Let’s get something straight about IA

I’ve written a lot of stuff over the last few years about information architecture. And I’m working on writing more. But recently I’ve realized there are some things I’ve not actually posted publicly in a straightforward, condensed manner. (And yes, the post below is, for me, condensed.)

WTF is IA?

1. Information architecture is not just about organizing content.

  • In practice, it has never been limited to merely putting content into categories, even though some very old definitions are still floating around the web that define it as such. (And some long-time practitioners are still explaining it this way, even though their actual work goes beyond those bounds.)
  • Every competent information architecture practitioner I’ve ever known has designed for helping people make decisions, or persuade customers, or encourage sharing and conversation where relevant. There’s no need to coin new things like “decision architecture” and “persuasion architecture.”
  • This is not to diminish the importance and complexities involved with designing storage and access of content, which is actually pretty damn hard to do well.

2. IA determines the frameworks, pathways and contexts that people (and information) are able to traverse and inhabit in digitally-enabled spaces.

  • Saying information architecture is  limited to how people interact with information is like saying traditional architecture is limited to how people interact with wood, stone, concrete and plastic.
  • That is: Information architecture uses information as its raw material the same way building architecture uses physical materials.
  • All of this stuff is essentially made of language, which makes semantic structure centrally important to its design.
  • In cyberspace, where people can go and where information can go are essentially the same thing; where and how people can access information and where and how people can access one another is, again, essentially the same thing. To ignore this is to be doing IA all wrong.

3. The increase of things like ubiquitous computing, augmented reality, emergent/collective organization and “beyond-the-browser” experiences make information architecture even more relevant, not less.

  • The physical world is increasingly on the grid, networked, and online. The distinction between digital and “real” is officially meaningless. This only makes IA more necessary. The digital layer is made of language, and that language shapes our experience of the physical.
  • The more information contexts and pathways are distributed, fragmented, user-generated and decentralized, the more essential it is to design helpful, evolving frameworks, and conditional/responsive semantic structures that enable people to communicate, share, store, retrieve and find “information” (aka not just “content” but services, places, conversations, people and more).
  • Interaction design is essential to all of this, as is graphical design, content strategy and the rest. But those things require useful, relevant contexts and connections, semantic scaffolding and … architecture! … to ensure their success. (And vice versa.)

Why does this need to be explained? Why isn’t this more clear? Several reasons:

1. IA as described above is still pretty new, highly interstitial, and very complex; its materials are invisible, and its effects are, almost by definition, back-stage where nobody notices them (until they suck). We’re still learning how to talk about it. (We need more patience with this — if artists, judges, philosophers and even traditional architects can still disagree among one another about the nature of their fields, there’s no shame in IA following suit.)

2. Information architecture is a phrase claimed by several different camps of people, from Wurmanites (who see it as a sort of hybrid information-design-meets-philosophy-of-life) to the polar-bear-book-is-all-I-need folks, to the information-technology systems architects and others … all of whom would do better to start understanding themselves as points on a spectrum rather than mutually exclusive identities.

3. There are too many legacy definitions of IA hanging around that need to be updated past the “web 1.0” mentality of circa 2000. The official explanations need to catch up with the frontiers the practice has been working in for years now. (I had an opportunity to fix this with IA Institute and dropped the ball; glad to help the new board & others in any way I can, though.)

4. Leaders in the community have the responsibility to push the practice’s understanding of itself forward: in any field, the majority of members will follow such a lead, but will otherwise remain in stasis. We need to be better boosters of IA, and calling it what it is rather than skirting the charge of “defining the damn thing.”

5. Some leaders (and/or loud voices) in the broader design community have, for whatever reason, decided to reject information architecture or, worse, continue stoking some kind of grudge against IA and people who identify as information architects. They need to get over their drama, occasionally give people the benefit of the freakin’ doubt, and move on.


This has generated a lot of excellent conversation, thanks!

A couple of things to add:

After some prodding on Twitter, I managed to boil down a single-statement explanation of what information architecture is, and a few folks said they liked it, so I’m tacking it on here at the bottom: “IA determines what the information should be, where you and it can go, and why.” Of course, the real juice is in the wide-ranging implications of that statement.

Also Jorge Arango was awesome enough to translate it into Spanish. Thanks, Jorge!

