Negative-initial sentences in Old and Middle English.

Beside other things best left unnamed here, I share with Roger a love for historical puzzles and scenarios. Therefore, a small contribution on a little ripple in the verb syntax of the oldest English seems a fitting tribute in the context of this volume.

There is an intriguing and seemingly minor syntactic difference between Old English and Middle English. (1) In Old English negated main clauses, there are two principal word order patterns: in the first and very dominant pattern, the negated finite verb comes first, as illustrated in (1); in the second, the subject comes first, immediately followed by the negated finite verb, as illustrated in (2). It is, however, rather difficult to find examples where the negated finite verb is preceded by a topic, although (3) is one of the rare examples:

(1) Nolde se Hoelend for his bene swapeah hym fram gewitan not-wanted the Lord for his prayer however him from depart

AEHomP.XIV.199

(2) poet cild ne mihte na oa gyt mid wordum his hoelend gegretan the child not could not yet with words his Lord greet

AEHomTh.i.202.20

(3) oinra synna ne weoroe ic gemunende, ac gemun du hiora your sins not become I mindful, but be-mindful you them 'may I not be mindful of your sins, but you be mindful of them'

CP.53.413.20'

In early Middle English, all three word orders are frequently attested:

(4) Ne cam ic noht te bidden [LANGUAGE NOT REPRODUCIBLE IN ASCII]ew forbisne Not came I not to give you example VV.15.9

(5) Ich nat nawt pe time I not-know not the time SW.249.18

(6) Alle pine preates ne drede ich all your threats not dread I St.Katherine.2102

These facts suggest that in Old English, the occurrence of a topic is not readily compatible with a negative element in first position whereas in early Middle English, it is. (2) This touches on the nature of the first constituent position in Old English main clauses, which is related to issues concerning the Verb Second constraint. Since there is no independent evidence that there are any essential changes in the Verb Second constraint between Old English and Middle English (van Kemenade 1987, 1997), I will here explore the possibility that changes in the force of negation are responsible for the difference between Old and Middle English illustrated above. To see this, we will first discuss the issues concerning the Verb Second constraint in section 1. Then, we will look at the historical line of development of negation, and in section 2, we will consider negative-initial sentences in the earliest Old English of Beowulf In section 3, I will present an analysis for the historical development of negative-initi al sentences. The conclusion will be that the possibility of having a topic in a negated sentence in Middle English arose out of the weakening of the negative element ne in first position, which fits with the general trends observed in Jespersen's cycle (Jespersen 1917).

1. Verb Second in Old and Middle English

Old English clause structure cannot be considered without regard to the Verb Second constraint. While Old English has mixed OV and VO word orders, and can be reasonably analysed as being typologically an SOV language, it is also clear that the sentence has some satellite positions in main clauses that are reserved for some first constituent and the finite verb. An initial illustration of this is given by the following sentences:

(7) a. hwi wolde God swa lytles pinges him forwyrnan

why would God so small thing him deny

'why should God deny him such a small thing?' AEHomTh.i.14.2

b. On twam pingum hoefde God poes mannes sawle gegodod

in two things had God the man's soul endowed 

'With two things God had endowed man's soul' AEHomTh.i.20.1

Observe that these examples illustrate the phenomenon of subject-aux-inversion that we still find in Present Day English. In Old English, however, it is not restricted to interrogative and negative-initial contexts, as (7b) illustrates. Such constructions are analysed in van Kemenade (1987) as involving preposing of the first constituent and Vf to the satellite sentential positions in CP as in the following adapted shucture:

(8)

[FORMULA NOT REPRODUCIBLE IN ASCII]

But not all Old English main clauses conform to this pattern. In sentences introduced by an interrogative element, a negative element or the short adverbial pa, subject-verb inversion is canonical, as (7a) and the following examples illustrate:

(9) a. For hwam noldest pu de sylfe me gecydan poet ...

for what not-wanted you yourself me make known that ...

