The readability of scientific texts is decreasing | Hacker News Reader

The readability of scientific texts is decreasing

elifesciences.org

82 pointsgfredtech67 comments

I like'd PG's tweet about this: "Possible explanation: papers are becoming less how you communicate ideas, and more how you register work to get credit for it."

https://twitter.com/i/web/status/906075608181915649

The old problem "when you turn a metric into a goal, it ceases to be a good metric". Scientists are rewarded for churning out a large number of fragmentary (if you have a really good idea, it's good for more than one paper), shoddily written papers to collect citation points. "The stuff between the formulas" is not something authors really care about.

I noticed the awful writing in new (physics) papers when I was still in university and had to read a paper occasionally. I was always relieved to find a paper from before roughly 1980. They are so much better written. Older papers are often better written than most textbooks while recent papers are much, much worse. It is not just due to the subject matter changing.

Reportedly, journals also used to have paid copy editors. Today - well, you have heard about Elsevier and other publishers, right? They don't only seem to increase prices to increase their profits.

Writing ability is on the wane in general. In professional settings, IMO this is mostly due to the disappearance of secretaries and clerical staff, lousy education and more non-native speakers.

A long time ago as an intern, I had to re-typeset and format some reports published by a .gov since the 1910s for reprint and web publishing. You would see variations in style, but the reports started getting worse in the early 80s, and are almost incomprehensible today.

It may also be because university education used to be for those with a great interest in science, and I guess for some reasonably intelligent upper class folks with nothing better to do. The first group had the motivation, the second group the habit and education to express themselves well. Today it is for those who want to get a good job. The education for a job majority lowers the standards for everyone.

Btw I'm a middle class guy who did it mostly for job prospects and such, so...

> if you have a really good idea, it's good for more than one paper

One professor of mine spoke fondly of the "LPU", the least publishable unit...

> "when you turn a metric into a goal, it ceases to be a good metric"

Goodhart's Law: https://en.wikipedia.org/wiki/Goodhart%27s_law

PS: Ceterum censeo Elsevier(um?) esse delendum.

This seems inherently linked to the increase in complexity and precision of scientific ideas, particularly over the time scale investigated (back to 1881).

I certainly applaud efforts to manage this complexity (e.g. the article mentions possibly adding "lay person summaries" in addition to abstracts), but I think that increased complexity and depth of scientific results is the intended outcome.

It seems analogous to these insane computer-generated proofs in mathematics -- maybe we need new tooling and approaches to make sense of them, but the fact that they exist is proof that we're discovering things and moving forward.

I don't buy that explanation - there's also just a lot of bad writing out there today. And, I'd wager, you'll find (more frequently than before) attempts to impress (and even obfuscate), rather then communicate and elucidate.

Note also that your hypothesis was briefly addressed in the article:

> An alternative explanation for the main finding is that the cumulative growth of scientific knowledge makes an increasingly complex language necessary. This cannot be directly tested, but if this were to fully explain the trend, we would expect a greater diversity of vocabulary as science grows more specialized. While accounting for the original finding of the increase in difficult words and of syllable count, this would not explain the increase of general scientific jargon words (e.g. 'furthermore' or 'novel', Figure 6B). Thus, this possible explanation cannot fully account for our findings.

Is "furthermore" scientific jargon? I thought it is just a professional version of "moreover". Like, I would say "moreover" to a friend, but write "furthermore" in a letter.

I had no idea that 'furthermore' or 'novel' are scientific jargon. I have seen that frequently enough to think it is normal English word. I would use furthermore in paper, because that was the word people seemed to be using all the time.

Eh. I'll definitely say that at least one of my favorite authors (as in, scientists) to read is being deliberately obscurantist about how he writes his papers. Or at the very least, he's using all his own vocabulary and terminology in each paper to push his favorite philosophical interpretation of his own work. This makes him extraordinarily difficult for other people in his field to understand, which actually means people don't engage with his ideas when they rightly should.

But he certainly gets to publish a lot with his own favorite colleagues and push his philosophy!

> This seems inherently linked to the increase in complexity and precision of scientific ideas, particularly over the time scale investigated (back to 1881).

