Tag Archives: skeuomorphism

Skeuomorphism III and the Semiotic Ghosts

Here, in the third and final part of this exploration of skeuomorphism (which sometimes might have leaned more toward the skeuomorphish than the skeuomorphic), I’m going to look at how it relates to an idea I’ve thought about quite a bit – what I call “materialist indexicality.”

Skeuomorphism and indexicality have quite a bit in common. Neither can be understood autonomously, as they’re defined by their participation in a sort of generic historical narrative. They both pursue legibility through a kind of “rehearsing of the past.” Through the skeuomorph or index, the current design is understood in relation to a past process.

You could object that there’s an important difference in what past is being referred to: while Eisenman’s early experiments in indexicality produce forms that refer to the design process itself, skeuomorphism pursues legibility through reference to past, external precedents. But this assumes too narrow a definition of indexicality: Eisenman quickly shifts from self-referential, autonomous indexicality (e.g. House II) to archaeological indexicality (e.g. the Wexner Center). This evolution doesn’t stop there. “Historicity” is a rather arbitrary qualification once you’ve recognized external inputs as a potential influence on form. For example: if architecture can index the historical urban grid, why not reflect a mathematical idealization of flows of people across the site? At this point, software developments (Maya in particular) fuse with this conceptual expansion of indexicality, allowing for experiments in what Jeffrey Kipnis called “hyperindexicality” (“Performance Anxiety,” 2G no. 16 vol. 4).

I’ll come back to hyperindexicality, but first let’s finish this simplified history of indexicality. The next episode sees the index tethered to materiality. If Eisenman’s early exploration of the index can be connected to an interest in testing Derrida’s ideas in architecture (this is not to say that it was done without translation or inventiveness), materialist indexicality has its own practice-theory pairing in Reiser + Umemoto and Sanford Kwinter, epitomized by his introductory contribution to their Atlas of Novel Tectonics.

Kwinter begins by describing their work as “new materialism” and ends with the suggestion that it “may well be a new expressionism.” The reference (“expressionism” is the last word in the essay) can’t be taken lightly. Kwinter plays a rhetorical game here: the historical connotations of Expressionism, signaled by putting the word “new” in front of it, are completely antithetical to what he means by describing their work as a new expressionism.

Rather than attempting a real summary of “old Expressionism” here, it’s sufficient to merely point out a key, uncontroversial characteristic: Expressionism’s conceptual investment in the way art reflects the artist’s subjectivity. This could not be more different from Kwinter’s description of R + U as “chemical engineers” whose work “transmits” / “reveals” / “discovers” / “deploys” / “expresses” material potentials by applying diagrams to them as “reading devices.” Kwinter works hard to make sure his terminology avoids implications of authorship. In Kwinter’s telling, the design process sounds like a chemical reaction, and the architect-engineer watches while flows of matter-information shift from one metastable state to another.

Between the agenda of “revealing material potentials” and the conceptual debt to Deleuze’s antihumanism revealed by Kwinter’s rhetoric, we have the making of a decidedly materialist agenda. But what makes it indexical? This time, the answer is best discerned by looking at a point of continuity rather than a contrast: the shared reliance on the diagram.

In the formalist architectural analysis of Rudolf Wittkower and Colin Rowe, the diagram played the same role that rhyme scheme annotation plays in formalist poetry analysis. Between reading Terragni’s Casa del Fascio and designing his House series, however, Eisenman went from using the diagram to describe (etymologically “write down,” i.e. “record”) architecture to allowing it to prescribe (etymologically “write before”) architecture.

The diagram mediates between forces (commonly described as “flows”) and form: it’s an explanation, in a sense, but it also claims to determine the translation of inputs into a formal output. The determinacy of this process entails the collapse of description and prescription. The temporal superposition of product and process inherent in the collapse of description and prescription is the essential characteristic of indexicality.

The implicit objectivity and directness of the relationship between conceptual input and design output underlies Kwinter’s anti-subjective account of new expressionism. One of its more problematic secondary effects is that this “new expressionism” risks appearing as “architecture without subjective agency.” This is a misreading. Kwinter’s account does not remove subjective agency (let alone creativity) from design. Rather, by taking the implications of indexicality to their logical conclusion, it resituates agency as a series of curatorial acts: the selection of properties to express, of flows to index, of diagrams to accomplish the translation of the virtual into the real.

If we now turn our attention back to hyperindexicality, we see clearly what makes it such a radical break: it is indexical architecture without the goal of legibility. Inputs produce formal outputs, but the diagram that accomplishes this translation is too complex to permit the aforementioned temporal collapse of description and prescription. And yet – conceptually speaking – this collapse is still in place. We know that in fact there’s a set of inputs being mapped onto a formal output. We’re left with the diagram as a promise of this translatability and the building as an image that refers us to the design method and its attendant ideology.

