A conversation on the architecture of information

I posted my original draft in English on my blog and Christina Wodtke offered her own abridged version of it on Twitter. Then Matt Nish-Lapidus, Design Director at Canadian firm Normative and vice president for the Interaction Design Association (IxDA), posted a few very interesting comments and questions in reply and poked me on Facebook. What follows is my stream-of-consciousness reply to the major points Matt raised in there. I’m quoting them here in part for clarity.

On the shift from Postmodern to Pseudo-modern

bringing in pseudo-modernist theory to this discussion (…) it’s a very valuable viewpoint (although) I find post-postmodernism more resonant in many ways… the fact that we can have multiple equally valid and relevant critical movements and theories at once is very post-modern.

I’d say it is a legacy of Postmodernism that carries into Pseudo-modernism, very much in the same way that Postmodernism took a different side but framed problems the way Modernism did. This might be immediately obvious to anyone and hence unnecessary exposition on my side, but these boundaries are just cultural tools we use to clarify the mess that reality is. Things are never going to be as clear-cut as we make them, and especially transitional periods will see a lot of oscillations and a lot of gray areas.

Pseudo-modern is totally in line with the idea of multiple points of view: we could simplify and say it just makes its framing even more extreme by removing the idea of the ”author” altogether as part of a meaningful conversation around an artifact. This is a new paradigm, and worth paying attention to. This said, we have been exposed to postmodern thinking for 30 years now. Of course we find it resonant: it’s either that or opposing it fiercely.

On the primary artifact of information architecture being abstract

I disagree and don’t fully understand the notion of separating form-giving from sense-making. Aspects of the final form, including even ornamentations, enhance/degrade/affect the overall understanding and meaning of the thing and it’s information. If we’re truly talking about the architecture of information, then we still need a shape with boundaries that the people in the space can comprehend. It may take multiple forms, but the form is inextricable from the meaning.

Couple of different points here. First, I would love to see what you say above articulated in such a way that it can be discussed at length. There’s quite a lot to be unpacked in there. Second, I don’t separate form-giving and sense-making the way you seem to imply here and it’s decidedly my fault: that discussion probably deserves much more space that I was willing to give it within the narrative of this specific paper. The goal here was to establish, within an academic conversation, that when we discuss information architecture (or IA) we shouldn’t be discussing pure ”Polar Bear IA” if not from an historical perspective.

This said, how can we “make sense of information without considering its final form” is definitely part of the challenge. A first observation is that abstraction does not necessarily entails the complete elimination of ”form” or shape. A square, a rectangle and a rhombus are all parallelograms. ”Parallelogram” is an abstraction, but it still conveys a certain degree of ”shape-ness”. Information architecture has long dealt with these issues through categorization (in its broadest sense).

A second observation is that while (at least as far as I’m concerned) we have no definitive answer to offer currently, this disconnection between meaning and form is a very tangible problem the nature of cross-channel transactions has made a common eventuality. In ”Pervasive Information Architecture”, me and Luca Rosati play with McLuhan and argue that, today, messages have no medium. What we mean there is that when you are sending me a message, you have little if any control at all on the circumstances of its final delivery to me. Is a voice message inhabiting the same form as an email? And a Karotz-delivered update? I guess not: their support and characteristics are wildly different. They employ different rhetorics. They appeal to different parts of our brain. They imply different social constructions (informal vs slightly less formal vs not formal at all). They could be perceived as embodied or spatially localized (think of a mobile phone vibrating or any kind of ambient alarm). Their form changes.

Now, does that change their meaning entirely? I argue that it shouldn’t, and that in there is some of the work that information architecture is supposed to be doing. Sense-making is not a top-down process, especially in pseudo-modern settings, but the result of the intentions of the designers, the affordances of the system, and the conceptualization of actors. That’s intended meaning, constructed meaning, and received meaning. (I’m mostly basing this on Krippendorf and the work of Kazmierczak on meaning in design. I also usually add a homeopathic dose of Khun to get out of there alive). As “meaning” as it’s used in the paper originates in the patterning, gestaltic process that connects those “meanings”, it “does not reside in the individual signs for and in themselves” (this is Kamzmierczak). That’s what I mean when I say that the object of IA is abstract.

On form and meaning

The same abstract IA applied to a website or a physical building (like the museum) take on different meanings and have different impacts on the people who visit them. How can you ”make sense” of the information without also considering its final form?

