Theory often takes a back seat to practice in the field of web design—after all, it’s hard enough to keep up with the latest acronyms without trying to carve out time for navel gazing about the profession. As a result, innovation in the way we think about our creations has lagged behind the breakneck pace of change in the technologies we use to create them. Consider, for example, our craft’s foundational metaphor of “information architecture.” Since at least Richard Saul Wurman’s 1996 book Information Architects, architecture has been the primary metaphor for how “those who build websites” think about what we do. By adding a new metaphor to our theoretical toolboxes, we can gain a richer, more nuanced understanding of the way that we inhabit cyberspace. This enhanced apprehension of the medium should enable us to create websites that better serve our users.
At the law school where I work, I was hired in large part to lead a redesign of our nearly four-year-old site, which contains around 6,000 pages. As I delve into this overgrown information jungle, I don’t feel like an architect—I feel like an explorer, an adventurer into the darkest recesses of institutional memory. I am charting abandoned territory, noting the relative locations of long-deserted pages, the promising paths that lead to deadends, and the sheer cliffs of 404 pages. But I am also mapping the school itself, though not physically: I am mapping the intellectual histories of our professors, the aspirations of our students, and the ideas, people, and capital that flow in and out of the institution. This map will, in turn, become the blueprint for the architectural portion of the project—laying the groundwork for the new temple that will rise from the ruins of the old site.
Consider cartography as a metaphor#section2
I suggest, therefore, that the art of cartography might fruitfully be used alongside architecture. This might sound like an odd suggestion, since at first glance, cartography appears to be nearly the opposite of architecture. Common sense tells us that an architect begins with an abstraction—a blueprint—and creates from that abstraction a concrete structure existing in physical space. The cartographer, on the other hand, starts with concrete structures existing in physical space and creates from that an abstraction: a map.
For years, most of us have thought of building a website as being more like the former. We sketch out tree hierarchies and wireframes, and use them as the blueprints for the creation not of a physical structure, but an informational structure (and sometimes, if we’re feeling generous to our users, we’ll re-abstract those structures into “sitemaps” to aid their navigation around the site). What we often forget is that the blueprints from which we construct a site are themselves maps of processes and flows that already exist, from verbal dialogues to the exchange of money for goods and services.
What Lefebvre knew about space and social interaction#section3
The conception of web designer as information architect depends upon a vision of cyberspace not unlike the vision of physical space held by René Descartes (1596-1650). To him, and to most of Western civilization for hundreds of years, space was a void, a preexisting grid which remains empty until points are identified and paths plotted upon it. (Think of the digital plains depicted in the movie Tron.)
In the 1970s, however, Henri Lefebvre’s work The Production of Space turned this view on its head, arguing that space is produced through the enactment of social relations. Space, according to Lefebvre, is created by the flows and movements of relational networks—such as capital, power, and information—in, across, and through a given physical area. A building, in Lefebvre’s reading, is a map of the interactions of the people who inhabit it; an architect is not a builder in an otherwise empty wilderness, but an observer, chronicler, and shaper of the networks that exist around her—in short, a map maker. Websites informed by a Lefebvrian conception of cyberspace rather than a Cartesian one would provide truly user-centered design, by recognizing that it is the users themselves whose actions produce the website; the web designer merely facilitates that creation.
Websites as memory dwellings#section4
For a different angle on how cartography might inform web design, we may look to a source even more unlikely than French philosophers: the rhetoricians of ancient Rome. Like most websites, which aim to convince the user to take some sort of action—buy this product, apply to this school, hire this individual, post a comment—the goal of the rhetorician was persuasion. To the rhetorician Quintilian (35-100 CE), the key to persuasion was memory: the rhetorician had to be able to access quickly and accurately “the store of precedents, laws, rulings, sayings and facts which the orator must possess in abundance and which he must always hold ready for immediate use,” or his attempts at persuasion would be in vain. (Institutio Oratoria, Book XI, Chapter 2)
In order to do so, Quintilian suggests a remarkable mnemonic device: the orator should imagine “a place…of the largest possible extent and characterised by the utmost possible variety, such as a spacious house divided into a number of rooms” (ibid.). In each room he should place an object or set of objects associated in some way with an individual fact or argument. Each room can then be visited in turn in order to bring to mind the concepts associated with each object. Memory (and thus, according to Quintilian, persuasion) is reliant upon this act of spatial imagination:
We require, therefore, places, real or imaginary, and images or symbols, which we must, of course, invent for ourselves. By images I mean the words by which we distinguish the things which we have to learn by heart: in fact, as Cicero says, we use ‘places like wax tablets and symbols in lieu of letters.’ (ibid.)
