the environment w--rC (C = dental or h, here etymological), though failing to spot the significance of the fact that the r + consonant is a lengthening environment (Kitson 1993: 20 and references). That falling together is bodied forth in monophthongization pweores > pwyres, which the charter material shows to be geographically coherent (Kitson 1993: 18 = map 6), affecting Wessex from Wiltshire and north Berkshire south-west as far as there is evidence for it (there isn't for Devon and Cornwall), and Oxfordshire and parts immediately north and east of it north of the Thames. Most of Hampshire and some adjacent parts are definitely not affected. "Chaque mot a son histoire", and that is especially likely for a sound-change whose spellings are so mixed that they have misl ed the grammarians as to its nature. One would like to know whether it has comparable coherence in other words where it occurs, especially ones with a phonetic structure w--rC or close to that, and whether if it does its domain in them resembles that in pwyres, as the patchy incidence of y-forms for betweonan etc. (Kitson 1993: 15 = map 5) clearly does not. Conversely one would like to clear up as far as possible ambiguity caused by the potential effects of this change as to the dialectal origins of words.

Contribution to a festschrift invites us to think of worthiness; and words "worth(y)" are ones where this potential ambiguity occurs. It so happens that as with words for "wood-bank" and "root" (Kitson: in press), a pair (?) of topographic words are homophonous, in a good many grammatical forms, with a pair (?) of dialectal variants of (a) word(s) of quite different meaning in the literary language. The quirks of vowel-levelling in the environment w--rC, together with inadequacies in traditional accounts of Old English dialects, have led for the topographic words to errors in standard reference-works, and to inaccuracies by me in the past. My purpose here is to correct them, casting in the process a certain amount of incidental light on the dialectology of literary "worth(y)".

The element(s) weord or wyrd, appearing as "-worth" in modern place-names, and meaning roughly "tenant farm", are substantive elements in 63[44] charter boundary features. (2) Their distribution is: Warks S79; Worcs S1329, S1361, S64(ii), S1599; Gloucs S467, S553, S786(vi), S1551; Somerset: S563, S793; Dorset S419(ii), S423, S4459, S656(i), S710, S96; Wilts S364, S881, S899, S891(i) (x2), S275(i) (x 2), S540 (x 2), S393(i) (x 2), S229(i) (x 2), S1581(i), S1577(i) (x 2), S1583, S1579; Hants S412 (x 3), S1275, S693(i), S842(v), S1007; Sussex S108(i), S50(iv); Kent S316, S1434(i); Surrey S528(x 2), S847(i), S911 (vi), S911(vii); Herts S916(iii); Berks S529, S542, S605 (x 3), S614 (x 2), S567; Oxon S887(ii); Hunts S437; Lincs S782. There is also one wyrde hyrne 'corner of a "worth"' in Surrey S911(vi). The vowel is consistently y except for the following: eo in Worcs S1329, Gloucs S1551, S786(vi), Sussex S108(i), Surrey S528 (both features), S911(vi) (the feature that is not a hyrne), Berks S614 (one feature); ye in Dorset S656(i); u in Wilts S364, Berks S529, S614 (one feature); o in Kent S316, Wilts S1583, S1579; w- with no vowel in Wilts S1577(i) (both features). Simplex wyrd also occurs as a qualifier in Worcs S786(i) and in Surrey S621 (x 2, both with eo), and -wyrd embedded in a place-name qualifier in Somerset P30, Dorset S419(i) (-i-), Wilts S1582, Berks 5614 (-u-), S657, Hunts S1562 (x 2), Nhants S533.

In the far south-west the place of wyrd is taken by weordig, substantive element in Devon S405(ii), (x 2), S998, S1547, and S386, also Herefs S1561 (referring to a feature in Gloucs), Worcs S1272, 5786(i), Hants S376, Berks S713, and weordign spelt weording in Worcs S1272. The vowel in relatively early texts S1272 and S405(ii) is consistently eo; S786(i) mixes -eo- and -o-, S1561, S386, S376 have -o-, S998 and S1547 -u-, and S713 -y-. It seems clear that in weordign eo is the only underlying vowel, with o in S1561, S386, and S376 as visibly in S786(i) a variant of it, y and u products of "late West Saxon" levelling in the environment w--rC.

