As a PhD student it's embarrassing, but I still find journal articles intimidating. Something to overcome, surely. But the more I read, the angrier I get, because it shouldn't be like this.
Like, I'm a smart guy. I actually understand the topics I'm studying pretty well already. I learned by hands-on experience, pop sci books / articles, and thinking. But when I encounter familiar ideas in a journal, it's like they're written in code! The ideas aren't hard for me, they're just obscured by arcane jargon and math.
This style of writing is both lazy and disrespectful. Technical language is often necessary in science, but it comes with a responsibility to help the reader understand.
I realize scientists mostly write for each other, but where do you think new scientists come from? If we don't make an effort to invite people into our world, then we're keeping them out, making science an elitist domain where only the privileged few may enter. That's deeply harmful to society, and to science!
@LizEllisPhD@mastodon.social@ngaylinn As a post-doc I find the same. it's about maintaining that mystery of elitism imo. If you can't explain an idea succinctly and without loads of jargo, then you shouldn't be in knowledge creation or exchange.
@LizEllisPhD @ngaylinn It might be elitism, yes. But there is something else: Writing about science is hard.
This is because to learn, say, a mathematical framework well enough to use it for research, you need to reach a viewpoint from which it looks obvious or even trivial. And then, to write about it, you need to switch this standpoint off and again view the stuff you have found with the eyes of a beginner. And this is difficult because you may already have forgotten the beginner's viewpoint.
Referees help in this process a bit. The ideal referee is therefore not an expert in the paper but someone who is a bit at a distance from it.
If I am right, then badly readable papers result from the joint effort of an author who has forgotten the beginner's mind and referees who know too much. 
@mrdk @LizEllisPhD That's very astute, and I couldn't agree more. It also helps me empathize with the authors, which is important. They are people like me, who on top of all this are also busy and tired and worried about other things.
I deeply understand that explaining complicated, technical things clearly to a novice (and even remembering to do so!) is quite hard, because I spent 8 years as a manager on-boarding new software engineers. I know writing that's clear, respectful, and useful to all skill levels is hard, because I've been passionate about code readability and documentation for more than 20 years.
This is hard work! It's a skill to be honed. And in my experience, it's invaluable. It also gets easier the more you practice it.
Loving this thread
I'm having a similar/ related conversation on complexity with @axoaxonic here (https://mastodon.online/@themanual4am/111728382309126932)
I wonder if a solution is to formalise a means of 'separating concerns' (like refactoring code to reusable libraries), but for conceptual constituents (or embeddings?): inevitably plural generalisations, metaphors, commonalities, etc
To reduce the volume of 'purely domain specific expression'
@themanual4am @mrdk @LizEllisPhD @axoaxonic In my mind, the value of jargon and formal languages in science is precision. A scientist does need to say precisely what they mean, and common language won't always do. There are also times when having something in formal notation is useful, for comparing to a reference or entering into a computer program or something.
However, there are other times when they do not provide that value, and only serve to obfuscate. I would also argue that when they are necessary, they would always benefit from a brief explanation, metaphor, or diagram to make them more intuitive. Even to an expert.
@ngaylinn @themanual4am @mrdk @LizEllisPhD I'm wondering what causes the latter, like if people feel like they have to make their writing dense in order to be published or taken seriously. There are also people who really don't know how to write any other way, I'm sure, which could be helped by more community feedback, editors, and teaching clear and natural writing in prerequisite courses earlier on.
There are some examples of people who manage to use high precision, jargon-filled language while also writing in an accessible, real way, like in Piper Harron's amazing thesis https://www.theliberatedmathematician.com/wp-content/uploads/2015/11/PiperThesisPostPrint.pdf
@axoaxonic @themanual4am @mrdk @LizEllisPhD I certainly think prioritizing writing skill building would help a lot! I'm new here, though, so I wonder, too.
