Don Knuth letter about libraries increasingly unable to afford prices (2003) [pdf]
cs.stanford.eduHe is attending the wrong library. Library Genesis is the way, now. And Library Sci-Hub.
And I am only half-kidding.
All that stuff lands on LG/AA thanks to those who have subscriptions, i.e. libraries
I’m ok with library Sci-Hub, but less so with Library Genesis. Pirating books reduces the incentives for authors to write more books. Pirating academic articles won’t reduce academics’ incentives to write papers.
Considering certain technical books range between 70 to 300 dollars per copy, and libraries are unable to afford them, it seems to me that the incentive system is already broken.
The publishers are very comfortable with the high prices they are able to charge the quite expensive American universities.
In some academic fields, books are also just collections of technical papers on some topic, or one really long paper, which is probably not something to earn money from. I assume most are sold to libraries, similar to journals. Though Library Genesis also contains other, more commercial books.
Good. There no reason to rewrite undergrad calculus books every semester besides money grabbing. The math hasn't changed in 100 years and there shouldn't be an incentive to write an updated version of the book for the 50th time.
The language changes, but more importantly students backgrounds and facility with the English language changes. Not that dumbing down the text or replacing old timey words should cost much, but cost disease and all that.
Granted, but they definitely do not change enough to warrant a new edition every year. This can and has been experimentally confirmed by countless students even before Library Genesis simply by showing up to class with an old edition of the textbook and still learning the material fine.
Does it? It seems complicated. Higher availability of books means more readers for a start. Large numbers of people don't read at all and therefore aren't even in the market for books. Then you have to consider the absurd length of copyright. How does an author receiving money for something they wrote 50 years ago incentivise them to write more? Even worse, how does a dead author's estate receiving money incentivise anyone to write anything?
>How does an author receiving money for something they wrote 50 years ago incentivise them to write more?
Receiving money for something you wrote 1 week ago doesn't encourage you to write more. It's the knowledge that you will receive money for next thing you wrote that incentivizes you to write more. And a 50 year payout should obviously provide more incentive than a 6 month payout.
There's going to be a point of diminishing return where adding extra years doesn't provide enough extra incentive to justify the cost, but I haven't seen any convincing evidence that 50 years is over that threshold.
The same holds true for copyright extending past the author's lifetime. Especially if an author is older, the knowledge that their children will benefit is certainly an incentive.
I don't get this argument for writing or anything else. I do things now for which I don't expect to get paid for the next 50 years. What's my incentive?
The libraries where I live are seriously a joke anyway, and even if they were not, I doubt they could compete with a website where you can find monographs addressed to like 100 people, in seconds, in your pyjamas.
And annas-archive
Pro tip: If you are near a university, the free public wifi may give you access to its journal subscriptions.
Where have you seen this?
Every one I've seen requires login, and usually university membership for the good stuff.
Back in the day, when MIT still owned 18.0.0.0/8, at least one journal gave blanket access to users coming from anywhere in that netblock. And there were ways to use it even if you weren’t officially affiliated as a student/faculty. Unfortunately Aaron Swartz did not use an untraceable method of access.
That's honestly quite tragic as we don't evolve anywhere fast enough to not be hit by global depression as the digital solitude clashes with the innate human needs of real socialization.
You aren’t supposed to socialize at brick and mortar libraries.
That's not really true.
> Twenty-five percent who went to the library learned about political or cultural organizations or leisure activities taking place in the local community. More than 20% went to the library with friends or colleagues to work together on a common assignment or a leisure activity. Seventeen percent used the Internet at the library to contact friends via e-mail, to chat, or to participate in discussion groups, etc. Sixteen percent used the library to learn more about local matters, social or political issues, etc., that they are involved in. Fourteen percent participated in organized meetings, such as author's nights, lectures, meetings with politicians, etc. Ten percent used the library as a place to meet family or friends before going together to the movies, into the city, to do shopping, etc. (p.19)
https://www.ala.org/tools/research/librariesmatter/use-publi...
Libraries are not archives of books. They are community hubs. My local library has programs for seniors with early mental impairments to do puzzles with volunteers to slow the progression of their decline. Another has programs for street youth.
Where on Earth did you get that idea? Every library I've ever been to has offered all kinds of events geared to "socializing," from Group Crochet, to Book discussions, to AA. Not to mention how common it is for libraries to offer "teen spaces" where kids are free to hang out, work on homework together, etc.