Of wyrd spellings with vowels other than y those with -eo- in S108(i), S528, S621 and S911(vi) form a geographically compact group in the south-east, to which S316 probably ought to be added, since its Hodoworda is visibly Latinized in the ending -a, and -o- for -eo- would be plausible as another Latinization. (S316 keeps y in the name Lyminge, so if this is Latinization it is definitely for -eo- not -y-.) S1551, S786(vi) and S1329 are at the extreme west-north-west of the distribution. S786(vi) and S1551 refer to the same feature, S1551 being a composition probably as late as the second half of the eleventh century. S1329 spells with -eo- in 974 the same feature that S1361 does with -y- in 985. S1577(i), S1583 and S1579 belong to a group of texts redacted at Malmesbury in the late twelfth or early thirteenth century which tend to avoid the spelling eo, or y with back values. S1577(i)'s w- might represent either OE wy- or late wu-: cf. repeated wile in S1577(i)(ii) beside once wulle in S1577(u) for OE wylle. S1583's and S1579's -o- probably represents OE -eo-, since the vowel to be expected as a normalization of y before r is u as thrice in hurste in S1583; but this is a probability not a certainty. (3) If these probabilities are accepted as a basis for argument, and taking the weordig(n) spellings together with the wyrd/weord ones, it is possible to see a second compact area with original -eo- not -y-, comprising Worcestershire except its eastern part, Gloucestershire north and west from about Cirencester, and mid-north Wiltshire immediately south of there.

The ye in Dorset S656(i) might be expected in principle to be a reflex of eo not y (cf. Ek 1972), but it is in a fifteenth-century cartulary, in the same part of Dorset as most of the -y- spellings, so probably to be seen as just a scribal error. Wilts S364, in a fourteenth-century cartulary, spells y as u in not only this word but every one where it should have a back value (sigewunne, hulle), whether as part of a cartularial scribal process or just possibly as an extreme part of the Mercianizing spellings which are authorial in the survey. S614's -u- reflects "late West Saxon" levelling in w-rC.

What we have in all this when isolated "late West Saxon" and cartularial spellings are weeded out is a distribution where one large central part, "West Saxon" in the traditional grammarians' terms or "Thames Valley Saxon" in mine (Kitson 1995: maps 4-5), has wyrd consistently with y, contrasting with two or three peripheral parts where the normal spelling is -eo- tending to -o- in latish texts. Of these peripheral parts one is the grammarians' "Kentish", my "old south-eastern" (Kitson 1995: maps 3-4), one depending on what we discover about its wider affinities might be labelled either "Anglian" or "Hwiccean", and the one with weordig is my "south-western" (Kitson 1995: maps 3, 7), a category not recognized by the traditional grammarians. The element of doubt is whether the development of vowels was exactly the same in monosyllabic weord as in weordig where original secondary stress on the second syllable might be a factor inhibiting reduction of the main stress in place-name compounds.

The forms exhibited as normal here are not those taken as such by Smith (1956) or the OED. The OED traces Worth, sb (2) to "OE. worp (weorp), wurp", as if -eo- were a variant of -o-, which will not do. Smith gives the quadruple head-forms word, weord, wurd, wyrd and wordig, weordig, wurdig, wyrdig but operates with word and wordig except where quoting from particular texts that whose forms are different. It appears from the body of his entry that he has been influenced by the use of word in the Lindisfarne and Rushworth Gospel glosses, where it means something grander than it does in charters, regularly translating Pontius Pilate's atrium. But these are texts in the Northumbrian dialect, where there is well known to be a sound-change usually represented as "retraction of e to o ...between u and r followed by a consonant" (Campbell 1959: [section]147; Hogg 1992: [section]5.30 is similar). So their word corresponds to southern weord tending to imply that the proper label for the Gloucs and Worcs -(e)o- forms wa s "Anglian".

