David Denison’s keynote at Santiago-Leuven-Edinburgh seminar

Posted on April 4, 2017 by

Professor Emeritus David Denison is giving a keynote lecture at the Santiago-Leuven-Edinbrugh seminar on Grammatical Change and Variation. This two-day event – taking place today and tomorrow (April 4-5) at the KU Leuven – is intended specifically for PhD students, but added some impressive keynote speakers to its programme: Svenja Kranich, Benedikt Szmrecsanyi, and David Denison.

David’s talk deals with counterfactuals. Here’s the abstract:


David Denison

University of Manchester

If- and inversion protases are a promising candidate for diachronic variational study: (1) If I had known there was a charge for each, I would have … (CBC 14530)

(2) Had I known this man would come back to Lebanon, I would have … (CBC 14756)

(Examples from BNC.) There is a clear and almost monotonic decline of inversion protases over the last four centuries , both in overall frequency and in the number of verb forms able to invert. A modest count in the ARCHER corpus was made in Denison (1998: 300, Table 3.11), slightly augmented in a recent presentation. Do larger corpora add usefully to the picture, and what explanations can be advanced for the long-term trends?

In this talk I focus more on the kinds of question that can be asked and – perhaps – answered in a syntactic corpus study. What, for example, is the appropriate domain to investigate? It should obviously include the two variants illustrated above of ‘past time remote conditionals’ (Huddleston & Pullum 2002: 748ff.) or ‘modal pluperfects’ (Visser 1963-73: III-2 §§2030ff.)). A non-standard extra morpheme derived historically from have is probably confined to the if type:

(3) If you’d’ve just switched it on when I come in it’d’ve been totally different. (KCX 4646) (4) I mean if they’d of banned it, there’d of been a lot more people out of work (KD1 1019)

Is it legitimate to count (3) and (4) in with (1)?

The fact that open conditionals are overwhelmingly of the if type is probably relevant, but can we safely assume the same distinction between open and remote conditionals throughout history? Then there are conditionals with protases introduced not by if but by assuming, provided, in the event (that), suppose/supposing, etc.:

(5) Suppose Delia had become friendly with Angy and gone to her flat that afternoon She’d have been one of the last people to see her alive. (HNJ 1573)

Lastly, there are conditional-like constructions that don’t involve hypotaxis at all:

(6) ‘Call me a fucking thief one more time and I’ll smash your fucking face in!’ (A0F 833) (7) A millisecond before and he would have been trapped. (A6W 595)

Is there really a single variable in (1)-(7)? Judging the (partial) synonymy of potential syntactic variants is a familiar but vexed question.

Modelling the variation would be harder than a neat pairwise comparison, even discounting the difficulty of a systematic search strategy. Furthermore, register and text-type must be taken into account, and spoken examples are prone to mistranscription.

Such methodological issues in research on English counterfactual protases will, I hope, prove of wider relevance.

Denison, David. 1998. Syntax. In Suzanne Romaine (ed.), The Cambridge history of the English language, vol. 4, 1776-1997, 92-329. Cambridge: Cambridge University Press.

Huddleston, Rodney & Geoffrey K. Pullum. 2002. The Cambridge grammar of the English language. Cambridge: Cambridge University Press.

Visser, F. Th. 1963-73. An historical syntax of the English language, 4 vols. Leiden: E. J. Brill.


Here are some additional links if you want to have a closer look at the abstracts or the program.


photo credits: https://www.kuleuven.be/english/news/2015/times-higher-education-world-university-rankings/image

Posted in: Uncategorized