I wrote down my perceptions of academia as someone on the outside. https://www.todepond.com/wikiblogarden/academia/from/the-outside/

Disagreeing partially... First, my background: 1. I am an academic. 2. I am not in CS, but in physics and chemistry. 3. I have been in academia for 30 years, and found it a good place for me back when I joined, but in today's conditions, I'd probably choose a different career. Many academics, myself included, mostly agree with your criticism. A good part of them, myself included, is working towards change in academia. But changing a system from the inside also requires surviving in the system, and that requires unpleasant compromises. My partial disagreement is about "put it in the bin". In spite of academia's faults, I think the world is still better off with it than without it. Getting rid of academia would leave a vacuum that would rapidly be filled by profit-oriented corporations. Which is why I prefer reforming academia from inside, even if that it a slow process.

Getting rid of academia wouldn't necessarily get rid of the academics, they'd just do scholarly work a different way, outside of what are often toxic, aggressively hierarchical and disfunctional institutions that many academics readily complain are terrible places to do research. (Maybe worth pointing out that @jonathoda's original blog post was also anti-academic)

I was just kidding about putting academia in the bin, but that's definitely my gut reaction. I hope my perspective is useful to anyone trying to make academia better

I do love writing papers, and despite my snark on mastodon go to and organise conferences too.. Although I do worry about the environmental impact, and much prefer conferences that are either properly online (i.e. with dedicated and experienced team taking care of a/v camerawork and streaming in any in-person gathering), or more like festivals with hands-on workshops and public events with decent soundsystems.

some related experiences: I want to share an idea, onward is probably a good place, the fact that I have to dress it as a paper (because I'm bad at writing papers and I will be just trying to emulate a style) is the most discouraging part of it. From the other side, most papers I read are sharing an idea, I have to spend a lot of work to extract the core idea from the paper format sometimes deciding if following the formalism is worth or not and then most of the time having to learn an ad-hoc notation only once to parse the formalism. Things that would make it much easier to "grok" like examples, demos, videos, actual runnable code in a repo instead of pseudo-code are almost always missing.

My snark: https://post.lurk.org/@yaxu/112030885917399237

With the format in particular, I hate to read two columns, the repetition of the abstract in the introduction, the paragraph that is just an index of the sections but written in prose. PDFs that make it hard to copy and paste text, page length forcing people to codegolf stuff. Almost no color, no interactive stuff, no syntax highlighting. When reading the references having to visually parse the lines to tell appart the title of the paper from the authors and the conference (also I would much prefer a table I can search, filter and get clicable links)

@Mariano Guerra I really like writing papers because it helps me develop an idea. If I already know what I'm going to write it's a less interesting process for sure. To be brutally honest about myself, I don't spend a lot of time reading papers.

In some areas the "we tested it on users" and the users are 5 people from the lab, students from the class or students from the university and it never convinces me that the user test provides any extra value to anyone

Heh yes there's a lot of emphasis on the need for 'evaluation', where CS people cosplay as experimental psychologists with none of the training or reproducibility

Ink & Switch speaks with an ‘academ-ish’ voice. Its papers aren’t actually papers. They have too much branding and too much character. I like them a lot!

They’re still very long and slow and considered. But I dunno.

Hey reader, what do you think? Do you think this academ-ish approach is good? Yes or no? And why? Let me know. But first,

• Is the academ-ish approach good? • In that I think academic papers follow a form more concerned with asserting authority and gatekeeping than they are with conveying information, I think the academ-ish approach wears the same problematic coat as the approach taken from the academy, but it requires you to spin the wool yourself.

I think that was partly @jonathoda's point in his blog post.

