The World of English Historical Corpora

Abstract

Computerized corpora can be said to have revolutionized the study of the history of English. Corpora as such do not increase our knowledge of the language of the past; they are based on manuscripts and editions available, at least in principle, even in nonelectronic form. It is evident, however, that corpora give us an opportunity to master huge quantities of textual material, to collect and sort evidence with a speed and level of accuracy that the scholars of earlier decades could only have dreamt of. Corpora have freed us from months and years of painstaking pencil work and from the straitjacket of worn-out examples appearing in older dictionaries and historical grammars.This easy access to a wealth of texts representing various periods, genres, dialects, registers, and social strata allows us to declare more ambitious goals for our research than was previously possible. We can now begin serious collection of the jigsaw puzzle pieces that would yield an integrated view of the English language and its development, a description and analysis based on the internal processes of linguistic change, as well as on the influence of a spectrum of extralinguistic factors, be they textual, sociolinguistic, or regional.1 Quite obviously, a conclusive "ultimate explanation" of the development of English will remain as unreachable as any final truth in science or scholarship, but with improved access to evidence, we can now organize our research aims more coherently as parts of larger entities, see the connections between individual structural and lexical developments, and avoid the feeling of fragmentation that often diminishes the value of research in historical linguistics.The story of English historical corpora is brief but intensive. The first efficient electronic tools for the study of the history of English, the Dictionary of Old English database prepared in Toronto and the Augustan Prose Sample and the Century of Prose Corpus compiled by Louis Milic at Cleveland State University, emerged in the early 1980s. The 1990s saw the completion of the Helsinki Corpus,

ICE vs GloWbE: Big data and corpus compilation

“…However, the relevance of size of corpora has often been overestimated, as shown in Nurmi () and Hundt and Leech (); also, the number of text types included in ICE (Table ) is much larger than what can be obtained from general websites and blogs. In this sense, it could be claimed that ICE and GloWbE belong, respectively, to the two types of corpora identified by Rissanen (: 8): small size ‘multi‐purpose’ corpora (ICE) and larger but genre specific corpora (GloWbE).…”

Digital Editions for Corpus Linguistics: Representing manuscript reality in electronic corpora

“…Rissanen 2000). Comparatively less attention has been paid to a third dimension, depth, which could be defined as the extent to which the corpus represents the various features of the original texts.…”