Design Speaks

These are the messages conveyed by the design of most library services and buildings: “We are not you. We are not even like you. You have to think the way we do.” From the jargon on signs in our buildings to the unexplained options in our OPACs, libraries are indelibly stamped with librarian-think.

Modern libraries are the product of centuries of development, so when we design library services, for good or ill, we are conditioned by the weight of history and constrained by legacy designs. The design of the Library of Congress’s Jefferson Building is a paean to authoritative knowledge. Classical columns and sculpted marble staircases lead to aristocratic halls beneath ceiling-engraved memorials to great writers and thinkers. Elegantly appointed reading rooms, all scrolled woodwork and polished tables, await the pursuer of knowledge. Surely this is user-conscious design at its finest? What stony-hearted patron could resist?

Yet patrons are resisting in droves. The Jefferson Building’s patron areas are severely underused, although reference transactions by phone and email are thriving. At the same time, Flickr, Amazon, LibraryThing, and other design-conscious web properties attract information- and entertainment-hungry users by the hundreds of thousands.

Buildings fail, in part, when their design proclaims, “I am to be admired, not used.” Patrons have no place in such design; it discourages them from embracing, enjoying, and engaging with the library’s space and services. The same goes for the design of the web component of services.

Every design decision conveys a message, intentionally or not, that governs how people interact with the designed object or service. Good design communicates the message attractively and usefully to its recipients, and the message encourages them to think and do what the organization would like them to, whether in a building or on the web.

On the web, tagging and its cousin, folksonomy, though not without pitfalls, let users draw on their own experiences and mental models to organize their information worlds. We miss an insight when we argue endlessly about uncontrolled vocabularies, the tyranny of the masses, and the privacy concerns surrounding public hosted services. Offering users control and a sense of mastery, letting them carve out their own information landscapes individually and in groups, is a profoundly courteous and inviting design choice.

LiveJournal lets people describe their mood on each post. Amazon welcomes opinions on its wares, even heated opinions. These options acknowledge that people feel as well as think. Smart Internet properties that design ways for people to express what they feel, even in as simple a fashion as Digg’s thumbs-down “bury [this link]” function, keep them coming back.

Good design appeals openly to emotion. Google’s April Fool’s Day pranks, cute “doodles” for holidays, and famous birthdays endear the company to its users. Consider also the development of bookmarking services. The cleverly named del.icio.us (see LJ netConnect, Summer 2006, p. 16–17) started out as a boring, squarish, tone-deaf design that only a hard-core geek could love. After Yahoo! took it over, its horrid browser-default serif font was replaced with the friendly sans-serif Verdana, and subtle shading and font-size tricks increased its visual appeal without detracting from its simplicity. A similar service, ma.gnolia, despite being late to the party, made inroads on the userbase with a prettier, friendlier look (see also Product Pipeline, p. 14–16).

The huge number of search results touted on a Google results page, for example, is a loose estimate of what’s available at best; Google cuts off the results it will actually display relatively rapidly. Even so, the single search box and the McDonald’s-like “billions of web pages searched!” message figure prominently in Google’s design because they create a tantalizing illusion of totality. Would anyone believe “it’s all on the Internet” if design choices at Google did not make that claim subtly but repeatedly?

It is no accident that the lead developers and businesspeople at many hot Internet properties were the sites’ creators and first users. Nor is it surprising that many of these sites design in some form of open communication between users and service-owners—be they user forums, official or unofficial web logs, or chat rooms. They even design error messsages to emphasize the humanity behind the technology: the shrugging plumber at Bloglines alleviates the sting of service downtime with plain-folks humor, and Flickr’s anthropomorphized servers “get a massage” when they are out of service.

In libraries, patron control and convenience have been afterthoughts at best. This is not least because patrons have too rarely had direct access to underlying structures of the library, such as the stacks or the MARC data library catalogs. In addition, our skills and training distance us from our untrained users, which leads us to think in terms of our own needs, just as software programmers notoriously design software for their own convenience rather than ours. As a result, patrons defect to information sources that respect and augment their thought processes.

Our print stacks overcame the impression of being disorganized long ago; the basic principles of collocation and classification give our patrons many fruitful browsing experiences. Why are our online collections and services bewildering?

A good deal of this confusion can be laid at the feet of vendors: those that insist on slapping their patron-opaque branding all over everything; vendors that refuse to participate in meta search and link-resolving initiatives or offer APIs that let us create patron-friendly interfaces to their data; vendors that lock up libraries’ own catalog records so tight that we cannot create interactive services even when we want to.