For me, the best part of submitting a demo for LIVE at SPLASH was the feedback I received from reviewers. It was incredibly helpful! the submission wasnt a paper though - I just sent in some videos lol. at the event, some people told me they'd love to see me write a paper about it because they want to see more detail about the nitty-gritty of the demo. "if your only tool is a hammer" and all that. i think that papers can be great! the only claim I'm challenging around them is if they're the "only way" of having a long term effect. again I'm making these challenges as a form of feedback, to try to help people make academia better

1. I had some misgivings as well when I read that blog post, but trying to articulate them felt like setting out to sweep a desert. Thank you for pitching base camp! My bottom-line decision was, if you want to persuade academics, Jonathan may be right. But that doesn't cover 10% of my desired target audience, so I didn't bother to think about it too closely. 2. There's a lot of status tied up in there, isn't there? Makes me wonder if going to an academic conference forty years hence might be like going to the opera today. What would the equivalent of Queen in the 60s be then? I think part of it will be, "send us stuff you've already published. Wherever, whatever format. We're just picking out an agenda to discuss in a compressed synchronous setting." Maybe call for attendees, then call for papers? 3. The format of a paper is excellent for a thorough exegesis of something. The process of writing it is 90% of the value of submitting to a conference for me. It forces me to think through things carefully (and no, I don't have to follow a LaTeX template). However, a) there are a lot of papers that shouldn't be papers just like there are a lot of books that shouldn't be books. There's some subjectivity here, but I suspect also some deeper thing a lot of people can agree about (if we were all willing to be so unkind out loud). Also, b) I find I can't force the paper writing process. I somehow got a PhD with a single workshop paper and no end of angst about the fact. Now it's a relief to be able to decide the time to write a paper. When the project needs it, rather than my career or some funding agency's schedule. Most of the time I don't know what I'm doing thoroughly enough to be able to write a paper about it. Perhaps trying to do so will advance things, but I think it has just as much chance of kneecapping them. Oh, and c) these thorough papers are explicitly discouraged at Onward, which is unfortunately coupling life cycle phase with audience. Wait, why is there so much sand around here. Dammit, I managed to get lost in the desert again.

here u all go, submissions are open, submit anything you want https://www.todepond.com/gallery/

I think "no ideas" excludes me 😂 with my chronic condition of over-thinking things..

merging/connecting ideas is ok, just no new ones

Ha, that's a no-op! There are no new ideas, only connections between ideas.

hope this helps

ideas

None of my new ideas can be expressed in writing. I have to clothe them in old ideas, coats atop sweaters atop longsleeves atop tees atop tanks atop lacy whatnots atop bandages atop coats again. Once the idea is dressed to survive the elements—if it's even fair to consider this sort of intellectual audience _elemental—_the idea overheats and dies.

Ironic that I’m cast as the Establishment Man here

Could you share any examples @jonathoda of difficult or unpopular ideas that survived or swayed minds because of a paper at the right time?

A paper in Onward! gets an audience that would not happen for a blog post if you are not JE. Most of us are not. I probably would not have seen LW’s presentation if it was blog post. We need both. And podcasts, wikis, microblogs, chat servers and whatever wonderful new ways of helping humans communicate. Humans do more and better as a group. We don’t do well alone.

@jonathoda Yes I think there's some misreading of your blog post! But still, Onward! looks a lot like an academic conference and I think there are a lot of other possibilities for structuring this kind of event that are a lot more accessible to a much wider range of people.

Wow that sounds great

NIME (new interfaces for musical expression) is a really awesome academic conference, in my experience really well organised, thorough peer review of papers, performances and workshops, with great ethical and open access policies. https://www.nime2024.org/

Good question @Kartik Agaram. It’s not just one paper, it’s when there needs to be a series of papers over years to work out an idea. For example Functional Programming was an academic curiosity for over a decade. More recently CRDTs took a decade to get to their current popularity.