In this formulation, the rhetorician is both architect (the builder of a memory dwelling) and cartographer (the memory dwelling is to actual memory as a map is to the territory it depicts). His imagined building, in other words, is also a map.
Mapping memory collectively: no more digital monuments#section5
The redesign of my law school’s website is not unlike the creation of a rhetorician’s memory dwelling. Built from the collective memory of our past and current students and faculty, the website is an abstraction—a map—of that memory, created to persuade prospective students to apply, current students to collaborate, and alumni to remain engaged with the school. The crucial difference is that the law school’s memory dwelling, unlike Quintilian’s, has more than one inhabitant. Our users, by continually adding to and altering the school’s collective memory, contribute to the production of the website—the abstraction of that memory. Instead of imposing an architecture upon users from above, we should use the flow of their interactions with the site and with each other to determine the form of this memory map.
Just as Lefebvre leads us to see built spaces not as the expressions of a single architect, but rather as the production of the wide variety of human interactions that occur within them, so websites created by cartographers would cease being grand edifices of unidirectional communication and become instead the collective product of the individuals whose lives intersect within them. The rise of the social web demands that if we are to help shape meaningful online experiences for our users, we must rethink our traditional role as builders of digital monuments and turn our attention to the close observation of the spaces that our users are producing around us.
Today I watched a video of a man from ThemeZoom contradicting “The Master Plan”. He said the information architecture was all wrong, even though “The Master Plan” was endorsed by ThemeZoom previously. This article demonstrated really smart information architecture, rather than just another trick.
It’s rare to find a genuinely insightful article amongst the general hubbub of the web these days, this was yet another humbling article from ALA. I’d like to see more articles like these.
I have just one thing to add. To take the metaphor in a different direction, I’d liken the cartographer to a set designer. A well-designed set may be necessary to make the play happen, but in the end of the night it’s the story that you tell that people will take away.
We’re playing it fast and loose with the metaphors here.
Doesn’t this article propose that a web designer is a *rhetorician*, not a *cartographer*? It likens a website to Quintilian’s collective memory map – an imaginary construct with no physical form – not to a tangible, designed map. That tells me we’re making imaginary houses, not representations of those houses.
So what happened to cartography, which I understand to be the science and craft of making maps? Maps aren’t imaginary places, they’re representations of places, imaginary or otherwise. They invariably take some kind of fixed form.
And are we mapping memory or shaping social interaction? Because the idea of enabling social interaction inside a memory map strikes me as a terribly muddled metaphor.
I posted on Zeldman.com about this earlier – so I’ll paraphrase ..
What a great unique and refreshing article on web design! We often think we’re doing things and building logical structures in an unprecendented fashion – we are pioneers and are paving the way, etc, etc “¦ yet that article proves quite the opposite is the case.
I think a failure to see that there are greater forces at work than just the designer and client is what seems to lead to brand new sites being obsolete out of the box “¦ I work as a consulting SEO mostly – and find my role changing more to one of evanglising the semantic web than ever before. Not a role I mind at all – just not something I saw coming 2 years ago.
Thanks for a insightful article!
Regards, Lee
It’s rare to find a genuinely insightful article amongst the general hubbub of the web these days, this was yet another humbling article from ALA. I’d like to see more articles like these.
Very inspiring 🙂 How do you foresee putting it into practice?
Sounds like you’re advocating a user-defined architecture (sorry!) that evolves over time, otherwise the map you create for the current user interactions may soon become out-dated what with all the tectonic activity on the web these days!
Do you envisage sites as evolving within a type of tag cloud while providing fixed points of anchorage, for example, in your case a particular year or department?
As a geographer and webdesigner, I could not have explained this side of my work much better. A website is ontologically a map (even if it is hard to explain or understand). Thanks.
I really appreciate this article. I love reading something lovely that makes me think.
