Blockchain for publications [closed]

Question

Today`s publications lack of the possibility to be extended by others after being published. A data base paper could for example benefit when contributors would gain authorship by opening data with a meta description.

Blockchain is a technology which offers the inheritance of documents. Each block can be seen as a single owner.

In case of scientific publications that means:

• Paper x00 is written by x.
• Paper x00 is taken by z and extended to xz0.
• Paper x00 is taken by y and extended to xy0.
• Paper xz0 is liked most by q and extended to xzq. q finds the advanced paper xz0 better than xy0 or the original x00.

There would be no merge options (xy0 and xz0 are merged to xyz), but fields within science would develop interesting publication trees over time with self pruning capabilities. Most useful contributions will inherit, none useful extensions will "die out". You can still cite other sources.

I worked as an informatician in the barely digitized field of forestry and found myself citing a technology called optimization developed in the 50`s. Specifically the "Downhill Simplex" method developed by two authors Nelder and Mead. I also know that only the basic principle of the algorithm is used in today's libraries. Details have been changed. Multiple thousands of method clones exists published over half a century. The implementation I rely on differs from the original published one. I always wondered who contributed most over the decades. Might have been a lot of scientists who made advances on the algorithm which others did rely on. But I do now know.

What benefits and what negative effects would such an application have I do not think of? Energy costs could be another issue which I cannot quantify.

To clarify, the benefit I see is the expandable authorship.

Answer 1

Scientific publishing is certainly experiencing a lot of changes and is looking very different from what is used to, even, let's say, 20 years ago. The increasing role of electronic publications, open-access journals, mere publication capabilities, open preprint services (like arXiv), ability to attach media\data\source code to the publication – all that has revolutionized and will continue to change the way we publish, read and comprehend scientific publications.

Nevertheless, all those advances solved (or attempted to solve) crucial problems in publishing and/or significantly increased the quality of the experience, both for the reader and the author. The extension/correction of the existing work does not seem to me like one of them.

Nobody prohibits (and usually it is encouraged) subsequent publications that extend the subject knowledge. That's why we have references at the end of the paper. Nowadays, a lot of publishers offer easy access to referenced material (sometimes, even across the publishers) in a click of a button. In addition to that, there is often also "Citations" section which would show the papers that cite this paper.

Example (just took a random paper from my field, not implying by any means that IEEE does a perfect job):

• The paper "Error analysis for the truncation of multipole expansion of vector Green's functions" by J. Song and W. C. Chew in IEEE MWCL.
• IEEE provides references that this paper contains with links to the available full-texts, contexts, and several means of accessing it.
• Also, IEEE provides access to papers that cite it both from IEEE and other publishers.

All that ensures the continuity and availability of the research. I certainly do not see the value of creating a new revision of the same paper later to include new contributions from the same or different authors. That would be a new publication.

As for the corrections, with electronic publications, the authors themselves can correct their paper if the publisher policies allow that. Moreover, let's not underestimate the value of the "Correction to the paper" publications. While some of them are not so interesting, others might offer valuable content by showing what and why it was wrong in the original paper. Again, with electronic publications, these correction papers can be linked to the original publications.

To sum it up. While blockchain can be useful to create the system you have described, I do not see the need for this system. And this need, in my opinion, should be investigated and analyzed first in great detail. Also, see some of my points on blockchain usage for avoiding fraud in collected data. Here I would repeat the final thought:

I am very interested to see practical uses of blockchain in other areas (let's see where the smart contracts go in 5-10 years). Academia and research process does not scream for blockchain and its usability is totally unclear and hard to implement.

Answer 2

In general, researchers prefer to finish a paper, publish it in its final state, and then be done with it. This avoids dependency hell: in subsequent papers you can write "By Lemma 3.17 of [14], ..." and know that Lemma 3.17 won't change or be renumbered. It also makes it easy to prepare lists of publications for Department chairs, deans, grant-making agencies, and others.

If someone wants to build on a previous paper (someone else's or their own), they can write a new paper and cite the previous work.

Finally, the security features of blockchain aren't really needed: there just aren't a lot of researchers falsely claiming to have contributed to papers. But the Stacks Project is somewhat along the lines you suggest. It is a huge book on algebraic geometry -- with ongoing updates, additions, and improvements, maintained as open-source software might be. It currently stands at 6,600 pages with 351 contributors.