the memory of photography

the memory of photography [blog]

Photography is a primary example of what Régis Debray would call a technology of memory. These technologies include writing, printing, painting, sound recording, computer code, and all the other ways in which we inscribe images and utterances in order to preserve and/or transmit them. Photography has of course since its invention, to a much greater extent than any of these other technologies, also been conceived as a powerful metaphor for memory – and memory has often been likened to photography. (Recall Oliver Wendell Holmes’s description of the daguerrotype as a ‘mirror with a memory’ or look up the cognitive science concept of flashbulb memories.) As anyone who possesses a family album will acknowledge, photographs are important instruments in the work of memory. In this project, however, I am also concerned with how memory functions as a crucial tool when working with photographs.


Inspired by the evocative representation of a photographic library in Stephen Poliakoff’s Shooting the Past (BBC, 1999), and building on my own personal experience as a picture researcher, I’m investigating how memory is utilised by those whose job it is to classify, organise and retrieve photographic images. One of my contentions is that the type of ‘photographic memory’ which is a major asset when attempting to locate a given picture in a physical archival space is quite different from the type of memory required when trying to retrieve an image from a digital archive using a text-based search engine. The very exact visual and spatial memory that the term photographic memory refers to seems to be of little use when one cannot remember the precise keyword that will bring up the desired image on the screen. And likewise, it seems an impossible task to supply each image file with enough keywords, or metadata, to cover all the potential associations and connotations that the human brain might make from the totality of visual stimuli contained in one single image.

Computer scientists work to bridge this semantic gap between language and visual interpretation by the development of ever-more complex image recognition software, with a view to enabling so-called content-based image retrieval (as opposed to the concept-based format of the keywording system, which in effect is only a digitised version of the traditional index or card catalogue). In my project, however, I am less concerned with how computers might be taught to replicate human brain functions than with how the actual human beings working with pictures and computers negotiate, on a daily basis, not just the semantic gap between data and image but equally the sensory gap between image file and image object.

The notion of the photograph as a material object is critical to the other aspect of photography’s relationship to memory that I want to explore in this project. This is related to questions first raised in my PhD thesis, which engaged with the intimate relationship between personal, cultural and material memories as they converge in popular photographic representations of the not-too-distant past. Towards the end of that project, I became intrigued by how the memorial significance and poignancy of black-and-white photographic prints appears increasingly bound up with the pastness they figure by virtue of their very material support – that is, their physical manifestation on photo-sensitised paper or film. My current work proceeds, in part, from this premise that the photographic print is increasingly coming to represent a material memory of the medium’s own past.