'wherefore would you not want to reveal to me yourself that...' AESL.XXXIII.307

b. pa foron hie mid prim scpum ut

then sailed they with three ships out

'then they sailed out with three ships' ChronA.AD 897

c. Ne sceal he noht unalyfedes don

not shall he nothing unlawful do

'he shall not do anything unlawful' CP.60.15

The same is not true for main clauses introduced by a topic. While inversion is near-canonical when the subject is a noun, as in (7b), pronominal subjects occur on the left of the preposed Vf, as in (10):

(10) a. Foroon we sceolan mid ealle mod & maegene to Gode gecyrran therefore we must with all mind and power to God turn 'therefore we must turn to God with all our mind and power Blickling.97

b. Be ooem we magon suioe swutule oncnawan ooet... by that we may very clearly perceive that... 'by that, we may perceive very clearly that...' CP.181.16

On the basis of this and a wider range of evidence, it was argued by van Kemenade (1987) that Old English personal pronouns are clitic elements that are procliticized to Vf in topic-initial constructions. Such procliticization is blocked when the first constituent position is occupied by an operator-like element as in the examples in (9). The essence of this analysis is that movement of Vf is always to the C-position in the structure (8), essentially because the phenomenon of topicalization + verb fronting is restricted to main clauses. There is some topicalization in embedded clauses in Old English, but it is shown by van Kemenade (1997) that this is restricted to a well-defined set of constructions, including passives, impersonals and the like.

An alternative analysis of V2 and pronouns in Old English has been presented by Pintzuk (1991, 1993). The essence of the proposal is that structures as in (8) are restricted to examples where the pronominal subject is postverbal, i.e. interrogative, negative-initial and pa-initial constructions. In Pintzuk's view, topic-initial constructions represent a lower sentential level IP, with the topic in Spec,IP and the pronouns as a clitic adjoined to the topic. (11) represents this with (10b) as an example:

(11)

[FORMULA NOT REPRODUCIBLE IN ASCII]

We will not discuss in detail here the issues concerning the position of pronouns; our focus here is on the nature of the topic position. Observe that Pintzuk's analysis, in which the topic position is Spec,IP, predicts that a topic is completely incompatible with a postverbal pronoun. However, in the rare instances where a negated finite verb is preceded by a topic, the subject pronoun follows the finite verb. (3) above is an Old English example of this, and here are some more (and note that the early Middle English examples (4-6) also illustrate this)

(12) a. ponne ne miht pu na poet mot ut ateon of ooes mannes eagan then not could you not the speck out draw of man's eye

AEEHomP.XLII.153

b. For oyssere twynunge nolde we hreppan his orowunge for this doubt not-would we touch his passion 'because of this doubt we would not touch his passion' (i.e. left the story of the passion of St. Thomas untranslated)

AEHomTh.i.520.16

The fact that a topic can precede the negated finite verb, which on both Pintzuk's and van Kemenade's analysis is in C, would seem to indicate that the topic position is Spec,CP. This fits with the fact that we find genuine topicalization in main clauses only. This is further corroborated by the arguments presented in van Kemenade (1997). There is what looks like embedded topicalization in Old English, as in (13), but this is restricted to what we may broadly call unaccusative contexts. (13) gives two examples of passives, but this also occurs with impersonals, modals, ergatives and a few other constructions (van Kemenade 1997).

(13) a. poet eallum folce sy gedemed beforan oe that all people (D sg) be (sg) judged before thee 'that all the people be judged before you' Paris Ps.9.18

b. ponne aelce daege beoo manega acennede purh hys mihte on worulde on worulde when each day are (pl) many (N, p1) given birth through his power on world 'when every day many are given birth through his power on earth' AEHomp.VI.120

There is evidence in the Germanic and the Romance languages and their histories, that such constructions have a number of special features, so that their word orders are not to be considered representative of the canonical word order. We are left with the observation that there is no real embedded topicalization in Old English. This would seem to favour the analysis in (8), which predicts that topicalization is restricted to main clauses. I conclude that this is the correct analysis. (3)

The analysis of the topic position as Spec, CP allows us to approach the puzzle concerning negative clauses .in its proper light. As observed in the introduction, negated main clauses are overwhelmingly introduced by the sequence negation element + finite verb, as illustrated once more in (14):