I am not kidding at all. We have that government library app sort of thing. It makes you wait for a digital copy. Like there's a line of people before you, just like in an old school library.
Why? Just use Genesis
Not all of us are able to, or want to, use ebooks. Especially textbooks where one may be going back and forth through chapters fairly frequently. Simply saying "muh Libgen" ignores how most people , especially teenagers, study. For teenagers, any smart device will lead to distraction if connected online. They usually choose Insta over the PDF reader in their distraction.
Sending a working person a 14 page letter about anything expecting them to read it is wild to me. Perhaps it's the quality of my writing but my personal experience is that even being way more concise, most people wouldn't care.
> Sending a working person a 14 page letter about anything expecting them to read it is wild to me
His addressees were members of the editorial board. Surely the editorial board should be accustomed to reading lengthy prose?
Now, writing a 14-page letter, of a quality that matches published articles, is what's wild to me.
Well, how long do we work on some slide decks to convince management? How much we polish them aesthetically? This is a scientist working to convince a management board. So he puts effort in it and tries to make a convincing pitch.
Just a different media ;)
I guess you should take into account that you are not (one of) the most preeminent computer scientists in the world, right?
If Don Knuth sent me a 14-page letter, you bet I'm gonna read it. I might frame it.
I was fortunate enough to get one of the last physical reward checks which was accompanied by a printout of my e-mail, written on in pencil.
Fortunately, it survived a house flood which destroyed the book it was in, so now the envelope it's in is prominently displayed on a rack in my living room and would be one of the things I'd grab in the event of a fire.
Every so often, when a co-worker knows who Knuth is I bring it in to a workplace to show off.
Extremely envious of this! I don't own a copy of Art of Computer Programming, so I did spend a not-insignificant amount of reading through some of Knuth's other writings trying to find mistakes that no one else has found, and I came up empty handed. There might be mistakes ripe for the picking in there, but I'm afraid I'm not quite smart enough to find them.
Still, a guy can dream. If I ever pick up Art of Computer Programming I might give it a go again.
I was fortunate that Knuth wrote a book on a topic which I have studied deeply, _Digital Typography_:
https://www-cs-faculty.stanford.edu/~knuth/dt.html
My correction is for pg. 33, and there was a minor point of improvement which I can't find.
Very cool!
He just needs to write a book on obscure cartoon history, and I'll be shoe-in. Until then I think I will just have to deal with living in envy.
You are aware that his first publication was in _Mad Magazine_?
https://en.wikipedia.org/wiki/Potrzebie
(scroll down to "System of Measurement")
or see:
I was not aware of that. That's a fun fact that I did not expect to learn today.
I have a letter from Knuth as well as a letter from Erdos. I've been hanging onto these.
be a good person and post some pictures about both so we can refer to it someday. For science !
I would definitely frame it. I framed an email from Norvig and a note from Dijkstra. Get joy from small things.
I once emailed Dijkstra. I was 14, it was the mid 90s, and I had just connected to the real internet a few months ago. I had just come across his name as a prominent computer scientist. I wanted to be one just like him, so I asked him what I should do.
He replied. it was 3 full paragraphs. He told me to study algorithms and to learn python.
You're right. I absolutely should frame it.
I have a blog post by Norvig where he gently chides me for wasting enormous amounts of CPU trying to find a counterexample to the Beale Conjecture. Not exactly frameable, and not exactly something to be proud of, but I got a chuckle nonetheless.
If a rando from hn sent me a 14-page letter, I'd probably read it. I might give it to the police afterwards.
Do you listen longer when your best dairy cow moos?
If my best dairy cow was Don Knuth, my bones would be NP-Hard.
Please don't post kink fantasies on the hacker news forums.
I don’t know. Don’t cows moo constantly?
> Sending a working person a 14 page letter about anything expecting them to read it is wild to me.
You really cannot understand why Donald Knuth would expect the editorial board of The Journal of Algorithms to read his letter?
Editors not knowing who sent in a document is one of the pillars of the double-blind system.
https://www.combinatorics.org/ojs/index.php/eljc/article/vie...
> Herb Wilf has co-founded two major journals, Journal of Algorithms in 1980 with Donald Knuth ...