Since breaking of e to eo before rC is supposed to be a common Old English development, it is intrinsically likely that the Northumbrian sound-change was we->weo->wo- as in the charters of the Hwiccean region, but at a prehistoric period. This will be another sound-change that like combinative u-mutation (Stiles 1983, Kitson 1992) began in Northumbria and worked somewhat patchily south. The extant Northumbrian evidence is wholly indecisive as between that and the direct retraction posited by the grammarians, but occasional instances of the same change in literary texts in other dialects which they have to explain away as "untrustworthy" (Hogg 1992: [section]5.30) are easier to accommodate by this indirect route. The most conspicuous ones are precisely for the topographic word w(e)ordign, which appears as GP wordigna, DP wordignum, translating forms of platea in the Vespasian Psalter ("A") and derivative psalter-glosses. (The translator presumably understood platea as a French place rather than as a Latin "str eet" as most of his fellows do.) This does not justify postulating a separate original form with o as suggested by Sievers -- Brunner (1965: [section]113 n. 1). (4) The metrical Paris Psalter, which draws on an A-type gloss, naturalizes the word in one instance to weorpige, C to wurdigum implying original -y- or -eo-. (5) The Vespasian Psalter's is a west midland dialect, which whether native to Lichfield as S.M. Kuhn and others have guessed or to Wootton Wawen, Warwickshire, as I am inclined to, (6) cannot be very far away spatially from S1272.

The Northumbrian evidence clinches what the existence of weordig suggests anyway, that weord is the underlying form in at least some Old English dialects, and is what dictionaries like the OED and Smith's Elements which give priority to Anglian dialects should use as their head-form. There are then three possible ways to account for the observed distribution of wyrd. We can suppose (very much as for the items in Kitson 1995: map 5) that "Thames Valley Saxon" had a different ablaut-form from the beginning, and that observed distribution directly reflects stabilization between domains of that and competing dialect forms at the close of the migration period. At the other extreme we might suppose that weord were the only Old English underlying form, wyrd wholly a result of "late West Saxon" levelling in the environment w--rC. Or it may be that both those explanations have a measure of truth, that there were two forms originally but that the domain of wyrd has spread as a result of levelling in w--rC.

That there were two distinct dialect origins tends to be confirmed by divergence in gender. Weord/wyrd has accusative -e as if feminine over most of the country; nominatives Gloucs S467 wyrp, Sussex S50(iv) uuyrth show that the -e is an inflectional ending. But only in four or five surveys is feminine gender grammatically explicit, shown in Somerset S563 and Wilts S881 by articles, mid-Hants S1275 and N. Surrey S847(i) by adjectives of direction, and N. Surrey S911(vi) by a genitive in -e (this less certain than the others in view of S412 below). Explicitly masculine grammar is shown by an article in Berks S529, by adjectives in Berks S605 and two features in SE Surrey S528(i), in the latter reinforced by endingless accusatives weorp. Endingless accusatives occur also in E. Dorset S969 and mid-Hants S1007. N. Hants S412 has genitive -e in one feature beside what should be strong masculine accusative sudeweardne governed by the preposition to (which should govern dative) in another. This kind of slippage does occur in charters elsewhere (on in the same sense regularly governs accusative), so S412 may either be viewed as a bizarre mixture really feminine (7) or really masculine, with genitive like accusative just looking as if levelled from the feminine declension. This would cast doubt on S911(vi) as well, though given its geographical closeness to S847(i) that is probably still feminine.

The combination of rare masculine grammar with with rare -eo- in two features in S528(ii) is a strong indication that those two variables originally went together, and that the underlying contrast is between wyrd feminine "Thames Valley Saxon" and weord masculine "old south-eastern", "south-western", and "Anglian". This accounts for all the stragglers in a way that was not possible with the dialect models available for Kitson (1990: 195), the beginning of whose penultimate sentence I hereby retract. Masculine gender for Anglian finds confirmation in Rushworth Mark 1468 bifora done word. The separate dialectal position of "south-western" here means that it is not possible to argue from this distribution to what should fill the gap in the far south-west of the map of pweores/pwyres (Kitson 1993: map 18).