(14) a. Nolde se Hoelend for his bene swapeah hym fram gewitan

not-wanted the Lord for his prayer however him from depart

AEHomP.XIV.199

b. Ne mihte se deao him genealcecan, gif he sylf nolde

not could the death him approach, if he himself not-wanted

'Death could not have approached him, if he himself had not willed it'

AEHomTh.i.214.32

As observed, these sentences are analysed as involving V-movement to C. Note that the finite verb here is preceded only by the negative particle ne. There is lots of evidence in Old English that ne is procliticized to the finite verb. Ne is never separated from the finite verb by any material, and assimilated spellings like nolde in (14a) for ne + wolde occur with overwhelming frequency. The question then is whether sentence-initial ne as in (14) is a constituent. If it is, then these sentences have a negative topic. If it isn't, these sentences are V-first, and their very considerable frequency does not tally with the frequency of V-first sentences in general. Moreover, there is then the recurring puzzle: why is it that ne+Vf hardly ever occurs with a topic preceding it in Old English? And even more puzzling: why does this become a real possibility as from the earliest Middle English, at a time when, as far as we know (van Kemenade 1997), the workings of the Verb Second constraint were in essence the same as in Old English? The early Middle English construction was illustrated in the introduction (examples (4-2)), and some more examples are given here:

(15) a of his utgang ne cunne we iett noht seggon

of his exit not can we yet not say

'we can't yet say anything of its exit' PC.AD.1127.73

b. Pet ne seide lie noht

that not said he not KS.214.25

c. purh unweotennesse ne mei ha nawt sunegin

through ignorance not-can she not sin SW.255.33

d. pis ne habbe ic nauht ofearned

this not have I not earned VV.17.9

I think that account for these puzzles can be found in the general and ongoing weakening of ne as a negative element. We know that, following Jespersen's cycle, in the further course of Old English and Middle English, ne first came to appear on a large scale in conjunction with not. It was gradually weakened and lost. This is documented in a fair amount of detail in three articles by George Jack (1978a-c). Evidence that at an earlier stage, ne was in some sense a stronger element can be found if we go back further in history, and try to trace back the origins of neg-initial V2 sentences. This led me to look at negative-initial sentences in Beowulf. I devote a separate subsection to them.

2. Negative-initial sentences in Beowulf

The core observation I want to make with respect to negative-initial sentences m Beowulf is that we find, beside the Classical Old English pattern with negation+ finite verb at the front of the clause (illustrated in (16)), an alternative pattern illustrated in (17). Let us first consider these illustrations:

(16) Nolde eaorla hleo aenige pinga pone cwealmcuman cwicne forloetan not-wanted of earls protector any thing the kill-corner alive release 'The protector of earls was minded in no wise to release the deadly visitant alive'

Boewulf.791

(17) a.No he wiht fram me flodypum feor fleotan meahte, hrapor on holme;

no ic fram him wolde

Not he thing from me on waves far swim could, quicker in water;

not I from him wanted

Boewulf.541

b. no ic me an herew aesmun hnagran talige, gupgeweorca, ponne Grendel hine

not I myself in war-strength inferior count, battledeeds, than Grendel himself

'I do not count myself less in war-strength, in battle deeds, than Grendel does himself'

Beowulf.675

c. Naefre ic maran geseah eorla ofer eorpan oonne is eower sum, secg on searwum

never I greater saw man over earth than is yours' one, hero in harness 'I never saw in the world a greater earl than one of your band is, a hero in his harness'

Beowulf.247

The position I would like to defend here is that the coexistence of these two patterns, indeed, their near-complementary distribution, testifies to the rise of the verb movement dependency in negative-initial sentences. In other words, I propose a syntactic account for this alternation. This presupposes that Beowulf actually represents an older stage of the language than Classical Old English prose does: the pattern illustrated by (17) is not found in the prose and it is therefore tempting to view it as an older one, and the Classical Old English pattern as the grammaticalized version of the pattern as in (16), which we may hypothesize was an innovation at the stage when Beowulf was written up. This goes against the observation in Mitchell (1985: [section]3959) that the Old English verse texts do not contain word order patterns that are not attested in the prose. I have no quarrel with this as a general observation, but for the negative-initial pattern as in (17), this does not seem to me to be correct: (17) is not, as for as I can ascertain, attested in the prose. It is, I believe, uncontroversial that Beowulf represents the oldest extensive record we have of the language.

