Yeezy Taught Me

In a talk given during London Open Research Week 2021, Arthur J. Boston, the Scholarly Communication Librarian at Murray State University, described parallels between the music industry and academic publishing, especially as it relates to open science and early career researchers. Below, you can find both the recorded video and the transcription in-line with the slide images.

When I say that Kanye West explains scholarly communication, I don’t mean that he literally explains it.

Although I can point to his actual public comments that do directly speak to our concerns, like OER.

Or to his lyrics that speak directly to how I have sometimes felt trying to get published in peer-reviewed journals.

What I mean to say is that Kanye’s way of being as an artist, a creator of content that has transitioned from physical to streaming formats, as someone who has relied on a label to distribute his music to relevant venues. The way he expressly sought to be signed with the most prestigious label of the genre as a young artist. This explains scholarly communication. The way he tests ideas out with colleagues in private, tests them out further with colleagues in public settings, the way that he is not immune to criticism and online comment. Not immune to both competition and collaboration with peers. This all has helped explain scholarly communication to me over the years.

Hip-hop and rap creators are subject to similar structures as creators of academic scholarship. Whether it is reporting scientific findings or recording a track, both of these end products are fixed in a tangible format and are therefore under copyright. Both groups create for an audience, and if they want to maximize their audience size, often they will rely on a distribution channel, a middleman, or a label to do the work of making the content findable by the appropriate audience, to get put on the library shelf, the retail store space, to be findable by EBSCOhost, by Web of Science, on Spotify’s New Music Friday playlist.

In exchange for a label or publisher distributing this content to the maximally sized, appropriate audience, these creators are often asked to sign over their author’s rights, or license some part of their copyright, to a middleman, a middleman that uses this exclusive right to maximizes their profit, because this is their business.

In a previous era, pre-web, an aspiring artist or scholar did not have a strong practical choice in the matter, given the structure of the systems of distribution.

Today, an artist or scholar who wants to remain near maximally independent, while still finding an audience, can choose to upload their preprints to somewhere like bioArXiv or Humanities Commons, or their mixtape or loosies to SoundCloud, DatPiff, or Bandcamp, and then rely on social media and other less formal channels to get that message across.

We’re still in a sort of transition period where some scholars and artists are operating in a digital online world, but maybe began their career prior to the mainstreaming of all the online-ness, and they are finding ways to bridge the gap between those two points in their life, now that they can reclaim some control over their prior works.

Take Taylor Swift — you know, the famous rap star Taylor Swift — she signed to a label, Big Machine, when she was a teenageer, and her original contract specified that the master recordings of her music would be owned by the label. Recently, Big Machine sold these recordings against her wishes to Scooter Braun. But here’s the cool thing about music copyright. There is copyright over the recordings of the music, and there’s copyright over the songwriting, like the lyrics and musical composition. Swift’s deal had her retain copyright over her songwriting and Big Machine just got the actual recorded music. Think of Swift’s deal as a sort of proto-Rights Retention Strategy. So what you may have seen lately is Swift re-recording her back catalog and releasing these as “Taylor’s Versions.” She is re-recording her classic albums, largely in the hopes that the fans, the stans, and other groups who would seek to license use of her music in different formats will choose the Taylor’s Versions over Scooter’s versions.

I mention this for two reasons.

1. I think it’s funny to interrupt my Kanye talk with Taylor Swift content. Justice for Taylor. I’ma let you finish.

2. A big chunk of working researchers are, career-wise, sort of in this Taylor Swift generation of publishing.

Maybe they began publishing papers feeling beholden to chase authorship in outlets based in part on the prestige of those outlets and in consideration of how that might translate back to their career goals or at least the incentives they felt would get them there, but now find themselves wanting to take more control over the distribution, by making previously paywalled articles open access.

And rather than paying thousands of dollars in article processing fees to go Gold OA, they go the Green route, where they deposit their postprints into open access repositories.

The postprint, or author’s accepted manuscript, is the final version of the article that has been through peer review, but not yet been through the publisher’s copyediting, formatting, and the like. The author’s versions, much like Taylor’s versions, are versions that the creator is able to take back a degree of control over, especially in expanding how it is put out.