What I’d really love to see, actually, is an expanded version of this paper that take s the issue deeper: what are the implications of changing the metaphor by which we define ourselves? How does this change in our approach change not only what we create, but the relationships that stem from its making? What do we stand to gain by this shift in vision? And, most importantly, how will this shift in vision change our understanding and perhaps practice of digital information sharing?
So, in your copious spare time 😉 I’d love to see an expanded article. I could have kept reading this paper for several pages more. This was right up my alley; I loved it.
Websites are also the emergent property of the activity that can be mapped. If you consider a publishing system like a blog, or youtube, then the physical (sic) website emerges over time as a by-product of its users activity.
Users come and go, publishing and commenting, etc. This leaves behind html pages, words, content. The website is ‘actually’ (and solely) made up of these artifacts and it’s their nature and texture that shapes future activity, which leaves more artifacts and so on.
In some cases it may be fruitful to consider mapping the activity your artifacts facilitate. However, to scale, you must consider the artifacts left by the activity facilitated by the artifacts left by… ad infinitum.
Nice article I must say. Its not very often that someone discusses website architecture from a cartography point of view.
This article prompted me to run a check through my own website and trust me I felt as I was on a treasure hunt something. Did manage to correct some inconsistencies but the journey was way better.
Excellent article.
The mention of Lefebvre reminded me of the work of Guy Debord and the SI, esp. the idea of psychogeography. (http://en.wikipedia.org/wiki/Psychogeography) Specifically, I was thinking of what they referred to as the process of dérive which is defined as:
“In a dérive one or more persons during a certain period drop their usual motives for movement and action, their relations, their work and leisure activities, and let themselves be drawn by the attractions of the terrain and the encounters they find there”
Users do this all the time unconsciously when they browse through sites. Certainly, the architecture of the site is guiding them, but at the same time, users make their own paths through sites, esp. when exploring. Through the use of log analysis, information archtects can follow the flows of traffic from page to page within a site which parallels relatively closely the idea of following the psychological flow through a space.
Fascinating article, but what is the on-the-ground use for this comparision to Cartography? Would love to hear those thoughts.
As I have some background in cartography I started to compare cartography and web design.
There is a process that is conducted when creating a map content. It is called “cartographic generalization”:http://en.wikipedia.org/wiki/Generalization
This process is divided into 4 basic levels:
* Selection (Reduction): you select which objects will be shown on a map
* Simplification: you simplify shapes to increase visibility
* Combination: you comine some elements where their existence is mor important then their position
* Moving: you move elements apart so they don’t overlap
We can clearly find a lot of similarities in this older craft. I guess I was never thinking ot of the box and realized there are infact quite a lot similar principles in web design and cartography
Interesting article and I agree with Pete N. vision. As info architects or geographers (or both), I think we can give new inputs to users, not only to map their way to use, read and surf websites. Isn’t this the sense of mapping…mapping the present to suggest (or experiment) the next step?
Thank you all for your comments. I knew ALA readers were on the ball, but I didn’t expect 10 comments before I even got to work! I’ll try to respond to as many of your questions and comments as I can.
@ David Ramos – You’re right that I have played a bit with the idea of what cartography is. That’s the problem with metaphor: one system never translates directly to another. What I am arguing (and I probably could have made this clearer) is that a web designer is like a cartographer in the way that the ancient rhetorician is like a cartographer (and there fore the designer is also like the rhetorician). I think that the act of imagining a place is essentially the act of creating a map (a representation), just as creating a map is imagining a place. I disagree that a map has to take a fixed form — what is the fixed form of Google Maps, for example? Finally, I’m not sure I see why “enabling social interaction inside a memory map” is necessarily a muddling of the metaphor; it seems to me that things like MMORP games have been been doing that for years.
@ Shahar Hesse and Divya Manian – To be honest, I’m not really sure how to put these ideas into practice, but “a user-defined architecture that evolves over time” as Shahar sums it up quite well. Tag clouds are certainly one method that might prove fruitful. It’s possible that truly doing these ideas justice might require technologies that don’t yet exist — but I’d love to hear suggestions.
@ amber simmons – Whew, that would be a pretty big paper… maybe even a book. 🙂 Again, I’d love to hear ALA readers’ takes on the questions you raise.