I think that's a reasonable conclusion here. I mean medicine in that time frame seems to have gone from "Yeah just shelve some cocaine and your cough will go away" to MRI's so it makes sense that the research and results have gotten a hell of a lot more complicated.

I suspect that today scientific texts are more frequently written by non-native English speakers and that technical jargon can be more precisely defined and understood in terms of non-English languages. It is also worth noting that scientific terms of 100 years ago are often more main stream, e.g. quantum mechanics, relativity, uncertainty, etc.

I'm pretty sure this is not the case. Scientific texts are usually awfully bad in any language I speak, including 2 languages that are my native. In fact, "scientific paper style" language is so distinct, that it can be easily applied to a paper that is not "scientific" per se, essentially obfuscating it.

It seems to be very unpopular opinion here on HN, but it seems clear to me, that academia has become some kind of patalogical structure, which exists to exists — essentially an organism. It feeds on government and student's money, as long as it can seems credible and necessary to the outside world; and people within this structure can live as long, as they show that they are not "slacking" somehow — publishing papers, giving lectures — but they don't actually need (nor often want) to produce any result. Form has become more important than the essence.

So, here we have it: the form of a "scientific paper".

The casual nature with which you dismiss thousands of people who have devoted their lives to the pursuit of academic knowledge is frustrating and incredibly childish.

Like seriously? Do you think all of us PhDs just sit around and laugh about how dumb the government is while planning how we can make more useless papers? Do you think grad students just train to become better bullshitters? Like what kind of hole do you live in where any of that even appears close to reality.

Seems to me you just tried reading a couple papers, got confused, and decided to justify your own intelligence by claiming their hard work and advancements in the field were pointless.

> but they don't actually need (nor often want) to produce any result

Stop talking out of your rear end. I want to produce a result. My advisor wants to produce a result. We all want to have meaning in our lives.

EDIT: As just a casual example of the importance of academia, look at the Deep Learning boom. First researched around 1960, only now becoming practical and useful. It's almost as if the things academics study need time and a lot of work to actually come to fruition. Research level study is highly expensive for very little real gains. This has always been the case. It takes a long-long time for the investments into research to make sense, but when they do, the advancements are completely game-changing.

The ratio of PhDs awarded annually to tenure track positions that opens up suggests that a lot of activity in the sector isn't leading anywhere.

It's almost as if most research doesn't go anywhere.

If I put you in a dark room, and ask you to find a way to light it, how many times will you fail to find the switch? Oh wait, I lied. There's not actually a switch, but there might be a candle in that general direction. Maybe.

That's how little we have to go off of. That's how empty any information on your study is. I don't think people really understand how hard this type of work is.

Research is a purely exploratory and creative work. Like art, science requires time, patience, and many many failures. You don't just get a bunch of eureka-Einsteins that come out of MIT day one knocking down Theorems and discovering new physical laws. That's just not humanly possible and it's ridiculous if you think it'll ever work that way

I am perfectly willing to admit that the vast majority of research and PhD's went nowhere and will continue to go nowhere for a long time. But that doesn't mean they're useless. Research requires exploration. It requires trial and error. It is highly inefficient and does not fit well in a consumerist society like ours, but this is the way that works best for us so far.

If you have some better idea, then why don't you go out and do it? How are you going to get a bunch of incredibly intelligent, creative, and independent people funded and provided for in a society that values immediate results over long term achievements?

I'm not arguing the academia system is perfect. Any one knows it's screwed up beyond belief. But it's not out of malice. We're not purposely trying to make the world worse, or something ridiculous like that. You have many people who care a lot about the science, who need to eat and sleep and need money to do that. Sometimes the money goes to their heads, sometimes their massive egos get the better of them, but do you seriously think we don't care about science? You'd have to be an idiot to do a PhD for the money or the fame. Why do you think we put 7+ years of our lives into stuff like this?

So most PhDs don't actually get tenured positions. What are you actually trying to say? Are you saying there aren't enough tenured positions? Or are you implying that these PhDs don't deserve tenured positions? Do you have any evidence for either of those claims? Because otherwise you're just participating in more armchair speculation.

