of the American Society for Information Science and Technology   Vol. 28, No. 1    October / November 2001


Go to
Bulletin Index

bookstore2Go to the ASIST Bookstore



If this is information architecture, I need a plumber!

by Andrew Dillon

Andrew Dillon will be, until December 2001, at Indiana University (dillon@indiana.edu ). Beginning in January 2002, he will be professor and dean at GSLIS, University of Texas-Austin.

The validity of the term architecture for information work has been the source of some contention among both IAs and traditional architects. I have been forcefully reminded in recent weeks, through my dependence on information technology, just how limited the architecture of information spaces can be compared to the physical structures of our world. Just as a problem with the plumbing or the roof of your home tends to grab your attention and demand resolution, computers in their various forms can make demands on users that stretch the patience and emotional stability of even the most sanguine.

All well and good you say information architecture is just like physical architecture; we rely on it, live in and with it, and use it to improve the quality of our existence. Indeed this is so, or should be so, but what seemed to stretch this comparison beyond its natural elasticity is the manner in which information architectures break down. Normal deterioration through aging or natural disaster is the exception with IT.  Instead, random glitches, unpredictable crashes, dead links, incompatible applications and just plain bad user support are the more likely causes of problems in such spaces.

Installing an Airport card in my Apple and attempting to configure it for use with a Base station (yes, I am a proud Mac user, despite what I am about to relate) I was encouraged by the simple instructions that indicated that the process was transparent and largely automated. Imagine my (and my system manager's) surprise when I realized that this would not work as expected unless I upgraded my operating system first. No big deal you might think, except nowhere, I repeat, nowhere on the packaging, the instructions or the installation disk was mention made of this requirement. Of course, we figured it out eventually but what should have been a simple process became a frustrating exercise in information gathering, trial and error and e-mail exchanging before we solved it. Just one little sentence on the box or the instruction booklet was needed. Hey, even an error message at the end of the failed installation would have worked. Without it, the application was rendered inoperable. It works fine now most of the time, as long as you remain calm when, for no apparent reason, it fails to log into your ISP with the message that your (unchanged) name and password are incorrect!  Wasn't the design of appropriate error messages one of the earliest recommendations of HCI research? For Apple, that is rotten! It is hard to be active in any information space that prevents your entry at the doorway.

Of late I have been searching for a new home (see bio information), and naturally enough, I have been using the Web to help me locate properties. Thus physical and information architectures have combined for me in real time. Realtors have obviously started to take Web users seriously as there are many sites out there that offer descriptive details, 360 views and calculators with configurable algorithms to determine mortgage costs, etc. Now there are multiple stakeholders in this scenario and the really annoying sites from a buyer's perspective offer just enough information to tempt without giving an exact address for the house. Any interested buyers are forced to contact the realtor for that rather important detail. This is design, for sure, and no different than many other marketing ploys to channel customers in certain directions. What is more frustrating, however, is the lack of standards on such sites and the requirement of some that you load a specific plug-in to your browser just to view their listings. I experienced endless hours of 'fun' trying to download in Internet Explorer only to see my efforts collapse at the last couple of bytes for some unknown reason (the error message told me there was 'an error'!).  Having downloaded, I found that installation required more effort and rebooting than I had imagined any user would care for in a week, never mind a single session. 

I could go on, but these are the norms of early 21st century information spaces and are familiar to all. Such interactions have no obvious corollaries in the architecture of physical space. Can you imagine if your house behaved like such an information space? Would each room require a different kind of light fixture to work properly? How about going upstairs one day only to find that you couldn't because the stairs stopped halfway. What if you found that the spare bedroom was no longer available in your house, and you were told to come back later? And how about not being able to enter another building if it were built on a different platform than your own house?

Of course, with physical spaces we do not usually expect all occupants to be their own repair specialists. However, it is possible to be an occupant and not need such repairs for years at a time. Information spaces rarely afford such convenience. Worse again, information technologies grow more and more elaborate, regardless of user requirement, forcing consumers on an upgrade path that many do not wish for and which has interactive consequences with the other applications in information space. Imagine having that forced upon us in the physical world. Perhaps we need better building codes for information space.

Humans have a tremendous sense of space and location, and we have created physical architectures that reflect both our needs and (to some, perhaps insufficient, extent) the ecology of the environment we inhabit. Buildings provide more than shelter: they reflect our cultures, our resources, our images of beauty and function. "Buildings are much more unfriendly than they need to be," said Christopher Alexander in his wonderful tome A Pattern Language. But when I think how poorly developed are our information spaces in comparison, I can only conclude that information architecture is a very real problem and one that we are not yet even close to solving. That said, I am not sure there even is a solution but an ongoing need for refinement and improvement coupled with a greater awareness of human need and contextual resources. Information architecture, as a field, needs to address such issues and counter the onslaught of technical determinism that pervades the information technology world. While the use of the term architecture has both its supporters and its critics, I feel it really can be justified in the information domain and, more importantly, used for inspiration and insight. Let's rise to the challenge that the term invites, not dismiss it as self-aggrandizing. Only in this way can we hope to shape information spaces that are truly pervasive, robust, beautiful, usable and of value to users more than producers.

How to Order

ASIST Home Page

American Society for Information Science and Technology
8555 16th Street, Suite 850, Silver Spring, Maryland 20910, USA
Tel. 301-495-0900, Fax: 301-495-0810 | E-mail:

Copyright 2001, American Society for Information Science and Technology