If they are part of a system (as they supposedly should be, otherwise why should we even bother), “meaning” (as identified above) should not be different. “Meanings” will, though. A consistent, pervasive information architecture is there precisely to allow the construction of a stable mental model (or more precisely, of a sense of place based on a stable mental model) even while intended, constructed and received meaning oscillate and re-balance their relationship.

I need to give you a little context here. I was trained as an architect and industrial designer in Milan in the late 80s and early 90s. At the time Milan was the place where architectural thinking in Italy was most concerned with the social, cultural and economic impact of ”design”. As such, it was systemic in nature. Ernesto Nathan Rogers, a rather influential Italian architect (his firm BBPR designed the Torre Velasca in Milan), repossessed Gropius’ famous framing of the Bauhaus as concerned with anything “from the spoon to the city” and used it for his contribution the Athens Charter in the 50s, describing what a typical working day for a Milanese architect was: design a spoon, then an apartment, and then a skyscraper.

Rogers, who passed on in the 60s, started a long debate on these issues (with Gregotti and Rossi, and through the Casabella magazine) that spanned a decade and that was objectively the mold used for training us. He was an enormously influential voice within the conceptual foundations of the Politecnico back then. As a result, we were given extensive city planning foundations and we were brought all the way up to understand plastic and colors. From the city to the spoon. We were also encouraged to be holistic and unconventional in connecting disciplines to our core curriculum (my classes back then included Aesthetics, Urban sociology, Science of vision, Typography, Anthropology of the city among others).

Making multi-faceted “things” that span different scales and work as systems is a natural part of my approach and of my mindset and I naturally tend to associate the construction of meaning with a process, more than with a final ”form”. That’s how I frame things. In a building, I’m considering flow, rhythm, relationships, the meaning of spaces. In an object, its semiotic value over its shape. In information architecture, I favor sense-making and place-making over order: Wurman’s original ”architectures of policies” and architectural, systemic theories of space come before controlled vocabularies as a useful framing to understand and structure complex information-based, dynamically changing, user-participated artifacts.

On shape and boundaries

If we’re truly talking about the architecture of information, then we still need a shape with boundaries that the people in the space can comprehend. It may take multiple forms, but the form is inextricable from the meaning.

As I said, sense-making for me is a systemic process that involves all facets of an artifact (whatever this artifact is). But the boundaries in such a process are another aspect of the architecture: they are an outcome, rather than an attribute of something (the shape). What is the shape or the boundaries of a game of hide and seek? Difficult to say, right? But you certainly know if you are playing: that’s because the game structures its boundaries procedurally and negotiates them constantly.

This is true of architectures as well. Not all buildings are defined by walls (the Borobudur is defined by relationships, for example), and not all walls are physical (think of Saint Peter’s square). Not all walls are meant to permanently freeze-frame a specific configuration either (think of the traditional Japanese house). When information is involved (and that’s the primary material IA works with), a lot of this sense-making process (not all) has to do with the structure first and with whatever instantiation you interact with second. It is not completely abstracted from “form”, but it does not simply coincide with it.

One of my examples to explain this in lectures (derived from some old conversations across the various MLs and originally introduced by Jorge Arango, I think) is chess. You can play chess via mail, on a board, with people. These are different instantiation of a similar (allow me) information architecture, the “game of chess”. They are sure different, but still chess. Interfaces differ, their forms differ, the game doesn’t. The role of IA here is to work on the rules that make the game possible and enjoyable, not to dictate the ”form” of the final artifact. When I argue that we framed IA in an artifact-centered manner, I mean that we only considered the boardgame as the only possible instantiation, because of what we had at hand at the moment. Very much like saying that since movies are now moving to digital (with a whole lot of different rules and possibilities, both technical and semantical), they are not movies anymore (and you could sure argue that, more below, but then it’d be on you to have supporting arguments for drawing a new line).

If you want an example closer to home, think Twitter. I’d have a hard time considering Twitter’s “form” as essential in its sense-making process. A lot of people don’t even use the official Twitter clients. Some still text messages in, probably. Twitter has many “forms”, on many different channels, platforms and devices, and these are made possible / not possible by its information architecture. Some are more successful than others. Now, if by “form” here we mean the set of relationships and the semiotic / spatial elements Twitter uses to structure a conversation (identity, temporal aspectualization, proximity, separation), then I’d argue that’s not “form” as you imply above (“aspects of the final form, including even ornamentations, enhance/degrade/affect the overall understanding and meaning of the thing and it’s information”) but structure, and a structure that is conveyed differently depending on the actual artifact you use.

