Distributed research journals based on blockchains

by blackstampede2 min read21st Oct 20215 comments


Cryptocurrency & BlockchainReplication CrisisWorld Modeling

I know what you're thinking (I mean, I probably don't but I'm going to pretend that I do for a minute): Blockchains are synonymous with cryptocurrencies at this point so I'm probably talking about creating some sort of coin and using it to pay academics.

Neat, but no. What I like about blockchains is that they're:

  • Immutable
  • Distributed
  • Organized into a fixed chronological order

These all seem like features that would be great for some sort of distributed research journal:

  • Immutable: Once some academic work is published you don't want it to change. Even if later it turns out to be wrong, it's a record of your progress as a field and no one should be able to sneak in and tweak it after the fact.

  • Distributed: You want teams of researchers, academic organizations and individuals to be able to work together over long distances without doubting that they're all sharing the same base of knowledge.

  • Chronological order: Early work should be early and later work should later- and able to refer to earlier work in a static way without worrying about things being moved around.

These features seem like they could solve two persistent issues in academic publishing. The first is the cost of access. Journals tend to cost a lot, which means that unless you're associated with some academic organization, you're not going to be able to afford them. The second is that research which attempts to reproduce existing results or disprove some previous work isn't interesting to academics (trying to build careers) or journals (trying to sell access), which has led to a replication crisis.

Distributed journals would be free by default (I could imagine some sort of pay-to-access scheme, but it seems like a reach), which would reduce barriers to entry for individual researchers. The cost of hosting the journal blockchain could be shouldered by anyone (or any organization) who wants always-up-to-date access to the latest research, or that just want to contribute. Linux distributions, software and source code are often mirrored by .edu servers for similar reasons.

Distributed journals would allow research to be reviewed by peers drawn from a very large pool (everyone who is active in the journal) which would work in combination with the free-by-default point above to diminish the systems bias toward novel results. You could also measure the precise impact that your work has had on the field through automated citation mapping, which might encourage attempts at replication.

It's easier to work on non-cutting-edge research, I imagine, if you can present convincing metrics showing that you've forever altered the course of scientific inquiry.

So I have some ideas on how something like this could be made, but I wanted to validate the basic idea first. Is there something I'm missing here, something I haven't considered?


5 comments, sorted by Highlighting new comments since Today at 10:31 AM
New Comment

[not downvoted - I'm happy that people are thinking about and exploring options like this, even though I don't think this particular application is very compelling. ]

I think Elizabeth has given the primary reason that this won't work - neither of the problems you list (expensive and often not replicable) are really addressed.  Those are problems with refereeing and intentional scarcity/prestige mechanics, not with a lack of auditable sequence of events.

In fact, most of the papers I read are freely published via a very simple mechanism: preprints on the author's web pages.  I don't worry that they're not bit-for-bit identical with what the relevant journal published, nor that there's some fraud about when something happened.  

That's fine, I was mainly throwing this out there to see what I hadn't thought of. Can you expand a bit on what you mean by "refereeing"? I think I mostly understand what Elizabeth is talking about regarding prestige mechanics (although I wouldn't say no to a longer explanation of that as well.)

Is hosting or fear of post hoc editing a blocker to open academic publishing? My impression is no, and if it were there are cheaper solutions.

Current publishing norms are kept in place at least in part by prestige dynamics, do you have a plan for addressing those?

Disclaimer: I'm not an academic and I've never published. So there are certainly ideas and dynamics at play here that I don't understand. Any clarification is great.

I don't think a fear of post-hoc editing is a major reason why you might want to build a journal on a blockchain- the distributed consensus is the main benefit you get- it lets you do away with publishers and publishing fees for access (I should probably reorganize that first list of benefits.)

I think that more transparency on how an individual contribution affects the entire community could improve or change what is considered prestigious. For example, citation networks and patterns over time could help you spot the pivot points where the community shifts from generally-agreeing to generally-not-agreeing with some particular claim and identify the researchers that triggered the change.

arXiv already allows for free publishing today. Nothing you wrote about seems to provide a meaningful improvement on arXiv.