The secret life of information architecture: the space between tools and theories
What’s IA?
This is the written version of a talk I gave at the 2019 World IA Day in Toronto.
What comes first to mind when we think of IA? Spreadsheets, card sorts, tree tests: these are the visible tools of our trade. All this sorting, grouping, and naming is hard-wired into our brains. It’s one of the first things we’re taught to do to develop our cognitive skills, to teach us how to parse our experiences.
Family trees and org charts show us how we see ourselves. They impose a specific order and a structure. And they tend to be hierarchical, which should give us all a clue about the underlying organizing principle: power.
At a primal level, we humans fear chaos.
Confronted with disorder, most of us feel deeply uncomfortable. We say things like “What a mess. I don’t understand what’s going on.” As designers, we see a mess, and we’re pathologically compelled to want to order it. Unsurprisingly, Chaos was the first thing to exist in the Greek cosmogeny. From there on out, we’re just trying to manage the beast.
There are quite a few visualization of the tree of life. I particularly like this one: https://www.onezoom.org/. There’s also https://itol.embl.de/.
Trees of life are a manifestation of our attempt to IA our most primal questions. We want to know what we are and where we are in relation to everything else on Earth.
IA is about wayfinding & sense-making.
We use IA to find our way to make sense of stuff.
Like much of design, our craft is about more than our tools. IA is about more than labeling, classifications, navigation schemes. More than card sorting and tree testing. More than what to put into navigation menus.
Before we dive into design tasks and theories, let’s step back to consider the whole system. Beneath the taxonomy are actors and objects. We need to find out who and what they are by starting with some questions:
Who’s in it?
Who’s related to whom?
How are the whos related to one another?
What’s in it?
How are the whats related to one another?
How are the whos and the whats connected?
When we go to work, we’re still human.
Once upon a time, I was part of a team tasked with redesigning one of our company’s websites. Imagine today you went to work and your business was on the top level navigation.
Then the website got re-designed. And the design team comes along and shows you the new website. And you see that your part of the business is now buried a couple levels deep in the menu. And the designer proudly points out that the new navigation menu is customer-centric.
Meanwhile, your mind is racing. You’re thinking, “Why is my part of the business buried deep in the navigation?” And “But we can’t service our customers the way the navigation menu is promising.” “Am I getting re-org’d?”
How might you react to this redesign?
Visible hierarchies and labels are visual representations of underlying systems.
It should surprise no one that it took 18 months to get everyone aligned on what goes in the navigation menu. And that customer-centric IA didn’t make the cut.
What is going on?
When we’re asked to design navigation menus, we need to remember that it’s bigger than the design of the surface. The real design work of IA and navigation menus is in probing these questions: What’s the organization’s business model? Or if it’s a not-for-profit, what’s the funding model?
In crime shows and movies, they like to say “follow the money”. That’s solid advice for us, too.
A purely customer-centric IA might not align with how the organization is set up to work, making it hard or impossible for the IA to come to life. I’m not saying it can’t, but what you’re actually re-designing is not the navigation menu or the IA of the site: you’re re-designing the organization.
My argument would seem self-evident, but then I hear talks and read blog posts critiquing navigation schemes that make no sense because they’re org-centric. Well, before we critique, we should ask: did or could the designer address the underlying social, political, and operational structure?
When things appear nonsensical, it’s worth zooming out to see the broader context.
A few Fridays ago, my design team took a day off-site to plan our 2020 individual practice goals. One of the activities involved the jars in the photograph below.
What do you notice about them?
When I went to buy those jars the night before our off-site, all I saw were the colourful lids, got very excited, and grabbed two dozen jars. I quickly developed a mental model about organizing them by the colour of the lids. The next day, I’m setting up the jars in our workshop room.
That’s when I realized the colour of the lids wasn’t the only thing different. The jars also had DIFFERENT. SHAPES.
Part of the point of this story is that I anchored to the colour of the lids because I saw only the surface.
Information architecture embodies perspectives.
But the real moral of this story is a story of perspective, which is at the heart of information architecture. How do we or should we sort and organize? By colour? By shape? Colour first, then shape? Or Shape first, then colour?
These decisions depend on perspectives.
If you’re blind, and I’ve sorted the jars by colour, it’s meaningless to you. The “obvious” solution is to be inclusive: sort by shape first because everyone can tell them apart, then colour for those who can see colour.
Or is it?