@James Arthur – I wish I could have come up with this phrase: “Websites are also the emergent property of the activity that can be mapped.” It sums up nicely a Lefebvrian view of cyberspace.
Looking forward to continuing the discussion!
@ Blaz Grapar – That’s fascinating. I’ll admit to not having done any in-depth research on the actual processes of cartography, but it would be an interesting case study to compare and contrast the processes of a web designer and a cartographer as they move through a project.
@ William Selman – Very interesting. I’ve never read any Debord, but it seems like that might be a fruitful theoretical path to explore. Thanks.
Hi Aaron,
Yeah, it might be a book 😉 But more information and analysis, even thought and discussion, on topics that we haven’t already covered to death is almost always a good thing! We’re still an emerging field, and I think philosophical forays like these are important. I don’t worry too much about the practical purposes–lay the groundwork first, and the application will follow, I’m sure of that. I get the feeling you might have more to say on the topic, and my point is, I’d love to hear more
What a useful perspective, utilizing concepts of philosophers that couldn’t have intended their theories to be applied in this sense. But what a logical and reasonable extension it is!
This makes me wonder what Paul Virilio would think, when he discusses his ‘dromoscopy’ in which geography is destroyed at the point information can travel without any seeming spatial/time relation. Clearly, the web space has a geography of its own, in its file structure and further in its aesthetics that enable the user to achieve the goal provided by the function the designer enables and creates. This type of cartography is possible only through the collaboration of users and the conscience integration on the designers part.
Thank you for your coherent explanation and fresh perspective!
# [Cicero] _Ad Herennium_. Trans. H. Caplan. Loeb Classical Library, 1954, (esp. p. 205 on).
# Frances Yates. _The Art of Memory_. Univ. of Chicago Press, 1966.
# George Johnson. _In the Palaces of Memory: How We Build the Worlds Inside Our Heads_, NY: Knopf, 1991.
# Paolo Rossi. _Logic and the Art of Memory: The Quest for a Universal Language_. Trans. Stephen Clucas. Univ. of Chicago Press, 2000.
# Jonathan D. Spence. _The Memory Palace of Matteo Ricci_. Penguin, 1984.
Brilliant article! As an urban designer by trade and occasional web developer by hobby i often find myself using the city as a metaphor for web layout rather than just cartography. And I’ve written elsewhere about urban design and architecture becoming more and more like information architecture (google “q-dar quameleon” or “quantumcity press”)… and the link between memory and urban space, with the requisite nod to Lefevre. so thanks again for a wonderful and unexpected insight here at ALA, it proves yet again that our current worldview is coalescing more and more across disciplines.
If I’m interpreting things correctly, the paradigm shift we’re talking about here is asking: “How do users want to interact with my site” instead of “How do I think my site should be structured.” It reminds me of something Liz Danzico mentioned in her presentation at AEA SF last week. She talked about “residue lines” (you know, like paths beaten into the grass where people actually walk) as paths of things people want, but which the architects didn’t provide. It comes down to watching and listening to users closely, and then adapting to what you find.
In practical terms, we can go a long way by doing user-testing early and often. Perhaps it means creating a quick and dirty HTML or other “prototype”:http://www.alistapart.com/articles/paperprototyping/ that you can get into users’ hands while you’re still in the early stages of design. Perhaps there is user-testing built into multiple stages of the process. Or it could mean more initial research on the sites your target users actually go to, and how they use them. It may just mean putting yourself in your users’ shoes a little better, and figuring out where they expect to find things, and how they expect to get around.
I hope my view of this article hasn’t been too narrow, since a lot of these ideas aren’t necessarily new. I loved the article, and I’ll hope to see more discussion on its practical application here.
The right article just in time to push me in the right line of thinking about a problem I have been mentally chewing on for some time (nothing to do with websites as a matter of fact 😉 )
I am not a cartographer, but have been working with (visualisation of) geographical and geospatial information. Why the metaphor of the cartogrpaher is so relevant IMHO is that you can represent the exact same data in different ways: as a very detailed exact mapping of the real situation or a very abstract schematic network representation. Which representation is relevant to a user is depended on the context in which a user needs the information! You just serve them the right memory map.