However, an author’s version, or postprint, implies that there is a publisher’s version. A situation in which creators wish to exert further control that the publisher did not directly provide, implies that the author and publisher did not see eye to eye to begin with. If an author wants to control when a piece is made available or what goes into it, they may decide to preprint first. For authors who wish to publish in a journal that does not require payment either from them or the reader, an author may publish in a diamond open access journal.

One way I like to think about this is with the example of Kanye’s protege Chance the Rapper. In 2016, Chance released his third album-quality mixtape called Colouring Book. Colouring Book was available for free on sites like Dat Piff and SoundCloud. In fact, you couldn’t buy a copy of it, it was only available as a free download. It was not for sale. In a lot of ways, Colouring Book was just about as close to an open access album as possible. I think equally important, or slightly more important, than the fact that his album was free is the reasons it was free. It wasn’t a compliance issue from a funder or institutional policy. It was because he wanted it that way.

Now I’m going to read you two quotes. One is from Chance the Rapper and the other is from Randy Schekman, a Nobel winner who quit his post as editor at a closed access journal to start up a rival open access journal. What I think you will hear in these two quotes is the same vibe.

Schekman: “A paper can become highly cited because it is good science – or because it is eye-catching, provocative or wrong. Luxury-journal editors know this, so they accept papers that will make waves because they explore sexy subjects or make challenging claims. This influences the science that scientists do..”

“There is a better way, through the new breed of open-access journals that are free for anybody to read, and have no expensive subscriptions to promote. Born on the web, they can accept all papers that meet quality standards, with no artificial caps.”

Chance: “After I made my second mixtape and gave it away online, my plan was to sign with a label and figure out my music from there. But after meeting with the three major labels, I realized my strength was being able to offer my best work to people without any limit on it.”

“I honestly believe if you put effort into something and you execute properly, you don’t necessarily have to go through the traditional ways.”

Here, Chance says you don’t necessarily have to go through traditional outlets, not that you can’t. And Shekman didn’t protest the peer-review process, his own journal is peer-reviewed, he protests the business model which incentivizes editorial models that seek flashy, positive findings, and drive 90% rejection rates, which wastes everyone’s time, effort, and money.

Chance released his album for free, but that did not stop him from seeking success, he simply wanted to maximize his artistic freedom in the choices he made with his music and to not limit the number of listeners with paywalls. But none of this was out of alignment with his career goals. His album won a Grammy that year. But he did have to jump through some creative hoops to be eligible for that Grammy nomination. At the time of his album release, the Grammy rules said that to be eligible for nomination, an album must have become available for sale during the past year and, of course, this album was not for sale.

Chance split the difference. The solution he came up with was to put his album on Apple’s new Music subscription service exclusively for two weeks, where listeners could technically pay in order to listen to the album. But if a listener didn’t wish to pay, there were still plenty of legal and artist-approved routes to do so. What I want to emphasize is Chance’s creative approach to dealing with a silly rule, which enabled him to meet compliance, but still ultimately was able to stick to his original objective of artistic freedom and literal album price freedom. I think many of you can empathize with dealing with silly rules when it comes to research and the credit accrued for it.

I think researchers, especially grads seeking jobs or current faculty seeking to keep jobs, can do more to be like Chance. Publish in Diamond Open Access outlets, deposit your postprints. Do things to get credit from the gatekeepers if you must, but then do what you can to ensure the potential audience does not have a gatekeeper of their own, in turn. If you’re an editor or senior ranking academic with an ability to change things, move to be more accepting of researchers who post preprints first or only. Try out new models, like what eLife began last year where they no longer accept manuscripts for review unless they’ve already been deposited as preprints on an open access server or arxiv.

Okay, so at this point, you probably see some of the clear similarities between creators of scholarly content and the arts, such as music. But this talk is about how one particular rapper explains scholarly communication. Before I drill down to Kanye in particular, let me just take the moment to mention that hip-hop, of all music, is the perfect genre to use. They’ve known this in English departments for decades how sample-based music provides a perfect analogy to introduce the tradition and idea of citation and quotation in writing classrooms. Hip-hop is also very good especially these days because of rappers’ use of mixtape sites to freely circulate albums without label control. DJs and producers often release stem packs and drum kits, which are files full of samples and sounds they’ve cut or created live, that other DJs and producers can download and reuse however they see fit, a practice which I’ve compared to researchers putting their code, methods, and data sets in places like OSF, where other researchers can download, inspect, and re-run for themselves. I think the most common way people think about parallels with music is with piracy, where: peak CD costs in the ‘90s and Napster stand in for the serials crisis and Sci-Hub. If all that sounds familiar, it’s because we’ve talked that one into the ground, so I won’t dwell on it.