As some have reported, reviewing is a helpful aspect of submitting papers to journals and conferences. It also happens to be the most important mechanism of quality control in scientific research. Anyone setting up an academ-ish environment outside of academia should think hard about reviewing. The difficult part of reviewing is finding reviewers that are competent and willing to do the job. Journal editors spend much of their time finding reviewers for submissions. It happens that a submission is so exciting that people volunteer to review it, but that's the exception. The typical situation is journal editors contacting people they know (or strongly suspect) to be competent in the domain, and a small percentage of the contacted people agreeing. They agree out of a sense of duty, because they know that reviewing matters. And they agree because they want to remain "good citizens" of academia. Because journal editors will recognize their names when it's their turn to submit a paper. This is ultimately why academia is very much a social network, with all the bad sides that come with it, such as conservatism and lack of diversity and inclusion. In the not so distant past (up to, say, 30 years ago), you became a member in good standing of academia by cooptation. You did a PhD with a respected member of your discipline, then a postdoc with another one, and you depended on these people vouching for you to get a tenured position one day. Academia self-selected its offspring, and ended up recruiting people with very similar profiles. This has started to change, but at the price of weakening quality control and consensus formation.

part of the reason i make my posts so provocative is it brings a lot of the people out of the woodwork to tell me why they think I'm right/wrong it's a very helpful review process, and it helps me get my work watertight. i think that some people don't realise how valuable this is. i like to "try on" different viewpoints and see what happens. when i hit a nerve, that's when i know it's worth going deeper!

@Konrad Hinsen I'd disagree with a lot of that I think. The problem with peer review isn't that there isn't enough nepotism these days, it's because academics are incentivised to publish too much. Half-baked work gets published by being submitted until it's accepted, increasing the load on the conventional peer review process to its current breaking point. Also the motivation to peer review for me is not to win favour with an editor - that would be straight unethical. It's to read a paper or article, find something interesting in it and have the opportunity to feed back and improve it. I always accept peer reviews if I can, as long as I feel qualified to review and the publication is fully open access (preferably diamond/platinum). These are still in the minority, though. In my opinion no-one should do volunteer work for extractive publishers like e.g. elselvier. A lot of academics have rightly pledged not to do that kind of thing. I'm also interested in the open peer review process, where reviews are also published. I've not explored that yet though. Academia has indeed literally self-selected its offspring. I've been surprised at how many academics hold soft scientific racist views about how the children of professors are more likely to become academics, as though DNA was the only force at play.. In my view the high percentage of privately educated people in UK academia undermines most of its aims.

@Alex McLean I pretty much agree with your ideas about how reviewing in academia should work, but what I described is how it actually works, judging from my experience. Open peer review is a big step forward, in my opinion, and was a major design decision in the small journal that I cofounded (ReScience). It's not without problems though: for early-career researchers, reviewing the work of a senior researcher who may one day be an employer means taking a risk. As for nepotism, I definitely do not want to defend it. But in the early days, up to the 19th century, when science was a lot smaller and had less economic and societal impact, it was an efficient strategy with few downsides. Much like kinship-based selection is an efficient strategy for structures of tribal size. My point is that keeping academia working, and stable over decades to centuries, requires mechanisms that may well be in conflict with its missions. It's not obvious to come up with better mechanisms to replace them, so slow but steady change is probably the best approach to pursue.

Working for who? I enjoyed this podcast (via the one in the same series on Rawls that @Jack Rusher share) and essay, for early 20th century perspective on how academia hasn't worked for most people in the UK. https://www.listennotes.com/podcasts/past-present-future/history-of-ideas-4-virginia-zcmJWo72uAw/ ReScience looks interesting, thanks for the link!

there's something about 1929!

@Alex McLean Working in the sense of maintaining a process over long time spans. In opposition to crashing, dying, or any other way of ceasing to exist and function. The purpose of a system is what it does Related: the <https:/www.jerrypournelle.com/reports/jerryp/iron.html|Iron Law of Bureaucrac> explains how such systems come to exist.

@Konrad Hinsen I completely disagree with your opposition of nepotism or death. Toxic institutions can undergo fast reform in the right hands. Or if they are so bad that they are better off dying, quickly replaced by something better. Indeed I think opinions like yours would not be able to stand for a minute in a healthier and less homogenous forum

We're still talking about whether papers are a good vehicle for information delivery, right? Haven't lost sight of that?

