Hans Martin Lehmann

Zero subject relative constructions in American and British English

This large-scale corpus study documents the use of zero subject relative constructions in spoken American and British English. For this purpose, it makes extensive use of automated retrieval strategies. It shows that zero subject relatives are still present in spoken American and British English, as represented in the British National Corpus and the Longman Spoken American Corpus. Moreover, there is a sharp difference between American English with 2.5% and British English with 13% of subject relatives with zero relativizer. Although zero subject relative constructions are frequently found with existentials and it–clefts they are by no means limited to these constructions. The social variables of the study (most notably age) come from speaker annotation which is used to provide the apparent time dimension.

Syntactic variation and lexical preference in the dative-shift alternation

In this paper we investigate the dative shift alternation on the basis of 580 million words annotated with a dependency parser. We explore the link between lexical and syntactic choice in a core case of syntactic variation. We show that the sheer amount of automatically parsed data permits new approaches toward studying not only lexis but also other language internal factors influencing the syntactic choice offered by the dative shift alternation. Our approach offers new possibilities by replacing intuition-based decisions inherent to traditional variationist methodology with decisions based on observation.

Parsing early and late modern English corpora

We describe, evaluate, and improve the automatic annotation of diachronic corpora at the levels of word-class, lemma, chunks, and dependency syntax. As corpora we use the ARCHER corpus (texts from 1,600 to 2,000) and the ZEN corpus (texts from 1,660 to 1,800). Performance on Modern English is considerably lower than on Present Day English (PDE). We present several methods that improve performance. First we use the spelling normalization tool VARD to map spelling variants to their PDE equivalent, which improves tagging. We investigate the tagging changes that are due to the normalization and observe improvements, deterioration, and missing mappings. We then implement an optimized version, using VARD rules and preprocessing steps to improve normalization. We evaluate the improvement on parsing performance, comparing original text, standard VARD, and our optimized version. Over 90% of the normalization changes lead to improved parsing, and 17.3% of all 422 manually annotated sentences get a net improved parse. As a next step, we adapt the parser’s grammar, add a semantic expectation model and a model for prepositional phrases (PP)-attachment interaction to the parser. These extensions improve parser performance, marginally on PDE, more considerably on earlier texts—2—5% on PP-attachment relations (e.g. from 63.6 to 68.4% and from 70 to 72.9% on 17th century texts). Finally, we briefly outline linguistic applications and give two examples: gerundials and auxiliary verbs in the ZEN corpus, showing that despite high noise levels linguistic signals clearly emerge, opening new possibilities for large-scale research of gradient phenomena in language change.

As and other relativizers after same in present-day standard English

This is a large-scale corpus study of relative constructions containing same in the antecedent. These differ from other relative constructions in that they permit the use of as as a relativizer and thus offer different possibilities of variation than other relative constructions, something that has not been well described in handbooks. We found that same-constructions occur much more frequently with relativizers having adverbial function, and that they also show a different semantic patterning than other adverbial relative constructions. The most common relativizer in speech is as with over 50%; in writing, as and that each account for about a third. A variable rule analysis showed that the factors independently favouring the choice of as were the function of same as antecedent head, the functions of as as adverbial or subject complement, and occurrence in speech. There are also some differences between speech and writing when as is the relative marker in adverbial function, in that the ranking is manner-temporal-locative in speech and temporal-manner-locative in writing. We discuss our findings in the light of the pragmatics of same-constructions and consider the history of as as a relative marker in English.