The possibilities of open science

Share this on social media:


Research Information spoke to four experts in the field about the ongoing move to open science, and the challenges that have emerged in an increasingly complex open-science ecosystem.

Open science is increasingly seen as a key part of ensuring that innovation and knowledge discovery happens as fast as possible and with as few barriers as possible. It is the best way of tackling some of the big challenges facing the world, whether climate change, global poverty or the next pandemic. But while the advantages of open science are increasingly acknowledged, getting to open has proved tricky as cultures and business models are slow to change. 

Open science or open access?

The slow speed of change to open science is reflected in the fact that much of the open science discussion is still focused on the open access stage of the research process. As Richard Gallagher, president and editor-in-chief at Annual Reviews explained, there are many opportunities beyond the traditional research paper:

‘At Annual Reviews we see the publication of the research paper as the start of what we want to do rather than the end of it. We want to layer on top of it information that makes the content accessible and useful to other parts of society in addition to the research community. We want to make it useful to policymakers, educators and anyone who’s got an interest in what current research thinking is on any particular subject. We want corporations to be able to act on information. Just making the research literature open is not quite enough, because not everybody can read and understand, or wants to engage with the very dry way that science is presented. The overarching opportunity is the opportunity to collaborate across different components of the scholarly publishing enterprise.’

A similar point was also made by Abhishek Goel, chief executive officer and co-founder of CACTUS, a technology company accelerating scientific advancement:

‘Most people aren’t likely to read a research paper in their lifetime, but lay summaries, directed at the public, could have an enormous benefit in the public understanding of science. This is why open access is just a small part of open science (and more broadly, open academia), although the terms are often used interchangeably.’

The potential of open science was most clearly shown during the COVID-19 pandemic, as Curtis Brundy, associate University Librarian at Iowa State University explained: ‘We saw what happens when we all move in the right direction during the pandemic – when the literature was opened up, the data was shared, the code was shared, the methods were shared – knowledge creation really did accelerate. That was a powerful lesson, and it was great reading about it in the new OSTP [Office of Science and Technology Policy] memo1. They talked about the response during the COVID-19 pandemic multiple times as evidence for why we need to move towards making all of this publically accessible immediately.’

The rapid openness of the response to COVID-19 is noticeably different; however, from the complex open-science ecosystem that most researchers have to traverse, where the complexities of the open access system could risk slowing the advance of knowledge.

A diversity of approaches

As publishers have sought to fulfil the competing requirements of authors, funders, and libraries, we have ended up with an increasingly diverse set of business models and an extensive bureaucracy to facilitate them. As Shelley Allen, Head of Open Research at Emerald explained, there is no one-size-fits all solution: ‘Those who are interested in open don't necessarily want to get there in the same way, and that's why Emerald have what we call being open to all. We have a diversity of routes towards open – zero embargo, green, transformative deals with those who want them, platinum journals – and that's just on the business models, let alone on working practices.

‘From an Emerald perspective, being an applied social science publisher our key stakeholders were quite happy with our approach to open for quite a while. Most were happy with our zero-embargo green policy, some customers were interested with gold, but as most of the research is unfunded in our areas there weren't too many comers for transformative agreements. That's definitely changed over the last few years. As a medium-sized social science publisher, it's really about having that diversity of approach, because we are very keen not to accidentally create barriers to publication while we seek to remove the paywall.

‘It is really complex and very expensive to keep managing all these different ways. The biggest challenges are the workstreams you need in place to manage these complexities. Data is a massive challenge, identifying the eligibility of authors, ensuring you let authors know what they are mandated to do, what they are entitled to, and take that work out for them, and then just generally trying to shift the needle beyond open access to open research.’

With such a diversity of models and workflows creating an increasingly large administrative burden for publishers, research institutions, and researchers, it is not surprising that there has also been a push for simpler and cheaper solutions. 