Author: Andrew Hinton

I use information to architect better places for humans. More at andrewhinton.com.

27 thoughts on “Let’s get something straight about IA”

  1. Yeah, baby! The stuff that makes me follow you on twitter.

    However, what you say often sounds like “designers with a linguistic twist”, but surely there’s more coming to pin it down and make it distinct from the evils of the designer moniker, right? In fact, your point 3.3 should be your 1.1 and the basis for everything; I like it the best.

  2. Brilliant and super clear, as always. I’ve been using the “spaces made of language” idea to explain what I do to prospects, and it seems to make a lot of sense to them. This is a very useful perspective into our work.

  3. As always clear!
    Especially “Leaders in the community have the responsibility to push the practice’s understanding of itself forward” could be considered a pillar of the issue discussed.

  4. Great article Andrew. I’ve said in the past if we put an IA, IxD, Graphic Designer, CEO, etc in a room and drew a picture of a product or a service all of these professions could build it. The only real difference, without diving into semantics, would the process they work through to get to that specific / desired end state.

    The effect of the drama that has been created has a fundamental issue that has hurt us all.. the opportunity to learn from other processes / approaches to problems. I’m an Information Architect… but the processes and tools that have shaped the discipline to this point are only but a few I use in my work as an IA.

    I’ve learned from so many different professionals within and outside the IA industry over the years and that has allowed me to be far more creative than I could have ever dreamed if I had not had such privileges; but you have to seek those out and want to learn; not just sit back and criticize others.

    As you say… move on. Move on to bigger and better things!

  5. brilliant post – spesh this quote: “…start understanding [our-]selves as points on a spectrum rather than mutually exclusive identities.”

    I often see myself with just my toes in the water of IA/UX (y’see what I did there – I lumped ’em together), as if there was this big lake of The One True Identity That Is IA. Whereas if we could all just identify that our experience, expertise and skills are moulded by the work that we each do and is therefore just a small part of what IA could be then we might all just get on a little more ;). Points on a spectrum (or stepping stones across the lake) allows us to broaden the scope of what we do for our clients but still make connections with each other to feel part of a community.

    The IA I do for law firms is nothing like (on a day-to-day practical/operational front) the work I used to do for an agency. But the IA concepts and practices I employ, and the pretty IA deliverables I create, are the same. I want to believe that still allows me to be in your gang.


    Oh! I also just wanted to support your evolution point above – legal services are also going through a bit of an evolution with the work they produce, the approaches they use, and the engagements they take on. Are they going to stop calling themselves lawyers? Not on your nelly.

    Perhaps the issue for us is just the umbrella term – the great thing for other professions is that they created the term ‘doctor’, ‘lawyer’, ‘architect’, that meant people really really (really) wanted to join the profession. We haven’t been so adept at creating The Profession Everyone Wants To Belong To (yet – as you reason above)…


  6. Brilliant post. Thanks for sharing, Andrew.

    I wanted to add one small note to #2 (under Why does it need to be explained), the one about the various views on IA, as that’s one of the things we try to formalize in the Pervasive Information Architecture book.

    We traced down and documented three different sources: information design (the Wurmanites as you say); information science (the Polar Bear), and information systems (coming from the 1980s, and currently mostly concerned with EIA).
    We see those as intertwining threads in a long-running (at least longer than we usually think, as we explain the book) development of what we could call classic IA.
    What you articulate above is the new, emerging idea of IA as a design layer across environments which has been shaping up in the past 4-5 years, and posts such as this one are definitely going to help clarify our vision.

  7. I really appreciate your caution with language, Andrew. You are a semantic wizard.

    Unlike your peers above though, I feel you wasted your time.
    There is nothing new here. All you’ve done is express the argument for BigIA instead of LittleIA. It’s the same thing you could have written back when Morville originally did all those years ago.

    The difference is that the ship has sailed. DESIGN has come in and taken over as the single paradigm for everything leaving IA in the dust to just be little IA once and for all, b/c any slippery slope is in fact that, slippery and just leads you down to the well of everything is creation and HOW you do that creation is what defines you: engineering, business, design, etc.