We are not, however, without guilt. When our web site designs pile link on disconnected link like a yarn-basket that the kittens have played in, our patrons come to believe that the information we offer is just as tangled. When our OPACs and journal-search services canvass only our own holdings and are too stupid to analyze patron queries to point them in the right direction, patrons think we do not have what they seek.

Our style of information-seeking can be regimented, joyless, and lonely. Yet play and experimentation are a key way human beings learn. The lone-wolf researcher is not by any means typical, and most work can be accomplished in different ways according to the skills and preferences of the worker. We too rarely tell patrons, “Well, there are several ways to go about this.” When was the last time a patron said that using an OPAC was fun. Our database search interfaces aren’t designed to encourage patrons to play around with search options. And our services are not designed to enable easy sharing of finds with a patron’s social world. When the design of our services isolates, frustrates, and bores our patrons, they will see us only as a last, desperate resort, if they see us at all.

Much information-literacy training falls into this trap, too. The mere idea that one must be trained to use library services intimidates some patrons, distancing them from the library. Other patrons who (however falsely) feel confident about their usual information behaviors find our insistence on training condescending or even controlling. Still others, intimidated by the inconsistency and complexity of our building and service designs, cling too closely to the specific keystrokes and floor plans we teach them and are left helpless in the face of change. We need to design our training with patron preconceptions and experiences in mind, weaving play and fearless experimentation into the fabric of our training sessions.

These design flaws and mistaken messages are hard to overcome, both for systems librarians trying to develop better-designed systems and for public-service librarians trying to help patrons navigate existing systems. The first step is to recognize the problems. The second and more difficult step is to refuse to defend them; tradition, authority, “the vendor does it this way,” none of these excuse bad design.

Fortunately, some librarians and library allies are designing services that communicate the same messages that users of hot web properties find so attractive.

The Ann Arbor Public Library’s (AADL) web site design (see “Catalog Annotations,” LJ 9/1/06, p. 32), with its clean layout, clear navigation, and low-key color scheme, sends an immediate message of helpful reassurance. Even better, the events weblog on the front page invites patrons to add their feelings and their knowledge to the site via comments—of which one post (about a gaming tournament) garnered 78! Nor is AADL the only library to join the so-called “biblioblogosphere.” Library and librarian blogs humanize and demystify the profession to our patrons and our funders; by design, they invite patron interest and engagement.

Casey Bisson’s WPopac, which filters the library catalog through blogging software, adds patron power to the design of the traditional OPAC. When patrons need to keep or share the record for a book, they just copy the URL. After all, everyone links to Amazon as a surrogate for book metadata because Amazon had reliable, persistent URLs long ago and OPACs did not. Patron comments on records? Any number of attractive visual designs? They’re right there.

WPopac offers users power over information retention and sharing and an opportunity to share opinions. Whether efforts like this suffice to revitalize the OPAC remains to be seen, but certainly the design sends patrons an inviting message.

As well as letting users share permanent links freely, we have made progress about opening up linking to the appropriate copy. Link resolving, in which servers pass citation information in URLs so that an appropriate digital copy of an article can be supplied to the patron, cuts through the thicket of database interfaces that separates patrons from articles they want. Until recently, link resolvers only worked from within database interfaces. Now, however, pieces of the infrastructure are moving onto the larger web, empowering librarians, bloggers, and enthusiasts to create article links that “just work” for any web surfer affiliated with a library with link resolving software. Although many details remain to be worked out, design progress on such pieces as OpenURL, COinS, and unAPI bids fair to smash the barriers between our resources and our patrons, sending a clear message that we have our data universe under control.

Zotero is an especially elegant barrier-smasher tool from George Mason University’s Center for History and New Media. This plugin open source Firefox browser automatically captures and stores metadata from many OPAC and database pages, and it allows easy cut and paste from pages the plugin does not yet understand. Captured citations can be sorted, annotated, searched, browsed, and shared with others. Loosely modeled after the design of popular music applications, Zotero makes citation management a joy instead of a chore.

The University of Pennsylvania’s PennTags social-bookmarking service inserts the library into patrons’ personal link caches. Librarians designed information-literacy training right into the service with a light, student-friendly tone. “All right, we’re librarians,” says the service’s tagging-tips page. “We can’t resist pointing out that we’ve organized a lot of information over the past few thousand years, so we have opinions about this kind of thing.” Even the cute tagged shorebirds at page top communicate humor and humanity to patrons.

All these efforts confirm that design isn’t about glitz and glamour. When done well, it communicates our values, our services, our beliefs, and our abilities to our patrons. We cannot always control the design of the services we offer, but we can ensure that the designs we do control convey the sincere, friendly, and inviting messages that people want.