When the founder of the journal writes the editorial board (not the reviewers of papers), it is probably expected to be read.
I bet the editors read it not once, but several times. And I bet it was even noticed years later and posted on websites where many others read it, were impressed, and commented on it. Don Knuth’s letters are special.
It's not a real letter. It's an (implicit?) open letter that is just a self published article, with the hidden threat that everyone will see it [1] so if they don't agree they will look bad.
[1] And nobody will read it, and everybody agree even the publisher. The publisher want to increase the price anyway, and everyone else want a cheaper journal.
yeah, you didn't rode the PS :)
Nope :) . I read it now. It's unrealistic. It will leak. Perhaps also he expected someone to read the 14 pages. Let's say it's a not very closed letter :) .
I think times were a little different then; when I had a company begin 2000s, we used to send long emails with details before and after meetings, and there had less and more productive meetings because everyone read these emails/docs. Now I try to send only 1 liners, because people tend to literally only read the first line and then ram the reply button and blurb some studied remark like 'ok'. And then these same folks drag out meetings asking, in a pathetic show of laziness, to 'go over the email'. If there is anything I find cringe, it's 'let's go over the email together'. It's just saying; 'I couldn't be arsed to read it and I want to waste everyone's time'.
ChatGPT was imitating us, after all
Some people think that with quantity they can paint over a lack of substance, others think writing more looks as if they did more work, yet others don't seem to think.. at all.
Bottom line up. Ususally what you need the other side to grok can be summarized in two sentences. Start with that. Give details after. If you have multiple topics seperate them into sections or make a list.
Any message is something the receiver needs to decode into actionable information. Often the receivers don't have any idea where your mind is, so you first need to being them there. The worst kind of message is one where you need to read all the way to the end to even figure out what the heck it is about and then read it again just to get what they want.
I think this might be true if the only goal is to convey information. Here it is more about convincing others and framing information. If you start with the condensed facts, readers who disagree might be put off already.
"I love my library and the other libraries that I visit frequently, and my blood boils when I see a library being overcharged."
Given the topic and my love for Knuth, I went into this paper ready to agree with him. But Knuth does a great job at stating his case.
This sentence caught my eye: "Elsevier, however, ignored my letter and did not reply" - who in their right mind would ignore a letter from Knuth?!
Because people keep throwing huge amounts of money and content to Elsevier while asking them to stop making money.
This isn't a defense of Elsevier by the way. The scholarly system is abysmal for publications and it's seemingly incapable of any meaningful change over non-geological timeframes.
But if you keep paying people to do something then they're going to keep doing it. If they stop, someone else will appear if that's the kind of thing you're funding.
What's insane to me is the biggest complaints come from three main groups - scientists, libraries/the universities that fund them, and funders. The content producers keep giving Elsevier content, the libraries keep buying it and the funders keep paying the content producers to give their content to Elsevier. Universities keep demanding academics give their content and their time for free to these journals else they don't get the progression they want.
Elsevier is a nasty symptom but a symptom non-the-less of this dysfunction. Those groups can absolutely change how things are done but the field moves glacially.
> Because people keep throwing huge amounts of money and content to Elsevier while asking them to stop making money.
Because the broken rest of the system (i.e. financials tied to "how many papers did you get published") incentivizes everyone to keep the status quo.
The entire academia publishing clusterfuck needs massive government intervention to dismantle.
For Elsevier, Knuth was just an editor of a Journal at peak. He is not a legend for them just one of hundreds and thousands of legends they manage.
Like a (legendary) middle manager which is in the company since the start and built it up but does not agree with the current business goals of the company. We know what happens with these managers.
They don't talk to anyone unless many libraries stop paying [0]. I wonder however if those open access deals will mean the death blow to printed copies and classical libraries.
I love the veiled threat in Page 4 about Journal of Logic Programming all editors abandoning Elsevier and starting a new journal (Theory and Practice of Logic Programming (TPLP)), noting the TPLP thrived after this and Elsevier's own restart having gone off the map. Especially with the balancing note at the end saying the cost per page of this new journal wasn't much cheaper.
The story goes that one of Niels Bohr's friends visited him and found him deeply engrossed in writing an application to a fund. Surprised, the friend asked why it took so long for such a prominent scientist as Bohr to write a simple application. Bohr replied, "I'm trying to make it short, but I haven't had time yet."
