Inspired by the corpus-building work of Randolph Quirk at UCL soon after his arrival at Lancaster, Leech pioneered computer corpus development.[5] He initiated the first electronic corpus of British English, completed in 1978 as the [Lancaster-Oslo-Bergen Corpus|Lancaster-Oslo/Bergen] (LOB) Corpus.[6] Later, in the 1990s, he took a leading role in the compilation of the British National Corpus (BNC). The Lancaster research group that he co-founded also developed programs for the annotation of corpora: especially corpus taggers and parsers.[8] The term treebank, now generally applied to a parsed corpus, was coined by Leech in the 1980s] The LGSWE grammar (1999) was systematically based on corpus analysis. Leech's more recent corpus research has centred on grammatical change in recent and contemporary English.[10]