It takes time to read all the prior research. You could grow old by the time you get through it all. Likelihood of contributing to the field declines with age.

It takes time to read all the prior research. You could grow old by the time you get through it all. Likelihood of contributing to the field declines with age. You might believe someone's proof of a conjecture and then be discouraged from delving any more into that rabbit hole. More often than not you will be reinventing something. But that's not necessary less productive than reading other people's work. In the former case, you're at least making something, if not new. So there are some arguments for being fresh in a well-trodden field with an ocean of research that you cannot boil all at once. On the other hand, there is the publish-or-perish pressure in academia, which requires original research. You could just keep busy and throw enough shit agains the wall such that enough of it sticks.