A PhD is effectively an apprenticeship towards becoming a tenure-track professor. It's not unreasonable to wonder why we are investing so much in training so many people when there is so little demand for that training. The vast majority of PhDs end up in industry doing something barely related, and are no more productive at it than those with only a first degree or even no degree at all.

> A PhD is effectively an apprenticeship towards becoming a tenure-track professor.

No it's not, plenty of people get PhD with the goal of ending as a researcher in the industry.

Or, even, shockingly, because they enjoyed doing research, even if they then don't spend the rest of their life doing research in academia or industry.

I didn't read krick as suggesting that malice was involved. He/she was talking about the kind of work that the institutional structure of science tends to favor.

Perhaps your experience was different from mine. I went to a well known, buy non ivy league school, and grad school for myself and most, not all, of those around me was more about faking it until you made it into industry.

Plenty of junk science, including a thesis I decided not to publish, while my advisor wanted the extra points on his resume. Many PHD students are stuck because they got a hard stem degree (like physics) but realized that job pickings were a little slim without further academic development. Or they are foreigners and cannot stay in the U.S. without studying.

Passion does not always last forever. Especially when you got off-ivy league schools. Didnt you hear stories in school about graduate students who were simply stuck for a decade because of lack of interest from them and/or their advisor? Not totally farfetched.

Yeah...I saw this about an hour ago, and bit my tongue. But it is dismaying that apparently the reverence for science is so high around here that it extends to downvoting criticisms of science as a process and sociological phenomenon to oblivion.

Maybe it will help these poor fellows if I say look, I have a PhD, I work in research, and I agree with everything krick says. Yes, papers are unnecessarily jargonistic and borderline illegible. Yes, there are massive problems with the incentive structure in science. We openly say science is "publish or perish"; how can we not expect that to incentivize lower paper quality, irreproducibility, and status signalling in the form of unnecessary jargon? Even if we assume the noblest of intentions for every single scientist, which is...idealistic.

spaceseaman is equating (IMO unfairly) criticism of the process of science in its current US manifestation with some kind of disrespect towards its obvious beneficial outcomes and motives. The whole thing IS full of inefficiency, and that's not solely because science is hard. Taxpayers have a right to demand that we don't waste their money and perhaps even to present our findings in a way they can understand with a reasonable application of effort (ideally not paywalled as well).

Well, the problem is that people so often go motte-and-bailey on it.

Motte: all the valid critiques of how institutional science works, all of which are well-known.

Bailey: full Paul Ryanism, cut the NSF and NIH to the fucking bone and tell scientists to go get "real jobs" in industry. Subject academics to yet more administration and reporting requirements that further incentivize bad science and just generally make everyone miserable.

We'd all be less tetchy about the motte if it wasn't used as an excuse for the bailey.

Thank you explaining my frustration in such a clear way. My original response was so feral because I interpreted his comment as a criticism of all scientists and a call to change the current University system to something more "libertarian" (because those are arguments I'm frequently exposed to, and the comment I responded to appeared to dog-whistle similar ideas).

I am perfectly happy to admit that institutional science is screwed up - really bad. Scientific texts are often un-readable, and the entire community has major systemic issues. But I react incredibly poorly to the opinion that scientists enjoy this system or even benefit from it. We hate the way academia is structured. It's just that no one can figure out anything better and the benefits for enough people are important enough that swaying them is incredibly difficult.

>We hate the way academia is structured. It's just that no one can figure out anything better and the benefits for enough people are important enough that swaying them is incredibly difficult.

It's also that everyone who wants to do science is held hostage to this system. There is no other institution focused on original research, beyond shipping a product within three to five years.

Academia sucks. There are no full-time research positions for good researchers anymore; grant funding has gotten as selective as the startup lottery. Even the people who "make it" have to work horrendous hours and spend all their time marketing themselves.

Even from the very outset, you're forced to put your heart on the line, declare science your calling and your passion, and then just suck it up when you can't find a tenure-track job.

Some fields don't have an "Exit" option that makes anything better, just a "Voice" option or bust. Besides which, every "Exit" is a betrayal of the social contract, a tiny declaration that society would rather rot and burn than fix problems like mature adults.