The underlying dichotomy just posited relates sensibly to evidence in continental Germanic languages. Continental place-names have two corresponding elements, werth mainly Frisian and Low German and wurth Old Saxon and Old High German, corresponding phonetically to weord and wyrd respectively. Werth in High German is reported as masculine by Kluge -- Seebold (1989: 787), neuter by Forstemann ([1967], 2: 1238-1240) and Bach (1952-1956, 3: 443), in Low German by all as masculine with variants neuter; wurth is mainly feminine according to Forstemann ([1967], 2: 1443-1444) and Bach (1952-1956, 3: 452), Kluge-Seebold (1989: 801) say masculine in Old Frisian and perhaps more widely. So this is in some sort a dialect agreement between "Kentish", "Anglian", and Frisian on the one side against "West Saxon" [="Thames Valley Saxon"], Old Saxon, and Old High German on the other, tidiest if where the German authorities contradict each other we believe Kluge -- Seebold for werth and Forstemann and Bach for wurth. (8) The f orms to be posited for Germanic are *werdaz m. and *wurdiz f.

Granted these origins, it seems likely that though levellings of gender and inflection have been considerable, the domains of weord and wyrd as phonetic entities have not actually changed very much since the end of the migration period. That of wyrd does not extend more than averagely into the south-east for a "Thames Valley Saxon" word (cf. Kitson 1995: map 5); the changes seen in S1329/S1361 and S713 look only marginal; a stage -y- does not necessarily have to be posited behind S998 and S1547 at all, though one is likely.

Let us turn then to the question of how these dialectal relations compare to those of forms for "worth(y)" in the sense "valuable, honourable". According to the standard grammarians this should be a non-question. The adjective weord and its derivative eWS wierde, 1WS wyrde to them are simply different words (Campbell 1959: [section][section]146-147, 149, Hogg 1992: [section][section]5.22, 5.31, Sievers -- Brunner 1965: [section][section]113-114, 107). How one might characterize the difference is another matter. An answer which works often enough to be useful as a rule of thumb is that weord, substantivized already in Common Germanic as "worth, value", is "valuable" (of things), whereas wyrde is "worthy" (of people); the OED notes that constructions "worthy of" and "worthy to" are normally only expressed in Old English by wyrde. But there are exceptions to that in both directions. The authorities do not explain the origin of wierde. One possible answer would be from a i-infix in compounding (cf. word 'word': h roedwyrde 'hasty of speech'); if so the amount of levelling to be posited would be quite large. Lexicographers of Old English have bunched the two words together: not only Bosworth -- Toller (1898) but more surprisingly Roberts -- Kay -- Grundy (1995, 1: 410, 1995, 2: 1508) do not give a separate entry for wierde. The Anglian form of the latter is supposed to be wyrpe wurpi wirpi(-) (Campbell 1959: [section]149, Sievers -- Brunner 1965: [section]114). So nominatives wyrp and singular weorpe are among imaginable forms that ought not to occur.

The material presented by the Toronto Microfiche Concordance shows that synchronically at least that is not the whole story. A full demonstration would be much too bulky for space and time to permit here, especially with the simplex words needing to be kept apart from the various senses of the noun weord and the verb pres. 3sg. wierd etc. 'becomes'. But we may broach some salient features of the distributions in compounds. The ones with the most useful samples are unwyrpe 'unworthy', arwyrpe 'honourable', and deorwyrpe 'precious'. The i-mutated vowel not -eo- is on the observable frequencies the norm in all of them, as would be expected for the first two but not the third by the rule of thumb about human/inanimate referents, for all three on the theory that - wyrpe originated as a composition-form. The vowels for the syllable -wyrb- in the various grammatical forms of these compounds, ignoring comparatives, superlatives, and other derivatives, of which especially for unwyrpe and arwyrpe there are a considerab le number, are: unwyrpe, -eo- 18, -ie- 3, -y- 40, -u- 30; arwyrpe, -eo- 4, -ie- 2, -y- 226, -u- 251; deorwyrpe, -eo- 6, -oe- 1, -i- 4, -y- 53, -u- 102. Both spelling-types are applied to both human and inanimate referents, often in the same texts. Nominatives unwyrp and unwurP are quite frequent, arwurp occurs occasionally; the final -e of deorwyrpe is not reduced. Unweorpe occurs at least once as a nominative singular (Boethius 27.62.18).

