Thanks for replying. these are all interesting points, and I'm intrigued by your prediction. If things work out that way I would be quite happy; pay walls in front of science suck!
> - A lot of papers end up in a lot of different official journals in some form. Publishing slightly varied versions of the same work across 3-4 different journals is a time honoured tradition.
Hm. Typically in CS these happens a different scales; e.g. a few papers with the same title (workshop, conference, perhaps another conference paper with a different application or an extension to the theory, ..., journal) but vastly different levels of detail/justification or significant extensions to theory/application. So same idea, but not the same paper and not presenting the same results.
Publishing the same exact results in multiple venues is, afaik, a form of academic dishonesty and a serious black mark. Hopefully it works the same way in other fields.
> - Actual journals do effectively the same thing that arXiv does, but theoretically with more 'experts' and rigour.
I'd remove the theoretically qualifier; getting into top venues in very difficult. Although I'm sure many worthy papers are rejected, the bar is still very high.
For your last three points, I'll cede to your knowledge on the matter.
Both thank you for the reasoned comments, and you're welcome. Also, I cede the "theoretical" point, as that was mostly me being snarky.
I agree that publishing the 'exact' same results is a no go, and that will usually get you a demerit (it's just too easy to discover these days). However, I would argue that the level of detail/justification that people change can vary dramatically. Some people are on the honest end of the scale, like you say, and change each version significantly. In fact, to step away from hyperbole, I would say that probably most trend that way.
However, I feel like the current system also incentivizes dishonesty, particularly with how brutal the associate professor and tenure tracks have gotten for many fields. For example, while CS is a relatively nice field, since its both new, and the job market for its PhDs is hot, others like Physics, Math, most liberal arts, ect... are awful, with 10:1 ratios in some cases between candidates and positions. In those, you'd better be perceived as a rockstar, or you're going to be flipping burgers. And the natural way to be seen as a rockstar? Publish. A lot.
Actually, now that I think about it, that would probably make a good sociology paper. Possible correlations of plagiarism, minor-edit multi-submissions, yatta... to the level of competition for professor spots.
> - A lot of papers end up in a lot of different official journals in some form. Publishing slightly varied versions of the same work across 3-4 different journals is a time honoured tradition.
Hm. Typically in CS these happens a different scales; e.g. a few papers with the same title (workshop, conference, perhaps another conference paper with a different application or an extension to the theory, ..., journal) but vastly different levels of detail/justification or significant extensions to theory/application. So same idea, but not the same paper and not presenting the same results.
Publishing the same exact results in multiple venues is, afaik, a form of academic dishonesty and a serious black mark. Hopefully it works the same way in other fields.
> - Actual journals do effectively the same thing that arXiv does, but theoretically with more 'experts' and rigour.
I'd remove the theoretically qualifier; getting into top venues in very difficult. Although I'm sure many worthy papers are rejected, the bar is still very high.
For your last three points, I'll cede to your knowledge on the matter.