GOMBRICH AND THE LANGUAGE MODEL OF DEPICTION

Two heresies have dominated the most interesting work to date on pictorial perception. According to the first heresy, a pictorial system is really an arbitrary sign-system or language for describing reality; the users of the system, both the senders and the receivers of pictorial messages, must learn the conventions for the pictorial sign before they can interpret it. Different cultures, different traditions, different artists and styles, even different media and techniques of representation, all carry with them distinct conventions which the intelligent decoder must learn before he is able to decipher the content of the pictorial symbol. I shall call this first heresy the semiological heresy.

A second heresy pervades art historical writing. It is the doctrine that the depiction of an object gives you an illusion of seeing that object. The pictorial artist strives to dazzle the beholder with a mirage of reality. According to this view, the progress of pictorial representation since the Renaissance rupture with medieval art has been a story of ever-increasing illusionistic effects. A claim which is a typical expression of this heresy is that Giotto's paintings achieve an illusion of space which is quite absent from the work of Cimabue. I shall call this heresy the illusionist heresy.

I regard the first as infinitely the more dangerous and misleading of these two orthodoxies – for indeed both heresies have now the status of pervasive dogma. The semiological heresy is the more pernicious view not just because, of the two heresies, it is the one which is more obviously incompatible with my own view, though indeed this is naturally a factor in my disinclination to accept it.