Good afternoon from cloudy Wimbledon.
I just read ‘The Scientific Paper Is Obsolete’ by James Somers in The Atlantic and I would greatly appreciate your thoughts on my perspective on this. I think this issue is of crucial importance for academic and human thought and discourse in general. I hope you will have a min to at least skim this message. And thank you for taking a look. I have tried to write this as skimmable as I can.
The article rings bells along the lines of Vannevar Bush’s 1945 article in The Atlantic, ‘As We May Think’ where he called out that “Professionally our methods of transmitting and reviewing the results of research are generations old and by now are totally inadequate for their purpose”. “A record if it is to be useful to science, must be continuously extended, it must be stored, and above all it must be consulted”.
In other words a call for *richer interactions* in connecting the future and past, connecting different pieces of information and connecting to the users minds.
Somers’ article starts out highlighting the vital importance of the development of the scientific paper in the 1600s, and the journals that curated and disseminated them; “journals made the chaos of science accretive. Scientists from that point forward became like the social insects: They made their progress steadily, as a buzzing mass.” The article then goes on to survey the active ‘paper’ of Brett Victor and Wolfram’s notebooks in Mathemathica as well as Fernando Pérez’s iPhyton which became Jupyter Notebooks. The main point of the article is that: “Perhaps the paper itself is to blame. Scientific methods evolve now at the speed of software; the skill most in demand among physicists, biologists, chemists, geologists, even anthropologists and research psychologists, is facility with programming languages and “data science” packages. And yet the basic means of communicating scientific results hasn’t changed for 400 years. Papers may be posted online, but they’re still text and pictures on a page.”
Delivering Richer Interactions in Scientific and Academic ‘Papers’/Reports/Articles
It concludes: “At one point, Pérez told me the name Jupyter honored Galileo, perhaps the first modern scientist. The Jupyter logo is an abstracted version of Galileo’s original drawings of the moons of Jupiter. “Galileo couldn’t go anywhere to buy a telescope,” Pérez said. “He had to build his own.”
To Pérez I say YES! We definitively need ‘richer’ documents to ‘see’ further (to use the analogy of the telescope instrument) but they must remain open and robust. To wit:
Out of the Sandbox, Robustly into the Real World
We are seeing this during a Cambrian explosion in video conferring collaboration systems where it is currently impossible to Zoom call someone on Google Hangouts or Skype someone on FaceTime. This is of course a big step backwards from being able to email anyone or call anyone on the phone. Do we want to enter private companies virtual phone booths to make our calls, without being able to choose the service? We only recently climbed out of the morass of Microsoft Word documents (all of Office in fact), controlling collaborative documents (word can be edited, PDF cannot easily, is what I mean), primarily through Google Docs. We definitively do not want to live in the pockets of a private software company which can hide code or simply die. How many HyperCard stacks are available today? How easily is it to view Flash interactions?
To make papers richly interactive, I really, honestly and deeply think that openness and robustness for preservation (for millennia but also for days and weeks and for those with different software today) is absolutely crucial, otherwise we are building commercially owned toy sandboxes only.
Meta & Interaction is as important as the ‘Content’ right?
If the meta and interaction is really as important as the ‘contents’, then it should be afforded the same level of presentation and protection right? I think so, and I don’t think that is a controversial position to take these days.
I therefore think we should make some sort of version of visual-meta standard: http://visual-meta.info And by this I simply mean add an appendix to the last page of a document which specifies what part of the text should be interactive and in what way, as well as citation information, formatting information and so on.
A normal book affords this respect to the meta by providing all the citation and production information on one of the first pages of the book, before even the actual text of the book, yet with computer documents there has been a history of hiding the meta and interactions and only displaying the tip of the iceberg. Most documents don’t even have anything below the surface, they are not icebergs, they are superficial floating snow drifts….
The visual-meta approach is to first add the means of production and then a BibTeX citation block, so that the document ‘knows’ what it is and can allow the user copy as a citation. It then can list, in JSON form, all the headings, so that the reader software can allow the user to use the headings for different views. It does this by listing every heading and stating what level the heading is.
What I am proposing here is that we add any and all interactions the author would like to have in the document, also in this format. This would allow the user to specify that ‘next Monday’ is actually Monday the 22nd of June 2020 and thus this can be done operations on. Scientific calculations could also be in the paper and the interactions specified in the visual-meta appending, to allow for Brett Victor like slider interactions, Mathematica loo-ups and so on. How this would work in the long run I don’t know, it is something we need to figure out as a community. What is clear is that this method works: You can see it in action and try it yourself on www.liquid.info
Btw, I have blogged this so if you want to blog a reply, or just email or simply comment, you can do that:
Again, THANK YOU for looking at this. How we extend our minds matter. This should be a wide debate :-)