As project lead for the recent launch of article sharing on, I’ve been working with a great team at NPG, Digital Science, ReadCube and Macmillan Science and Education. It’s an exciting role, one that gives me a comprehensive view of the how the pieces fit together, and the idealism and intentions of the people behind this experiment.

Researchers and the public are clearly benefitting. Since we launched on Tuesday, many thousands of readers are accessing shared NPG articles each day, either through links shared by peers or via media outlets.

Of course any new venture, particularly one as bold as this, will attract critics. Some have offered constructive and thought-provoking input. As Steven Inchcoombe and Timo Hannay clearly noted in our press release and in media interviews, this is an experiment, designed to garner feedback from researchers, publishers and other actors in the world of scholarly communication. So we welcome suggestions and critiques that lead to improvements, and will be acting on them in the coming months.

However, it seems there has been an unusual amount of misunderstanding, as well as a number of outright errors. Some of these have been addressed in recent posts from Digital Science and NPG. However, given my perspective across all aspects of the project I thought it might help to offer a single consolidated response to what I view as the key misperceptions.

This is a step backwards for Open Access – This was never communicated by NPG as an OA initiative, nor is it a move to replace or supplant any of our existing OA efforts. Rather it’s an additional offering to make it easier for researchers to share subscription content with colleagues, collaborators and the general public while giving institutions and the publisher more information about what type of content readers are interested in. My colleagues Timo Hannay and Steven Inchcoombe address this in more detail.

You must download or register ReadCube to read an article This is not correct. ReadCube runs in the browser: it is a javascript library used to mimic the PDF layout but is rendered as HTML5. If you have a reasonably modern browser, you can view shared articles on If you want a permanent copy of the shared article, you can download the free ReadCube Desktop app, and save a copy to your ReadCube library, as a read-only version to keep as long as you like. But this is entirely up to the user.

ReadCube is simply a DRM tool for publishers and is trying to lock users into a closed system – This is particularly misleading. ReadCube first and foremost develops software for researchers to facilitate research organization, discovery, and interactive reading. With over a million unique readers each month, it has a large, loyal user base. As ReadCube has expanded into providing software tools for institutions and publishers, it was in a position to respond to NPG criteria for sharing of content on We expect other publishers, platforms and software vendors to offer similar solutions. In the meantime, ReadCube is powering an initiative that’s expanded access to people who hadn’t had it before.

Specifically regarding DRM on the shared content pages, as my colleague Timo Hannay noted in a response to one such critique: “There’s no DRM on  in the cryptographic sense that that term is usually used: the text is in the web page, making it searchable, indexable and accessible to screen readers. In order to make the pages consistent with NPG’s sharing policy, we’ve used HTML5 to disable copying and printing.” This latter restriction does not prevent people from reading, sharing or accessing the content via screen readers. DRM, as employed by the music and film industry, relies on encrypted content that can only be used when unlocked. This is not that.

Macmillan are using ReadCube to snoop on scientists – All Macmillan businesses are fully compliant with EU data privacy regulations, which are stricter than those in the US. That means no snooping on individuals, among other things. Neither Macmillan nor ReadCube collect personal data unless the user accepts our terms of service.

It’s worth further addressing this in two parts – data from NPG content sharing, and that collected by ReadCube through its Desktop and mobile apps.

NPG content sharing – For the purposes of this experiment, we will of course collect aggregate anonymised data. This will help NPG to learn how researchers respond to sharing and make informed decisions about user workflows and its business. This is necessary for wider sharing of subscription content across the industry. Most people understand that in exchange for allowing their paid content to be shared, publishers will need to broadly understand the extent and nature of sharing activity – it’s reasonable, widely accepted commercial behaviour. NPG’s experiment and the data it provides will hopefully help the STM publishing industry avoid the mistakes made by the the music and film industry in trying to keep too tight a grip on content

ReadCube Apps – Again, you do not need to register for ReadCube or download their software to read an NPG shared article. Should you decide that you like their reference management tools, you can choose to register and download their free Desktop app. If you choose, you can add your own articles and documents to the Desktop app. Note that ReadCube will never sync files to the cloud unless explicitly selected by the user. Any data ReadCube gets from these users is not shared with any publisher, including Macmillan.

Shared articles can’t be accessed by blind or partially sighted users  – Because the text is unencrypted and rendered as HTML within the browser, shared NPG articles should work with most screen readers. Our testing has confirmed this but we encourage anyone having problems with this issue to contact

Shared articles can’t be viewed on mobile apps – ReadCube currently has iOS and Android apps for mobile devices. Support for shared articles is on their development roadmap and is expected to be available in 2015.

NPG content sharing prevents machine-readable access – This was designed to expand human-readable access. Enabling mass text and data mining is a valid issue that needs to be addressed by the publishing industry, but is outside the scope of this experiment. But it was certainly not meant to prevent the expansion of machine-readable access.

Nature already enables sharing – Nature enables sharing of the article link, but until now unless the recipient was also a subscriber they would only see the abstract and paywall. Now anyone receiving the link from a subscriber can read the complete article.

This is merely a PR stunt for NPG and ReadCube This was not done as a PR stunt, but rather to increase access to researchers and the public at large. The numbers of articles shared and accessed since the launch shows we’ve done that. Further, this is a tangible benefit for NPG authors, who can use this new feature to easily and widely share their work. It’s also worth bearing in mind that for NPG this initiative does entail potential commercial risk, which is not usually a characteristic of PR stunts. This is one reason it’s been launched as an experiment. Lastly, if you really think content sharing is just a PR ploy, ask current users if they’d like us to switch it off.

NPG content sharing could indeed increase the number of ReadCube users (though again, you don’t not need be a registered ReadCube user to view shared articles). We of course hope ReadCube will become the favourite of researchers for its user experience, rich content recommendations and cross-platform functionality. But as the publishing industry moves toward more widespread sharing, we anticipate, and welcome, other platforms offering sharing alternatives to researchers, institutions and publishers. Again, this is simply a first step, one we were in a position to take.

It’s early days yet. While results thus far suggest users are benefiting significantly from this experiment, we’ve still much to learn about how to best enable content sharing in a way that benefits researchers and is sustainable for NPG. We welcome reasonable feedback, and if there are substantive concerns, we will address them. In the meantime, I hope this clears up matters around how content sharing works, and our intentions around it.