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

6. Conclusion

We set out in this diachronic-synchronic study of (there’s) no question with two descriptive research questions. Regarding the clausal structures, mainly existential ones, our aim was to reveal the paths of change that led to their current idiomatic and grammaticalized uses. For the adverbial, we wanted to find out if it developed via ellipsis from the clausal structures.

Diachronic study of the adverbial forms with question revealed that the adverb emerges later than the clauses. However, the absence of clauses with negation invalidates the hypothesis that the adverb is the result of ellipsis of the matrix clause. The first occurrence of the phrase no question is in fact in the adverbial form in Early Modern English. We suggested that no question, without question and out of question probably emerged by analogy with adverbials like no doubt, without doubt and out of doubt, as well as no way and no wonder, which were already entrenched in Early Modern English.
The first attestations of question with complement clauses in Middle English and Early Modern English were in composite predicates such as make question and lexicalized clauses such as there be question, which first meant ‘ask’ and then ‘challenge’. In addition, there be question also acquired the meaning ‘be at issue’. In Late Modern English, these two idiomatic patterns specialized for negative contexts with the senses ‘be unchallengeable’ and ‘not be at issue’. It was these idiomatic patterns that grammaticalized, resulting in positive modal content clauses and negative modal markers respectively. The existentials with no question thus clearly show how lexicalization can be a crucial step towards grammaticalization, not only structurally, but also semantically and pragmatically.
This diachronic reconstruction also showed the importance of distinguishing lexicalized from grammaticalized uses, a theoretical issue high on the agenda in current grammaticalization studies. In this debate it is essential, we argued, to spell out differences both in syntagmatic and paradigmatic patterning. On the syntagmatic axis, we followed Boye & Harder’s (2007) recognition criteria for distinguishing lexica(lized) from grammaticalized uses on the basis of their having primary or secondary status in discourse usage. On the paradigmatic axis, we proposed that a lexicalized item imposes lexicosemantically motivated collocational and colligational relations, while grammaticalizing elements come to express more and more meaning options, with their typical interdependencies, from grammatical systems. The grammaticalized no question clauses are a good example of Halliday’s claim that the dissociation of associated variables and the recombination of the independent variables is an important ‘semogenic’ process, i.e. a process of change creating coding possibilities for an enriched semantic resource (Halliday 1992: 27).