3 ms·
I think the problem is requiring excessive polish in academic papers. This causes two problems: 1) It's easier to polish something which is incremental. 2) I
by CJefferson 6y ago
I think the problem is requiring excessive polish in academic papers.
This causes two problems:
1) It's easier to polish something which is incremental.
2) It's easier to achieve the look of polish by purposefully hiding issues and smoothing over edges.
I'd personally like every paper about a technique or algorithm to include as many problems where it does terribly as where it does well -- this is annoyingly uncommon. Also, people shouldn't be afraid to discuss all the nasty unfinished bits but in practice if you mention them reviews say "well, go fix that before acceptance".
- chalst 6y agoI think this is close to the truth but misses an important factor: journal editors are, as a result of increasing submissions, asking for more refereeing. As a result, an increasingly high proportion of academics don't referee at all and those that do tend to referee more papers in less depth. With less time spent per referee report, easily appreciated evidence of polish matters more where papers that pay attention to careful issues of methodology, tackle deep problems, etc., require time on the part of the referee that they are unwilling to spend.
- bjornsing 6y agoYes. I’m into ML/AI and would love to read more papers on cool ideas studied in isolation, but most papers out there are more “we combined these fiftytwelve techniques and look it’s SOTA on this dataset!” (It should also say “we have no idea why” but they tend to leave that out.)
- Cd00d 6y agoI don't know if it's the cause, but the "polish" requirement is horrendous. When I left academia one of my most significant thoughts was, "at least I won't have to argue over the 21st draft of a 3 page paper ever again".