The largest single difference between these samples and those for the topographic word(s) is the significant incidence of late -u- for -y- or -eo- (probably mainly but probably not exclusively for the former). Adjectival compounds, including ones in un-, have main stress on the first syllable, so this is partly a matter of accentuation, front rounded vowels whatever their quality being less well preserved in syllables which could at most have secondary stress and must in late Old English for practical purposes have been unstressed. It is also partly a dialectal matter. The individual author from whom most -u- forms are cited is AElfric, whom there is reason to place dialectally in the bight of the Bristol Avon in north-west Wiltshire not far from S1577(i). (9) The rest mainly, though not exclusively, come from texts there is reason to place north of the Thames, where there is next to no sample of weord/wyrd. That may or may not apply to the Hatton manuscript of Gregory's Dialogues, where in more than fifty in stances arwurp(-) is apparently the exclusive form, having been a significant minority beside majority arwyrP(-) in the Cotton manuscript. AElfric's tendency to -u- forms in these words is reminiscent mutatis mutandis of his strong preference for betwux (Kitson 1993: 17). It contrasts with his complete lack of them in pwyr(h)- derivatives (Kitson 1993: 21), to impress on us that levelling in w--rC is not a unitary sound-change and "chaque mot a son histoire".

The distribution of -eo- forms where they occur resembles in its general shape that of topographic weord against wyrd. Unweorp(-) is cited from the translation of the Benedictine Rule by Aethelwold of Winchester, from the Boethius of King Alfred whose dialect belongs fairly far south in Wessex (Kitson 1993: 23-30), from the Cotton MS of Gregory's Dialogues, some sort of west midland "Anglian", from The Wonders of the East whose language ccntains some "Mercian" elements (Sisam 1953: 73, 94), from a gloss, and from the Orosius, which is to be located at Bristol (Kitson 1996), whose dialect anyway has a strong Anglian component. The combination of south Wessex with "Anglian" is much like that of weord against wyrd, taken at that level of abstraction. In terms of detailed geography it is not so like, since though Winchester might be argued to be an outlier of the south-eastern weord area, Alfred's dialect fits most likely to Wilton where the topographic word should be wyrd. Forms in arweorp(-) are cited from the Life of St. Machutus, which belongs dialectally to north Gloucestershire (Kitson 1993: 35-40), from S1281 a grant by a bishop of Worcester, from the Vercelli Homilies which belong somewhere in the southeast north of the Thames, and from a canticle in the Bosworth Psalter, which has not been closely placed but whose vocabulary agrees predominantly with the west midland "A" group. The word as well as the form may well be absent from distinctively south Wessex authors. When it comes to deorweorp(-) the form of the second syllable is liable to be affected by dittography from the first. This I take it is the reason for one such form cited from AElfric! Apart from that texts cited are south-west Midland again, Wulfstan, the Martyrology, Lambeth Psalter, Letter of Alexander to Aristotle perhaps east Midland (Sisam 1953: 88-95), with diorwiorpestoen beside diorwiorp- forms in the Canterbury Psalter to provide a south-eastern presence.

Forms in wierp- are almost completely confined to the Cura Pastoralis. For unwierp- they wholly are, as for ([+ or -]un-)taelwyrd- '(un)blameworthy' (-le- 3, -y-4) and stoelweorp- 'stalwart' (-le- 3, -y- 4), the only other words for which -ie- forms strike the eye from the Microfiche Concordance reverse-spelling frequency-list. The Cura Pastoralis notoriously has a good deal of io for etymological eo (Campbell 1959: [section]296), and wierpe might well be a secondary dialectal effect of the same kind (ie rather than io presumably conditioned by the following e) for weorpe which was presumably the authorial form since the main author of this translation was King Alfred. A single wierpne in the Orosius beside plentiful unweorp(-) may also be open to suspicion of being scribal rather than authorial. Still arwierpra from the Parker Chronicle for 716 and wierpne in the Laws of Ine do seem to be unambiguous evidence for the doctrine of Campbell (1959: [section]149) that West Saxon did not participate in retraction of *wirdi to *wurdi and for the citing of wierpe not wyrpe as the form for "early West Saxon" in general. In my terms "West Saxon" here means specifically "Thames Valley Saxon": the evidence of unweorp tends to show that whatever the underlying development south Wessex did not share it.