    What I constantly read from the IAI folks who maintain a very close position like yourself, above, is the syndrome of “I have hammer. I have a GREAT hammer. What can I do with it?” There are a ton of things that one can do w/ the skills of an traditional (little IA) that can apply to many facets of life. That does not mean that what they do with those skills is always the output of information architecture. Which is the way of expressing that you can define the discipline by its activities or you define it by its output, or more importantly you can define it by the overlap of activities and activities. Once you separarte the activities from the output you enter the realm of the slippery slope you’ve entered in your own admitted semantic debate.

    Now before you go all hog wild about this IxD person treading here and doing exactly what I’m claiming you do. Notice above I did not place any juxtaposition with IxD. I can care less at this point. Design is transdisciplinary end of story. There is no single design practice, discipline, or even community that does everything, and doesn’t do everything. The same slippery slope argument can be easily applied to IxD, ID, GD, Arch, etc.

    Instead of defining activities, I have been concentrating on output. What makes a GREAT interaction? What is interaction? What are the aesthetics of interaction.

    THAT is where I would challenge IAs to go. Define information? great information? meaningful? understood? transformative? influential? etc.? What is a “right” IA? Is saying “that works” enough? The how you get there is so influenced by so many different sources that there is no way to say that THIS practice owns anything any more.

    I teach IxD b/c I believe that all forms of designers need to understand the intrinsic properties of behavior. I teach IA because information, narrative, semiotics, etc. are some of the containers that behavior transcends.

    there is no way to create any experience without either of these (and other) attributes.

    So in the immortal words of Mr. Anderson (aka Neo): “There is no IA.”
    In other words, like the Matrix itself it is just a metaphor, that we can only apply so long as its extension doesn’t break. It is the end of semantics. We’ve reached the point as designers that NONE Of our paths are meaningful except as communication tools. They are limited, and we have to let the limitations go instead of holding onto them.

    — dave

  8. @ Alex
    Thanks. I don’t think “designer” is an evil moniker though. It’s a great term (and information architecture is certainly a design discipline).
    As for nailing it down more – see the links I embedded in the first sentences of the post — I’ve done (I think) a better job of explaining these points in those writings.
    In sum: yes, information architecture is design, but it’s not all of design — it’s a specific area of design that’s become more essential since the explosion of the networked information layer we live in.
    Honestly I think it’s the sort of thing that, once someone sees what I mean, it’s clear — but it’s awfully hard getting the frame to shift for people who don’t see it yet.

  9. @ Jorge thanks! very encouraging πŸ™‚

    @ Antonio I agree that’s a big part of moving things along — and I’m trying to do my part with it as well.

    @ Jeff it’s a privilege to work in such a multidisciplinary, rapidly evolving field: designing for people living with these history-shifting technologies. It’s a huge thrill to learn from people who come at the work from different angles & perspectives.

  10. @ Kate
    I do believe that all practitioner communities go through identity shifts & evolution, or else, frankly, they’re probably dead as practices anyway. One sign of a living community is the ongoing conversation about the nature of what it is they do. Even, as you say, lawyers and such.
    I think a lot of us have anxiety about what we call ourselves in our professions because we want respect and recognition, but we also want people to understand what we do. So much of what we do in this field is still too new for that to happen just yet. (Although, honestly, even though we hear “doctor” and give it some respect, how many of us know what a given doctor does other than a vague sense of “medical stuff”?) [I wrote some more about this identity issue back in March, btw: http://www.inkblurt.com/2010/03/26/what-am-i/%5D
    Thanks a lot for the comment!

  11. @ Andrea: thanks! Can’t wait to see the book — it’ll be an important step forward.

    @ Jennifer: I really appreciate the encouragement, thank you πŸ™‚

    @ Dave: I disagree of course. There’s a lot of really bad design happening out there because of an ignorance of what IA practice can bring to the table. (e.g. Google is getting sued over Buzz because why? They made bad assumptions about the contextual experience & appropriateness in connection of email vs chat vs publishing vs “friend” etc). If you teach IA as an important element of design, then you recognize it as existing in some form. Sounds to me like we mostly agree. Your beef seems to be more with someone going the extra step to think of IA as a professional identity … that we should all just call ourselves “designers.” But that’s a little like saying “don’t call it an apple … it’s just food!” or “don’t call yourself human, you’re just a mammal!” Doctors don’t call themselves “doctors” to one another — they’d be perceived as imbeciles by their peers. Scientists don’t tell other scientists “I’m a scientist” — they say what kind of scientist they are. People have a need to identify with a subset — it’s only natural. You can pretend as if everyone should just call themselves “designers” but it ain’t gonna happen (even if they do it, they’ll likely be assuming everyone knows what subset of Design they’re referring to). Sounds like you’re the one who may have some letting go to do?