The idea that there is a syntactic account for the alternation illustrated by (16) and (17) is not a novel one: in a generalized version, the phrasing in Kuhn (1933) allows of such an interpretation, as Stockwell and Minkova (1994) point out; (4) and this is espoused explicitly in Hock (1986: 198-196), which in turn is adopted in modified form in van der Wurff (1990). While it would go beyond the scope of this contribution to go into the details of these proposals, it seems to me that the observations here pertaining to negative-initial sentences do not readily allow of generalization over all potential verb fronting patterns. Rather, I think that in a very real sense, the precise nature of the first constituent, in the case at hand a negative one, plays a crucial role in the process of 'attracting' the finite verb. The negative constituent, like the interrogative one, has a special force in this respect, as they have had throughout the history of English, and the details concerning the 'early' stages of Verb Second with other constitiuents simply remain to be sorted out; the aim of this contribution does not go beyond making some suggestions for the early historical development of negative-initial sentences.

The initial negative constituent as in (16-17) is, of course, a prime example of an unstressed particle in the sense of Kuhn (1933), and as such would be subject to the metrical account proposed there and by many others. (5) It seems to me that there are several arguments that speak against a metrical account in the specific case of the initial negative constituent. First, in the overwhelming majority of cases, sentential negation in Beowulf is expressed by a negative element in sentence-initial position, as illustrated by (17). This negative element typically occurs immediately following the halfline break; it is never preceded by any other elements (in the pattern (17), it should be emphasized). It seems to me to be reasonable to suppose that this is because it must have scope over the whole sentence, and this is a motivation that can have little to do with rhythmical considerations. There is quite a variety of such negative elements: no; na; noefre; n(e)alles; noenig(ne); nooer, and although they never al literate, it is rather difficult to read them all as being completely unstressed. (6) The second argument is that, in those cases where there is verb fronting (pattern illustrated by (16)), the finite verb invariably picks the negative element as its host. This is true for all types of finite verb. The pattern with verb fronting is most frequent with copulas, with other auxiliaries next in line. However, examples with lexical verbs are by no means infrequent; an example for each is quoted in (18):

(18) a. Noes hit lengra fyrst, ac ymb ane niht eft gefremede morobeala mare ond no mearn fore, foehoe ond fyrene; woes to foest on pam

Not-was it longer first, but after one night again performed murders more and not mourn for, violence and malice; was too bent on that

Beowulf.134

b. Ne meahte ic oet hilde mid Hruntinge wiht gewyrcan, peah poet woepen duge;

Not-could I in fight with Hrunting anything achieve, though the weapon was good

Beowulf.1659

c. Ne gemealt him se modsefa, ne his moeges laf

not failed him the spirit, nor his kinsman's sword

'his spirit did not fail him, nor did his father's sword'

Beowulf.2628

The fact that lexical verbs take part in fronting of the finite verb, speaks clearly against a metrical account: lexical verbs are not the typically weakly-stressed verbs that play a role in a metrical account; in fact Kuhn (1933: 100) suggests that the motivation for fronting of the lexical verb was analogy to the copula and auxiliary. It is, of course, hard to disprove such a statement; but it should be noted that the fronted finite verb gemealt in (18c) has primary stress, since it alliterates with modsefa and moeges. The fact that all types of verb pick the negative constituent as a host seems to be significant as well: in other types of Verb Second clause, the finite verb follows any pronouns in the sentence-initial string. This further reinforces the idea that it is the negative character of the first constituent which triggers fronting of the finite verb. In other words, it reinforces a syntactic account, at least for the case of negative-initial clauses. An interesting piece of circumstantial evidence for this is afforded by an example in Beowulf of a topic-initial negative clause:

