library mgmt tech anguish

we need to have a talk about “censorship”

It’s not news at this point that Midwest Tape (MWT), parent company of digital library resource aggregator hoopla, has a major content problem. Back in February, Library Freedom Project (LFP) and Library Futures wrote a Medium post demanding accountability for the company’s platforming of white supremacist, fascist, homophobic, and disinformation-filled invective (for samples of the titles in question, see the #VendorSlurry hashtag on Twitter). Vice Motherboard, LibraryJournal, and WGBH covered the situation, and a group of public library directors put pressure on Midwest Tape for a meeting with leadership at the Public Library Association (PLA) 2022 conference in Portland, Oregon. At the Massachusetts Library Association (MLA) 2022 conference held just a few weeks ago, company leadership was back to speak at a one-hour session hosted by MLA’s Intellectual Freedom and Social Responsibility Committee (IF/SRC). I wasn’t at PLA, but I was able to attend the MLA talk, and let’s say I was not impressed by what I heard.

Before I get to what happened at MLA, some additional background: LFP members not only worked on the Medium statement that went out into the world and spoke to media about the problems but also sent letters directly to MWT/hoopla CEO Jeff Jankowski. We received no response, but found that library directors who initiated conversations with the company about potentially unsubscribing did; this isn’t really surprising, given they aren’t missing out on money from people like me (wrong market) and LFP as an organization. The directors who spoke to company reps were disappointed by what they heard. It seems like it’s been a hodgepodge of hoopla folks falling back on their position that they must be “neutral” and work in accordance with the ALA’s Library Bill of Rights, a bunch of promises for functionality that both have no timeline for implementation and don’t address underlying concerns, and an oft-repeated desire to provide as much content as is possible for the platform to do. What was presented at MLA was all of this, plus a weighty tone of defensiveness that managed to make things even more off-putting.

At MLA, we heard from two customer relations types about the discovery of this content on the platform from their perspective and then a whole lot of half-explained statements about how they might fix the problem. This took up 2/3 of the allotted time in the session, which was a failure in and of itself as the room clearly wanted far more time for Q&A. I can’t recap everything that was said verbatim, but I did snag this photo of one of the slides in their deck:

This says “As a partner to public libraries, we aim for neutrality,” showing a screencap of the aforementioned Library Bill of Rights. On the one hand, this shows an impressive lack of understanding of current professional discourse; on the other, we need to have a serious chat as a field about whether or not we want to keep this document in its present form. In particular, we need to talk about these bits regarding “views”:

  • Materials should not be excluded because of the origin, background, or views of those contributing to their creation.
  • Libraries should provide materials and information presenting all points of view on current and historical issues.
  • A person’s right to use a library should not be denied or abridged because of origin, age, background, or views.
  • Libraries which make exhibit spaces and meeting rooms available to the public they serve should make such facilities available on an equitable basis, regardless of the beliefs or affiliations (views) of individuals or groups requesting their use.

Applying this inclusion of “all views,” as hoopla insists they do, means you get books in their collection on Holocaust denialism (Debating the Holocaust), COVID denialism (Fight COVID with Melatonin), conversion therapy (Attack on the Family), and defenses of the alt-right in their own words (A Fair Hearing). They acted to remove at least some of the Holocaust denialism texts from their platform, but hundreds if not thousands of titles in these other categories of xenophobia and disinformation are still on there, and they seem to believe it is the responsibility of individual libraries to report these books when they see them. That said, there are very limited collection management tools for library staff to use to actually deal with this trash – the default is for it to be included in hoopla’s library – and in the words of the hoopla reps at MLA, over 20,000 new titles are being added to the platform every month.