  12. Andrew, brilliant – you should write one of these every year … oh you have πŸ˜‰

    The Google Buzz example you gave in your response to Dave is very powerful. More examples of people applying (or not) this IA thinking to good (or bad) effect would be great, do you have any more?

  13. @ Peter: thanks! yeah definitely was a hairball I needed to clear. *cough*

    @ Richard: Facebook has historically been full of things similar to the Google fiasco. Flickr is an interesting success-then-oversight example: making a simple, tiered structure for friends, plus the folksonomy aspect for organizing one’s pictures (and exposing it for discoverability), and easy ways to get photos into the site (email them, upload them from just about anything, etc) … all contextual-connection wins. But then Twitter seemed to catch them flat-footed, and a dozen picture services have outstripped them for casual picture sharing on the fly.
    I know I’ve run into others, but can’t recall at the moment — maybe we should start a list someplace? πŸ™‚

  14. I think we are speaking at extremes & that isn’t helpful.

    Information Architect is a real job.
    Information Rchitecture is a real discipline
    IAs (this is the important part) do MORE than Information Architecture.

    This last point is the important 1.

    But I will extend further…
    1. In the grand scheme of things there will never be a comparative critical mass of as titled IAs or even UXers compared to Graphic/Visual/Communication Designers. Or IxDs compared to Interactive, Web, Industrial, etc.

    I have always been of the opinion that it would be more powerful to build the disciplines (IA & IxD alike) over the professions. The disciplines by their nature have to be transitive to a host of practices beyond either’s core practices & communities.

    If you really want to so,ve the bad IA problem, then teach IA to those doing a bad job of it, instead of creating more IAs. It will beinfiniteky mire effective towards our mutual goals of designing kick ass experiences that both meet the needs, wants, desire of human beings while fitting their cultural, social & biological realities (blah blah blah).

    Something like that.

    This isn’t to say that there is not going to continue to be core practicing IAs or IxDs. Of course there will be.

    – Dave

  15. @dave I get what you’re saying there … personally I’m not interested in making more IAs (i.e. people who self-identify as IAs). In fact I’m happy to see a lot of people change their self-identification over the years to UX-whatever or IxD or Content Strategist, etc … because then the people left calling themselves IAs (i.e. that’s their center of gravity for professional identity) are the ones who hopefully get some of the stuff we’re trying to articulate. But, again, whether they do or not is less my concern than doing good work.
    My primary aim in continuing to articulate this is not to win people into a camp, or gain market share (or, god forbid, “land grab” … ugh) … but to get us on the same page about a very important area of practice that needs to be better understood so the stuff we make is *better.*
    I admire, for example, the Content Strategy folks who are doing a great job of promoting, explaining and shaping the understanding of a very important practice that is evolving and coming into its own before our eyes … of course it’s not the only thing those people can do or ever do … but it’s a big enough issue that it needs a name, and needs its contours described well enough so that people can point at it and say: look, your Content Strategy sucks! Get someone who knows how to do that! That person may not self-identify as a content strategist … but they hopefully care enough about it that they have deep understanding & expertise there. Those lovely people are beating the drum explaining the value of CS, what it *isn’t* (which is very important — i.e. it’s not “copywriting” on a larger scale, etc).
    I’m just trying to do the same for information architecture. Not necessarily for “information architects” — whether people call themselves that or not is frankly none of my business. But if they do, all the better — it makes it easier to find people with similar obsessions, so we can drink together and not annoy everyone else with our DTDT ramblings πŸ˜‰

  16. So, really, we’re talking about clarifying the practice of IA vs. the title that some people give themselves vs. the domain and context such a person might dip their feet in.

