No question: lexicalization and grammaticalization processes in the development of modal qualifier meanings

Lexicalization and grammaticalization: the development of idioms and grammaticalized expressions with no question¹


1. Introduction

In Present-day English, the string no question is used mainly to modally qualify a proposition, i.e. with grammatical meaning, either as adverbial (1) or as part of a clause (2)-(3). As pointed out by Kjellmer (1998), there are two modal clausal structures, expressing opposite polarity values. There’s no question followed by a finite clause is always positive, as in (2), which can be paraphrased as ‘he definitely must stay’. There’s no question of + gerund conveys negative polarity, as in (3), which means ‘Her Majesty will NOT turn up unannounced on your doorstep’.

(1) You have to be mentally strong but too many of our players jacked it in. There will be changes on Saturday, no question.” (WB)²

(2) He's better than our two former managers ... The best clubs have continuity at the top. There is no question he must stay. (WB)

(3) There’s no question of Her Majesty turning up on your doorstep unannounced. (WB)

Less commonly, no question is also used lexically, as in the semi-fixed idioms illustrated in (4) and (5). We identified two idioms with the form there be no/never any/etc. question. In (4), the expression means ‘not be challenged’ (OED VIII: 48), while in (5) it means ‘not be an issue’ (with question used in the sense of ‘subject of discussion’, OED VIII: 48).

(4) ... for Tolkien, Rickettsia quintana proved a life-saver. The army was notoriously suspicious about any attempt to `cry off sick', but there was no question about Tolkien 's condition. (WB)

(5) We also wanted to know if he wanted anything out of it, and because there was no question of payment, ..., we felt reassured that he really was doing it for us. (WB)

The synchronic presence of idiomatic and grammatical layers suggests that processes of both lexicalization and grammaticalization took place. This article sets out to reconstruct these processes on the basis of diachronic and synchronic data described in Section 2.

Lehmann (2002) has proposed that lexicalization often precedes the grammaticalization of periphrastic expressions. As we will show, the development of the clausal structures with (no) question is a case in point. We will reconstruct the formation of the two idioms, ‘be (un)challengeable’ and ‘be (not) at issue’, and we will examine how they relate diachronically to the two grammaticalized modal patterns, the positive and the negative one (Section 3). This requires us to spell out the recognition criteria of lexicalized versus grammaticalized uses in the concrete analysis of our data (Sections 3.1, 3.2, 3.3, 3.4). In these sections, we will focus mainly on criterial syntagmatic features, taking a functional-structural approach as developed in Langacker (1999), Halliday (1994), Hopper & Traugott (2003) and Boye & Harder (2007). The history of the clausal patterns with (no) question also forms the occasion for us to contribute a more personal position to the current theoretical debate about the difference between lexicalization and grammaticalization. Taking our inspiration from the (neo-)Firthian tradition (Firth 1951/1957), we will propose that lexicalization and grammaticalization lead to different types of paradigmatic organization. A lexicalized item imposes lexicosemantically motivated collocational and colligational relations, while grammaticalizing elements come to express meaning options, with their typical interdependencies, from grammatical systems (Section 3.5).
We will also study the origin and development of modal adverbial no question. Simon-Vandenbergen’s (2007: 32) study of there is no doubt and no doubt found that the chronology in which these expressions emerged is compatible with the hypothesis that the adverbial developed by ellipsis from the clausal structures. As ellipsis could also be considered the possible origin of no question, we will investigate whether this adverbial resulted via ellipsis from the clausal structures (Section 4). Conclusions will be formulated in Section 5.


2. Data

The noun question occurs from c.1300 in English, which made it necessary to collect diachronic data from Middle English on. We used the Penn-Helsinki Parsed Corpus of Middle English (PPCME), the Penn-Helsinki Parsed Corpus of Early Modern English (PPCEME), and the Corpus of Late Modern English Texts (CLMETEV). As we wanted to trace how the strong association of question with negative quantifier came about, we included all nominal hits, manifesting all the early variation in the expressions with question. In total, this diachronic dataset contained 5,289 tokens.

The synchronic dataset was compiled from WordbanksOnline corpus. For reasons of comparability with the diachronic data, we extracted data from written British English sources only. As our focus in Present-day English was on the uses of the highly entrenched expressions with no question, we took a random sample of 250 hits, obtained by the search string no question.