Live data from Hacker News

Time to break academic publishing's stranglehold on research

newscientist.com

311–312 of 312 posts

Re: Time to break academic publishing's stranglehold on research

#311
post #305

Earlier quoted context omitted.

You should elaborate. In my experience academia is largely a status structure where people are groomed to be hirable at schools X, Y, Z and publish regularly in conferences/journals A, B, C. You're saying this is fertile ground for hardcore innovation? No way. The extreme majority of the work is incremental and predictable; small low-impact results. (Disclaimer: My experience is in CS.)

Alright, I'll elaborate. My dissertation was based on a highly speculative new treatment, in a field where mathematical models are not widely used. One of the potential answers was "Yes, this should entirely change the course of care for a particular disease." My current advisee's first paper is, essentially, "Has the field been doing this wrong this whole time?" Note that I publish regularly in the good society leve…

That's cool, congrats. In my field, computer science, it's far from your experience. Over and over I've experienced PhD students being pushed away from more radical ideas (opening "risky" but potentially really promising new areas of inquiry) toward incremental work on what's currently "hot" in the field. Yes I concur that I used "incremental" in an unnecessarily disparaging way. Some incremental work can be revolutionary, but in my experience in my field, most of it is humdrum.

Re: Time to break academic publishing's stranglehold on research

#312
post #307
post #287

Earlier quoted context omitted.

>how about a model like stackoverflow? papers are published like questions, and reviews are the answers. readers upvote good papers and good reviews, so that the most upvoted and most reviewed papers float to the top. The problem is that long-form works that require substantial analysis by experts (such as scientific papers read by competent scientists) don't work for a voting system like Stackoverflow. Internet voti…

right. votes would just be for the general popupation, to show which papers are popular. if votes are used at all. scientists would focus on actual reviews and citations. the only challenge i see is how we motivate people to write reviews. perhaps as some sort of a trade. for every paper uploaded you are expected to review 3 other papers. so that each paper can get at least 3 reviews. there is also a potential for ma…

> right. votes would just be for the general popupation, to show which papers are popular. if votes are used at all.

The last thing we want is the general public voting on scientific research. Any results that don't fit their world view will be voted down.

Post reply on HN