A couple of minor points re: word list
1. The link to the concordances of a particular frequency of a particular lemma does not correspond to the frequency of the lemma but of the word form. So the lemma 'joe' taken from a 'cinema' subcorpus may show a frequency of 3062 but if you click on this number, you only get 9 hits as most of them are 'Joe' rather than 'joe'. Equally, 'iii' is high but you only get a few lines as most are 'III' as in 'Rocky III'. 'manage' will not have 'manages' etc.
2. Is there any way to rank the list by salience - this would be useful but perhaps too complicated as the subcorpus list would have to compared with the whole corpus???
