The occurrence of not-contraction (e.g. she isn’t) in three genres in CONCE is examined in detail. In an overall quantitative analysis, not-contraction is compared with uncontracted forms as well as operator contraction (e.g. she’s not). Other potential factors accounted for include the operator itself, gender, word order (e.g. is she not vs. is not she), and no-negation as an alternative negating strategy. A multifactorial, variationist analysis of contexts where not-contracted and uncontracted forms are the main variants demonstrates the importance of factors such as genre and tense. The results provide solid evidence of colloquialization in Drama and Fiction, where not-contraction becomes more frequent; they also reveal that women may have used contraction less than men, which is expected considering the stigmatization of not-contraction, and that the uncontracted is not she pattern in questions may have been used as a more acceptable way of rendering spoken contractions in writing.

Participle clauses as postmodifiers in noun phrases (e.g. the window broken by the thief) have been discussed as a potential indication of densification, as they are shorter than their most obvious alternative, that is, finite relative clauses. A careful examination reveals that only a subset of tokens of participle and relative clauses can be considered equivalent and exchangeable, and to reflect this finding both variationist and text-linguistic analyses are carried out. There is some evidence of densification in newspaper language and, to some extent, scientific texts; history texts, in contrast, develop in the opposite direction, underscoring the importance of considering subgenres of academic writing. Among the newspapers considered, the Poor Man’s Guardian, which was aimed at working-class readers, shows no tendencies towards densification; this may be due to journalists’ perceptions about the paper’s readership. The issue of whether variationist or text-linguistic approaches are more suitable is discussed. Non-restrictive participle clauses are shown not to indicate densification; instead, they function as a characterizing or backgrounding device in narrative texts.

In this chapter, the idiolect-centred approach to language change is reconciled with historical corpus linguistics as a methodology. The issue of what corpus-based studies can and cannot tell us is discussed in terms of the granularity of the analyses. The two types of change in focus in the case studies, colloquialization and densification, are introduced. The two main ways of operationalizing frequency in syntactic studies, namely variationist and text-linguistic approaches, are contrasted. Finally, the two corpora on which the case studies are based, CONCE and CNNE, are introduced; CNNE and its relationship to the newspaper market in nineteenth-century England is discussed in some detail, as CNNE is a new corpus and as the newspaper trade underwent far-reaching changes during the period covered by the corpus. The importance of considering a wide variety of genres owing to the increasing linguistic genre differences that characterize Late Modern English is emphasized.