Code4Lib 2017 and The Cobra Effect
I have recently returned from Code4Lib 2017 in Los Angeles (and a few lovely post-conference vacation days in sunny southern California). This was my first national Code4Lib, and I have brought back much more than a great t-shirt to our Dublin, Ohio, office. My head is swimming with new ideas, and I relished the opportunity to meet many of you in the Developer Network community and to talk about OCLC APIs (among other things, like where to find LA’s best taquerias).
I found the opening keynote by Andreas “Dre” Orphanides of NC State University particularly thought-provoking on the topic of systems design and thinking. These are terms that get thrown around a lot, and I think many in the audience, myself included, lacked a good understanding of what they actually mean. Dre defined systems as models of reality, models being surrogates for things, e.g., a map standing in for a city. Therefore, systems are fundamentally “wrong” in the most basic sense, but in order to solve problems, they must be wrong in useful ways. In particular, they must not ignore the humanity of the actors participating in them.
As an example, Dre cited the cobra effect, which originated from an anecdote set in colonial India. As the population of venomous cobras began to grow, the British government decided to put a bounty on them. After some initial success, the officials realized that entrepreneurial locals had begun breeding cobras, only to kill them and cash in. So, the colonial overlords had designed a system to reduce the cobra population, but it failed because it didn’t account for the human motives of its participants.
How does this lesson apply to library developers? One example is the historical (and to some extent current) practice of libraries measuring success using simple statistics such as circulation figures and gate counts. While these “hard” numbers carry some weight, they fail to account for any value (or lack thereof) that the user experiences as a result of these checkouts or visits to the library. The library developer community can help fill this void through projects like the Harvard Library Innovation Lab's Awesome Box, which allows patrons to declare returned items “awesome.” Is this a perfect measure of end-user value? Clearly not, but tools like this, which wouldn’t exist without the insight of coders like us, are starting points toward new ways of evaluating library success.
While Dre honed in on the unintended consequences of the systems we create, OCLC's own Karen Coombs focused on asking the right questions during the software creation process. Karen encouraged us to reframe traditional requirements and feature lists as end-user problems that the software needs to solve. This will shift focus away from existing system designs and capabilities and toward more creative, innovative solution paths. If the library developer community makes a concerted effort to do this, we will surely avoid many instances of the cobra effect coming back to bite us (pun possibly intended).
To hear more on topics like this, I would encourage you to attend OCLC’s inaugural DEVCONNECT conference, which we’re hosting in Dublin, Ohio, on May 8 – 9. Many developers from the community will be sharing projects they have created using OCLC APIs. OCLC staff-led sessions will include a view into new OCLC products and enhancements and will cover best practices for building and delivering services, taking a user-centered approach, and addressing the social and cultural issues that impact library developers. It’s a great opportunity to network with fellow developers. And, the conference includes a reception in our newly renovated headquarters, a tour of our state-of-the-art data center, lunch on Tuesday, and, of course, a sweet t-shirt. At only $35, it's the best value in conferences anywhere. I hope to see you there!