In nearly every one of my graduate reading seminars, at some point, we were asked to read Professor Elke Duncker’s 2002 Joint Conference on Digital Libraries (JCDL) presentation, “Cross-cultural usability of the library metaphor” [1]. It’s an excellent academic work that my professors typically used to kick-off discussions on “quick-and-dirty” ethnography in HCI research, and value-centric design.

The JCDL is a highly-regarded, international joint conference of the ACM and IEEE. As such, the proceedings are behind a pay-wall. With some Google Scholar searching, though, you’ll turn up a bunch of free copies helpful results. So, huzzah!

I’ve actually made it a point to read the paper several times outside of class because of how compelling I find some of the ideas it presents. I remember while studying for comprehensive exams I kept thinking it would be amazing to get a value-centric design question so I could let loose with all the pontificating I had done inspired by Professor Duncker’s work.

In the paper, Duncker analyzes the culture and values of an indigenous population in New Zealand, the Maori. The community has Polynesian origins stretching back over a millennium, and is rooted in tribal genealogically-focused traditions, especially the sharing of oral histories through story-telling. Duncker presents the notion of “tapu” (i.e., sacred) traditions, which dictate the audiences, environments, and rituals wherein Maori histories can be communicated and shared. These traditions are highly guarded, and the sharing of tapu knowledge outside of very specific contexts is disrespectful, and a source of great potential conflict in contemporary New Zealand.

In this context, Duncker discusses the perception and use of libraries and museums by Maori people in New Zealand’s westernized urban centers, specifically the cultural incompatibilities and difficulties encountered by Maori university students.

In true ethnographic style, Duncker drops a handful of really great quotes from Maori students reflecting on encounters with stodgy and uptight “Pakeha” (i.e., white) librarians.

Rules against chatting inside of libraries, the perceived “clinical” nature of librarians, and seemingly arbitrary organizations of resources are opposed to traditional Maori models of knowledge sharing and archiving. At best libraries are viewed as difficult to navigate, and at worst they’re seen as an affront to the tapu nature of certain information. The overall result is the cultural alienation of Maori people in New Zealand, a notion that Duncker implies could underlie the socio-economic difficulties experienced by segments of the Maori community.

Following an ethnographic overview of the Maori people, Duncker presents an experimental methodology for assessing the usability of digital libraries and organizations of information, through a Maori value-centric lens. Duncker argues that the design of digital libraries based in western traditions is implicitly burdensome on other cultures that may not share an understanding of the metaphors inherent in such designs. From this perspective, Duncker argues that levels of cultural abstraction are necessary to the design of libraries and the information they archive if they are expected to be globally usable.

Professor Duncker’s analysis is a powerful and rigorous one, but in her presentation of design implications she makes one point that I can’t help but be diametrically opposed to. Specifically, Duncker states:

"Culturally relevant digital libraries for indigenous cultures would have to offer a number of features: Firstly, they need to accommodate locally specific needs for tools such as access restriction tools for Maori content." [1]

Most of my classmates had no problem with this statement, but I found it deeply troubling. In fact, with that on one sentence, I realized that Duncker’s paper, which I love, was perhaps conflating cultural value-centricity with what I ever-so-studiously called “value-amelioration” (i.e., pandering) and, in doing so, making some potentially disastrous recommendations.

In my mind, value-centric design is a proclamation of the need to be cognizant of cultural biases in order to build robust and usable systems, not a call to build systems that necessarily promote those cultural ideals. Take for example, the discussion Duncker presents of theft of library materials considered to be sacred, by Maori people. These thefts are motivated by cultural opposition to the sharing of certain information outside of the Maori community, and are partially in retaliation for what is viewed as the unfair appropriation of Maori culture by westerners. Plainly speaking, in such a context, I would be advocating for the design of anti-theft systems and mechanisms for enhancing the availability of such materials, not the design of access restrictions around certain content. Maori cultural values be damned.

My opinions on this matter were unpopular among classmates, and I think my ideas were construed as aggressively insensitive to our world’s diversity of cultures and deeply-held beliefs and values. I immediately checked myself, fearing that I was being close-minded about what is “right” and “wrong”. After all, I don’t fancy myself a moral absolutist in any sense. But I couldn’t find any resolution.

I’ve been Godwin-ed a number of times throughout grad school… Probably less because my views are actually super controversial, and more because I can really get into some rants on certain topics (big surprise, right?), and Godwin provides a really effective out when it’s a nice day out and you’re sick and tired someone in your cohort going on and on about just about anything.

I was quickly Godwin-ed for ostensibly arguing that my values were morally superior to anyone else’s, and the class moved on.

I fully appreciate that my views can be seen as high-and-mighty, in the sense that I do believe restricting access to certain information on the grounds that it may offend a specific cultures’ sensibilities is patently wrong and contrary to universal progress. I don’t believe in much, but I do believe that when it comes to information, the flatter the structures that dictate its availability, the better. But because I recognize the value in fleshing out these types of things until they are clear in one’s mind, I offer the following thought-experiment:

Middle-eastern cultures are notoriously and aggressively patriarchal, and it doesn’t take much searching to find reports of various human rights violations and institutionalized misogyny in that part of the world. Deeply ingrained cultural practices explicitly devalue women. The use of cheap mobile phone social networking applications and encryption has radically changed the ways in which young people, especially women, communicate in an otherwise stifling environment, but are looked down upon a great deal by other segments of the community at large.

A value-centric designer of a system meant to support either side of this cultural schism would find themselves in a tricky situation; one would be explicitly denying the validity of long-held cultural values by designing systems to empower a traditionally marginalized segment of the population, but building systems that perpetuated such cultural values would be complicit in that continued marginalization. Does then a cultural lens that could focus such a blurry situation even exist?

I don’t think so. No localized culture, anyway. I hope that eventually our flattening world will spawn a truly global culture with shared ideals and universally accepted values… But then again it’s that type of thinking that usually gets me Godwin-ed during class discussions…

Again I come back to the notion of value-centricity over value-amelioration. The idea is to be aware of the biases concomitant with any cultural world-view, not to necessarily design towards their satisfaction. Any culturally specific design will inherently be at odds with some other culture’s values; I mean, it’s those dividing lines that even made those cultures distinct to begin with, right?

Furthermore, a call for value-centric design is not a proclamation that all cultures are equally valid. In fact, it’s the opposite. Value-centric design recognizes that cultures are loaded with biases and fundamental discrepancies (some of which may even be relatively morally objectionable), and insists that in order for designers to design well they must be conscious of that fact. Does this potentially leave designers vulnerable to making unfair judgments? Yes, definitely. But again, in the absence of an as-of-yet non-existent global culture, what wouldn’t?


  1. Duncker, E. (2002, July). Cross-cultural usability of the library metaphor. Proc. JCDL 2002.