I mean you're totally 100% right, but I don't think he wrote that with the intent of harm. I think in his post is a more serious issue. Academic papers are incredibly hard to understand. I received an undergrad degree in CS and spend a lot of time teaching myself theoretical CS. I often struggle to read CS papers, and find it hard to imagine many of them being useful to anyone without a PhD. For instance I was reading a paper on computer vision which used a Bayesian Network to predict a depth map from a regular 2D color image. After spending several weeks, I felt like I understood the article almost well enough to implement it. Thing is, there were some steps in the learning algorithm which relied on a specific kind of optimization (I forget which), but they didn't specify which parameters were used, so it wasn't obvious how to implement it to me. My friend has a PhD in EE (on the theoretical side, his education focused heavily on statistics and optimization) and he only looked at it for 30m-1hr, but he wasn't sure what they meant either.

I can always just fall back to gradient descent instead of the optimization parts of the algorithm I don't understand, but that impacts performance, and I got the feeling I'd spend hours implementing this only to have something which can't perform in real-time so I gave up. It was just a side project. It's frustrating after devoting weeks to this. Reading and attempting to implement a research paper in my own field is incredibly hard.

I feel like if research papers were written in a more accessible, less dry manner and specified details important to implementation, the world could be a better place. If it were easier to implement these things as side projects, you might have people experimenting with these things at home and then starting businesses out of them.

Failure to fully specify the methods is a problem, but it is an institutionally wide issue related to a number of complex issues. In honest open areas, this lack is due to issues with space in journal formats. In other fields, it may be to protect IP...ugh

I suppose I sensed more malice than might have been there, and I agree with you. I also find many papers difficult to read and especially lacking important implementation details.

I've found that many researchers are very open to questions via email, however. Most of the time I'll get a response that can clarify a thing or two.

I personally feel that researchers just need to be better writers though. Most of the time, the writing is just plain bad. Perhaps there needs to be a way to enforce a bare minimum in writing quality. Personally I wouldn't be surprised to see more researchers employ editors for themselves.

> After spending several weeks, I felt like I understood the article almost well enough to implement it.

How long do you think it took the researcher to develop the idea? Probably longer than a few weeks. Even if they could have communicated better, it's still the case that you, a non-expert (my apologies if this is untrue, I am assuming based on the context), can decipher and implement the method in a few weeks!

If I was you, I'd definitely reach out to the author(s). They're usually surprisingly responsive. And they love to hear from people interested in and using their work.

As a PhD student doing computational stuff, I agree with this fully. I'd say that in many, probably majority, of cases it is not possible implement the algorithm based on just the article. Many corner cases needed are omitted from the paper and if a pseudocode is given, it may have huge steps handwaved just giving a line like "optimize this functional".

It's one of the most mindboggling things (the publishing racket is perhaps worse) about academia that CS papers introducing an algorithm aren't required to publish their implementation, even though there necessarily seems to be one doing eg. simulation studies.

I don't really understand the rationale behind omitting the implementation. Maybe people write such crappy code that they're ashamed of publishing it. Or it's the more sinister scenario that the algorithm is actually crappier than the paper claims.

Latter has happened quite a few times in experience. Algorithm performed as designed on the training or example set just to fail terribly on real life data.

Or someone handwaved important things like having an information function be available (impossible, used only in proofs of correctness, way of estimating it is critical).

Or a key assumption on input was just mentioned somewhere in the depths of the paper.

Or a very specific way of measuring the result hides the deficiencies. (Similar to p-hacking or misusing stats in medicine.)

About 50% of the journal articles I read in grad school (engineering) were straight-up BS. Unfortunately, it took me about three years to figure out which 50%.

It's sounds like the paper you struggled with was designed to obfuscate the particular conditions under which its results were produced. This may be beause one of the authors thinks there's money to be made from the secret sauce, or because they're misrepresenting how broadly applicable the result is. Or because they're sloppy. Or all three.

At a certain point, you have to stop blaming yourself for not understanding the literature. Sometimes it wasn't meant to be understood.

