Accelerated Grammaticalization in The Peterborough Chronicle
Elly van Gelderen
Arizona State Univerity
September 2006
Introduction
It has often been said that the Peterborough Chronicle is the beginning of Middle English (e.g. Clark 1970: lii). In this paper, I examine aspects of this change by looking at two kinds of grammaticalization, namely prepositions becoming complementizers (i.e. for, till and at) and adverbs being reanalyzed as aspectual markers. In the last part of Peterborough Chronicle, there is a tremendous increase of complementizers (and sentence embeddings). There is also a fast loss of verbal ge- and an increase of other aspectual markers. The grammaticalization of for accelerates between 1135 and 1154, and ge- virtually disappears after 1130. This symbolizes the true start of ME.
The aspectual changes around ge- and the introduction of the complementizers for and till both involve grammaticalizations resulting in `higher' position in the tree: P to C, inner to outer aspect, verbs to auxiliaries. I will argue that an Economy Principle `Late Merge', as in van Gelderen (2004), accounts for both developments.
The outline is as follows. In section one, I provide some background on the general dialect of the Peterborough in comparison with another version of the Chronicle, namely the Parker Chronicle. In section two, the increase in complementizers are discussed, and in section three the loss of aspectual ge- is as well as strengthening by adverbs. Section four is a conclusion.
1 The Peterborough's Linguistic Character
In this section, I will briefly compare the Peterborough Chronicle, which originates in the Danelaw area, with the more southern (West Saxon) Parker. As we'll see, established dialect differences do not neatly differentiate these two versions. This is relevant to the accelerated grammaticalization that is my focus since grammaticalization could be due to language contact (see Heine & Kuteva 2005). I'll first look at the phonology and then the syntax (see Clark 1970 and Bately 1983 for more) and show that there is not much evidence for a radically different language, even though the final continuation has a few northern characteristics.
1.1 Phonological differences
Phonologically speaking, in West Saxon (hence WS), a low front [æ] becomes low back before another back vowel. Thus, the singular dæg `day' alternates with the plural dagas `days' (and the dative dagum). In Mercian, these words are deg and dægas, and can be shown to have changed from the WS forms (see Dresher 1980). Comparing the Parker and the Peterborough shows that there is no difference: both have dagum and dagas, i.e. the WS forms. There is no increase in northern forms in the continuations either.
Kuhn (1939) suggests that this fronting to deg/dæg doesn't occur when the liquid l plus a consonant follows the a (Dresher 1980: 53). In West Saxon, æ is broken to ea before [l+C], but not in Mercian (eall, healdan vs all, haldan). This should predict that the Parker (being WS) would have more breaking, and that is not the case for hand I (entries up to 891) but is true for later scribes. In the Peterborough Chronicle, the unbroken form occurs at the beginning (which is expected) but also in the end, as in ald mone.
Another characteristic of West Saxon is that ie is used for Mercian e (e.g. gieldan vs geldan `pay') but no forms of g(i)eldan occur in the Parker or the Peterborough. The Parker has three forms of feld, again showing it isn't clearly southern, and the Peterborough has five. Neither has WS field.
Another dialect characteristic is that a should correspond to o in northern texts in pairs such as hand/hond, land/lond, and man/mon. This is again not borne out in the two versions: in the supposedly West Saxon Parker, mon is more frequent than man(n) (namely 31 and 15 occurrences respectively) whereas man(n) is more frequent than mon in the Peterborough (namely 278 and 2 occurrences respectively). The same is true with hand/hond: 16 instances of hand and 0 of hond are found in the Peterborough; in the Parker, there are 57 of hand and 6 of hond. The forms of land and lond are almost as frequent in the Parker (42 and 35 instances respectively), but land is the only form in the Peterborough (namely 272 and none of lond). This criterion shows the Peterborough to be a more southern text than the Parker. The different forms are spread more or less equally over the different time periods of the Peterborough, so the difference is not due to the scribe.
