English relative clauses in a cross-Germanic perspective

Abstract

The article talk examines the distribution of relativising strategies in English in a cross-Germanic perspective, arguing that English is quite unique among Germanic languages both regarding the number of available options and their distribution. The differences from other Germanic languages (both West Germanic and Scandinavian) are primarily due to the historical changes affecting the case and gender system in English more generally. The loss of case and gender on the original singular neuter relative pronoun facilitated its reanalysis as a complementiser. The effect of the case system can also be observed in properties that are not evidently related to case. Specifically, choice between the pronoun strategy and the complementiser strategy is known to show differences according to the Noun Phrase Accessibility Hierarchy. While English shows a subject vs. oblique distinction in this respect, matching its nominative/oblique case system, German dialects show a subject/direct object vs. oblique distinction, matching the nominative/accusative/oblique case setting in the language. The particular setting in English is thus not dependent on e.g. a single parameter but on various factors that are otherwise present in other Germanic languages as well, and it is ultimately the complex interplay of these factors that results in the particular setup.

Subject‐Object Asymmetries and the Development of Relative Clauses between Late Middle English and Early Modern English

“…The results thus show that the emergence of who , just like the spread of the wh ‐strategy, is related to a subject/object or a subject/non‐subject distinction (see Bacskai‐Atkari 2020b on how this relates to the case system).…”

“…This method has the advantage that the same loci can be compared, so that the observed differences are not due to highly different contexts or even registers; this allows for some quantitative comparison. In order to extract the relevant Middle English data, I adopted the methodology of Bacskai-Atkari (2020a, 2020b for Early Modern English: the hits for 'who ' and 'whom' in the New King James version (1989) were taken as the basis and the corresponding elements in the Middle English translations were examined. The New King James version is fairly reliable on the [AEhuman] distinction, only very few nouns showing variation, such as people or country, which can be taken as 'borderline cases' (beyond the classical 'sanctioned borderline cases', see Herrmann 2005, quoting Quirk et al 1985: 314-318, 1245-1246.…”

“…van Gelderen 2014: 173) so that its conservative features are not surprising; still, the overall distribution regarding wh / that holds in this text just as in the speech‐related text types of the period, making it altogether not strikingly different as far as relative clauses are concerned (cf. Bacskai‐Atkari 2020b).…”

“…A further implication concerns the occurrence of relative clauses in a given language: the subject function can always be relativised, while lower functions can only be relativised if all the functions ranked higher are. Relative pronouns are generally more likely to appear in lower functions (Herrmann 2005 for English dialects, Bacskai‐Atkari 2020a, 2020b for Early Modern English); this is due to processing reasons (Romaine 1984, Kirby 1996, Fleischer 2004a, see the Filler‐Gap‐Complexity Hypothesis of Hawkins 1999) and it is in this respect similar to the distribution of resumptive pronouns. Moreover, wh ‐based relative pronouns started to appear also in the lower functions first (Romaine 1982, Gisborne & Truswell 2017; see also van Gelderen 2004, citing Dekeyser 1986 and Allen 1977: 197–199).…”