Saying “we need to present all views” is a close relative of the old chestnut “I don’t agree with everything on the shelf in my library!” which I heard a panelist announce at another MLA presentation last month. Both show a woefully facile understanding of the relationship between intellectual freedom and social responsibility, something that underscores the ineffectual and performative nature of diversity, equity, and inclusion statements and commitments to anti-racism that suddenly appeared for the first time in the summer of 2020. It makes me seethe when I hear librarians defending shelf space for this propaganda and hate speech while they also prattle on about their buildings being welcoming and accessible to all. Some of them – and hoopla is also along for this ride – will also go on about how we need to keep this stuff somewhere for a mythical public library patron who needs it for “research.” Yes, there are disinformation researchers in academia; no, they do not rely on hoopla’s collection targeted at public libraries (and in many cases, targeted at K-12 schoolchildren, as they depend on public library resources where school libraries have ceased to exist). Next door to this, you also get the librarians accusing each other of censorship by removing titles like the #VendorSlurry from their collections; I’ve been told I’m a censor for weeding, and how we need to keep weeding as “bias-free” as possible. Well, frankly, fuck that, because I am biased against xenophobic lies. I’ve said this before and I’ll say it again: if that makes me a censor, I’m happy to be one.

Zooming back out, the Midwest Tape folks talked about a few of the actions they’re taking in response to their content crisis, in particular the establishment of a content review board and the improvement of their “algorithm” (in particular, it sounded like adjusting how relevancy ranking works). I have to say, it’s always remarkable to me when I see tech companies operating in 2022 like they have paid absolutely no attention to the ever-growing body of research and recommendations coming out of critical technology/data studies, or even just what’s happening in the industry in general. (I feel like MWT would deny they are a tech company, even as they are acting an awful lot like the toxic bros of surveillance capitalism.) Facebook has a content review board that is famously a useless front. Our presenters informed us that 40% of the board would be made up of library-adjacent people – not even actual librarians, per se, but people who have had some sort of experience in the field, whether it be taking one class or working at another library vendor. The other 60% sounded like a random assortment of people at Midwest Tape. When pressed for details, like whether we could know who is on the board, we were told probably not because hoopla is concerned for their “safety.” This is quite the sidestep. Even if one assumes that means safety from fascists (and not safety from librarian activists, which is honestly what I think they mean), they are putting librarians making content decisions in the line of fire as they are public servants whose identity is necessarily and unavoidably known.

So, this leaves us with the technical fixes they did a whole bunch of hand-waving about. hoopla loves telling everyone about how their content selection happens via a mix of “human and automated processes,” but will never give anyone a lick of detail about what either the human or the automated processes in question look like. Now, one could argue this is the result of some NDA enforcement, but if that is the case – if we truly cannot know what these processes look like – that means libraries should not be using them. Just as you wouldn’t let the patron who got their umbrella stuck in the book drop suddenly take over your Ingram ordering, libraries cannot rely on random unknown actors, whether they are “human” or “automated,” to do collection development. hoopla should understand and respect the need for transparency on how these processes work; a library’s collection development policy and the staff resources spent on selecting and deselecting are core to the operations of the institution, and some might say a library isn’t a library (and is instead just a meaningless collection of junk, which is also what adding 20,000 titles per month to a digital library will get you) without those things in place. Until they can wrap their heads around this and stop with the “we’re improving our processes” crap by muttering some sweet nothings about search ranking, there’s nothing to be done here. And, guess what, guys – we aren’t concerned about whether Debating the Holocaust is number #2 in the results or number #2000. What we’re concerned about is why the shit it’s on the platform in the first place.

The MWT/hoopla situation is awful, no doubt, and Library Freedom Project, MLA IF/SRC, and our various allies are going to keep the pressure on. But librarianship has got to have a reckoning about this whole censorship/intellectual freedom/neutrality debacle. If our colleagues are defending hoopla’s selection of A New Nobility of Blood and Soil by a literal former SS Captain, a book not even sold on Amazon, in their obsessive quest for free speech absolutism, where do they draw the line? Books that encourage pedophilia, rape, suicide? I’m not sure I even want to know, considering how disappointed I am in this field already. But if they say they want all views, hey, well, those are views, too. 🤢