A last phonological difference that I'll mention is that between micel and mycel, swiþe and swyþe, and cirice and cyrice. Campbell (1959: 132-3) says that by lWS, an i around labials or before an r becomes y. Again unexpectedly, the entire Peterborough patterns more with the south than the north in that it has more mycel than micel (namely 87 and 45 occurrences respectively for the uninflected form) and the Parker has only micel. However, if one looks at the distribution of these forms throughout the manuscript, the expected forms (micel and circ(e)) occur in the last part of the Peterborough.
1.2 Morphological and syntactic differences
According to Clark (1970: xlvii) the morphology in the Peterborough Chronicle "shows the East-Midland basis ... even more clearly than phonology does". For instance, there are some plural present indicative forms in -en. She provides some other morphological evidence. Dekeyser (1986) examines relative markers in the Peterborough and shows that the demonstrative option is lost, that þe is receding and þat and the wh-relatives are increasing, especially in the final continuation.
As far as the syntax is concerned, Kroch & Taylor (1997) argue that Verb-second is strong in the north and that clitic fronting and V-3 occur in the south. If correct, there should be a difference between the Parker and the Peterborough, and there is some, as in (1), where the Parker has V-3 and the Peterborough V-2:
1. a. Her on þysum geare for se micla here
Here in this year went the big army ... (Parker, anno 893).
b. Her for se myccla here
Then went the big army ... (Peterborough, anno 893).
Many other constructions exhibit no difference, however, as in (2), and even pronouns are preposed in the Peterborough, as in (3):
2. a. Her todælde se foresprecena here on tu (Parker, 884)
b. Her todælde se forsprecena here on twa (Peterborough, 884)
`In this year the above-mentioned army split into two'.
3. a. Her hiene bestæl se here into Werham (Parker, 876)
b. Her hine bestæl se here into Wærham
`In this year, the army stole away to Warham' (Peterborough, 876).
Sentence (4) is actually the opposite from what we would expect, with the southern text having V-2 and the supposedly more northern one having more V-3:
4. a. Her gefeaht Ecgbryht cyning ... (Parker, 833).
b. Her Ecgbriht cining ge feaht ... (Peterborough, 833)
`In this year Ecgbriht fought ...'.
In this section, I have provided some general, mainly dialectal, characteristics of the Parker and Peterborough Chronicles. Even though the latter is geographically located in a more northern area (in the East Midlands Danelaw area), the typical characteristics don't often bear that out. I will nevertheless show that some northernisms occur in the last part of the Peterborough, e.g. the use of till and the demise of ge-. These are the result of grammaticalization and such change is compatible with a higher degree of language contact as well. I will add to this the picture the sudden increase in for complementizers and aspectual markers in the Peterborough Chronicle.
2 Complementizers in the Peterborough Chronicle
I'll turn to the grammaticalization of the complementizers for and til in the Peterborough Chronicle (from now on PC). As is well-known, grammaticalization involves the loss of semantic and phonological features and an increase in grammatical function. Well-known examples involve verbs being reanalyzed as auxiliaries and prepositions as aspectual markers and complementizers. I'll first discuss the data and then discuss Late Merge, a structural reason behind grammaticalization.
2.1 Prepositions in change
In the PC, for is used as a preposition of causation, as in (5) and (6) <1>:
5. þa luuede se kining hit swiðe for his broðer luuen Peada. 7 for his wedbroðeres luuen Oswi. 7 and for Saxulfes luuen þes abbodes
`Then loved the king it much for love of his brother Peada and for his pledge-brother Oswiu and for love of the abbot Saxulf' (anno 656).
6. ouþer for untrumnisse ouþer for lauerdes neode ouþer for haueleste ouþer for hwilces cinnes oþer neod he ne muge þær cumon
`either from infirmity or from his lord's need or from lack of means or from need of any other kind he cannot go there' (anno 675).
Forðæm also functions as `because', as in (7), but there seem to be no instances of a locational for (which does occur in OE):
7. forþam Trumbriht wæs adon of þam biscopdome
`because T had been deprived of his biscopric' (anno 685).
The earliest instance of for as a finite complementizer in English is in the PC, if the OED is correct, and is from the entry for 1135, as in (8). There are two others from the entry for 1135, as in (9) and (10):
8. for þæt ilc gær warth þe king ded
because (in) that same year was the king dead (PC, 1135, 6)
9. for æuric man sone ræuede oþer þe mihte
because every man soon robbed another that could
`becasue everyone that could robbed someone else' (PC, 1135, 8).