Let’s talk about Kanye’s rollout for his seventh album The Life of Pablo. First, he played a version of his album on a Friday at Madison Square Garden. He performed slightly different versions of two tracks the next day on Saturday Night Live. Then early on Sunday, a still slightly different version of his final album was released. But even after the official release, he kept tweaking the album. It seemed fine and normal to present alternate versions of songs ahead of the album release, sort of like how a scholar might present a draft of a paper at a conference and use feedback gotten there to improve the final product. But Kanye was adding Frank Ocean verses here, and dropping Vic Mensa verses there, and adding new songs after the final album was said to have been released, and it got kind of messy, especially for those talking about the album in the culture, on blogs, and in professional music criticism. If you wrote about the album on day one, you’d often have to go back a week later to note the changes, or otherwise risk a disconnect in the minds of future readers.

This really spoke to me about issues in publishing like when a journal might issue a change or withdrawal to an article which perhaps may have already been cited by subsequent papers. Earlier, I spoke about posting preprints, and even this has the potential to get messy in a similar way, such as when an author has preprints that sit alongside publisher versions of articles, which might vary in their content. Some people have suggested that in these cases, the best practice is to cite only the publisher’s so-called version of record. But others, noting that sometimes people are choosing to cite a preprint version, want to keep a more comprehensive scholarly record intact, and rather than obsess over version of record, they say that what is needed is the development of mechanisms to deal with the Record of Versions. There has been some good momentum in this space since; bioArxiv now allows you to “follow” preprints, which will send you alerts anytime a change takes place, and Zotero now alerts you whenever any article you’ve saved in your library becomes retracted. And so on.

So right now I have a minor piece in a journal that has been available since September 7, as an “in-press, corrected proof.” And somehow, it’s not technically published? On the one hand, I’m incredibly relieved to have it out there, available, but on the other hand I still am waiting to get to claim it for the year. I should mention this is not a diamond OA outlet or library published journal, but I do have my postprint ready to roll. Instead, this is a closed-access journal with a journal impact factor listed right there on its home page.

Journal Impact Factor, of course, is that number reflecting the yearly mean number of citations of articles published in the last two years in a given journal, as indexed by Clarivate's Web of Science. (I like referring to it Clarivate’s Web of Science; it’s very exact, sort of like saying Dr. Frankenstein’s monster, rather than just Frankenstein.)

Anyway, the more citations a journal’s articles receive within that set period of time, the better it works out for their impact factor to rise. A lot of folks, people supportive of DORA and Leiden, would probably prefer this thing would just disappear because of its potential and proven ability to distort the scientific record by giving journals an incentive to favor the publication of article types that are prone to citation, leaving out other types of worthy content from the scholarly record. But to return to my article, which is available to read without technically being published, I like to call this The Hotline Bling strategy.

One of the other big rap albums to come out in 2016 was Views, by Kanye’s frenemy Drake. Just two years prior to its release, the Billboard 200 had changed their algorithm to better account for streaming. The Billboard 200 is sort of like Impact Factor for music. Sure, an album may be a top ten record for the week, but does that mean it’s any good? So Billboard changed their algorithm so that when tracks from an album received a designated number of plays, that would be counted as “one album equivalent sold.”

So Drake, a guy who likes to be number one puts out a very bloated 17-track album, and to top it off, he tosses on the nearly-year old single Hotline Bling as a bonus track. Hotline Bling was ubiquitous for a while, and previous to its inclusion on this album, on its own, it had accrued 573 million streams, so by the time Views dropped the album had no choice but to go number one on Billboard.

So, thank you to my journal for putting my article out well ahead of its actual publication, but I know what you’re doing, why you're doing it, and it’s greasy. Just think, if a journal puts out articles for months in advance of their publication, they’re creating artificially perfect conditions to boost the number of citations that occur in the eligible two-year time window that Impact Factor calculates.