I think academic CS papers are hard to understand because computer science is hard. I mean Knuth is an excellent writer and doesn't use a lot of jargon and yet, The Art of Computer Programming is extremely difficult for most people due to the subject matter. This despite much of the subject matter not being at the cutting edge of computer science.

>The casual nature with which you dismiss thousands of people who have devoted their lives to the pursuit of academic knowledge is frustrating and incredibly childish. Like seriously? Do you think all of us PhDs just sit around and laugh about how dumb the government is while planning how we can make more useless papers? Do you think grad students just train to become better bullshitters?

Many people with PhDs and experience of academic publishing and the academic world can attest to exactly that. In fact many had -- some even published studies about it.

http://journals.plos.org/plosmedicine/article?id=10.1371/jou...

And what does the "devoted their lives" has to do with anything? When someone has a PhD all that tells us is that the pursued a career in academics (and sometimes not even than). Nothing about them having noble goals and "devoting their life to the pursuit of knowledge" etc, except in the tautological way that working in research implies it. They might just like the paycheck, the prestige, or whatever else. Heck, they might even have been pushed by their parents. Or they couldn't think they'd make it as entrepreneurs and didn't fancy getting into the enterprise job market.

When doctors are wined and dined by drug companies to promote crap drugs, or are complicit in ordering thousands of needless operations to pad their profits, as they often do, have they not devoted even more years that the average PhD holder? They did, but people can still cheat and take advantage of a system, whether their credentials or position.

Nobody gets a PhD for the damned paycheck, because there basically isn't one.

I think the reason that people prioritize the form is because of a kind of cargo cult. The form is followed because that's how it's always been and it has worked so far. And people are scared that if the form is changed it will have unexpected consequences.

All this discussion doesn't consider the fact that papers are written for specific audience: scientists in a very narrow field. What seems to be gibberish for you is perfectly understandable for people working in that area, otherwise it wouldn't be even published in the first place. Now, it is true that it would be nice to have this translated into laymen words, but the problem is that scientists usually have better things to do...

I found your post pretty condescending. I have read and understood many papers. Even written one.

The facts are the opposite really. Today we are collecting so much data that we don't have enough time to actually write it all up. That is why we beyond busy.

And to your others assertions. Utter hogwash.

Academia is absolutely necrotic. Between the fact that good scientists spend most of their time writing grants, the fact that most scientists spend their life as eternal postdocs, the rigged publishing game, the increasing balkanization of fields, the fact that most people are trying to protect the status quo rather than overturn it, and the goal of publishing rather than actually discovering, things look very grim.

I think that government needs to cut back on grants, and instead provide better incentives for corporations to do open access basic research. This would still have the problem of producing somewhat obfuscated science, but in general it would push things towards being more interdisciplinary and functional, since corporations could increase profitability by leveraging synergies between basic and applied research.

Once upon a time a top scientists could contribute to biology, chemistry, and physics, yet today this is extremely unlikely. All the knowledge we accumulate and build on need names, structures, and nuance. You can't just throw that all away and describe what you are doing or thinking limited to a basic highschool level vocabulary.

The problems we face are harder, for complex, and more esoteric than ever before, and it is amazing.

Albeit things like "managerial summary" (not ELI5 but quite close) are very much needed. Abstracts tend to not deliver this. More importantly, often key limitations are overlooked in both to make the research look way more groundbreaking than it really is.

Caveat lector: speaking mostly for computer science and medicine.

It is interesting to finally realize that a large part of the audience in this web site is really anti-science. News that in some way attack mainstream science are very much commented in a positive light, as if scientists were secretly trying to make their own work less available and obscure on purpose.

Or maybe some are defensive of bad science and can't stand to see news that attack bad practices in mainstream science to be lauded?

Being anti- the bad versions/practices of something is not the same as being anti- that something. (The same way a whistleblower cop is not anti-police).

There's this story Brecht wrote about a guy telling another: "I'm an enemy of newspapers. I want them closed down". To which the other guy replied, "I'm an even bigger enemy of newspapers. I want better newspapers".

>as if scientists were secretly trying to make their own work less available and obscure on purpose.