If you’re not blind, what if you want to find jars of different shapes but all with yellow lids? By making shape the top hierarchy, I’ve made you work just a bit harder to find what you’re looking for.
Now, in digital space, we can tackle this with faceted navigation, filtering, and sorting.
These solutions give agency back to us as users, so we can explore the catalog of jars using whatever organizing principle we like.
In meat space, our options are different than in digital space.
The Cotton Library: a Story of Contexts, Constraints, and Mental Models.
Sir Robert Cotton lived during the Renaissance, working in the courts of Queen Elizabeth I and after her death, King James I of England and VI of Scotland. One of his claims to fame is his personal library of manuscripts of some of the oldest English writing that we know of.
Now, the 16th and 17th centuries were a long way before Melvil Dewey invented the Dewey Decimal System in the 19th century. Cotton lived during the Renaissance, when Europe was just emerging from the so-called Dark Ages, rediscovering the learning of the Ancients.
When he designed his library, the best mental model he had was formed with three things: the physical space of his personal library, which measured 26 by 6 feet; some bookcases; and the busts of a bunch of Roman Emperors.
His IA strategy was to place the bust of a Roman Emperor on top of each bookcase, assign a letter of the alphabet for each shelf, and a Roman numeral starting from the left.
People are still arguing about what the logic could possibly have been. Check out this post from author Matt Kuhns, who wrote a book on the Cotton Library, to start your journey down this rabbit hole: https://www.mattkuhns.com/2017/05/cottons-memory-palace.
To our modern eyes, the IA of the Cotton Library seems…well, “idiosyncratic” is a word often used. It’s not unlike our reaction when we look at the navigation menus of business websites. Now, we can sit back smugly with our Dewey Decimal System to SMH and LOL at Sir Robert Cotton and his weird IA. But we need to remember: the man was still counting with roman numerals (like our friends in the NFL with their Superbowl LIV). His context and constraints presented different possibilities. When we make IA decisions, we need to understand the contexts, constraints, and mental models that govern the social and political of the organization.
Best solution Solution that best fits all the conditions we need to balance.
As designers, we sometimes have the luxury of designing and solving for a narrow set of conditions. More often than not, we have a complex set of conditions to balance. We have to ask “Whose perspective do we need to consider?”
And we need to be aware that problems and opportunities that exist in one set of contexts and constraints might not be available in another set of conditions. We live and work in a world that’s too complex to allow us simply to “design beautiful experiences that people love to use.”
People often frame the conversation about design decisions in terms of “compromises.” I dislike that word. It implies that if you could get rid of some constraints, then you’d have something you didn’t have to compromise on, that you could have “best” or “perfect.”
But guess what? Some of those constraints are people. You can’t get rid of people.
So, I prefer to think about it this way. It’s less about finding the best solution and more about finding solutions that BEST FIT all the conditions we need to balance.
What contexts and constraints are we balancing here? What’s the best fit design?
If we hope to create change, we must know things as they truly are.
Broadening our perspective beyond designing surfaces, beyond our fixation with the tools and theories of IA is important because if we hope to create change, we must know things as they truly are. I see the act of architecting information as a cultural act, a statement of values, an expression of power. Those with the power get to define the hierarchy, and if we have that privilege, we need to zoom out to get a broader perspective before making decisions.
Consider this leaf.
If we were thinking of designing just this leaf, we’d miss seeing the broader context of where this leaf fits.
So let’s zoom out a bit. So now we see this leaf lives in this forest.
What if we zoomed out further?
How does the design of the leaf matter when seen in the context of a clear-cut forest? What other questions does this zoom level open up?
People often confuse my bald statements of current reality as being negative, so I want to be clear: I make these statements not to be a Debbie Downer but to acknowledge reality because we can’t move forward by looking at the world through a narrow lens. Or by putting on rose-tinted glasses and sweeping reality under the rug.
Progress and change comes from acknowledging what needs to change, to understand why it needs to change. Otherwise, we might be solving the wrong problem, using the wrong tools for the job, or missing the real opportunity altogether.
I’d like us all to consider zooming out of the details: look at the big picture before diving into the tools and theories because it gives us a chance to frame the problem or opportunity space more broadly, and it gives us more options for the way forward.
IA is fundamentally about way-finding to achieve sense-making. And we won’t find the way if we see only the details. It's worth zooming out away from the trees to see the forest.
