Tag-Archive for » web «

other uncanny valleys

It seems that nearly every article on technology and culture I read these days references Masahito Mori’s concept of the uncanny valley at some point. It’s entered the intense late-meme phase which makes me think it’s about to be obliterated by some hot new techno-trope. But I think the underlying idea of the uncanny valley is actually a robust way of looking at other forms of cultural dissonance and estrangement that have nothing whatsoever to do with robots.

For instance, before I even came across the term, I had identified for myself something that looks to me like an uncanny valley of history: the period between the end of living memory—for practical purposes, about 100 years ago—and the beginning of what we are comfortable thinking of as history because there is substantial scholarship on the primary documents—roughly 125 years ago. In that uncanny valley we can’t quite find our familial connection to history any longer, but it’s close enough that it doesn’t yet feel completely autopsied by conflicting lineages of critical scholarship. Right now, that valley falls roughly between 1890 and 1915. Others might place the valley slightly differently, or even disagree that it exists at all, but it certainly exists for me.

I’ve recently identified a second, more culturally bound, uncanny valley pertaining to digital culture. This valley exists in the 70 years between 1923—before which nearly all printed matter is out of copyright in the United States—and roughly 1993—when the generation of early web adopters began putting the documents of their lives online. Although a good deal of writing and other media from this uncanny valley has been uploaded, a lot of it is behind paywalls of one kind or another (including the copyright-driven paywalls of Amazon.com and its ilk). Huge amounts of pre-1923 media have been uploaded through Project Gutenberg, Google Books, YouTube, major libraries and archives, and oddball projects like BiblioOdyssey. At the near end of the valley, you can find important or just interesting writings, artworks, and documentation by all the digerati online for free (as well as some behind paywalls). For a lot of what’s in the uncanny valley itself, you have to buy the materials online, pirate them, or go to a traditional dead-tree library. I’m not saying it’s not worth the effort to do this—far from it!—but that as one reads, skims, surfs, browses, and devours one’s way around the web these days, it can feel as if Rudyard Kipling is closer to us than William Faulkner, R. Austin Freeman than Sara Paretsky, D.W. Griffith than the Marx Brothers, Julia Margaret Cameron than Margaret Bourke-White.

What’s your uncanny valley?

Category: Uncategorized  Tags: ,  Comments off

Cartoon We

In an idle moment the other day (yeah, I really shouldn’t admit to having those in academia), I loaded into my Firefox an add-on I found on Facebook called MyWebFace that promised to let me make a cartoon image of myself.

An aside: I might add that I found this app only because I was searching the Facebook ad board looking for an ad that I would be willing to  ‘thumbs up’ to appear alongside my page. Not only did I find no such thing, I discovered that Facebook now does not allow users to ‘thumbs down’ an ad. Possibly this is my fault, since I relentlessly thumbs-downed (down-thumbed?) dozens of ads in my first weeks on Facebook, hoping to help skew the ad pool toward public service ads and away from shopaholic ads (the bulk). It always gives you a good feeling as an adult citizen of a republic when your options are: you may approve of this, or you may approve of this.

cartoon-meBack to MyWebFace. It’s set up as a kind of simple Identi-Kit, allowing you to construct your face from mostly predictable parts: noses, eyebrows, lips, etc., all of various shapes. Size, color, and placement of most objects can be adjusted. After I was finished, I ended up with what you see here. Anyone who knows me will see that this is not a good cartoon of me; the reason for this lies in the kinds of choices the software gives—and just as importantly, withholds. Eyebrows, eyes, and mouths come in a fair range of shapes (not enough noses, though). Skin color is wide open– a matter of picking from a palette of millions of shades.

For face shape, however, you appear to be stuck with a default upper half of the face and a modifiable chin. Result for me: entirely wrong face shape. (Oddly, you can choose various kinds of ‘blush’ for your cheeks, as if that is more important than basic face shape.) In addition, the general body type is  too skinny for me and not modifiable (this head shot is a detail of a full-length image). Hair style: no options match my admittedly idiosyncratic style. Accessories: no glasses frames match the ones I wear. Hair accessories and hats: nothing matches what I wear; the earphones were the best option because I use them when playing online games. Clothing: nothing really matched in the limited choices, largely because almost all the clothing styles skewed at least a decade younger than I am. In general, registering aging was not an option. You can make your hair all blond/red/brown/black, or all gray, and nothing in between—no one in MyWebFace’s world is grizzled. There is one pane for wrinkles, but they’re so lightly drawn and weirdly thought out that they don’t make the face look older so much as scribbled on. Basically, you can’t make yourself look older than about 25 with this software (or to be fair, much younger either; no kids need apply). Can you say target demographics?

Filippa HamiltonCoincidentally, as I was halfway through writing this, I stumbled on the image below over on boingboing.net, in a post rejoicing in the wonderful  title “Ralph Lauren opens new outlet store in the Uncanny Valley.” They credited a favorite site of mine, Photoshop Disasters, where the image has since been made to disappear by a DMCA-wielding Ralph Lauren; see this Huffington Post entry for a similar threat against boingboing. Ralph Lauren has admitted to their “poor imaging and retouching that resulted in a very distorted image of a woman’s body.” Yet model Filippa Hamilton was fired by Ralph Lauren last April; she says they told her she was overweight. So fashion logic has finally created the inevitable impasse for itself, in which the cartoonized emaciation of this image is “very distorted” but a 5’10″ model weighing 120 lbs is “overweight” (although one might want to take models’ claims about their height-to-weight ratio with a grain of salt; given the constraints of their job, they have every incentive to modify this figure to gain professional advantage.) In other words, at ground zero of the American female body image, the concept of “just right” has at last shrunk to a complete null set.

My reason for making this post, though, has less to do with the general problem of body image than with the propagation of this cartoonization through software—through Photoshop, which allows it (but does not require it; you can just as easily make yourself fatter and older in Photoshop), and MyWebFace, which actually requires it. What has previously been reinforced through consumption of media created by others, we are now made complicit in reinforcing through the software we ourselves use and the objects we ourselves make.

Will the semantic web be gendered?

Via maybe maimed but never harmed, Read Write Web reports researcher Corrina Bath’s cautions of gendered ontologies in the Semantic Web.

Too often, “binary assumptions about women and men are not reflected [upon] or the (gender) politics of [a particular] domain is ignored. Thus, the existing structural-symbolic gender order is inscribed into computational artifacts and will be reproduced by [their] use.”

Bath cites Bowker and Star’s example of how phone books in the US were first arranged indexed by the husband’s name, reflecting assumptions about the use of infrastructure and truth of American social life at the time.

In the original, longer interview at Austrian Semantic Web, Bath expands that the stakes of feminist ontology in the semantic web are two fold. First, what kinds of relationships between knowledge objects will be formalized, how will minority interpretations be handled, and what room for contestation of knowledge obects will there be? Second, and relatedly, will the semantic web recognize the contingency of truths and the situatedness of ontologies?

Category: Uncategorized  Tags: , , ,  Comments off