In a previous posting (here), I discussed the relative frequency of 'my research' and 'my researches' in a large corpus of contemporary American English.
Just recently, Google Books have made available a search facility based on 5.2 million digitized books, in a facility they call 'Books Ngram Viewer' (here). (For a detailed discussion of it, see David Crystal's DCblog.) And this allows us to compare the occurrence of research with that of researches since the year 1800.
In the following plot, the blue line shows the relative frequency of 'my research' while the red line shows that for 'my researches' from 1800 till 2000.
This suggests that my assumption that researches is a modern development (which may be influenced by usage in places such as Brunei and Singapore) is not quite correct. In fact, it seems that researches was overwhelimingly more common until about 1940 when the use of the word as a non-count noun became more usual.
I love the development of on-line utilities like this that allow one to investigate things oneself at the touch of a button.
Peeving and changes in relative frequency
8 hours ago