Hacker News Reader

Comment by yummypaint | Hacker News Reader

A finer point often missed on HN is how different publishing is across fields. In physics i have always been able post preprints to the web/arXiv and leave them up indefinitely. Their scientific content is identical to the journal formatted version and they show up in search often before the paid version. This is true even for the historically worst offenders (i.e. Elsevier)

Last I heard the situation in chemistry was still completely different. The ACS is also known for bullying cash strapped college libraries into exorbitant and unnecessary journal subscriptions.

View on HN

I had dinner with the arXiv creator once (https://en.wikipedia.org/wiki/Paul_Ginsparg). What I understood is is that before the arXiv, all "good" universities would each mail each other copies of preprints, and so researchers there would have access to the newest research developments. Eventually, they'd get published, but that takes a while, so anybody not in this circle would be far behind and have a hard time being relevant in his field (theoretical particle physics).

So the arXiv was a more modern, scalable, democratic replacement of what was already popular within the culture of the field.

I'm a medical physicist. I have articles on arxiv, and articles in clinical journals like Circulation (which have a huge impact factor, are required to get the techniques "out there" and demonstrate their value, but basically necessitate putting all the detail and maths in the SI, if not in a separate paper).

Your statement is very, _very_ true. Physics and maths are ahead of the curve, and in astronomy in particular they hammer papers out at a rate of knots where the arxiv is what matters. Biology, biomedicine, and chemistry are totally different. Everyone uses Word (which I hate) and everyone's formatting looks like shit as a result. The concept that you could have "camera ready copy" is alien. Moreover than that, stuff like a data supplement increasingly needs to be hosted by someone else the further down that train you go. You or I would be happy slapping something up on a static University-run website or linking to it; that gets harder the less technical you get. Moreover, the culture is far less collaborative (in my experience) with a lot more competition. The thing I loved about particle physics was that everyone helps. Mathematicians give talks thanking others honestly for their contributions, and name-dropping people who helped in odd ways. That's very alien to the "We ran this clinical trial" talk I hear a lot of in my world.

Heck, there are multiple groups "competing" in my field on diseases like prostate cancer, which sometimes makes the "losing" group unable to publish their results in a "good enough" journal, and it's not easy to coordinate or organise -- the funding bodies encourage competition. Physicists tend to take the attitude that "there is only one universe to study" and collaborate sensibly; sure, Fermilab and CERN are nominally in competition, say (or ATLAS and the CMS, say) but their competition is more about validation and they explicitly pool resources and ideas such that this is the case. I did my masters' in particle physics and am acknowledged in a textbook chapter I wrote with my then-professor -- take it from me, it's friendly. (I ultimately moved to biology because I felt I was too stupid to progress far in particle, and the medical problems were -- to me -- far easier to solve, or at least make progress on. Doing maths and building things in a hospital is much more of a "USP" than doing maths and building things in a particle physics lab, though I do still love it a lot).

The net result is that publication practices are cultural more than anything, and frankly in my experience only really the mathematicians and physicists have enough freedom and balls to do their own thing. I wish the other fields would follow suit, but there are a lot of vested interests trying to prevent that.

I agree; I get the impression that physics and math still do things in The Old Way of science; if we still communicated that way, modern physicists would be writing letters to each other critiquing work and offering advice. Somehow the life sciences became stuck in the “publish or perish” mentality.

Just out of curiosity, you mention everyone uses Word in the life sciences—what is used in physics? LaTeX? I also have to use Word, and hate it.

Not the OP, but in my experience, when you are doing math, physics or CS and don't use LaTeX, you'll be looked at sideways. There is virtually no alternative when it comes to typesetting advanced maths.

I have friends who used MS Word to write their thesis in CS undergrad.

Surprisingly it is possible to write math in a LaTeX-like fashion and they show up quite well. I have had more trouble aligning things with Word than with LaTeX so I kinda looked at them sideways.

The math support in MS Office is sufficient for some tasks and it's actually very usable then. But I find it woefully underpowered when it comes to typesetting advanced stuff. I come up against its limit quite regularly (I need to make my slides in Powerpoint for $REASONS). Also, annoyingly, it slows down significantly when there are lots of nested blocks (e.g. subscripts, superscripts, which I need often).

MathML / very neat handwritten formulae are decent alternatives.

Yep. "The authors don't use TeX" is also sign #1 on Scott Aaronson's "Ten Signs a Claimed Mathematical Breakthrough is Wrong" https://www.scottaaronson.com/blog/?p=304

> This simple test (suggested by Dave Bacon) already catches at least 60% of wrong mathematical breakthroughs. David Deutsch and Lov Grover are among the only known false positives.

Without precision, recall and a proper problem formulation,”60%” is pretty bad, if not outright meaningless.

Yes, LaTeX is what’s most widely used in physics.

Perhaps unfairly, it's hard to take physics papers not written in LaTeX seriously. At least within particle physics and astrophysics/cosmology... I don't really read papers from other fields.

Note that, arXiv makes you post LaTeX sources if it detects that it's a LaTeX document, and not everyone remembers to redact perhaps not meant to be public comments :).

Most of my biology collaborators have switched in the past 5 years to putting all their new submissions on biorxiv. It seems like medrxiv is now getting popular since COVID started too. But, as in math and physics, I guess a big problem is articles that predate this change are still paywalled in many cases.

I would add that even within physics there are massive differences, which also shows up in how grants are written. While the fundamental physics areas (like particle and astronomy) do publish a lot on arxiv etc. In the more applied areas it can often be quite different (which also applies to the use of word unfortunately). The metrics you use for your grant applications are also very different, the more applied areas often have to show impact through patents, spin offs etc., while this is less so in the fundamental sciences.

