Heather Lane wrote:
That's a good idea. I know one or two translators who don't work with the programme, and a heap that do, so that would be plausible.
May be taking the same text, or couple of texts, and having people like you suggested translate this/these same text(s) would indeed produce some interesting results.
... that any translator worth his/her salt would revise a "CAT translation" afterwards to sharpen the flow, and polish it particularly from the point of view of segmentation (i.e. in addition to what any professional translator not using a CAT tool would do).
So, I'd suggest you'd need to compare the two translations in their raw (i.e. most basic, pre-revision) form to get reliable results on the actual impact of segmentation.
Best of luck with your studies