Submitted by richard on Tue, 2006-08-22 18:32.
Colour Separation:

is constructed from photos of over one hundred people who are related in some day-to-day way with the core Mongrel group. Using "HeritageGold" software it transforms their images into eight un-glamourous sterotypes of black/yellow/brown/white men and women.
These are people that never existed. These images wear the masks of the other stereotypes. The masks are spat on. On the cover we have a white man wearing a black mask covered in spit. We have no idea who has done the spitting. Is it a white man fed up of his friend pretending to be black? Has he been spat on because he is a black-masked man who is white underneath?
"In this work as in the rest of society we perceive the demonic phantoms of other 'races'. But these charecters never existed just like the nigger bogeyman never existed. But sometimes... reluctantly we have to depict the invisible in order to make it disapear."
Harwood
Colour Separation was distributed through various outlets in London, the UK and abroad.

This project produced a series of forty images featuring the head and skin from the project's contributing artists, also the artists' friends and family, and from groups local to the community in were we worked.

Each artist, with their friends and family, represents in their genetic heritage, one or more of the following geographical regions - Africa, East Asia, North Europe and South Asia - indicated respectively by skin pigmentation ('black', 'brown', 'yellow', 'white') and biological difference.
The head images will then be merged digitally, made anonymous and separated again to arrive at four distinct racial types, both male and female.
The eight composite anonymous portraits were then masked. (Masking appears in the project in a manner similar to this example: Masking the face of an anonymous portrait encourages the viewer of a poster to be caught up in a racial dichotomy. The viewer is unsure why the base face has the skin of another face sewn onto it.)
Further, with spit added, the user can detect signs of racial abuse but not identify who is the abuser or who is the abused.