I am a big fan of old corpora. Of course, I do also appreciate XXL corpora compiled semi-blindly from the Web. But, comparatively speaking, the older corpora have the kind of spick-and-span internal structure...
Delving into corpus linguistics, its theoretical relevance, and its applications
- Characterizing discourse genres with prosodic features in a reference treebank of spoken French
- Statistical analyses of Spoken French Prosodic Structures
- PreechVis: Visual profiling using multiple-word combinations
- It is important to note that" partially productive patterns may count as constructions
- Can word vectors help corpus linguists?