Oh ffs Ivan

in fairness, my original post does say this :)

It’s a straight-white-male-dominated world obsessed with intellectualism. The whole ‘paper process’ saps out all the fun and feeling, devoiding your work of any emotion. “All we need is rational thought and reason” making no room for expression and empathy. No wonder it’s so cold.

Some people reading this will be thinking “What’s so bad about that?” which is… uh nevermind

so feels on topic, but also i'd like to request that all responses are now in poem form. iambic pentameter or haikus are fine

Heroic couplets seem like something HN would revel in

> part of the reason i make my posts so provocative is it brings a lot of the people out of the woodwork to tell me why they think I'm right/wrong Funny to juxtapose this with the recommendations in Ink & Switch's Academish Voice, which was released publicly after discussion around your post.

yes! and its also been fun to bring my slightly provocative style to internal ink & switch discussions too. its why im thinking a lot about it at the moment. and i'd love for ink & switch to lean more towards the second group of bullet points there. thats what makes me love their work :)

I see a lot of value in trying on extreme positions, letting an idea take you to weird places and bounce that off other people. But I'd argue that such practice requires the counterweight of demarcating a zone within which rigor reigns, and half-baked ideas are not welcome. We need both.

i'd encourage you to try out a more provocative style yourself to challenge your assumptions about what it can and cant be! happy to help you explore that

I can assure you that my style is plenty provocative lol

ok cool well my offer is there :)

Not in metre, sorry: I’d also ask why it is considered provocative to state a personal position? Why stating that sort of thing (often, not always, but I think especially in technical groups) invites folks to dispute it

for example, i disagree that it equates to extreme positions and half-baked ideas

I wonder if we agree that the provocative mode conflicts with "avoid hyperbole", "make no claims without support", "avoid absolutist language"? I know it does for me when I'm in provocative, discuss-extreme-positions, explore-half-understood-ideas mode.

for me, "being provocative" means challenging the status quo and monoculture of our world. i havent thought about how it relates to those style guides. i guess you could probably be compatible with those while still being provocative in other ways? in my experience its sometimes the most basic boring and simple positions that can be the most provocative haha im not trying to convince anyone of this. again im just sharing my perception of all this because it might be valuable to some people trying to make academia/tech/etc better for people on the outside. i know thats something some people want to do

@Alex McLean > I completely disagree with your opposition of nepotism or death. That's not what I said. I said that, in the early days of science, cooptation was an efficient strategy to keep the system alive, with no major downsides. This has lead to something like nepotism today (it isn't strictly nepotism because it's not based on kinship relations), and that is something we have to change. But changing it requires first understanding how it came to be and what role it played in the past.

a self-parodying haiku

here is my feedback

about academia

put it in the bin

The authors would aim to educate rather than impress. They would not

need to be defensive about their work, and would introspect about their

learnings and their thought processes.

Some would, others wouldn't. People are driven by different motivations. The fundamental issue I see with today's submit-review-publish pipeline is that it no longer fulfills one of its traditional roles: quality control. The reasons are multiple and include (1) metric-based evaluation (2) instiutionalization of prestigious journals (i.e. they defend their existence and prestige rather than serve their original purposes of communication and quality control) and (3) increasing competition among an increasing number of researchers in an environment providing less resources. What we see is evolution in action: the survival of the fittest in terms of resource extraction, i.e. the supremacy of people who are good at convincing non-experts to give them money.

Honestly I think CS academia is a joke. Most research happens inside industry (including the open source world btw) and companies often self publish either code, their own papers, patents, or all three.

No idea about graphics personally. For CS generally, I’d there say there’s patches of good work on both sides (industry and academia), and also lots of boring stuff on both sides too.

I guess location matters too; I do see a lot of good papers out of Glasgow for some reason. But CS academia inside the US seems to have pretty much died.