The argument for using blockchain to secure scientific research

Blockchain may be known as the computer technology behind cryptocurrencies such as Bitcoin, but at its simplest, blockchain is a technology that offers a way to share information safely and protect against fraud and hacking. That’s why Joris van Rossum of Digital Science, a U.K.-based company that provides services to researchers and publishers, argues it could be good for managing scientific information as well as rogue, online money systems.

A blockchain is a distributed database, a sort of digital ledger. A user creates a block of data and then applies a mathematical function that creates a so-called hash of fixed length unique to that block; the underlying data cannot be changed without changing the hash. The blocks of data are strung together into a chain, which is sent to multiple computer servers, making it next to impossible to change the existing hashes without the alteration being obvious. All the entries into the ledger, including who made them and when, are set in digital stone. For cryptocurrency, that means if a person pays for something with a bitcoin, that same bitcoin cannot be used to pay someone else.

Similarly, van Rossum says, blockchain could be used to secure scientific data and even establish who carried out an experiment first. He is leading an initiative at Digital Science to test the use of blockchain to establish more trust within the peer review system. Neil Savage talked with van Rossum about the findings of a report he authored last year on using blockchain for scientific research.

Where does your interest in blockchain come from?

I’m a publisher that has worked in the area of innovation for many years, always trying to see how we can adopt new technologies to improve the way we service the academic community. That’s also where my fascination with blockchain comes from. It’s a revolutionary new way of thinking about exchanging value, storing data, connecting people, and allowing them to exchange information that they wouldn’t be comfortable doing without this new technology.

What are the problems in science that blockchain could address?

I think the industry is plagued by a few challenges. There is the reproducibility crisis. Scientists increasingly find that results published in scientific journals cannot be replicated. We don’t record how research is taking place—all the steps from idea to experiments to study design—so we only know the end result, which is that the researchers create an article that describes what they did. Their notebooks aren’t publicly accessible and, in the worst case, can even be tampered with.

There is a commercial crisis around business models for scientific publishing. When I started in the industry, open access was a new thing, but it led to its own problems. People from poor countries had challenges getting published because they can’t afford open-access fees. Predatory publishers exploit researchers’ need to get published and lure them into submitting articles to journals that are only launched to make money out of publishing fees.

There is a crisis in metrics. The way that scholars build a reputation is based on more or less two data points: the number of publications and the number of citations. But we all know there are many other activities that are as meaningful as those but not recorded.

The final problem is peer review. There’s the problem of too few reviewers being available, especially as the number of manuscripts submitted from countries like China has risen. I think one of the biggest problems with peer review is the lack of transparency. Transparency means multiple things. On one level, what did the reviewers submit? On a more basic level, did peer review actually take place? Everything that went into the research and everything that went into reviewing the article is not recorded publicly, making it a black box, which is of course very much contrary to the modern spirit of transparency.

How might blockchain help with some of these issues?

We could create a decentralized data store that captures all scientific events. So whenever a researcher performs statistical analysis, uploads data, does peer review, submits a manuscript, or gets cited, that’s all captured. We would know more about researchers’ activities, and we could create better metrics to evaluate their performance. Everything that went into a scientific article would be stored somewhere. It’s opening the black box—creating more ways to verify—which should make the research more reproducible.

Another benefit would be that blockchain technology could support the so-called tokenization of science. The idea is that if you provide tokens to researchers when they perform certain tasks, this creates incentives for them to perform duties such as being part of a committee or editorial board or reviewing grants or papers. If the researcher could then use the tokens to buy services, like paying for an open-access fee, we could create an economy around science that would solve some of the motivational issues that we see today.

Wouldn’t such data storage lead to worries about being scooped by another group or losing intellectual property?

What blockchain does on the most basic level is the hashing, which conveys that a person uploaded this data at this particular time. That hash is a string of code unique to the data set. But you can never go back to the original document through the hash. So although I can show this document was uploaded by me, you can never, by looking at the hash, actually look at the information itself. That’s still up to me to share or not. The most important thing is that blockchain, without doubt, establishes the novelty and shows that a specific person had a specific idea or specific data at a specific time.

Would blockchain mean extra work for chemists?

It depends on what we’re going to use it for. It could mean that, for example, publishers start to share peer review information (without revealing reviewers). In that case, researchers wouldn’t need to do anything extra. I think a very important principle is that we shouldn’t burden the research. Whatever comes out of it should make the work of researchers more rewarding, more efficient, and if that’s not the result, then probably we’re not looking at the right solution here.

How might using blockchain affect the business of scientific publishing?

If you look at blockchain initiatives such as Project Aiur or Scienceroot, they say that the system should be overhauled, we should disrupt the entire industry, and blockchain is going to be the fall of publishers. That is one model. There are also those that say publishers are an important partner in scholarly communication, but we need to change things a bit. You can think about applications of blockchain that make publishers’ contribution to scholarly communication more efficient. You could allow publishers to build a new business model which is not open access or subscription based but based on small payments whenever a paper is downloaded.

Blockchain is not one technology with one solution, and the extent to which it will be disruptive depends on how it will be implemented, if it will be implemented at all.

Neil Savage is a freelance writer. A version of this story appeared in ACS Central Science: cenm.ag/vanrossum. This interview was edited for length and clarity.