(19) wide sprungon / hildeleoman. / Hreosigora ne gealp / goldwine Gearta; / guobill geswac, / nacod cet nide, / swa hyt no sceolde, / iren oergod. far leaped/ war-flames./ Of-famous victories not boasted / goldfriend of Geats; / war-sword failed, / naked at battle, / as it not should, / iron long-famous. /

'the goldfriend of the Geats boasted not of famous victories; the war-sword, naked in battle, failed, as it should not have done, the long-famous brand'

Beowulf.2583

This word order pattern looks like a standard Old English one; the topic Hreosigora has primary stress, as is evident from its alliteration with hildeleoman. This is followed by the negated finite verb and the nominative subject goldwine Geata; the finite verb gealp seems to be on the fourth stressed position of its line. It is hard to see a metrical motivation for the verb position here.

I conclude from the above discussion that it is the negative character of the first constituent that motivates attraction of the finite verb in negative-initial sentences. This points to a syntactic account for the rise of this particular sub-pattern. There is some further crosslinguistic evidence that it is the nature of the host constituent that motivates fronting of the finite verb. The above-discussed alternation between fronting and non- fronting of the finite verb in negative-initial clauses is likewise found in Gothic, as the following examples from Ferraresi (1991) illustrate:

(20) a.niu jus mais wulthrizans sijuth thaim Mark.6.26

not you much better are (than) that one

b. jah ni sijaith baitrai withra thos Colos. 3.19

and not be bitter against them

c. unte fist unmahteig guda ainhun waurde Lucas.1.37

because not-is impossible God no thing

Similar facts are reported for Classical Latin by Adams (1994a, b). Adams shows that negative non acts as what he calls a "second position" host, beside quantificational adjectives (omn-, magn-, max-, sum), demonstratives (huius, hanc, hoc), antithetical elements, bringing out clearly that it is the grammatical status of the host that determines the position of the verb. (7) Some examples from Adams:

Alternation Pred-non esse/non-esse-Pred

(21) a. redditurum non esse Caes.Gall.1.36.5

go-back (fut. pple.) not be

b. non esse uenturum Caes.Gall.1.42.4

not be come (fut. pple.)

As in Beowulf, the pattern (21b) is most frequent with copulas, with auxiliaries next in line.

3. Discussion

Let us go back to the puzzle with which this contribution set out: why are topics so rare in negative-initial constructions in Old English, and considerably more frequent in Middle English? On the basis of the discussion in sections 1 and 2, I want to suggest a historical scenario for the development of sentence negation in early English, which is fully in line with Jespersen's cycle, and yields a plausible angle on the grammaticalization process involved.

My suggestion for the earliest stage of the language, represented by Beowulf, is as follows. Sentence negation in Beowulf is regularly expressed by positioning a negative element in the first position of the clause, and on the basis of the discussion in section 1, I suggest that this position is Spec,CP. This constituent does not seem to have any level of stress that is metrically significant. However, its spelling (na, no, etc.) does suggest an unreduced form when it is not immediately followed by the finite verb (pattern as in (17)); the reduced form ne is found almost exclusively in those instances where it is immediately followed by the finite verb (pattern as in (16)). The alternation between the two patterns evinced in Beowulf bears witness to the rise of V-movement to C. The nature of the first constituent is of crucial importance, as discussed above. I suggest the following analysis for this alternation: the development from the non-V-fronting pattern to the V-fronting pattern represents a weakening of the negation element, and I suggest that the nature of the weakening represents a development from a stage where the negative element is an independent constituent, to a stage where the negative element becomes a part of the extended projection of the verb. This involves that the negative element acquires a V-feature (a verbal feature that attracts the finite verb to a certain position). Thus, in the non-V-fronting pattern, the initial negative constituent is an independent constituent, whose function in Spec,CP is to take negative scope over the whole clause. In the V-fronting pattern, the negation element has weakened (this is visible in the spelling as reduced ne); the effect of this is. that CP now has a verbal feature, which must attract the finite verb to the C-position. We thus get the V-fronting pattern as in (8) in Beowulf which is also the standard Classical Old English pattern. Let us now turn to the question why this pattern does not readily tolerate a topic in Old English. My suggestion for this is that, while in this pattern initial ne is weakened phonologically, and almost certainly procliticized to the finite verb, this weakening is essentially phonological. That is, in the syntax, ne is still a Spec,CP element, Thus, Spec,CP is filled in negative-initial sentences, and no other constituent can occur there. Let us now turn to Early Middle English, a stage where topics are found in Spec,CP, even though there is a ne + finite verb sequence. I suggest that by this time, ne is weakened even further in the sense that it is now syntactically weak as well. Initial ne has lost its negative force, which is also evident from the fact that it is, on a large scale, reinforced by a second sentential negator, as documented by Jack (1978a-c). (8) My hypothesis is that it no longer is a Spec,CP element, but rather a phonological relic, soon to disappear altogether. Therefore, in the Middle English equivalents of the Old English negative-initial sentences, Spec,CP is available for another constituent, such as a topic.