What anyway is the relation between the weorp(-) and wyrp(-) forms? There are two possible general lines of explanation, not necessarily mutually exclusive, which both involve analogical levelling of some kind. One would have it at the level of words or morphemes, with some subdialects reshaping compounds etymologically in -wyrde by analogy with simplex weorp. (Campbell 1959: [section]324 n. 1 says "wyrd- for weord- in compounds is influenced by wyrde"; it would be the other way round here.) The other is that in some subdialects the earliest West Saxon product of i-mutation of eo, which according to Campbell (1959: [section]202) was io, did not in these words develop to le but was levelled back to eo as in leornian 'learn', gepeode 'language', and other words cited by Campbell. If this applied to the midland texts with -weorp- it would make them for this particular dialect feature a kind of non-"Thames Valley Saxon" "West Saxon", implying a distribution vaguely analogous to that of hlidgeat in Kitson (1995: m ap 9). Perhaps it is simpler to suppose that the phonetic levelling applied in south Wessex but that in the west midlands analogy was at a verbal level.

As often with dialectological inquiries, the details are rather resistant to tidy summary. If we are to have a resounding generalization to end with, it may be that topographic and non-topographic words are shaped by the same general sense of what sounds right in a dialect. I expect the honorand will be able to provide a better one by and by. May he count this a Lasswyrdne dael.

(1.) Cf. Christ 678-679, Fates of Men 21-26, and Meritt (1945).

(2.) Conventions of reference are as in my previous publications, based on Sawyer (1968) and Pelteret (1990), elaborated as Kitson (1990: 186-187; 1995: 50-51).

(3.) If the repeated spelling wode for OE wudu were as some scholars would have it a sign of graphic avoidance of the sequence wu-, then that might apply here as well. But I think (Kitson 1992: 9-10, 22-23) that wode reflects opening of the vowel u in an open syllable, and is not likely to be relevant to this closed syllable.

(4.) Hogg's derivative reference to "EWS worpig 'enclosure', found once in CP(H) alongside CP(C) weorpig," is slightly more misguided. What reason is there to count any form found only in the Hatton manuscript of the Pastoral Care as "early" West Saxon at all?

(5.) The full conspectus of pertinent psalter-gloss forms is: [17.sub.43] AC wordigna, B (Junius) wordina; [54.sub.12] AB wordignum, C (Cambridge) wurdigum, P (Paris) weorpige; [143.sub.14] A wordignum, E (Canterbury) wordignum P wordum.

(6.) From a balancing of indications explored in Kitson (forthcoming: ch. 9).

(7.) Especially bizarre were weardne a blend with weak weardan as formerly surmised by me (Kitson 1990: 195).

(8.) I have not seen Valtavuo (1952), referred to by Kluge -- Seebold as discussing the etymologies of these words inconclusively.

(9.) This closer placing than in my (Kitson 1993) is argued for in Kitson (forthcoming: ch. 7).

REFERENCES

Bach, Adolf

1952-1956 Deutsche Namenkunde. 3 vols. Heidelberg: Winter.

Bosworth, Joseph - Thomas Northcote Toller

1898 An Anglo-Saxon dictionary. Oxford: OUP.

Campbell, Alistair

1959 Old English grammar. Oxford: OUP.

Colman, Fran

1992 Money talks: Reconstructing Old English. Berlin, New York Mouton de Gruyter.

Colman, Fran (ed.)

1992 Evidence for Old English: Material and theoretical bases for reconstruction. (Edinburgh Studies in the English Language 2.) Edinburgh: John Donald.

Duncan-Rose, Caroline - Theo Vennemann (eds.)

1988 On language: Rhetorica phonologica syntactica: A festschrift for Robert P. Stockwell from his friends and colleagues. London, New York: Routledge.

Ek, Karl-Gustav

1972 The development of OE y and e in south-eastern Middle English. Lund: Gleerup.

Fisiak, Jacek (ed.)