In the case of chess, it could be argued that rules and procedures (structure) somehow place constraints on the final form(s) anyway, for example by imposing the use of 32 pieces on a 64-cell gamespace. This I agree with. That’s what I meant by saying I don’t see form-giving and sense-making as totally separated processes.

City planning is probably a good example of this and the closest I can come to if you consider established fields and the physical environment: how the city functions and is experienced is sure a consequence of point-to-point decisions on the style of the buildings themselves, the materials being used, the social behaviors being encouraged, but the underlying structure allowing / constraining that is the plan and the rules it set down through far more abstract indicators such as density, ratio of free spaces / built spaces, green areas, shopping / living areas, and so on.

An important point is that there is no value judgment here. Both levels exist independently and interconnectedly, and they both contribute to the final “experience”. But they pertain to different gravitational areas, and when projects scale up they might end up to be the practical (not conceptual) pertinence of different specialized fields/roles. Now, I’m trying to clarify aspects of a conversation that comes with a lot of baggage, some good, some not so good. These lines we draw are fluid, digital is only making them even more fluid, and when working along a continuum (or around different milk pails as Dan Klyn argues, pick your metaphor), if we want clarity we shouldn’t be confusing who we are, what we do, and what is the discipline we are building / applying.

Those are different things and I think you tweeted something touching on this very recently. While I understand the importance these issues might have in the practice and in the market, and the constant need for self-reassurance or self-reinvention that somehow goe with that, I’m personally not really tackling that part of the conversation. I was taught systems, I don’t like fences: I want to understand how the framings we have move from one side of the continuum to the other and what are the distinctive elements that make our fields / disciplines what they are, in order to ground them better and consolidate their educational and research position. That’s my contribution to maturing our practice(s).

Professionally, and I’m speaking for myself again here but I doubt this is uncommon, I’ve been doing work that surely qualifies as interaction design, or information design, visual design, product design. I also designed spaces. And of course I designed information spaces. I teach. And I honestly can’t see a problem with this.

But if I design a chair (I did), my approach to the chair is informed by my being me, that is, an architect (and many other things, of course). I don’t call that chair “architecture” (unless I see an architectonic purpose in it and want to stress that point like if, say, I’m designing a throne), but I’ll be working in a very different way from someone coming from carpentry or the visual arts. My process, and my framing, will be different. I need this to be clear to me if I want to reflect on my practice and escape circular reasoning and self-referentiality.

I detailed what I mean here a tad better in the PIA book when explaining imprecise vs precise design, and I also have a blog post on my website that tries to explain it with an interior design example based on one of those “Extreme House Makeover” tv shows we had here in Sweden. Does that mean the final artifact will be radically different also? Not necessarily, but it usually will be. My gravitational pull moves me closer to the end of the continuum where the city plan or the building reside. I’m looking at the system and balancing my priorities accordingly.

You could argue I gain some in the process, and lose some, and you’d be certainly right, but that’s the nature of the game. These layers (me, the field, the process, the artifact) are part of the systemic nature of the design process: they move independently, but as one changes, the others do as well. In this picture, which is of course my picture, information architecture gravitates more towards architecture, systems and structures and, if you will, imprecision.

The big question

Is this still information architecture? (yes, this is intentionally provocative).

I can’t answer that question for everyone, of course, but yes, it’s definitely IA from my perspective and in my (scientific) narrative.

More pragmatically, I see no real discontinuity, as I argue in the paper, but rather a repurposing and the acknowledgement of the epistemological mistake (if you can call it that) we were making by considering photography to be synonym with film + optical camera, as I write there and as I suggested here with the movie example above. It’s still photography, right? Even though you don’t really have ”graphein” anymore, as that implies writing on something (glass, metal, paper) and film is gone and printing out a secondary and non necessary corollary. What we are concerned with is the epistemological statute of photography, and that is still there.

Same with IA: my (scientific) reading maintains that ”website” is an incidental artifact we have now expanded / modified / commoditized / evolved, and that ”structuring information” (not going into lengthier descriptions here) is the real core epistemological statute we are concerned with. As a consequence, this is a new stage in an ongoing process, not a new process.

Hope this is useful, even if it’s a tad messy and fairly unpolished. Cheers, A.

(Flight 93 Section, 9/11 Memorial, New York 2.0. A remix of InSapphoWeTrust, http://www.flickr.com/photos/skinnylawyer/6451064881/ CC BY-SA 2.0 2011)