As many of you know, I’ve been a fangirl of/advocate for persistent identifiers (PIDs) and their metadata since May 2015, when I joined ORCID as Director of Communications/Community Engagement. The organization was about four years old at that time, and growing quickly. We signed our first consortia agreements in 2014; the Crossref and DataCite auto-update functionality was launched in October 2015; and we reached 3.14M registered ORCID iDs just in time for Pi Day 2017. Fast forward 10 years* and a whole lot more progress has been made. At the time of writing, per the ORCID stats web page, there are close to 10M active ORCID records, with users in every country; membership has grown to almost 1,500 organizations in over 60 countries (many via the 29 ORCID consortia); and a whopping 6,250 systems are actively integrated with ORCID. All of which means that ORCID is now effectively the de facto identifier for researchers.

But – and it’s a big and important but – it is not by any means the only researcher identifier, and nor should it be. There are many use cases for these identifiers and, while ORCID meets many of them, it can’t possibly be the first choice for everyone. National, disciplinary, and proprietary identifiers all also have an important role to play. Fortunately, ORCID is designed to work with these other identifiers – in fact, you could argue that’s what it does best. Unfortunately, however, this isn’t as well known as it should be and, in some communities, there’s a mistaken belief that a choice has to be made between ORCID and their preferred researcher identifier. As a result, even if they have registered for an ORCID iD, researchers working in these communities aren’t able to benefit from the true power of a trusted global, cross-disciplinary, PID-connected ecosystem.

As ex-ORCID colleagues and fellow PID enthusiasts, Josh Brown (also of MoreBrains Cooperative) and I jumped at the chance to write a white paper about this topic. The paper was commissioned by ORCID, who provided us with some of the data and background information we needed, and we certainly share their view that “it’s ORCID and, not ORCID or.”  However, the paper very much reflects our own observations and recommendations on the topic. We included several mini case studies to demonstrate some of the current opportunities and challenges, and we hope our paper will be both interesting and helpful to the users and developers of a wide range of researcher identifiers – and PIDs more generally.

National identifiers are a great place to start: there’s a clear need for them because research is so often shaped by where it is conducted, funded, and evaluated (including the priorities of whoever happens to be in charge at the time – as we are seeing, to our cost, here in the U.S.). National identifier systems are, therefore, helpful for collecting information about research in a national context including supporting the local language(s).

However, research is a global endeavor in terms of its funding and its communication – authors can publish anywhere in the world, for readers anywhere in the world. So, in addition to continuing to use national researcher identifiers, it makes sense to also make use of ORCID, which is a fully global system – and the Portuguese CIÊNCIA ID is a great example of this. CIÊNCIA ID enables everyone involved in Portuguese research – researchers, teachers, administrators, technicians, and others – to access national science services and to share and reuse information across reporting and administrative systems. This allows data to be shared across multiple Portuguese repositories, and enables researchers’ ORCID records to be populated with this data, making it easily available for re-use anywhere in the world. At the time of writing, there are close to 112k ORCID iDs connected with CIÊNCIA ID, 83k of which have been automatically updated by the system, including the addition of 3.33M work items.** Interestingly (especially given our work to develop cost-benefit analyses of the savings from PID integrations at the national level), this ORCID integration contributes to annual savings of more than 154 hours per researcher of time spent on data entry and rekeying information.

Another approach to using ORCID alongside a national researcher identifier can be found in Germany, where researchers can connect the curated ‘authority file’ model provided by the Deutsche Nationalbibliothek (DNB) via their Integrated Authority File (GND) with their ORCID record, enabling the two systems to exchange information.

Next up, disciplinary identifiers for researchers. Just as different disciplines work in different ways, with different techniques, tools, and norms, they may also have their own infrastructures, which help to address discipline-specific challenges or practices. The arXiv repository, which enables preprint sharing in physics and related fields, is a good example. However, changes to disciplinary boundaries, along with an increase in interdisciplinary research, means that these discipline-focused systems also need to work well with multidisciplinary ones. Again, ORCID is a good option, since it spans the entire research landscape.

Our case study here is INSPIRE HEP, which has been the main information hub for the high-energy physics community for almost 50 years. This community is known for its ‘hyper-authorship’ of papers, so INSPIRE HEP contains numerous articles with thousands of authors, many of whom have similar or identical names and/or the same affiliation. Its 100,000 author profiles are connected to over one million articles! It has been linked to ORCID since 2013, enabling works in INSPIRE HEP to be linked to the authors’ ORCID records – and vice versa – making these contributions visible, shareable, and connectable to other systems that use one or both identifiers.

Last, we looked at proprietary research identifiers. Elsevier’s Scopus ID and Clarivate/Web of Science’s ResearcherID are the two biggest players here, with millions of researchers using one or both systems. However, while these identifiers are public-facing and free for researchers to use, they’re not fully open; rather, they are designed to be used exclusively in each companies’ suite of products, and to optimize the user experience for those products. Their utility is therefore limited – on their own, they can’t integrate data with internal systems or those from other providers. But, by connecting with ORCID, they can! ResearcherID, for example, is designed to help match authors across all Clarivate’s products, solving author identity issues, building accurate links between authors and their publications, and sharing information across services to build up more complete profiles and improve the accuracy of analytics such as citation metrics. All good stuff – and because ResearcherID and ORCID are integrated, author records can be matched to publications using either system. As well as increasing the accuracy of these matches, this also means that, where an ORCID iD is linked to a ResearcherID, the information in both records can be exchanged. At the time of writing, the Web of Science integration has resulted in close to 790k connected iDs, and about 505k updated records, including 9.43M work items.**

Elsevier’s Scopus Author ID ORCID integration provides similar benefits. And interestingly, like Thomson Reuters (who spun Clarivate out as a separate company in 2016), Elsevier was an ORCID launch partner organization. Why, you might ask, did these two large commercial companies help to create an open, community-governed competitor to their own services? The answer is that, as with other shared scholarly infrastructure (Crossref, NISO, etc.) both of them recognized that having a globally adopted, trusted researcher identifier would support a more interoperable information landscape, creating a shared foundation on which they and others would be able to build better services.

So far, so good. But what can or should the research community – including ORCID itself – be doing to help ensure that the metadata attached to all types of researcher identifiers can flow easily between systems? We made three high-level recommendations in our white paper:

1. ORCID should proactively engage with providers of national and disciplinary identifiers. Interoperability between ORCID and the identifiers currently preferred by some geographical and subject communities will benefit everyone. Engaging with those communities to better understand and address their concerns about integrating with ORCID is, therefore, critical.
2. ORCID should work with other researcher identifiers to improve their integrations. These are not actions for ORCID alone; they will also require investment from partners to unlock the potential value. Some of these integrations were developed a long time ago and, in order to be of value to their user communities, they need updating. Examples include the International Standard Name Identifier (ISNI) integration with ORCID, which was developed as part of an EU-funded project, and is currently not functional; and the recent integration of ORCID into Github, which is an excellent first step, but needs to be fleshed out into a concrete plan of action
3. Providers of researcher identifiers should increase the presence of trust markers in ORCID records. Trust markers are data that has been added to ORCID records by a trusted party, i.e., an ORCID member organization. Integrating with ORCID will be more valuable to other researcher identifier systems if they understand how the trust markers they add to ORCID records are being propagated throughout other systems used by their communities. Continued investment by ORCID in ways to simplify and support this process will be required. In addition, it will be important to continue to lower the barriers to participation in ORCID, for example, for communities with fewer resources and/or less capacity.

Integrations between ORCID and other types of researcher identifiers benefit everyone, by enabling trustworthy information – validated in one or both systems – to flow seamlessly across the many online tools and platforms used by researchers and their organizations. ORCID has an important role to play in this work, but it can’t – and shouldn’t – try to solve every problem associated with researcher identification. It will definitely (continue to!) take a village to increase adoption of ORCID by other researcher communities but, in our view, it’s worthwhile. ORCID is more powerful in combination with other researcher identifiers than it – or any single identifier – could ever be.

*I left ORCID in late 2019

**Data provided by ORCID