If that gives them an advantage (e.g. publishing lots of BS papers and advancing their careers) then they are (and we know they are, meta studies and experienced academics say and show so). There's nothing of the "Area 51/Illuminati" kind of conspiracy thinking about this, if that's what you imply.

Rather it is the classic self-advancement BS that goes on since the world started, where people exploit loopholes and cheat to get ahead. That includes scientists, especially in today's publish or perish climate.

This. I am against bad science and bad scientific writing. There is plenty of both (just follow @RealPeerReview on twitter if you need convincing, but be prepared to suffer).

That doesn't mean at all that I'm against science. In fact, the opposite is the case – I'm against bad science, pseudo-science, and bad science writing because I'm so ardently in favour of good science.

And who is to define bad science? People who, by their own words, were not able to read scientific papers because they are "difficult to understand"? I am not saying that the specific language used in scientific literature is an advantage, but also doesn't constitute a problem in itself. Unless you are able to read those papers and point out where the "bad science" is, this is just a vacuous statement.

Also, pointing at failure points in the peer review process is ridiculous. This is not religion where you need to uphold every word. Science is made with lots of ideas that, looked from far ahead, are incorrect. "How did that ever cleared the review process"? Well those reviews only cover the minimum necessary for something to be published, it is not a guarantee that the contents are correct. It is the social process of science that takes care of that.

I don't think it is anti science, I think it is anti-academia (which is a sentiment I mostly agree with).

Of course, there is also a growing trend against scientific hubris (and against "experts" in general) because of highly visible expert failures in the fields of nutrition and economics, for example. I don't so much think this is anti-science in general (people still love the fruits of science, after all), but anti-arrogance.

"Furthermore", "novel", "distinct" are scientific jargon that reduce readability? Seems like a non-problem to me.

Usually it is the omissions that hurt. Rarely bad sentence structure.

Harder jargon sometimes does so too and especially custom underdefined notation.

I'm building a site (not launched yet) to try to help with this problem: https://www.wikipaper.org

It's not just readability that's a problem, it's also that relevant code, data, etc are scattered all over the internet.

Another problem is that many people struggle with english which is the defacto language for research. This introduces two problems: first, one must learn to speak english, then, one must learn how to trawl through academicese.

There have been various similar projects to wikipaper in the past, but one of the reasons that they fail is that not many researchers have the spare time to contribute to a project like this. I spoke to the guys running Google Scholar and they told me this is the biggest problem. I believe I have an innovative way to solve this problem however, anyone who wants more details please get in touch.

I believe that making it easier to understand research going on in neighboring fields will dramatically increase the rate at which research is performed.

The concept of Research Distillation could be a solution to make the texts more accesible to scientists.

https://distill.pub/2017/research-debt/ via https://news.ycombinator.com/item?id=13932806

If you consider that current areas of research is both broader in overall and narrower/deeper for a given publication, it follows naturally that the language of the text targets a smaller audience. Is it conceivable that we can maintain general comprehensibility and continue to expand and refine knowledge?

I could see that it's possible to write in a style with more analogies or illustrative language but I can't tell from the article if this factors into the observed trend.

There's strong selection for obfuscation. If you make your research easy to understand it's easier for peer reviewers to poke holes in it. If you make it difficult to understand then peer reviewer don't want to look dumb by asking too many questions. Just make sure you sound smart.

We need a neural network to translate from a scientific to normal writing style. Only half-joking here... It can train itself on latest Deep Learning papers too!

With what source material for the normal writing style? If it trains based on news coverage all it would do is pick phrases from the abstract and maybe use a thesaurus, with no accuracy guarantees...

As is the reproducibility. Mere correlation or is there a causal link?

http://www.bbc.com/news/science-environment-39054778

It looks like they only looked at abstracts. I'd argue that they should be more complex and that abstract readability don't correlate with the readability in the text.

You can communicate the big ideas and importance in the necessary context in the abstract. These are built on top of many small simple ones. Back in the day 'We measured this thing' was plenty for a journal paper because there was so much disagreement that each result was novel and interesting.

Now I still believe that the quality of writing has decreased, but my point is that you can have a complicated abstract for an idea that is extremely clear in the text.