Another way of addressing this break between indexicality and hyperindexicality is by trying to relate the latter to skeuomorphism (as I did with the former at the beginning of this essay.) To restate: skeuomorphism justifies itself on the grounds that a legible relationship between past and present increases functionality. Hyperindexicality is interested in the mere existence of this relationship, independent of whether or not it can be discerned. The translation of this idea into skeuomorphism dramatizes the difference: it would be as though a designer argued that historical references were productive even if they couldn’t be recognized as references.

Stated this way, we see that indexical legibility has a fundamentally semiotic dimension (something which is not necessarily  the case for hyperindexicality, which isn’t interested in this sort of conceptual legibility.) Since we can’t literally have the past over again, indexical (or skeuomorphic) legibility depends upon reference. House II is not literally its process of formation; it uses semiotic techniques to evoke this process. This extends to the way material Eisenman’s used materials: when he clads the Aronoff Center in EIFS, he’s not deluding himself that “EIFS isn’t a material”; he’s thinking “EIFS will contribute to my desire that the building be understood as not ‘about’ material.” The semiotic dimension of indexical legibility is admitted and instrumentalized.

If we look at hyperindexicality, we see a sort of (often, presumably, unintentional) referentiality: as I argued before, these projects refer to the design process that bore them in a way that might be described as “ideological,” but they also engage in this sort of “cross-scale inauthenticity” where the aggregation of assemblies at one scale refers to its prior, virtual existence as pure, plastic geometry.

This is where, in my view, things start to get prickly for indexical materialism. This referential operation is forgotten in material indexicality, which pretends that the result (e.g. a building) is a direct consequence of the process, rather than a system of signs that refers to that process. A “fold” in contemporary architecture is not an actual fold, but an assemblage intended to look like a fold. In other words, it’s skeuomorphic: these “material expressions” consist in the reproduction of a material property which is “authentic” in one context in a new context in which is it “inauthentic.” The digital reproduction of a leather texture has the same relationship to real leather that the façade-scale reproduction of a fold does to a bent soda can.

While we can see how Kwinter’s description of R + U’s design process as akin to a chemical reaction could maintain the designer’s importance as a curator of material expression, it is difficult to see how this “mere expression” can be reconciled with the fact that the building is not the direct result of this process, but an image that refers to this process. Moreover, the “actual” process – likely far more complex and dependent on such concrete and messy things as trucks, screw guns, plotters, city officials, laborers, etc. – is elided.

Just as I argued that inauthenticity was not necessarily a problem for skeuomorphism, I don’t think it’s necessarily a problem for materialist indexicality (though it does force on to question whether the philosophical implications of “materialism” are appropriate in this context.) But there’s something bothersome about the fact that the architectural version of skeuomorphism believes itself to be free of semiotics while producing designs that rely heavily on metaphors. Ultimately I think this cognitive dissonance reflects a disciplinary problem. Even in the absence of a “reading subject” who beholds the architectural object, semiotics often plays an unacknowledged role in the design process. “Conceptual architecture” – since Eisenman at the very least – has never been content to have a concept, it must make the concept legible.

To close the gap between “authentic expression” and “legibility,” these conceptual projects make recourse to a kind of “semiotic overstatement.” This is why Banham writes that Archigram’s genius was to understand that, even if a computer city might look like nothing, it had to look like something. This is why architecture generated by algorithms intended to maximize the efficiency of its flows is curvilinear. It’s why Greg Lynn moves from buildings whose rectilinear geometry indexes the virtual curves used in the design process to actually building those curves. Architects interested in exploring materiality produce hyper-articulated details: it’s a way of not just having materials, but of drawing attention to them. It’s why we take smug pleasure in knowing that Zumthor’s walls at Vals are not actually composed of enormous stones, and it’s why architectural minimalism is a design aesthetic, not a construction method.

Skeuomorphism 2: Authenticity

There’s an interesting point of critique in the last line of the Metro tagline/manifesto that I posted before: by declaring their designs “entirely authentic,” the anti-skeuomorphism designers are implicitly declaring skeuomorphism’s inauthenticity.

As discussed in the previous post, skeuomorphism’s “inauthenticity” comes from taking relationships that were intrinsic to the way something worked (for example, the volume knob on an old stereo reflects its underlying mechanical basis) and recreating them such that their visual presentation is functionally extrinsic (there is no such underlying mechanical basis for turning a virtual knob to adjust volume.)

This opposition between “authenticity” and “inauthenticity” is perhaps clearest in the disagreement over fake materiality. It’s easy to read the flat, super-saturated colors of Microsoft’s “Metro” interface as an attempt to find visual elements without real-world analogs. Metro apps assert their virtuality, just as skeuomorphic interfaces on iOS assert their materiality.

