Pathetic Fallacies & Category Mistakes: making sense and non-sense of the (near-future) Sentient City
“Sentience” as a concept has long been a flash point of controversy between the humanities and sciences when non-human actors are concerned. Ruskin coined the term “Pathetic Fallacy” in 1856 to signify any description of inanimate things that ascribes to them human capabilities, sensations, and emotions. His translation of the Latin phrase “natura abhorret a vacuo” (nature abhors a vacuum) is widely known, and even deployed (relatively) recently by a military general as a reason for NATO’s swift entry into Kosovo following the withdraw of Serbian Forces in 1999.
Within literature anthropomorphism is by now an accepted literary device, yet within the natural sciences, for example, it is still considered a serious error in scientific reasoning if taken literally. Latour suggests that the difficulty lies in describing agency in the absence of anthropomorphic actors, that there is a lack of accepted vocabulary to address the non-human agency of “things,” technological or otherwise. His “actant” is but one contribution to building such a lexicon. Anne Galloway’s recent post on “lacking words to describe non-humans, and what this means if we try to account for relations between humans and non-humans” is essential reading.
As Keller Easterling points out in her essay Only the Many (Log 11, Winter 2008), the term Category Mistake, introduced by Glibert Ryle in The Concept of Mind, describes a seemingly nonsensical mixture of logics. “For instance, one mistakes a part for a whole, or inverts levels in a hierarchy”, she writes. “Or a child thinks a division is a smaller part commensurate with a battalion or a squadron, when it is the overarching category for those of smaller divisions”.� Beginning with Jesus and extending to messianic characters in general, she shows how category mistakes are markers for dominant logics with universal claims, but goes on to suggest how they can also serve as an escape hatch out of the monotheisms of logic and discipline. “In order to find the trapdoor into another habit of mind, one would not quarrel with, but gather evidence in excess of” these dominant logics. She closes with a quote from Gregory Bateson’s Metalogues (in Steps to an Ecology of Mind):
D [daughter]: But you said that if we always talked logically and did not get into muddles, we could never say anything new. We could only say ready-made things. What did you call those things?
F [father]: Clichés. Yes. Glue is what clichés are stuck together with.
D: But you said “logic,” Daddy.
F: Yes, I know. We’re in a muddle again. Only I don’t see a way out of this particular muddle.
Non-human sentience is no stranger to the science fiction community. From Clarke and Kubrick’s HAL (sentient machine) and Stanislaw Lem (and Tarkovsky’s) Solaris (sentient planet) to DC Comics’ Ranx the Sentient City, Gibson’s sentient cyberspace as portrayed in Neuromancer, the sentient programs of the Matrix, or Bruce Sterling’s spime (to name but a few), science fiction has imbued a range of inanimate “things” of all scales with forms of agency that do not map neatly to those of ordinary humans.
Five years ago, the Economist published an article – “The sentient office is coming” – that described then current research in augmenting computers and communication devices with sensors to enable them to take into account their environment and adapt to the changing conditions of their use. Here the aim was to create “convivial technologies that are easy to live with”. Yet as the article points out, cohabitation with sentient things is not without dilemmas. What happens when we the toaster in your home gets bored of always making toast, or the fax machine in the office thinks the tone of your fax doesn’t jive with that of the firm?
Achieving “sentience” in the domain of Artificial Intelligence (AI) research is both a serious research agenda with a long history as well as a good ploy for securing corporate sponsorship. ATT/Cambridge University’s Sentient Computing project (1999), which attempted to “combine sensors and computers to monitor resources, maintain a computational model of the world, and act appropriately.” Combining sensors and computers was at the time nothing new, but the broad attempt to “maintain a computational model of the world” proved daunting. As of 2006, the project was re-focused on tracking and location systems for “sentient” vehicles and sports.
Today the focus appears less on trying to maintain a proprietary computational model of the world, and more on using the world itself as “model” and letting ordinary people contribute to its making. Recent projects in Urban Computing and Locative Media have focused on crowdsourcing metadata by enabling people to markup and annotate digital maps with notes, images and media objects geocoded to specific places. A recent article by Kazys Varnelis and Leah Meisterlin of Columbia University’s Netlab explores some of the implications for designers in an age of such “intelligent” maps.
With the specter of peak oil looming globally, the general public is finally beginning to register that as a planet we need to negotiate our way of life with those of the various actants and ecosystems we depend on, be they environmental, political, economic, social or technological. Crang and Graham’s recent paper “Sentient Cities: Ambient intelligence and the politics of urban space” does a great job at outlining how corporate and military agendas are driving the technological ecosystems we’re likely to cohabit with in the near-future. Yet Crang and Graham’s analysis of the role of artists working with Urban Computing and Locative Media as one of “re-enchanting urban space”–of making visibile the invisible traces of things past, a “haunting of place with absent others”–renders artistic practice in relatively conservative and familiar terms, casting art in a reactionary role vis-a-vis technological development. What other roles might artists, architects and designers play in shaping how we inhabit the near-future Sentient City?