> It looks like they only looked at abstracts.

From the text: "We then validated abstract readability against full text readability, demonstrating that it is a suitable approximation for comparing main texts."

And they perpetuate bad science by not saying how they validated it...

(Either they just ran the algorithm on some full papers, which is just as worthless, or did "by inspection", which requires additional qualification and exact results they assigned.)

Are readability scores generally a good metric to follow when authoring content? For example, would you pay attention to it for a landing page or a blog post?

Discussions of the relative merit of academia's current state, and specifically the publication process, pop up from time to time around here, so I won't rehash those points here.

Additionally, I think the article's author (and other neighboring posts here) bring up valid points regarding the escalating complexity of science and potential correlations between that complexity and the written complexity required to communicate it. I think the article about the ABC conjecture [0] posted earlier today [1] is a perfect example of this.

However, I would like to pose another suggestion that may play a role in this effect.

It is easy to see how a paper's acceptance in a journal or conference serves as an evolutionary pressure on the author's style; in other words, one of the reward functions for a paper's style is defined by its ability to be published (since higher publication count correlates with higher funding availability, for better or for worse).

With such a function in place, it makes sense that papers will start to exhibit evolutionary traits (styles) that promote survival irrespective of their practical or functional benefits. Let us also consider the committee review process as part of our environment: several humans must decide whether your paper will be published or not, based on its domain novelty. There are 4 possible outcomes:

1) Paper is novel, reviewers understand it; outcome, publication (weight=1).

2) Paper is novel, reviewers don't understand it; outcome, possible publication (weight=0.5).

3) Paper is not novel, reviewers understand it; outcome, no publication (weight=0).

4) Paper is not novel, reviewers don't understand it; outcome, possible publication (weight=0.5).

Therefore, if you're publishing something, and either (A) you know it's not very novel, or (B) you're not sure how novel other people will think it will be, it's in your interest to obfuscate your paper as much as possible.

Additionally, for Cases 2 & 4, the weights probably trend even higher. Human vanity may produce an outcome closer to "I don't understand it, therefore it may be over my head; I will therefore convince myself it is a good paper. Weak recommend!" at a higher rate than "I don't understand it; I will ask for clarification from the author or the rest of the committee, at the risk of appearing foolish in front of my colleagues."

If these interpretations are true, then the parent's article's results are not particularly surprising, just depressing (from the perspective of "academia as human progress engine").

[0]: https://en.wikipedia.org/wiki/Abc_conjecture [1]: https://news.ycombinator.com/item?id=15206540

The simplest explanation is that we've built up large bodies of specialist knowledge that can't just be reduced to "Rock falls down, gravity. huh."

Take biology, and look at this depiction of metabolic pathways: https://img.scoop.it/B6Bp6lXGshGZEPRnL8_P4YXXXL4j3HpexhjNOf_...

How do you bring that complexity down so that someone coming out of highschool can avoid having to learn a lot about it first, and still transmit something useful to other specialists? Otherwise, what we'd get is another episode of NOVA

I have been told that a good paper is simple and concise but the best paper passes the "Norwegian Test" whereby we suppose that if you were handed only the abstract and plots in your paper you could understand the entire development presented within the paper.

The complexities reality contain do not necessitate the complexity of the details or developments as presented by a single paper.

I've always favored summary books assembled from smaller papers as opposed to large and extremely complicated papers.

Survey papers are worth mentioning. They are somewhere between original research papers and textbooks. For these, readability does still seem to be important, so if there is no textbook, you'll want to start with a survey paper.

Give a big picture overview, categorize similar things, focus (omit unnecessary detail), mention similarities to known systems, don't assume highly specialized knowledge just to save a little space - the usual techniques to explain things!

Edit: If you look at a text, you can easily see whether or not the author can or wants to make themselves clear. Papers from after roughly 1980 just don't give that impression anymore. You also know the difference between good and bad software documentation when you see it, if that is closer to your area of expertise. I think your explanation is plausible, but wrong.

That is fine for reviews. I 'm not sure that is appropriate or even possible in empirical reports.

Reply on news.ycombinator.com