Removing barriers

The model Gallagher and Brundy are promoting, Subscribe to Open (S2O), may be seen as a more natural evolution from the original subscription model. S2O is a subscription model where, assuming subscribers participate, the publisher will make the year’s content available as open access. As Gallagher explained: ‘We want the people that are supporting the publishing enterprise – libraries, readers, reading institutions – to continue to be the ones that support the costs, whereas article processing charges (APCs) and read-and-publish agreements move that obligation to the producers of the content. It doesn't cost us any more to not have the paywall than to have the paywall, and if we can continue to get subscription income, then we don't have to charge any more for it. It's much more equitable than the other ways to open access. There's been some studies and some complaints that open access is really playing into the hands of the wealthy. S2O works well on all areas of research – it's equally available to researchers from different institutions, different countries and different fields of research.’

Brundy is supportive of the model from the perspective of the library: ‘Our library has implemented several different kinds of OA models, and S2O is really easy to do. There's no article level workflows, it's really simple to go from one year you're paying for a traditional subscription to the next year paying for something that's delivering open access. Equity is the most powerful and important aspect of the model, especially as we learn how inadequate waiver programs and discount programs have been in trying to achieve equity with the APC-based models. There is not a publisher out there that hasn't centred on diversity, equity, inclusion, and if you're going to adopt a model that doesn't have equity considered at the front end you are very soon out of alignment with your own organisational values.

‘From a library perspective, we've been investing in a system of scholarly communication that slows down the advancement of science and introduces issues around reproducibility. A core value for libraries is information equity, so we have also been investing in a system that is out of alignment with a core value of our profession.’

In comparison to other open access models S2O is new and experience is limited, with just 95 journals from 12 different publishers, and it will require far more experiment from a wider variety of publishers to satisfy some sceptics. 

Alternative visions

While S2O is pushing the focus back on the readers of the content, it is not the only possible route out of today’s complex and expensive ecosystem. As Gallagher noted, the potential of overlay journal on preprints is still a big unknown: ‘They have the potential to really shake the current way scientific publishing is done, faster, much cheaper, but also somewhat less structured.’

This is the route that Goel sees as becoming increasingly important, with the creation of new markets for new services: ‘The obvious challenges of stagnating library budgets and increasingly large APCs are not sustainable for an industry. There will be an increase in the use and uptake of ‘free’ services, such as preprints, and a different approach to serving researcher needs via personalised service. These would encompass not just publishing, but everything on the continuum from idea to impact.

‘We believe that the end-user, who has been excluded from most business models in scholarly communications, represents an enormous opportunity to personalise services and offer support and content at an affordable price point. So, a B2C model, thus far largely absent from the market, could be the biggest change over the next few years.

‘There is also a huge opportunity in the area of commercialisation of research. Traditionally, the bridge between academia and industry has not been very strong; a huge opportunity exists there, helping ease the financial burden for the governments and institutes.’


It’s now over 30 years since the arXiv preprint server was made available online, and for all the talk of the importance of open science we are still struggling to find an efficient solution to the problem of even open access. As Gallagher explains, ‘There’s a sort of valley of death between giving up your existing model and getting the new model to where it needs to be.’

Unsurprisingly, with the perceived progress of humanity at stake, tensions can run high and adversarial positions can be taken, but it’s always going to be a gradual change rather than a quick fix. As Allen explained: ‘It's not just about policies and mandates and business models, it's also about the culture, it's also about incentives, and the support, the structural support to help change people's behaviours in how they publish. That can only be achieved in a collaborative way across the ecosystem, and we do see that happening in pockets, but it would be good to see more trade bodies talking to other trade bodies. That is the only way to solve this, through collaboration, cooperation and conversation about the problems and what we can do. I don't think it's possible to change it overnight, and it certainly isn't possible without that collaboration.’

It seems unlikely there will be a one-size-fits-all solution to open access or open science in the near future, but if such a solution is to be found, it will undoubtedly be found quicker together.

SPONSORED: Stop chasing your tail, choose AI summarisation

SciencePOD’s new research amplification solutions, a boon for open science

Why are so many scientists still chasing their tail? They can never find enough time to read all the relevant past studies they need, let alone keep an eye on the latest research. On the upside, this tail-chasing dilemma creates an opportunity to introduce new business models for services that help scientists keep up-to-date and amplify the impact of their research. On the occasion of the 2022 Frankfurt Book Fair, SciencePOD is introducing ScioWire beta, a new artificial intelligence (AI) solution designed to empower every scientist – and their publishers – to increase the impact of their work. How? By creating summaries of individual published papers automatically and on a large scale. This double-edged solution provides both a newsfeed of the latest research and on-demand AI summarisation.