Yet if we think carefully about this division – skeuomorphism as inauthentic and anti-skeuomorphism as authentic – things get more complex. This complexity becomes apparent when we notice that, directly adjacent to the battlefield where the skeuomorphism debate is being fought, there’s a swath of strategic territory that has gone wholly uncontested.

If “fake materiality” characterizes the majority of skeuomorphic design elements, we might think about this uncontested territory as the land of fake spatiality. And yet, while criticizing the former, Metro designers are explicit in their praise of the latter. According to the Building Windows 8 blog, “every piece of UI comes in from somewhere and goes somewhere when it exits the screen.” This isn’t unique to Metro, of course. Neither side in the skeuomorphism debate is making any move back toward entering commands line-by-line or designing static interfaces.

But there are no “somewheres.” In a digital interface, places and spatial relationships emerge gestalt-like out of the consistency and coherence of the visual cues that refer to them. When Metro designers decry the digital simulacrum of a volume knob, their criticism is based on the fact that there’s no underlying connection between function (which is now computational, not mechanical) and presentation. But this line of criticism is never extended to space and movement, which are no less arbitrary in relation to the underlying computational functions.

The “justification” (if we want to call it that) for the space-and-movement paradigm adopted by both sides in the skeuomorphism debate is not “authenticity,” but the functional benefit of using consistent, familiar metaphors (which sounds suspiciously like the justification for skeuomorphic design as a whole). It’s not that certain techniques aren’t more or less functional than others, it’s that we shouldn’t confuse “functionalism” with “authenticity.”  The former admits experimentation and debate (e.g. “Is anachronism functional or not?”). “Authenticity,” on the other hand, admits no further criteria. “Do you like this painting?” is an interesting question. “Is that a real Picasso?” is not.

Unable to define merit in terms of authenticity, we’re left not only to judge the value of particular techniques, but even to determine what ends these techniques ought to pursue. For example, skeuomorphic design might be more or less functional – but is it good taste?

It might be easier to see the value of bringing in new criteria of judgment if we look beyond digital interface design. Recently, Lebbeus Woods wrote a blog post about Michelangelo’s designs for fortifying the walls around Florence that speaks to this potential:

For all their practical purpose, these drawings have uncommon aesthetic power. Of course, this is because they are made by one of the greatest sculptors, and a self-taught architect—an “amateur of genius,” as he has been called—but it is also because the bastions required had too short a history as a building type to have ossified into a rigid typology. Michelangelo was relatively free to invent strong new forms and didn’t hesitate to do so.

Another way of saying “too short a history as a building type to have ossified into a rigid typology” is to say “too short a history for there to be clear linkages between techniques and criteria for judgment.” Michelangelo’s freedom here isn’t actually the product of new goals (or at least it doesn’t have to be): even if all he had in mind was “create functional fortifications,” he didn’t have sufficient precedents available to him to know what design techniques would accomplish that goal.

In the Letter from the Editors for PLAT 2.0 (an issue of the journal I co-edited with Eileen Witte), we wrote something that seems to overlap with Lebbeus Woods’ observation: that technological development can change the standards of judgment (or their hierarchy). Our example was Auto-Tune: by making it relatively easier for vocalists to sound like they’re singing on key, Auto-Tune increases the relative importance of other criteria for judging performance – say, physical appearance.

It’s obvious that technological development can produce innovative design, but we tend to consider only the direct advancements due to new materials and machines. The Michelangelo and Auto-Tune examples point to a different mechanism: the novelty that emerges from disrupting existing relationships between techniques and criteria for judgment. But this mechanism suggests that such experimentation is not contingent upon technological development: Peter Eisenman’s House series provides another example where changing the criteria of judgment (in this case, via a linguistic theory that wanted to make architecture that served as a meaningful text rather than functional machine) proved liberatory.

Any discourse where “authenticity” and “legitimacy” are the criteria for judgment tends to imply that the value of its techniques is universal. In the prevoius part of this discussion of skeuomorphism, I concluded by saying that usability which hinged purely on familiarity tended to produce homogeneity and close down innovation. Here I’ll extend this to say that design which justifies itself in terms of “authenticity” has the same result.

Skeuomorphism 1: Anachronism

First: I’ll just confess right now that this didn’t end up going where I thought it was going to. Nothing about “type” here.

Second: a story.

Once upon a time, there was a device named A.

A performed function X.

Because of the way A did X, it had a set of visual characteristics, B.

One day, a new device showed up named C.

C also performed function X.

Although the way that C did didn’t require it to have visual characteristics B, it had them anyway.