This is good research but terrible writing. Why are the instances so out of order? I understand starting with the first English occurrence, but why not go from there to the oldest known occurrence in any language?
And the outcome was ... ACM Transactions on Algorithms (TALG)[1][2]
[1] https://home.cs.colorado.edu/~hal/s.pdf
[2] https://en.wikipedia.org/wiki/ACM_Transactions_on_AlgorithmsYup:
> In 2003, the entire editorial board of the Journal of Algorithms resigned to start ACM Transactions on Algorithms with a different, lower-priced, not-for-profit publisher, at the suggestion of Journal of Algorithms founder Donald Knuth. The Journal of Algorithms continued under Elsevier with a new editorial board until October 2009, when it was discontinued.
https://en.wikipedia.org/wiki/Elsevier#Resignation_of_editor...
This sometimes works and sometimes it doesn't. A similar thing happened with Machine Learning (Springer) and for the same reason: the entire editorial board abandoned it and founded the Journal of Machine Learning Research (https://jmlr.org/). But Machine Learning recovered and is still a dominant factor -- and still charging high fees.
It's interesting to note that in the context of this letter about the Journal of Algorithms, that there is now an open-access journal called "Algorithms" which looks like it launched five years after this letter:
Take a look at the length of the "controversies" part of that publisher's wikipedia article.
https://en.wikipedia.org/wiki/MDPI#Evaluation_and_controvers...
With a different set of problems imo. MDPI is a for-profit publisher with a "pay-to-publish" model, charging authors $1800 per accepted paper. And also a mixed reputation on quality, verging on a "paper mill" where they accept anything just to maximize their publication fees. Not a great model imo. There are non-profit Platinum OA journals that don't charge publication fees, like JMLR (https://www.jmlr.org/), which is the only kind of OA I consider interesting.
Not a rhetorical question. Are they more affordable now in 2024 ?
How much does the existence of open access journals affect the affordability overall ?
No, the price of ebooks is insane and getting worse. Libraries are somewhat of a captive market for publishers and they set insane costs & limit the number of times an ebook can be checked out before the library has to buy it again. IMO the cost of electronic materials to libraries is one of the biggest issues in our society that no one talks or knows about.
I never understood the overpricing of ebooks...
A price of a book is a price of the "data" inside (copyright) + price of paper, printing, binding, packing material and distribution
A price of an ebook is a price od the "data" inside (copyright) and the price of cheaper digital distributon.
So, you save on paper, boxes, trucks, you often also save on retailers (if you sell directly), and you want more money for that? No paper for me, and you also take resale options away.. and for more money?
Just the resaleability is a scam... if you bought it, you own it. If buying isn't owning (including reselling, reading on any device, etc.), then piracy isn't stealing.
I'm sure government regulation could solve both problems, but they're more interested in screwing the "normal people" instead.
What's there to understand? It's capitalism 101. Producers will charge the maximum the market will bear.
Sure, but who the hell buys a digital copy they can't even resell for more money than a physical copy?
I usually prefer digital books for a lot of reasons, they're convenient to bring with me places & I can change my mind about what I will read; I don't get wrist pain from holding the book open; I can read in the dark with just the ereader's backlight on; I have control over font & font size, so I don't have to get used to differences every book I read; there's no chance of papercuts; you can use ctrl+F.
The downsides of ebooks are that I can't write in margins (if it's nonfiction, I would never write in a fiction book), and the ereader chassis conducts heat better than paper so it's extremely cold at first in the winter. Also footnotes are annoying as shit compared to footnotes in paper media, and tabbing back to a known earlier page (e.g. a map for fantasy novels) is worse UX than in a physical book.
Cost aside, ebooks win almost always, although sometimes I'll buy a nice edition of a book I love to decorate my bookshelf, and when I read paper books I enjoy the things they're better at than ebooks.
(And of course there's audiobooks, where physical media makes little sense)
I don't want to have to deal with reselling books... Plus I'm a hoarder so I really don't want to tempt fate and accumulate physical things I don't need.
Yep. Basically, people value ebooks higher.
Sadly this is true for me. Much as I wish ebooks had the freedoms that come with paper, I value reduced physical clutter more than the ability to share and resell.
