Database-driven

Built using PHP and MySQL, Turnstile’s database randomly served up a selection of images from an archive of just over 10,000 images. Each one was carefully keyworded to ensure that every web surfer would be given a unique browsing experience.

Behind the scenes was an extensive, yet easy to use, content management system linking keywords, images and snippets of text to provide an idiosyncratic viewing experience.

The site also featured text, thousands of snippets of overheard conservations — Earwitness — and observations from the street not caught on camera chip — Bystander. The database was designed to serve all content randomly so that no one knew what to expect, me included.

Photo play

Every photograph on the site was taken by me on my jaunts, either in my neighbourhood, on a trip or while on residency. My obsessive process is centred on the granular: a stuck to the pavement Band-aid, a sodden plastic bag, discarded or forgotten underwear and an endless list of detritus.

Every page featured a randomised list of keywords, along with a grid of images. Clicking a keyword would take you to a more focused grid, for example, a grid of Port-a-Loos, spills or oil splotches, or to an overview of images of a particular place: Berlin, Paris, Sydney or Canberra among many others. An A to Z link provided a way to navigate by topic and presented a growing group of typologies.

Turnstile was originally an element of my PhD research titled Cultural Mulch. My thesis considered the significance of mass-produced objects and their collectibility. The final work was a Wunderwall comprising 3,254 images and was titled: There are too many things in the cupboard.

Connected to Turnstile was Tokens, a blog featuring curious things.