This story highlights the fact that “skeuomorphism” is a status conferred by participation in a specific historical narrative. The critique of skeuomorphism responds to the consistency of these historical relationships: it’s not that visual characteristics B are inherently bad; it’s that keeping them around in device C is inappropriate and inauthentic. The skeuomorph is necessarily an anachronism; the anti-skeuomorphism camp gives this fact a normative addition, judging that anachronism is bad design.

There’s a story in architecture with the same structure: it’s the classic tale of how a once great style with a political and/or social agenda became an empty aesthetic. (The canonical example is the shift from early modern architecture to the International style.) At some point the style’s techniques (expansively defined) were legitimately connected to a social and/or political agenda; eventually, however, we find architecture which seems to consist only of the superficial reiteration of those same techniques, devoid of ideological effect. These critiques share in common the idea that the legitimacy of a technique is historically contingent.

What about the other side in this skeuomorphism debate? To return to the variable-filled narrative from earlier, those in favor of skeuomorphism argue that visual characteristics help us understand that C does the same thing did. They, too, implicitly recognize the historical contingency of technique: as a simulacrum, the skeuomorph doesn’t retain the function of the original design element it references. In fact, a skeuomorph always makes the same performance: it reestablishes visual continuity between devices whose functional basis has become discontinuous.

The long-term (and wholly unsurprising) consequence of repeating a technique for the sake of establishing familiarity is that we forget the origin of these echoes. The skeuomorph disconnects from its historical (potentially functional or ideological) referent and becomes part of a homogeneous field of equally untethered brother skeuomorphs.

Since I’m starting to sound intellectually unfashionable, I should clarify something: the problem with skeuomorphs is not their lack of meaning or historical reference (pff!), but the tendency of this paradigm to equate “usability” with “familiarity.” Historical “familiarity” becomes ahistorical “consistency” becomes stylistic ossification. Applied to the architectural narrative, the problem isn’t that endless repetition of the International Style disconnected it from the ideologies underlying modern architecture, but that endless replication of the aesthetic tends to work against more complex engagements with new contexts.

To Skeuomorph or Not to Skeuomorph?

There’s a debate going on in the land of graphic design over skeuomorphism. What’s a skeuomorph? The OED’s got our back via Wikipedia (just being honest): “A skeuomorph is a derivative object that retains ornamental design cues to a structure that was necessary in the original.”

It sounds pretty straightforward until you try to pin down definitions for “object,” “ornamental,” “cues,” “structure,” “necessary,” and “original.”

I’m going to approach the term along a more concrete path: apps. Skeuomorphic design features you may have encountered:

  • Handwriting fonts in apps that use a keyboard for input
  • Spiral bindings at the top of your digital calendar app
  • Wood grain background for your ebook collection
  • Dials to control volume in a music app

At this point you’re either someone who knew about this debate and knows where this is going, have recognized that all of the above are characteristic of apps on Apple’s iOS, or you’ve never used an iPhone.

The debate over skeuomorphism could probably be had if it were only iOS providing an example of the “pro” camp. (The benefit of theoretical discourse in design is that we can fight about things without actually having the things.) But then the guys and gals in charge of the user interface guidelines for Windows Phone and the upcoming Windows 8 operating system decided to become evangelists for anti-skeuomorphic design. To the fake brushed steel, leather, paper, and wood textures of iOS, Microsoft responded with flat, colored squares and overlaid with consistent sans-serif typography.

Nothing says “first world” like watching the cliché and increasingly irrelevant Mac versus PC debate renewed in the form of a proxy war for an esoteric debate about graphic design. Well, nothing except interrogating the parameters of this debate from an architectural perspective.

There are three things I want to talk about in relation to skeuomorphism. They overlap a lot, so I’m going to split the discussion into three separate posts in the hope that it forces me to retain some argumentative organization. I’m listing them here so that if I exhaust a topic early or forget what I wanted to say before I get around to writing it, it’ll still be here as a provocation.

One: Traditionally, criticism leveled at skeuomorphic design makes reference to the fact that the work is anachronistic. I want to point to the fact that “anachronism” a tricky term in relation to design, but that this complexity might suggest new ways of talking about type in architecture.

Two: There’s a whole “class” of arguably skeuomorphic design elements that have gone unaddressed by the anti-skeuomorphism camp. This “naïve skeuomorphism” rests on a set of unquestioned architectural metaphors whose invisibility is performative in ways that have not been sufficiently explored.

Three: As we dig deeper into the debate over skeuomorphism, we discover that it arises out of the more generally problematic/complex intersection of metaphor and functionalism. Looking more carefully at the debate about skeuomorphism illuminates certain problems in architectural discourse that have been paved over or considered resolved.