It's hard to compare as the business model has changed, I believe that nowadays it's basically impossible for a library to get an individual subscription to a individual journal: subscriptions are bundled and (online) subscriptions are sold institution wide with contracts running in the millions. That's very different from what Kunth is describing, where individual libraries choosing what subscriptions they need (eg we need the JoA in our (physical) collection so we buy a subscription to that).
Aaron Swartz articulated the problems with gatekeeping of knowledge.
Information is power. But like all power, there are those who want to keep it
for themselves. The world's entire scientific and cultural heritage, published
over centuries in books and journals, is increasingly being digitized and locked
up by a handful of private corporations. Want to read the papers featuring the
most famous results of the sciences? You'll need to send enormous amounts to
publishers like Reed Elsevier.
There are those struggling to change this. The Open Access Movement has fought
valiantly to ensure that scientists do not sign their copyrights away but
instead ensure their work is published on the Internet, under terms that allow
anyone to access it. But even under the best scenarios, their work will only
apply to things published in the future. Everything up until now will have been
lost.
That is too high a price to pay. Forcing academics to pay money to read the work
of their colleagues? Scanning entire libraries but only allowing the folks at
Google to read them? Providing scientific articles to those at elite
universities in the First World, but not to children in the Global South? It's
outrageous and unacceptable.
"I agree," many say, "but what can we do? The companies hold the copyrights,
they make enormous amounts of money by charging for access, and it's perfectly
legal — there's nothing we can do to stop them." But there is something we can,
something that's already being done: we can fight back.
Those with access to these resources — students, librarians, scientists — you
have been given a privilege. You get to feed at this banquet of knowledge while
the rest of the world is locked out. But you need not — indeed, morally, you
cannot — keep this privilege for yourselves. You have a duty to share it with
the world. And you have: trading passwords with colleagues, filling download
requests for friends.
Meanwhile, those who have been locked out are not standing idly by. You have
been sneaking through holes and climbing over fences, liberating the information
locked up by the publishers and sharing them with your friends.
But all of this action goes on in the dark, hidden underground. It's called
stealing or piracy, as if sharing a wealth of knowledge were the moral
equivalent of plundering a ship and murdering its crew. But sharing isn't
immoral — it's a moral imperative. Only those blinded by greed would refuse to
let a friend make a copy.
Large corporations, of course, are blinded by greed. The laws under which they
operate require it — their shareholders would revolt at anything less. And the
politicians they have bought off back them, passing laws giving them the
exclusive power to decide who can make copies.
There is no justice in following unjust laws. It's time to come into the light
and, in the grand tradition of civil disobedience, declare our opposition to
this private theft of public culture.
We need to take information, wherever it is stored, make our copies and share
them with the world. We need to take stuff that's out of copyright and add it to
the archive. We need to buy secret databases and put them on the Web. We need to
download scientific journals and upload them to file sharing networks. We need
to fight for Guerilla Open Access.
With enough of us, around the world, we'll not just send a strong message
opposing the privatization of knowledge — we'll make it a thing of the past.
Will you join us?
Aaron Swartz
July 2008, Eremo, Italy
> P. S. Im sending copies of this letter to several friends who are interested in journal publishing but are not members of our board. But this is not an "open letter"; I would prefer not to have my remarks circulated widely. I'm emphatically not a revolutionary. I just want our journal to do the right thing.
:/
EDIT: as others have pointed out the letter is now hosted publicly by Knuth himself.
This was his thinking in 2003. Nowadays he is publishing this in his own website at Stanford (https://cs.stanford.edu/~knuth)
I think GP is pointing out Knuth did not want this letter to become public knowledge, yet here we all are reading it.
We're reading it from Knuth's website, where he shared it.
Ah yep I misread the post I’m replying to, my bad.
(2003) (@dang too)
I am only popping in to say: "wow! what a fantastic typesetting!" (but I shouldn't be surprised)
it's actually quite basic, when it comes to what (La)TeX can do
Can someone return this mans videos? https://cs.stanford.edu/~knuth/news03.html#videos
It's taking too long. At least 20 years now
I'm curious if anyone subscribes to journals they would recommend today?
Besides arxiv which is free, the last time I looked at papers properly was when I was a graduate student.
so, of course I had free access to virtually everything.
