Regional variation in late Middle English negative concord

Non-literary Late Middle English (LME) showed substantial variation in the syntax of negated quantifiers, whether in Subject or Object NPs (Ingham 2002). These, unlike ordinary NPs, could appear either in the canonical positions of PDE, or in displaced positions, e.g.:-

(1) There schulde no man sey nay to it Paston, D 647,45
(2) I may non leysour haue Paston D 182,48

Analyses presented in the present paper show some variation between Negative Concord (NC) and PDE any-series quantifiers within negated contexts, e.g.:-

(3) Mi hert, whether it lufe my god or noght, wate no man bot god, for noght þat þai may se me do.
Rolle, Form, Ms Cambs Dd V. 64 Horstmann ed. 39,5
(4) þat nane suld reprehend þam in any thing þat þai do
Rolle, Form, Ms Cambs Dd V. 64 Horstmann ed. 5,33

It is known that present-day Scandinavian languages show tendencies towards negative argument displacement and also have lost NC. We sought to understand whether these two types of variation in LME were related to contact influence with Scandinavian.

Results from the examination of private correspondence and religious prose data show that variation in LME negated quantifier displacement seems not to have been strongly associated with geographical area, though there is a perceptibly greater incidence in East Midlands than in more Western 15th century sources. In any case it is doubtful that the variety of Scandinavian spoken settlers in Eastern England had distinct syntactic possibilities for negated quantifiers. OV with ordinary objects continued to be productive in older forms of Icelandic (Hróarsdóttir 2000) until recent times. The distinctive syntax of negated argument displacement in LME is thus hard to identify with contact influence.

Variation between NC and any-series forms in negated contexts showed a clearer geographical association. Prose writers of Northern origin, particularly Rolle and Wyclif, made substantial use of any-series items where Southern writers, e.g. Trevisa and Chaucer, had virtually exclusive NC. Verse texts from the same period present a similar picture, with northern verse being less strict in following NC than southern/midland.

Crosslinguistically, Rowlett (1998) links NC to the existence of a negative head element. As shown by Jack (1978), the ME negative head particle ne was retained longer in the South. Old Norse between the supposed dates of composition of the Poetic Edda (c. 900-1050), and the Prose Edda (13th-14th centuries), showed a decline of the Negative head -a(t) in favour of the adverbial negator eigi ('not') (Eythórsson 2002). It is shown that the vast majority of negated quantifiers in the Poetic Edda are already unsupported by a negative head. It is concluded that contact with Scandinavian language varieties featuring negative quantifiers unsupported by a negative head could have facilitated the demise of NC in Northern varieties of English.