The use of arxiv etc is slowly making its way into the more applied areas, so things are moving to become more open. On the other hand in quantum we are currently seeing the opposite trend, so much research is now within startups who are absolute flush with money, but typically don't have a path to commercialisation. So they do research, but only very little of that research gets published anywhere, so the scientific process is becoming much less open.

Why do the different disciplines have such different cultures?

The Web itself got started at CERN, not in a zoology or animal testing lab. It takes a lot of time for successful innovations to diffuse throughout a broad, worldwide subculture. Just do the math.

I would guess money plays a part. The commercial prospects for discovering a new fundamental particle and discovering a cure for prostate cancer are quite different.

Well, it depends. Is the cure for mice, or actual humans.

I would wager the nature of work. It might be because maths and CS involve more work that doesn't require fancy equipment, there is lesser involvement of financial risk as compared to fields like biology or chemistry, where equipment is so expensive.

You just said it. Scientific fields are not differentiated arbitrarily; they drift apart as their experimental needs diverge. Chemistry and particle physics used to be one field, back before the cost of particle accelerators diverged from the cost of chemistry labs.

Right -- so how do the different experimental needs of each field inform their different cultures?

Return on investment pays a role. The higher the profit margin, the more competitive and "locked down" the field is. But things like national prestige and space race type international politics comes into play. Very rarely is lots of money spent on pure science - there's almost always a catch.

I think there is a lot to learn from what you said, but I just want to remind everyone else reading your comment that it is anecdotal.

I have experience with both research in physics and chemistry in multidisciplinary teams, and my experience is the exact opposite to yours in terms of competition and the willingness to help others. Heck, the last time I released some results on arXiv a group that is notoriously unfriendly suddenly reached out and was all friendly. It turns out that after discussing with us about our results they released a half-finished draft on something similar that they have been working on, and presumably tried to stay ahead/dilute the significance of our results by getting it published in a peer-reviewed journal first. Also, during the time we were doing our work, nobody responded when we were asking if we could access data or algorithms. I don't blame anyone and we're definitely not entitled to anyone's help, but that's just a counter example and obviously depends on how well you know the field, everyone else in the field, reputation, etc.

I personally enjoyed using LaTeX for writing manuscripts for physics-oriented topics, but I don't mind Word for chemistry manuscripts. In fact, I would argue that writing chemistry manuscripts in LaTeX sucks for most chemists (maybe except for theoretical and some physical chemists) simply because editing chemical structures, reactions, equipment setup and process diagrams is easy with an embedded chemical editor object in a Word document.

I'm not familiar with how they work internally at publishers when getting manuscripts ready for publishing, but most journals require you to use a Word template with ready-to-use style rules and adhere to certain easy-to-follow formatting rules for everything else. Remembering my early days of using LaTeX (when I was already familiar with using Word for manuscripts), I could say the same about how cumbersome it is to do certain things in LaTeX without even talking about chemical structures. Particularly when it comes to journals that charge subscription fees, I'm not sure why you seem to suggest that it's on the authors to get the manuscript "camera ready". If nothing else, I have seen just as many crap-looking papers formatted with LaTeX than those with Word, and the same can be said about good-looking papers. So... I guess it just depends on the authors themselves and/or the editors.

For much the same reasons as what I have said about competition, I think your last comment is just as biased against your own experience than everything else in my opinion.

> Moreover, the culture is far less collaborative (in my experience) with a lot more competition. The thing I loved about particle physics was that everyone helps. Mathematicians give talks thanking others honestly for their contributions, and name-dropping people who helped in odd ways. That's very alien to the "We ran this clinical trial" talk I hear a lot of in my world.

That’s probably because physicists and mathematicians have less to lose by openly talking about their results and methods, while in bio research, there is a lot of proprietary secrets about lab techniques and hard-earned experiment data, which many researchers see as competitive advantage they justly reap in exchange for the labor they put into obtaining it.

I am a particle physicist and would like to say that Fermilab (where I'm working on one of the experiments there) and CERN don't compete. One of the best examples is the construction of DUNE which is a new experiment which will be the largest neutrino experiment so far is done between CERN and fermilab closely. The prototype is even there. Fermilab main focus is different a little bit from CERN. Also Fermilab is part of every CERN collaboration (except maybe small experiments here or there). Usually in particle physic we tend to confirm results with different experiments and sometimes different techniques and because of the nature of statistics and probabilities involved, this always encouraged collaboration and people trying to help each other. I understand that this is completely different from the competition in other fields.

Exactly -- and that's the point I was trying to make about competition. Fermilab arguably is CERN's biggest "competitor" and it's entirely complementary – e.g. look at the work done with LEP and Tevatron. In contrast, I've had other academics politely ask me not to present my work at a particular conference because so-and-so might be there and that might mean that they'd see it and their paper on x+∂x might come out before our paper on x-∂x, and that would make it harder to get into a particular journal, etc --- and I didn't. It's pretty stupid really. I wish we'd pool resources sensibly, and divide the big ideas ahead of time.

> In physics i have always been able to post [to] arXiv

I'm fairly sure Sci-Hub (and book scanners like CZUR's) made DOI useful, even (and perhaps especially) for working and teaching physicists at large research universities with excellent librarians.

Sometimes in the moment one wants to read something sufficiently old or inconveniently rare cited in an arxiv preprint or in one of its easy-to-click references. Or if one is Reviewer Number 3 or that person's victim.

What do you think of fields like medicine or biology where preprints can become a source of misinformation. For example a lot of conspiracy around unproven treatments for covid, vaccine efficacy, and vaccine safety directly source preprints that never make it to publication.

Reply on news.ycombinator.com