It’s not really cost effective to subscribe to an individual journal and the topics across an entire journal are not as focused as you might think.
there’s a lot of irrelevant stuff published compared to your specific research (the purpose of these papers is to publish very specific articles and you read them to help you with your own very specific research.)
Alls to say it’s an odd question because an individual wouldn’t really subscribe to a specific journal, like you would with a magazine or periodical.
Well, there's _TUGboat_, the journal of the TeX User's Group --- free with a membership (or membership is free w/ a subscription), but it's rather specialized, and a non-profit thing.
say the person with the most expensive book for sale.
There is a common irony of authors selling overpriced books complaining about things being too expensive. And I wish it applied to Knuth.
Unfortunately, Knuth books are not the most expensive, by far. Sure, TAOCP is about $300, which is expensive, but it is not a single book, and it totals to about 3000 pages, which is 10 cents per page. It is actually below average for textbooks.
In his letter, he complains about articles selling at 50 cents per page, so 5 times more than his "most expensive" books. So for me, it is true that Knuth is "part of the system", and also that buying TAOCP is probably a waste of your money (because there is >90% chance you will never read it...), but here, he has a point. So much that today, many don't even bother paying for articles at all, if it is not open access and it is not on sci-hub, it gets ignored.
Some rants from the foxhole.
American Chemical Society is the one of the main publishers in molecular sciences. Researchers at Finnish universities haven't been able to access articles published after 2023 after failed negotiations [0], greatly hindering one's - and collectively the nation's - ability to progress in these fields. It's quite frustrating, shocking, and eye-opening to have this rug pulled beneath you.
Finland is not a poor country, and the situation is surely worse elsewhere. Nonetheless, our economy and the academic funding situation is quite crappy and getting crappier. In 2022, Finnish university library consortium spent ~25M€ for subscriptions [3]. Last year, the negotiated sum for seven main publishers was ~16M€, inc the failed ACS deal. One can easily imagine better ways to use the dozens of millions.
Science is expensive and inequalities between countries/uni's/wherever are a n unfortunate fact of the world. Not every player can pay millions to get the 10M€ Cryo-EM machine, and thus can't compete in advancing knowledge frontier in this.
To some extent, constraints cultivate creativity. One can still participate through collaboration, theoretical and computational work, creative crafting of experiments with already existing equipment (& with fascinating DIY low-cost open-science hardwarex stuff!)
However, one must know the giants on whose shoulders one stands on, and the game played by the behemoth publishers attacks this fundament. The consequence - inequality in accessing knowledge is deeply disgusting in its artificiality.
Meanwhile, people at eg. MIT are able to get the whole ACS corpus in sweet delicious machine readable XML [3]. In the same time it takes for the "poor" researcher to get one email requested watermarked pdf with detached figures that they excitedly share to their group, a Boston grad student can curl terabytes and science-of-science/NLP/RAG the shit out of it.
Gap exists and grows with the arbitrarily increasing costs. Something needs to change, but for now, I'm cynical. Strong will get stronger and so on.
Thank god for open science movement living on github and *rxivs, and for the risky work taken on by shadow librarians.
[0] https://finelib.fi/sopimus-acsn-kanssa-paattynyt/
[1] https://www.kiwi.fi/display/finelib/Vuosikertomukset
[2] https://finelib.fi/kustannukset-saatava-kuriin-tiedelehtien-...
[3] https://libraries.mit.edu/scholarly/publishing/text-and-data...
I've recently resumed research I initially started with Indiana University (and stopped for life reasons). I've been out of academia since 1995.
Back then, I had the resources of Indiana University for support. Today, lacking institutional affiliation, I couldn't do research at all - the typical price per paper is $35 from publishers, which is on average something like $2 per page. And why not? They don't have competitors for any given paper, and papers are not fungible.
I'm not independently (or even dependently) wealthy, so I can only do this research using the shadow archives. And the irony of the situation is that the access I get through shadow archives is far, far better than what I had at Indiana University in 1995, where I had to physically go to the library and hope nobody had checked out their copy of a given book, and where I paid 10 cents a page for Xerox copies. (I recently went through my files and found every single article there in PDF online - and burned all the 10-cent pages I'd copied in the 90's. It was ... weird.)
Knuth's right (well, of course; he's Don Knuth, what did I expect?) - the commercial academic publishing industry is holding us all back.
this letter is from 2003.