I think the web is slowly moving into a direction of being able to pick the representation of the same data that best fits the use at the moment you access the information (let the user pull up one of his stored memory maps). The role of the webdesigner is to be able to indentify the different uses and come up with different presentations that fit the intended use, instead of the one-design-fits-all approach we see on the majority of sites (even on the supposed web 2.0 sites).
How to get there is probably a way of experimenting and having people around like the author of this article.
This reminds me of how sidewalks are often built on college campuses:
First, a set of sidewalks are built based on what the architects feel will be good and logical paths between buildings.
Later it becomes apparent, as tracks are worn through the grass, that students have other ideas about which paths are good and logical. They can’t be stopped, and the dirt paths are unattractive, so the dirt paths are soon paved over.
As destinations rise and fall in prominence, new dirt paths appear, and the process starts again.
Here’s where the web should have an advantage: the campus is stuck with all their old sidewalks as the green spaces become crisscrosses of concrete, but the web is agile enough to respond and update unused pathways.
Insightful, but I think now that you’ve enticed us with an interesting metaphor and a great classical reference you should spend more time on concrete things we can take away from this theory. What does it suggest about how we should approach structure? Can you use it to point out common flaws in design? I hope you develope the concept more.
Wow, this was really nice article to read. I’m in very similar situation to author: I’ve been hired to redesign the website of our Uniiversity of Applied Sciences. That job is about done now, but my greatest challenge is yet to be achieved. Next in line is our intranet, which needs heck lot of redesigning. Your article was very inspiring, so thank you for that!
Love these metaphors of mapping and safari. When you have 6,000 pages of information I think that definitely qualifies as an adventure.
Interesting. It certainly gives a different perspective from the traditional architect metaphor.
Whether it applies quite so well to a simple, static brochure site as it does to a 6000 page community site is debatable, but certainly worth considering nonetheless.
I think the lesson in either case is to pay more attention to the end-user and be less precious about our own grand ideas.
Thanks again to all of you for taking the time to read and comment on this admittedly very abstract article. Mulling over your comments during the past week, I’ve had a chance to think a bit more about the practical implications of adding a new metaphor to describe what it is that we do when we make websites.
The most basic practical implications of this article, I think, have nothing to do with its content and everything to do with stirring up the kind of conversation we’re seeing here. The occasional questioning of a discipline’s foundational assumptions is necessary for that discipline to grow — from conversation and argument grow new ideas that can push the discipline in new directions. Frankly, I rather wish more folks who commented here disagreed with me. For the sake of the profession, anyway. 🙂
In any case, looking back at the article again, I think the practical core of the article is a call for user-centric design. This is certainly not a novel idea, but there can never be enough reminders that we should be building websites for our users, and not for ourselves or our clients.
More specifically, it is a challenge to change the process by which we as web designers determine a site’s structure. I love the idea (raised by a number of commenters) of building a website like paths on a college campus, allowing users to create the map, charting the natural flows of information the way one would chart foot traffic through the quad. At first glance, this ideas sounds a bit absurd when applied to an informational website — how, and why, would you build a site with no structure at all and then expect people to be able to find anything? I think some sort of underlying structure would be nescessary — after all, on the campuses in question, classroom buildings, dorms, and other landmarks are presumably built with some sort of relationship already implied.
The folksonomy approach suggested by some commenters addresses the aim to allow users to determine structure, but I’ll confess that I don’t think I’ve yet seen a truly successful folksonomy-based site (if you know of one, please set me straight). Unfortunately, folksonomies get very unwieldy very quickly (I’m imagining billowing thunderclouds of tags), which can impede their usefulness. So I wonder: are there ways other than tag clouds to implement a folksonomy system? I suppose sites like Digg.com might be considered a folksonomy of sorts, in which users’ voting up or down of content partially determines that content’s relationship to the content around it. Conceivably a system like that might not only help determine which content users found most useful, but also enable them to have a say on how content should fit together on the site.
I guess I’ve started wandering into slightly more theoretical territory once again, but I remain very curious as to what the readers of ALA see as the practical import of all of this theoretical talk, if any, and I throw the question out to you: how might seeing yourself as a cartographer as well as an architect change the way you work?
