In reply to Pewtle:
Yes, i had a thought around that. But if you do a research about, for example, 'which are rounder, pears or apples', and you research consists exclusively on measuring the roundness of pears and apples, the only results you'll get is:
- Apples are rounder than pears
- Pears are rounder than apples
- Pears and apples are the same roundness
- We can't establish if there are roundness differences between pears and apples
Not much scope for unexpected discoveries, is it?