Colouring Book, Views, and the Life of Pablo. These are all examples from 2016. But we’re in 2021 now and all of the things I’ve mentioned so far still hold up well. I’ve been thinking about things in this way for a while, and so I put that lens on everything I come across in music now. Kanye held three different listening parties in advance of his most recent album Donda, all of which featured pretty different versions of the album. His in-house producer, Mike Dean, noted how Kanye was using the live feedback of the audience to help guide decisions. And it’s stuffed with 27 tracks, of which he’s already been cutting and adding features on again. The live stream events were Apple Music exclusives and both the events and the album had, at the time of release, broken Apple’s streaming records for the year.

Bjorn Brembs had this great tweet in September, where he said “citations are not a proxy for quality. If they were, I would have become a better researcher in 2018. Instead, I published a controversial paper as one of 70 authors.”

I really appreciated Brembs’ candor, because it made me think about Donda’s streaming successes, about all the factors that are helping to boost the quantitative measures that come together to give Donda this picture of success. The way it’s been built for streaming, with its 27 tracks, 30 featured artists, and with controversial figures attached to the project.

At the second listening party for Donda, fans went nuts because Jay Z was featured on a song called Jail. The two artists had not been on a track together in a long time. But at the third listening event, Kanye dropped Jay Z from the song and added DaBaby who had just made some pretty horrific homophobic comments at a concert, and Marilyn Manson, whose been credibly accused of abuse by several women. I know Kanye was trying to feature these two figures in keeping with the album’s theme of redemption, and I do think folks should be allowed the opportunity to seek public redemption, but these two have not yet sought to earn this in any meaningful way. Kanye has to know this sort of controversy probably contributes to the album’s overall streaming numbers, but it leaves a sour taste in my mouth on an album I otherwise quite like.

Now, on the final album that is on streaming services, the version of Jail featuring Jay Z is the one that’s listed on the album proper. At the end of the album, the version of Jail featuring DaBaby is included and named Jail, part 2.

I would have preferred this alternate track did not exist at all. But I’m somewhat thankful that if it must, it’s been done in a way that is very convenient to omit from my head space.

When the final track on Donda plays before Jail Part 2, I stop and move on to something else. In this way, I don’t hear DaBaby, and more importantly perhaps, Spotify knows I’ve not played it, the record label sees data that it’s not been played when the rest of the album has, and the algorithms and decision makers get some sense, from my account, that DaBaby is not being reward, which in turn means his music is less likely to appear as a recommendation for me on Spotify, on the overall listening charts, and on Spotify curated playlists.

The point is, for me, the way that song plays operate in a music industry now mostly centered around streaming, are really quite like the way that citations work in mostly-online publishing and algorithmically-based discovery systems.

The more citations an article receives earlier on, the higher it will display in a Google Scholar search result, and thus gain further citations still. As scholars, we have a choice about who and what to cite in our own work. And these choices have real ramifications we should consider, ramifications about whose voices are counted, who is left out of the conversation, whose positions of power are reified, and the knowledge types we accept or overlook.

I won’t give a lecture on citational politics here; it’s outside of my expertise. I simply want to point out that for audiences who have not thought about, or not thought very deeply about citational politics, an example like whether to stream someone like DaBaby or R. Kelly or any number of musicians may be a good introduction into that topic.

Rather than diving right into the scholarly weeds, this topical sort of discussion might more quickly move things toward understanding. And really, that’s sort of the point and purpose behind all of the examples we’ve gone over today. Even the field of discussion for newcomers to this space, whether it’s students, early career researchers, the public, whoever. Scholarly communication can be a pretty intimidating and overwhelming space to enter into, but it doesn’t have to be.

The conversation doesn’t have to be about Kanye, or rap, or even music. It can be video games, pro wrestling, news reporting. All of these spaces have creators, mediators, incentives, economics, metrics, and anyone who has a deep personal interest in these spaces is bound to have intuitions, ideas, and experiences that they can use as an entry point into the scholarly comm space, and more importantly, bring different sort of thinking to our longstanding, evolving, and brand new dilemmas.

I’m very curious to know what are the other domains that you all have drawn on when thinking about things like publishing, open access, responsible metrics. What are the industries that you think have gotten things right, that scholarship distribution still struggles with? Who are the creators and artists you admire and why? What can you do in your own work as scholars to better achieve ideals you admire in others?