To augment “Daniel Potter above”:http://www.alistapart.com/comments/mappingmemory?page=2#20 I’d point to:
1. “The Timeless Way of Building”:http://www.amazon.com/Timeless-Way-Building-Christopher-Alexander/dp/0195024028 by Christopher Alexander
2. “At Home in the Universe”:http://www.amazon.com/At-Home-Universe-Self-Organization-Complexity/dp/0195111303/ref=pd_bbs_sr_1?ie=UTF8&s=books&qid=1220360674&sr=1-1 by Stuart Kaufmann
A short conceptual masterclass that can’t but teach you how to create living websites!
Digg.com is the obvious practical implementation of this effect. Actually, what’s the work of a cartographer here if the paths are laid out automatically by the users?
On the statistics site, we would need more than the page-by-page overviews and show actual flow (animations even, realtime even).
If we are able to provide an infrastructure that changes itself over time based on user interaction then we’re done. Maybe we’ll really be making maps of what’s there. Oh no of course, not, it will change too fast…
It’s funny you say we should add Cartography as a metaphor. I thought it already was a part of information architecture? I started my professional career working on signage and wayfinding and the first thing you figure out is a a sign is more useful if it is part of a system, a hierarchy or some kind of informational architecture (not to mention actually pinning sign locales on a real map). It was from sign system planning that Web design seemed both similar and natural. This work continued to UI/UX design and I still “map” out a site or application first. In fact, I find these hybrid tree diagrams + wireframes to be the most useful way for everyone to discuss an application or Web site; designers, developers, marketers and executives can see the lay of the land at glance and see where patterns and problems occur.
On the statistics site, we would need more than the page-by-page overviews and show actual flow (animations even, realtime even).
The comments on ‘residue lines’ was very insightful. I think one practical thing to take away from this article and the discussion is that designing your site to be easily adaptable to be important. You really never know how a website will work until you actually put it out in the market. From a business perspective being able to respond quickly to market reaction is what separates winners from losers.
From a technical point I think tags and a content weighting system could be practical ways to achieve some of the ideas discussed in the article.
That was unclear to me. My father’s an engineer and my mother is an artist, so maybe it’s my somewhat equally-brained legacy that makes it hard for me to understand why we want to make this such an arcane process.
“Don’t Make Me Think” needs a sequel called “But Stop Trying to Think For Me.”
The worst mistake you can make in my opinion is to try to guess ahead of your users. We adapt rapidly to systems. When systems attempt to adapt to us, I feel like we’re missing the point.
The idea of forgetting about the architect sounds great, hell, forget the builder too; we are here to find reality anyway right? The only problem is that Lefebvre’s idea didn’t turn Descartes on his head; he (like all post-moderns) built on the foundation he was given. That’s because you don’t have to be as smart a Descartes to understand that you can’t have anything “between”? until you have “discrete things”? (ideally, at least two). The “relationships”? are between the objects described by Descartes.
It sounds to me, Aaron, that you did have a fun time exploring someone else’s architecture and I encourage you to not shy away from the task of building now that it is your turn. I ask you to at least accept your own agency and influence in the world and especially on the site you create. You see, there is no history “per se”?, there is only what we remember and someone has to choose what that, is so go ahead and make something beautiful. The cartographer, like the architect, chooses what is important, decides what to map, what is of value. If you think a cartographer’s job is to present everything, show all experience without evaluation then you would have a life size road map that is as useful as just walking down the street to see where it goes. Quintitian understood this, that is why he meticulously designed his eternal space, he is not exactly the poster of your idea.
Fascinating post. I’m still trying to wrap my head around the idea of web design as cartography … this will give me something to mull over today.
Thanks!
@Harland: I think you’re reading a little more radicalism into my post than is really there — I’m not calling for *forgetting* the role of architect, but rather for enriching our understanding of the role of the designer by adding to it the metaphor of cartographer. It’s not a choice of one or the other. Web designers are BOTH architect AND cartographer, because websites are both map and territory (Korzybski never saw a website).
Nor am I advocating the abdication of the designer’s responsibilities in favor of a purely user-created topography — if we take the example of the paths on a college campus, for example, the users/walkers are only allowed to choose the paths between the buildings, not the placement or purpose of the buildings themselves.