Let us summarize in schematic form the scenario outlined in this section:

(22)  Beowuif
      The non-V-fronting pattern
      [Spec,CP no [.........finite verb.........]]

(23) Beowulf and Classical Old English
     The V-fronting pattern
     [spec,cP ne [C finite verb [.........]]]
     citicization

(24) Early Middle English
     the topic-ne-finite verb pattern
     [Spec,CP topic [c ne + finite verb [.........]]]

This scenario gives some formal substance to the steps by which ne evolves historically from an unreduced initial constituent (22) via a syntactically meaningful proclitic (23), and a syntactically meaningless proclitic (24) to ultimate disappearance. Its syntactic effect is visible in the details of negative-initial clauses and topic-initial negative clauses. If correct, it also has interesting implications for the grammaticalization process at work in the development of negation elements, in the sense that the grammaticalization takes place via discrete steps that allow of a formulation in structural terms.

(1.) I thank Donka Minkova and Robert Stockwell for their help. When I first presented this material at the XIIth ICHL conference at Manchester, August 1995, I called initial no in Beowulf a topic. They shrieked at the metrical implications of this, and I followed them up. This has helped. They are not to be held responsible for the ensuing story.

(2.) This evidence is discussed in detail in van Kemenade (1997, forthcoming).

(3.) A further refinement of this analysis is proposed in van Kemenade (forthcoming) and Hulk and van Kemenade (forthcoming). In this analysis, the topic position, on the same evidence as quoted in the main text, is Spec,CP. Evidence from multiple sentential negation shows, however, that the position of the finite verb in topic-initial constructions is to a position lower than C, although higher than the position I proposed by Pintzuk.

(4.) It should be emphasized, however, that this is not Kuhn's aim: he is, of course, concerned with establishing that accentual/metrical relations have played a very important role. It seems to me, however, that metre cannot overrule the limitations of the syntax.

(5.) For a critical discussion of the issues involved, see Stockwell and Minkova (1994).

(6.) Although it must be admitted that this is a controversial point, some sorting out of the metrical evidence is still to be done here.

(7.) The term 'second position' is slightly confusing, since the aim of Adams' articles is to show that what is at stake is not 'second position', but rather attachment to certain types of host.

(8.) There is same evidence of multiple sentential negation in Old English, as discussed in van Kemenade (forthcoming); Hulk and van Kemenade (forthcoming). But this is a minor pattern.

REFERENCES

TEXTUAL SOURCES (LISTED BY ABBREVIATION)

AEHomP = Homilies of AElfric

1967 John C. Pope (ed.) (EETS 259.) Oxford: OUP. (Quoted by homily and line number.)

AEHomTh = The Homilies of the Anglo-Saxon Church

1846 Benjamin Thorpe (ed.). London: The Aelfric Society. (Quoted by volume, page and line.)

AESL = AEifric's Lives of Saints

1890/1900 Walter Skeat (ed.). (EETS. O.S. 94, 114.) (Quoted by Life and line number.)