@axoaxonic @ngaylinn @mrdk @LizEllisPhD
Distinct from my above point (which is more to do with resolving knowledge – map to territory), there is a secondary cause of that style of communication, ASD – though I can be more specific
Some of us create isolated interpretive contexts for interpreting/ modelling specialist (often technical) concerns (note that the default interpretive context is the self)
Initially, these specialist technical contexts don't contain *any* language
\
@axoaxonic @ngaylinn @mrdk @LizEllisPhD
We can compose highly precise (constrained) stimulable representation directly (world building), which we can't initially describe (language is lossy serialisation of arbitrary-dimensional graph)
I say more here: https://neurodifferent.me/@themanual4am/111133442195724875
Initially, each statement/ sentence requires a distinct simulation run, to serialise, as if describing a walk through a physical space
\
@axoaxonic @ngaylinn @mrdk @LizEllisPhD
Domain-specific context is the simulated terrain of the specialist graph, so the first communication pass will be domain-specific (as this is closer to the graph)
Initial conversations are with other specialists; and any pre-serialised statements are available without simulation, so available more generally
Note new statements require simulation (a quiet space to disassociate from meat space and use those faculties as if in that specialist terrain)
\
@axoaxonic @ngaylinn @mrdk @LizEllisPhD
All other language terms are unavailable, and must be integrated (associated, by split context described in above link)
Crucially, one can't always just link in simple terms; often one must simulate and literally refactor the graph to separate concepts into constituent structural/ behavioural fragments, which then match simple terms
Those fragments may exist; but might need to be generalised (like matching/ tuning embeddings)
\
@axoaxonic @ngaylinn @mrdk @LizEllisPhD
This all takes time: to be persisted, embedding work is an offline edit; edits consumes resources, which must be replenished (via cerebrospinal fluid)
Offline edits require sleep, or sometimes simply shifting back to default context (or context less) will do (resources permitting)
Increasing heart-rate replenishes resources (assuming body availability)
At times it feels like one simple term a day! (but that feels like epiphany, so is v rewarding)
@axoaxonic @ngaylinn @mrdk @LizEllisPhD
Lastly,
Complexity still exists! the terrain is still there, just more explicit (like well refactored code, relative to 'big ball of mud')
And these refactoring are high-dimensional too: the same generalisations/ metaphors might appear all over; across plural distinct aspects/ perspectives of the graph
Each simplified term needs domain context to be added explicitly, as necessary (like wiring up interfaces)
\
@axoaxonic @ngaylinn @mrdk @LizEllisPhD
Meta:
This interplay, between complex and simple terms, is reminding me of more is different
Individually simplified terms lack context, and plural simplified aspects are essentially other domain constituents
But (like more is different) when we *communicate* only by simplified terms, what is lost is structural relation between related constituents (which is part of what must be added explicitly) – the *why and when* each detail is relevant
See it?
@themanual4am @ngaylinn @mrdk @LizEllisPhD
I'm hyperlexic and a lot of the experiences I have with insufficient speech are because I can't find ways to express my thoughts in ways that wouldn't confuse the neurotypical people. But I want to be able to communicate with anyone, so I translate or mask as much as I can, but it's hard to do that in real time. Often I just say nothing or give a minimal response
With writing, if I want to reach a wider audience with diverse neuro types and styles of parsing info, I can edit, listen to my own drafts with a TTS program etc. It's still very difficult to even imagine how someone's going to parse my thoughts though, let alone figure out how to accurately package those thoughts in accessible language. Writing classes can teach nice ways of writing, but overcoming the "double empathy problem" in sci com would be a whole different challenge.
I think the majority of inaccessible academic writing, though, comes from weird old traditions in academia, lots of authors emulate and repeat aspects of the same technical pseudo-regal style that's been going on for centuries
@axoaxonic
I agree; this is not about ASD. Precise writing does not need to be illegible; most of the illegibility in my field comes from extreme abstraction, which is imo the opposite of precision. To write clearly and precisely requires thinking clearly and precisely, and that is hard work; it's easier for many not to write clearly, but to imitate a style steeped in status whose inaccessibility is an implicit gesture of hierarchy.
@themanual4am @ngaylinn @mrdk @LizEllisPhD
@independentpen @axoaxonic @ngaylinn @mrdk @LizEllisPhD
Hello
Interesting -- it's remarkable how directly your words apply to software development, which is also writing (but perceived v differently)
1. precise code does not need to be illegible
2. excessive abstraction affects legibility; which is very different to precision
3. to code clearly/ precisely is to abstract well (is circumstantially appropriate)
4. use of terse/ short variable-name style: implicit gesture of hierarchy/ status
\
!@themanual4am
Code is so interesting
@axoaxonic @ngaylinn @mrdk @LizEllisPhD
@LizEllisPhD@mastodon.social@independentpen
There was an academic who basically refused to write using formal academic styles and he reported much greater engagement and impact of his work. I lost the link though, but I think about this a lot.
Hard agree that precision and clarity are really difficult. I stuggle with this personally.
@axoaxonic @themanual4am @ngaylinn @mrdk
@LizEllisPhD @independentpen @axoaxonic @themanual4am @mrdk Yeah. One perspective on this is that the only thing preventing us from writing however we like is us. We can just change our behavior. Easier said than done with group dynamics, of course, but in some sense it really is that simple.