1995 Mediaeval dialectology. Berlin, New York Mouton de Gruyter.

Forsteman, E.

1900-1916 Altdeutsches Namenbuch, Vol. I: Personennamen, Vol. II: Ortsnamen. (ed. H. Jellinghaus.) Bonn.

[1967] [Reprinted Munich: Fink; Hildesheim: Olms.]

Hickey, Raymond - Stanislaw Puppel (eds.)

1997 Language histoy and linguistic modelling: A festschrift for Jacek Fisiak on his 60th birthday. Berlin, New York: Mouton de Gruyter.

Hogg, Richard M.

1992 A grammar of Old English, Vol. I: Phonology. Oxford: Blackwell.

Kitson, Peter Robert

1990 "On Old English nouns of more than one gender", English Studies 71: 185-221.

1992 "On the chronological and geographical spread of Old English combinative u-mutation", Studia Neophilologica 64: 3-23.

1993 "Geographical variation in Old English prepositions and the location of AElfric's and other literary dialects", English Studies 74: 1-50.

1995 "The nature of Old English dialect distributions, mainly as exhibited in charter boundaries, Part I: vocabulary", in: Jacek Fisiak (ed.), 43-135.

1996 "The dialect position of the Old English Orosius", Studia Anglica Posnaniensia 30: 3-35.

1997 "The root of the matter: OE wyrt, wyrtwale, -a, wyrt(t)rum(a) and cognates", in: Raymond Hickey -- Stanislaw Puppel (eds.), 127-142.

forthcoming A new foundation for Old English dialect study.

Kluge, Friedrich -- Elmar Seebold

1965 Etymologisches Worterbuch der deutschen Sprache. (22nd. edition.) Tubingen: Niemeyer.

Lass, Roger

1978 "Mapping constraints in phonological reconstruction: On climbing down trees without falling out of them", in: Jacek Fisiak (ed.), 245-286.

1988 "The 'Akzentumsprung' of Old English eo", in: Caroline Duncan-Rose - Theo Vennemann (eds.), 221-232.

1989 "System-shape and the eternal return: Front rounded vowels in Old English", Folia Linguistica Historica 10: 163-198.

1992 "Front rounded vowels in Old English", in: Fran Colman (ed.), 88-116.

1993 "Old English fricative voicing unvisited", Studia Anglica Posnaniensia 25-27: 3-45.

Luick, Karl

1914-1940 Historische Grammatik der Englishchen Sprache. 2 vols. Leipzig: Tauchnitz.

Meritt, Herbert Dean

1945 "Beating the oaks", American Journal of Philology 66: 1-12.

Mic(rofiche) Conc(ordance) = Venezky, Richard L. -- Antonette diPaolo Healey 1985 Pelteret, David A.E.

1990 Catalogue of English post-Conquest vernacular documents. Woodbridge: Boydell.

Roberts, Jane -- Christian Kay -- Lynne Grundy

1995 A thesaurus of Old English. 2 vols. London: King's College.

Sawyer, Peter Hayes

1968 Anglo-Saxon charters: An annotated list and bibliography. London: Royal Historical Society. [An early draft of a second edition, under the general editorship of Dr. Simon Keynes, has been circulated and is available from Dr. Susan Kelly of Newnham College, Cambridge.]

Sievers, Eduard -- Karl Brunner

1965 Altenglischen Grammatik nach der angelsachsischen Grammatik von Eduard Sievers. (3rd edition.) Tubingen: Niemeyer.

Sisam, Kenneth

1953 Studies in the history of Old English literature. Oxford: OUP.

Smith, Albert Hugh

1956 English place-name elements. 2 vols. (English Place-Name Society vols. 25-26.)

Stiles, Patrick

1983 "The attestation of early Old English wudu 'wood': A note on the evidence for the date of combinative back-mutation", Neuphilologische Mitteilungen 84: 415-418.

Valtavuo, T.

1952 Der Wandel der Wortstamme in der Synonymik fur 'Hugel'. Helsinki.

Venezky, Richard L -- Antonette diPaolo Healey

1985 Microfiche concordance to Old English. Toronto: Pontifical Institute of Mediaeval Studies.