1922 Friedrich Klaeber (ed.). D.C. Heath and Company.

Beowulf = Beowulf and the Fight at Finnsburg

Blickling = The Blickling Homilies 

1967 Richard Morris (ed.). (EETS 1967.) Oxford: OUP. (Quoted by page number.)

CP = King Alfred's West-Saxon Version of Gregory's Pastoral Care

1885 Henry Sweet (ed.). (EETS 45, 50.) London. (Quoted by page and line number.) St. Katherine = The Life of St. Katherine

1978 Eugen Einenkel (ed.). (EETS 1978.) New York: Kraus reprint Co. (Quoted by line number.)

KS = An Old English Miscellany

1872 Richard Morris (ed.), 26-36. (EETS O.S. 49.) London. (Quoted by page and line number.)

PC = The Peterborough chronicle 

1970 C. Clark (ed.). Oxford: Clarendon Press. (Quoted by AD and line number.)

1868 Richard Morris (ed.), 245-269. (EETS 1868.) (Quoted by page and line number.) [1969]

SW = Old English Homilies and Homiletic Treaties

VV = Vices and Virtues, Part I

1888 Ferdinand Holthausen (ed.). (EETS 89.) (Quoted by page and line number.)

OTHER WORKS

Adams, Jim

1994a "Wackernagel's Law and the position of unstressed pronouns in Classical Latin", Transactions of the Philological Society 92: 103-178.

1994b "Wackernagel's Law and the placement of the copula esse in Classical Latin", Proceedings of the Cambridge Philological Society.

Ferraresi, Gisella

1991 Die Stellung des Gothischen Verbs im Licht eines Vergleichs mit dem Althochdeutschen. [Unpublished thesis, University of Venice.]

Hirschbuhler, Paul - France Martineau - Maria Luisa Rivero (eds.) forthcoming Negation: Syntax and semantics. Amsterdam: John Benjamins. Hock, Hans Heinrich

1986 Principles of historical linguistics. Berlin: Mouton de Gruyter.

Hulk, Aafke - Ans van Kemenade forthcoming "Negation as a reflex of clause structure", in: Paul Hirschbuhler et al. (eds.). Jack, George

1978a "Negation in later Middle English prose", Archivum Linguisticum 9: 58-72.

1978b "Negative adverbs in early Middle English", English Studies 59: 295-309.

1978c "Negative concord in early Middle English", Studia Neophilologica 50: 29-39.

Jespersen, Otto 

1917 Negation in English and other languages. Copenhagen.

Kemenade, Ans van

1987 Syntactic case and morphological case in the history of English. Dordrecht: Foris.

1997 "V2 and embedded topicalization in Old and Middle English", in: Ans van Kemenade -- Nigel Vincent (eds.).

forthcoming "Sentential negation and clause structure in Old English", in: Ingrid Tieken-Boon van Ostade et al. (eds.).

Kemenade, Ans van -- Nigel Vincent (eds.)

1997 Parameters of morphosyntactic change. Cambridge: CUP.

Kuhn, Hans

1933 "Zur Wortstellung und -betonung im Altgermanischen", Beitrage zur Geschichte der deutschen Sprache und Literatur 57: 1-109.

Mitchell, Bruce

1985 Old English syntax. Oxford: Clarendon Press.

Pintzuk, Susan

1991 Phrase structures in competition: Variation and change in Old English word order. [Unpublished Ph.D. dissertation, University of Pennsylvania.]

1993 "Verb seconding in Old English: Verb movement to Infl", The Linguistic Review 10: 5-35.

Tieken-Boon van Ostade, Ingrid -- Gunnel Tottie -- Wim van der Wurff (eds.) forthcoming Negation in the history of English. Berlin, New York: Mouton de Gruyter. Stockwell, Robert P. -- Donka Minkova

1994 "Kuhn's laws and the rise of verb-second syntax", in: Toril Swan et al. (eds.), 213-233.

Swan, Toril - Endre Morck - Olaf Jansen Westvik (eds.)

1994 Language change and language structure. Berlin, New York: Mouton de Gruyter.

Wurff, Wim van der

1990 Diffusion and reanalysis in syntax. [Unpublished Ph.D. dissertation, University of Amsterdam.]