    Is it meaningful to call yourself IA when you’re doing project-wide design? Is it right to call yourself a UX person if you’re doing the structure and labeling of pages in a website? Is it correct to call yourself content strategist if you, indeed, do copywriting? If I don’t do wireframes nor website structure but focus more on the semantic landscape in linguistic vs. semiotic expressions of an intranet (working with organisational change, for example), am I still an IA?

    I think there’s a myriad of “things we can and do do” that belongs in all of the various design camps, but claiming a specific title means something about the focus of our work. Some times I’m a UX guy (usually if I’m whipping existing stuff into usable shape), other times IA (working with more semantics and structuring of information), and then project manager, designer, content strategist, usability researcher, usability designer, information manager, interaction designer, UI designer … the list goes on. I think these titles we choose is not for describing the job we specifically do to the client, but simply is a moniker we use amongst ourselves in describing what part of a bigger problem we’re focusing on a little bit more than all the other things we probably do as well.

    Everything is IA. Everything is UX. Everything is, really, content strategy. All we’re doing is trying to pry the “designer” moniker away from the graphic designers, probably unsuccessfully.

  17. @Alex I see where you’re coming from … I think … I have to say though that I really don’t look to pry anything from anyone (I do think “design” would do well to be … well, liberated from the assumption that if you say you’re a “designer” it’s a specific subset of people who design things … but that’s more of a wish-list item than anything for me). Anyway I’ve done some thinking/writing before about the identity vs role vs activity etc …
    Here: http://www.inkblurt.com/2006/03/29/a-layer-model-for-the-ia-profession/
    And Here: http://www.inkblurt.com/2008/05/07/a-model-for-understanding-professional-identity-and-practice/
    And Here: http://www.inkblurt.com/2010/03/26/what-am-i/

    Wow, evidently I think way too much about this stuff … I need a new topic …

  18. This needed to be said. Thanks, Andrew, now I have a page to bring all the haters to.

    Is the polar bear book really that down to earth? 😦 (haven’t yet read it)

  19. @ Alex D
    The PB book is excellent, and has an enormous wealth of how-to information as well as great advice on how to deal with various work/design scenarios.
    Actually I think it does a great job at the scope it handles; but I think the authors would agree with me that it doesn’t cover *everything* someone doing IA as their main practice should understand. Especially with cross-media/cross-context design work. Perhaps the next edition will evolve in that direction. I’m definitely looking forward to the Pervasive IA book that Andrea & Luca are working on to help move things forward even more. Oddly there is a lot of writing going on about post-web-1.0-IA but hardly anyone is calling it IA (Kuniavsky, Greenfield, et al) … though that’s coming around lately … folks are beginning to reclaim the label a bit (like Gene Smith’s recent blog post http://nform.ca/blog/2010/10/mind-your-inward-paths and Peter Morville’s talk on Ubiquitous IA http://www.slideshare.net/morville/ubiquitous-ia ) IAs have been talking about all these things for years, but were somehow afraid to call it IA … for fear of confusing people, maybe? Hopefully we’re going to change that.

  20. Thanks for a great post; I don’t agree with everything here, but there’s a lot of goodness in it, as well as in your heart–something missing from many of the discussions (or, rather, discussers) in this space over recent years. I hope to dig in further soon.

    As far as another edition of the polar bear book? No way! Three was one more than was necessary. πŸ˜‰ And no, it doesn’t cover everything; 500 pages is already too long for most books. (That’s not counting the 100+ pages we threw away, BTW.) That was the problem with writing the second edition–there was already too much to squeeze into a book.

    Finally, I still think we spend too much energy defining the damned thing (DTDT). The most interesting work any of us does–and the most growth in any of the UX-related fields–has little relationship to or dependency upon definitions.

  21. Andrew – really nice post. Thank you.

    Back in 2005 I wrote a post that asserted companies needed to employ interaction designers and information architects and engage them to support the development of cross-channel experiences –> specifically ones that transcend the digital medium. Why? Because most of the rest of the corporation doesn’t understand interaction design…information architecture, path and task analysis… etc. I believe IAs and individuals skilled in interaction design bring a unique understanding of how to organize the content and the experiences so they align and flow together. This is not a skill embodied in traditional IT, marketing, customer service, product development teams.

    Perhaps this is a whole other post entirely – but addresses some of what Dave Malouf is talking about above. Pithy stuff.


Comments are closed.