I think the main exception is privilege. I'm sure some people are given much more leeway to bend the rules and do as they like than others. I bet I can guess who, also.
@LizEllisPhD@mastodon.social@ngaylinn
I'm in the process of writing a paper and I've deliberately made the tone conversational (partly due to shame and embarrassment about previously mentioned pretentiousness!). Be interesting to see what reviewer 2 thinks once it's done!
@independentpen @axoaxonic @themanual4am @mrdk
@LizEllisPhD @independentpen @axoaxonic @themanual4am @mrdk Definitely. Good luck! I'm curious to hear how it goes. :)
I think this important conversation and change must motivate without invoking shame and embarrassment; and in fact explicitly persuade against that kind of self-judgement
Isn't it ok:
- to have written to satisfy circumstances as they appeared on some prior occasion?
- to learn that other circumstances exist, and view accommodation in positive light, as growth?
- to not load up guilt if all accommodations cannot be made, on occasion?
@LizEllisPhD @ngaylinn @independentpen @axoaxonic @mrdk
And I'm not sure if it's a good thing if scientists in less-well-or-un-funded situations are personally judged for asymmetric privilege/time/resources?
This almost feels like a optionally secondary process (1. make it work 2. make it right); maybe distributed to students to assist (win win?), so a system change, with an ideal, albeit eventual outcome?
@themanual4am @LizEllisPhD @independentpen @axoaxonic @mrdk I feel like you're describing some desirable outcomes, but I'm not sure how we'd get to a system that promotes those outcomes.
@themanual4am @LizEllisPhD @independentpen @axoaxonic @mrdk I keep coming back to the comparison between science writing and readable code / tech writing. In some ways they have very similar requirements regarding precision, technical detail, and simultaneously targeting an audience of experts and novices (if not the general public).
Yet, the outcomes seem different to me. I feel like coding projects are on average more accessible, and there are popular systems, tools, and processes to encourage that with no parallels in academia.
If that's so, then perhaps we can learn by comparing these two domains. What makes them different? My first instinct is incentive structures, but I haven't thought that through yet.
@themanual4am @LizEllisPhD @independentpen @axoaxonic @mrdk Certainly, software teams and companies feel more ownership over their code, which they must maintain for years even as people come and go. By comparison, people come and go in academia, and their works are just "out there" in the literature, rather than being integrated into a curated body of knowledge maintained by the field.
Also, the value of documenting code is making the lives of engineers easier. Companies invest in good documentation to save their employees time. By contrast, academics write for survival and for status. It's a task that's required and may make or break a career. This changes the dynamic from being more altruistic to being more selfish, from being more about the audience to being more about volume and making an impression.
@ngaylinn
I think that's it. The different nuance in purpose between making a thing that functions versus having to play a status game for personal survival and advancement is my hunch. There's also cultural forces. Not to say dev is perfect (misogyny comes to mind), but there is a startling dysfunction to academia - a climate of siloes, perfectionism, overwork, and lack of support, in addition to established writing norms - that may also play a role
@themanual4am @LizEllisPhD @axoaxonic @mrdk
@independentpen @themanual4am @LizEllisPhD @axoaxonic @mrdk Yeah, I'm not one of those "academia should just follow the example of big tech" folks. It's definitely not so simple.
You're right, though, about academia's complex of dysfunction. It's really weird to see it, coming from big tech. It has very serious problems, but at least the tech industry is thoughtfully designed to serve a purpose and follow certain values (especially the profit motive, sadly). By comparison, academia seems chaotic, arbitrary, unfair, wildly inefficient, etc.
I'll also note that (from my experience at Google) participants in a tech organization are way more engaged in shaping and participating in that organization. In academia, there's an enormous gulf between research and administration that is never crossed.
Hi @LizEllisPhD,
Apologies for the above toot. I meant to edit/ correct at the time, but something came up, and it's taken a minute to get back
FWIW, the sentiments – push-back on unchecked guilt and shame (from this project), and other notions from other separate conversations – were not meant to be directed at you specifically; & all were poorly worded
As-is, with your name up top, it reads as if directed, and preachy af. So sorry (it's haunted me)
I hope that your edits went well!
@axoaxonic @ngaylinn @mrdk @LizEllisPhD
Apologies for the abrupt disappearance, something unavoidable came up – that aside, I have been chewing over this topic/ broader conversation since
I found a deep mapping between the accessibility of science writing and software code, based around the idea of domain translation (see images)
Though, I feel the accessibility issues you refer to here are more 'expressive style/ adornment'?
Is that right? Can anyone provide examples?
\