10. for agenes him risen sona þa rice men
`because against him soon rose the powerful men' (PC, 1135, 18).
This locates the first use of complementizer for with Scribe II, i.e. during the Final Continuation. Between 1135 and 1154, the use increases dramatically compared to the period before 1135, as (11) to (23) show. (The years 1136, 1139, and 1141 to 1154 do not have entries):
11. for he hadded get his tresor
because he had got his treasure (PC, 1137, 3).
12. for æuric rice man his castles makede
`because every powerful man made his castles' (PC, 1137, 13-4).
13. for ne uuæren næure nan martyrs swa pined alse hi wæron
`because never were martyrs as tortured as they were' (PC, 1137, 20).
14. for nan ne wæs o þe land
`because none was in that land' (PC, 1137, 42).
15. for ouer siþon ne forbaren hi nouther circe ne ...
`because nowhere did they forbear a church nor ...' (PC, 1137, 46).
16. for hi uueron al forcursæd
`because they were all accursed' (PC, 1137, 53).
17. for þe land was al fordon mid suilce dædes
`because the land was all fordone bysuch deeds' (PC, 1137, 54-5).
18. for he wart it war
because he was of it aware (PC, 1140, 3).
19. for his men him suyken 7 flugæn
`because his men deserted him and fled' (PC, 1140, 17).
20. For þe king him sithen nam in Hamtun
`because the king took him afterwards at Hampton' (PC, 1140, 38).
21. for he besæt heom til hi aiauen up here castles
`because he besieged them till they gave up their castles' (PC, 1140, 52).
22. for he was an yuel man
`because he was an evil man' (PC, 1140, 56).
23. for he dide god iustise 7 makede pais
`because he did good justice and made peace' (PC, 1140, 77).
There may be an earlier finite complementizer in Chronicle F, as in (24), but I have not looked into F in detail:
24. far hi ne brahtan nan ge writ
`because they brought no letter' (Plummer 1889, p. 82, anno 995).
Excluding the verb for `went', there is a total of 101 prepositional and complementizer occurrences of for in the PC. Of these, 16 are finite complementizers recorded during the last few years, as in (8) to (23), and one is possibly infinitival, namely (25). I'll give more of the context here since it would be the earliest in English, and not listed in the OED or work on for such as van Gelderen (1998):
25. Hit ofþuhte naþema ealle Frencisc 7 Englisc oc se kyng hit dide for to hauene sibbe of se eorl of Angeow
`However, all the French and English thought ill [of the wedding] but the kind did it in order to have peace with the count of Anjou' (PC 1127, 10-2).
Apart from for, there are other prepositions starting to function as complementizers. There are two instances of til, as in (26) and (27). According to the OED, the use in the PC is the earliest use. Again, it is in the recordings of Scribe II. Interesting is that till occurs only as a complementizer, not as a preposition. This may mean that use was borrowed rather than grammaticalized out of an existing preposition:
26. til hi iafen up here castles
`till they gave up their castles' (PC 1137).
27. til hi aiauen up here castles (PC 1140, 52)
Mustanoja (1960: 408-9) states that the use of till is "typically northern". There is no til in the Parker, but then the Parker doesn't go past 1070.
The first instance of at as non-finite complementizer in the OED is from 1280. If (28) were ME, one could argue at was an infinitival C. In the PC, seon `see' is never weakened this much and therefore (28) has a different analysis, namely with at expressing the dative, and se a D:
28. He geornde at se kyning þet he scolde for his luuen freon his ane mynstre Wocingas het
`He desired of the king that he should for his love free this one monastery called Woking' (PC 777, Thorpe p. 92).
In this section, I have shown that the use of complementizer for blossoms in the Final Continuation, in particular after 1135. This is a sign of accelerated grammaticalization, typical for ME. The use of till (though not at) fits with this but is much less frequent; it is also a sign of northern influence. I will now discuss Late Merge and show how this Economy Principle is relevant to these data.