Beyond open-access publications

The ScioWire launch comes at the right time. It is taking place as open access (OA) is increasingly shaking up the old business models of scholarly publishers. Dwindling subscription revenues leave them with no other option but to explore new business models that better serve the scientific community. This approach makes scientist-centric solutions increasingly attractive. Specifically, new product opportunities are emerging for the post-publication stage – the stage associated with dissemination, discoverability and ultimately, research impact. For new solutions, we can thank advances in natural language processing applications in the field of automated summarisation.

This new context calls for new business models centred around the base-unit of scientific record – the research paper. Innovations associated with the nature of the scientific record and its wider availability cover micro-publications, open peer-review and preprints. Meanwhile, technology is also increasingly dictating the way scientific data is recorded,  in line with FAIR principles, to allow for machine readability and cross-analysis of the entire scientific output. Despite these innovations, wider scientific progress offering cross-fertilisation opportunities remains out of reach for most scientists, who remain siloed—often self-confined to their field of expertise due to the massive volume of information to digest.

New ways of consuming research

As AI technology matures, now is the time to re-think how scientists consume research. There is an opportunity to make hay of the benefits of natural language processing to increase the discoverability and dissemination of research studies and amplify their impact. AI summaries are quickly generated, easy to digest and share – they can act as a preview for individual research studies.

One way to productise AI summarisation is by designing a newsfeed to help scientists monitor latest advances in research. That’s what SciencePOD has done with its new solution: ScioWire beta – delivering a keyword-customised newsfeed of the latest OA research. Each feed item is made up of automatically generated summaries of the latest OA papers, complete with context. ScioWire is an affordable productivity tool designed to help scientists select relevant studies more quickly.

Why summarise?

Before ScioWire, scientists interested in monitoring the latest advances had to rely on abstracts to separate the wheat from the chaff. The trouble is abstracts often give a subjective view of the study – more often than not, with the intention of supporting the authors’ next funding application. By contrast, summaries reflect the key findings of each paper. What is unique about SciencePOD’s own AI summarisation tool, SUMMSci, is that it goes one step further by delivering contextualised summaries. Through a partnership with Juan-Manuel Torres-Moreno, senior researcher from the Laboratoire d’Informatique d’Avignon at the eponymous French Université, SciencePOD’s solution reflects the years of experience of this world expert in summarisation. The summaries of academic studies are generated automatically within seconds. And they come complete with context: keywords, acronyms elucidated and lay definitions of technical terms. In addition, they feature a second type of ‘quick-to-consume’ information – research highlights that answer key questions: Who did the work? Where and when did they publish? What did they find? How did they find these results? What are the future research steps?

Attention economy

Summaries represent the building blocks of research amplification strategies. They are proven to draw the attention of readers to the original research, yielding much higher impact than the full-size article on its own. However, AI summaries are by no means a substitute for the full text. Instead, they act as a ‘teaser’ to help readers decide whether or not to read the details of the full paper.

These snackable, digestible versions of scientific papers fit the evolving content consumption habits and preferences of scientists who, on average, only have 5- to 15-minute slots of time throughout the day to consume content. 

The new solution can thus enhance researchers’ productivity – when preparing a review paper or doing background bibliographical research, for example – which is critical to the success of their work. That’s where ScioWire goes one step further in providing scientist-centric features, by offering AI summarisation on-demand and research monitoring functionality in the form of a customisable newsfeed. 

Impact at scale 

Research impact matters to scientists as well as scholarly publishers. In addition to relying on SciencePOD’s global community of science and medical writers to create summarised research highlights ‘by hand’, publishers are exploring how AI summarisation technologies can help disseminate large volumes of article summaries linked to the original source research paper.

In addition, SciencePOD will help with distribution by offering publication to its own ScioWire web magazine and make its content available through a personalised widget embedded in online portals hosting communities of researchers. Combining attention-grabbing AI summaries and distribution networks, SciencePOD’s new solutions will empower publishers to offer greater value to authors and to draw wider audiences to the latest research studies they publish at an unprecedented scale.

Sabine Louët, founder and CEO of science content creation company, SciencePOD

Find out more about ScioWire on the SciencePOD website: