BY FRANK RITTER
ORCID is an ID number authors can request from a consortium, primarily created by for-profit publishers. It is nominally free and intended to be unique to individuals. It is designed to be a database primary key to tie all records of an individual together.
Using ORCID is a conflict of interest because publishers often do not clearly disclose why and when they require it. One has required it without telling the journal’s society. Associated costs are pushed onto authors and downplayed.
ORCID solves some problems, like providing a universal contact across a career, but that is not sufficient to require it. ORCID mostly complicates scholarly publication for authors and fixes problems that can be solved by due diligence. It is similar to requiring a social security number (SSN) for non-tax-related activities.
The arguments for ORCID, so far presented, are often misleading. For example, Wilson and Fenner note some strawmen: “Have you ever tried to search for an author, only to discover that he shares a name with 113 other researchers? Or realized that Google Scholar stopped tracking citations to your work after you took your spouse’s surname a few years back?” No, I have not searched for an author that shares a name with 113 other researchers because I use their affiliation and field. With Google Scholar, you can create an account and use that to link your publications. Managing Google Scholar takes about an hour or less a year. But we do not require authors to use Google Scholar. ORCID’s use appears to support decreasing community ties and due process by editors, whereas in the past the authors and reviewers were people, not numbers. Not noting up front that a journal requires an ORCID number is a type of dishonesty.
I examine here a longer paper on ORCID, addressing a few of its claims (a more complete treatment is available) to show that they are misleading or wrong. Taken as a whole, I believe these arguments are not correct or honest. I think that publishers really want ORCID as a primary key, like a SSN, to aggregate and sell author data.
Claim 1. ORCIDs protect your unique scholarly identity
Wilson and Fenner assert that author name ambiguity creates problems for everyone in scholarly communication. This is not true—unless I am not part of everyone. It has not created problems for my students, who I would like to have included in everyone. Folks without unique names can and do make their own pen name, like David Cranfield Smith or Yu April Apple Yan. A sixteen-digit number is not easier to learn.
There is indeed no easy way to identify the publications of a university’s faculty and students. You have to read the papers to understand what and where scholarly work was done. This understanding does take effort and some sense about who is involved and who is not. We have seen the problems and warping effects of numbers-based research assessment, and other reporting purposes are too vague to take seriously. If information is unavailable because it takes too large of an effort to locate or identify, perhaps it is not valuable. Researchers who would use an ORCID number to find collaborators are using a far weaker tool than Google Scholar.
Does ORCID “help authors get credit for all …publications”? Penn State used the “Pure” system to summarize my work, without asking or even telling me. Thus, Pure was used by Elsevier to get credit and payment for my work
Claim 2. Creating an ORCID identifier takes thirty seconds (and is free).
While true, this claim is misleading. It will take additional time to maintain and use an ORCID identifier. It will take hours for you to do the work of building connections that they can sell.
Your university also has to pay to provide these numbers. Not a large amount, around $5,000 per year. Just for the US’s two thousand colleges and universities, this is $10 million a year, all to solve a problem for for-profit publishers.
Claim 3. ORCID is getting big fast.
ORCID is popular because for-profit publishers are requiring it, not because most academics are choosing to get one.
Unintended consequences
ORCID is really an attempt to make a primary key for for-profit publishers’ databases. This key will support a wide range of activities, mostly benign, but misuses of ORCID will lead to similar ills as that of misusing SSNs and driver’s licenses.
There is currently no protection from a US agency or foreign government from using these numbers to track who publishes inappropriate literature, cites malcontents, or to create dossiers of your work. This can also lead to diluting what you think is your best work when summaries of your work, such as for a grant, are automatically generated.
Also, the tie to ORCID will help for-profit publishers to find your papers on the net and then require the papers to be taken down. So, you may get less credit unless someone pays.
Summary
Thus, I claim an ethical and personal privacy objection. ORCID is an attempt to provide a universal identifier to support both appropriate and inappropriate tracking. It takes more than it gives to authors. I do not submit work for publication that requires ORCID. It is ok if you want one, but requiring it is not acceptable. When is something so good for authors that it has to be required?
Frank Ritter is professor of information sciences and technology and of psychology at Penn State. He is also an associate editor of Human Factors, a journal that does not require an ORCID number to be an author but allows it.
Really? I don’t know about the costs – maybe they are bad – but otherwise ORCID seems like an obvious good. It can track not only publications but also grant reviews. This makes it so much easier to evaluate people for hiring, tenure, and promotion because everything is verified and available in one place (as opposed to trying to read saved emails thanking people for reviewing grants, etc.).
And yes, name confusion absolutely is a problem, as are changed names. I know academics who don’t change names upon marriage simply because they don’t want to lose their publication history. It’s fine to not change your name, but it should be a choice, not forced on you by your job.
Is your recommendation to solving name confusion literally to make up names? Was that serious?
People can already track down those publishing stuff they don’t like. As you yourself pointed out, it’s easy to track down scholars and what they publish. You’re contradicting yourself here, since you earlier argued that ORCID doesn’t make it any easier to track work done by scholars.
thanks for reading.
Really. Please let me address the points you raise. I do not claim that there are not advantages to ORCID, but that its use is in many cases onerous, it represents gray work aka unfunded mandates, there are unpleasant secondary effects, that the arguments for it are often specious. And, if it is so handy—it should sell itself and not be required.
There are some subtleties that are missed I believe in your post.
The costs can be real, you should compute them. The costs are pushed onto authors to maintain the links. If the costs were light, then let the reader or the publisher pay them, and not require them of authors.
To the extent that people can be tracked, it will, for example, require small conferences now to work with ORCID to give the credit. It may be easier for you, but not for the small publishers or workshop organizers.
Academics who won’t change names upon marriage are not fully helped by ORCID, unless people only search by ORCID number. I will still have to know people’s names. And, yes, as David Canfield Smith (https://en.wikipedia.org/wiki/David_Canfield_Smith) showed, you can get a unique author name if you wish when you start with a common name. I have moved from Frank Ritter to Frank E Ritter as an author to be slightly less common. I am less memorable as 123-45-6789, I suspect, to people but not to machines.
As for tracking, as you yourself note, people can already track down those publishing what they do (or don’t) like, and what they do or don’t review. But it is more difficult as current events are showing. You are contradicting yourself in noting that people need help in hiring when they could just do the work to understand the candidates.
But, there are clearly different use cases and different users and stakeholders. I’m just making the case that large publishers are pushing work onto authors and researchers, and that the use of ORCID should not be a requirement, particularly if it is so handy—it should sell itself. Brandolini’s law also applies.