Herbert Simon situates his essay as a communication between the humanities and the sciences and in so doing opens up a space for writing their emergent border histories, in particular, border histories of digital machines. Once opened, traffic across the borders questions the neat and safe division of the two cultures and plays with Simon's "unabashed," opening embrace and his missionary position.
My response locates itself in the space of this middle-passage. I invite Simon and readers to leave the "bridge" that spans the essay to set sail in its waters. Together we will visit places and times that contributed to building the architecture of the von Neumann computer, the contested but nevertheless dominant and dominating computer architecture of the post-War period up to the 1980s. Simon's model of memory, a model crucial to his arguments for "assigning 'meaning' meaning," re-presents this architecture crafted in the post-Holocaust, post-Hiroshima, post-colonial days of the 1950s. This computer architecture encrypted and sealed over imperial histories of its design and thus foreclosed the anxiously intimate process of re-memoration and mourning needed to tell emergent histories of its fabrication. Such a mournful voyage is worth making, since it can help to multiply the possibilities for other computer architectures, other memories, different technocultures. Donna Haraway has passionately described a vision of this knowledge:
We are also bound to seek perspective from those points of view, which can never be known in advance, which promise something quite extraordinary, that is, knowledge potent for constructing worlds less organized by axes of domination. In such a viewpoint, the unmarked category would really disappear-quite a difference from simply repeating a disappearing act. (Haraway, 1991: 192).
Both the sciences and the humanities collaborated to "invent" the von Neumann architecture in the the crucible of colonialism and imperialism. It was a long-time in the making. We can find rusted fragments of its assemblage in heterogeneous spaces such as medical meetings held in late-nineteenth century London; in census reports for India distributed in the Imperial Gazeteer of India (1881); in the anthropology lectures of Lévi-Strauss given in the École Pratique des Hautes Études of the 1950s; and in the travel books of cyberspace written in the 1980s. As we sail to these places, our wake will trouble the waters of "evocation" of meaning (a key trope of Simon's paper) in an effort to produce a passage of meaning.
In the middle of things in 1881 we find ourselves in London at the Seventh International Medical Congress. A group of experts ponder over the damaged head of a dog in a suitcase and the paralytic movements of two monkeys from which parts of the brain have been removed. They decide on the basis of this evidence that brain function is localizable, that X can mark the spot for memory, speech, visual functions, etc. The territory of the brain was now officially open for imperial claim and different scientists competed in the grab. At the same moment (1881), an array of heterogeneous "mapping" activities for India conducted by the British (studies of its geology, meteorology, botany, zoology, ethnology, castes, languages, and population-the first synchronous census of the population of Indian Empire was taken in 1881), as well as "official" maps of Imperial India came to be gathered together in nine volumes and published in England as the Imperial Gazeteer of India. I wish to join these two gazeteering efforts, one for the territory of the brain, the other for the colony of India, as problematic exercises co-constructed by the humanities and the sciences. The desire to locate, fix and hold Empire in order inscribed itself on the "natural" surface of the brain and on the "anthropological" surface of India.
On these supposedly "natural" surfaces it was possible to erect "history" (Simon's inverted "T"(20) comes to mind) as a past separated from the present. Around this division, notions of memory could be organized and normalized. "National" memory was split into the long-term memory of the past, which Simon calls "the cocoon of information, stored in books and in long-term memory, that man spins about himself"(Simon, 1981: 127), and into the "stimuli" of the present. The work of "accessing" long-term memory fell to the historians. At the same time the branches of neurology came to normalize "memory" in its biological (natural) domain, with implications for culture and history. It is hard to tell the historian apart from the neurologist in this project. Not surprisingly, history and neurology came to institutionalize themselves with the creation of the journals-Brain in 1879 and the English Historical Review following in 1886. To sail quickly on to other waters-the divisions of nature and culture, history and anthropology, past and present, and long-term and short-term memory required certain revisions in the late 1940s and 50s, as Britain "lost" the Raj and Algeria fought against the French (1954-1962). The fantasy of the "intransigence" of the subaltern translated into "intransigence" in a different register. For politically concerned humanists it now seemed improper to divide the world into the civilized and the non-civilized. Societies once relegated to anthropology, "the people without history," pressed so hard on the limits of the "societies with history" that their making of history could no longer be denied or disavowed. In 1954, as a sign of the times, Claude Lévi-Strauss, the renowned anthropologist, retitled his course at the École Pratique des Hautes Études from "Religions des peuples non-civilisés" to "Religions comparées des peuples sans écriture." Through a complicated nostalgia, Lévi-Strauss conserved the division of labor between anthropology and history, nature and culture, civilization and tradition, and long-term and short-term memory by refiguring these divisions onto speech and writing. His reconfiguration occurred at the time that von Neumann was engaged in developing and writing about computers. Lévi-Strauss' Savage Mind (1962) could write von Neumann's The Computer and the Brain (1958) and vice versa (note also that the first conference on Artificial Intelligence convened in 1956). The oral, according to Lévi-Strauss, is timeless ("the characteristic feature of the savage mind is its timelessness," Lévi-Strauss, 1966: 263), and therefore spatial, locatable, a kind of short-term memory, just as the written is the archive, long-term memory, "the embodied essence of the event"(Lévi-Strauss, 1966: 242), but also spatial and locatable. Indeed, Lévi-Strauss concludes The Savage Mind with a discussion of Information Theory (which is always capitalized in the text):
But the idea that, theoretically at least and on condition no abrupt changes in perspective occurred, these two courses [savage mind and Information Theory] were destined to meet (Lévi-Strauss, 1966: 269).
Lévi-Strauss and Von Neumann worked as relays for this destiny in the 1950s.
So deeply inscribed is the imperial machine and its empire of memory that it can be found reterritorializing itself into cyberspace, most notably in the travel books of William Gibson (Gibson, 1984; Gibson, 1987; Gibson, 1989). Gibson has claimed that "on the most basic level, computers in my books are simply a metaphor for human memory; I am interested in the hows and whys of memory, the ways it defines who and what we are, in how easily memory is subjected to revision."
In a recent study I have tried to show how nineteenth-century locational notions of memory shape these travel-books and how memory works in them, once again, as a space of terror, which forecloses the possibilities of rememoration, of writing and reading emergent histories.
The timely question remains then-why this strong, ongoing investment by the humanities and the sciences in imperial machines and empires of memory? How is it that there is an abiding longing for this dream: "The entire process of human knowledge thus assumes the character of a closed system"(Lévi-Strauss, 1966: 269)-Von-Neumann architectures of oblivion? There were other competing computer architectures (the emerging history of the death and rebirth of neural nets between the 1960s and and 1980s begs to be told) and other ways of thinking of memory and rememoration. Perhaps this is where we should have sailed first, to those sites of failure, to those sites excluded from production in the 1960s and 70s. It seemed important to me as a historian, however, to point out some of the rusted fragments of these imperial machines and to consider the disavowed and anxious intimacy of the sciences and the humanities in guarding the empire of memory over the past century.
The time for the voyage in has come. Underway, we can keep these words from Beloved in mind, as the humanities and the sciences sail that course together:
As long as the ghost showed out from its ghostly place-shaking stuff, crying, smashing and such-Ella respected it. But if it took flesh and came in her world, the shoe was on the other foot. She didn't mind a little communication between the two worlds, but this was an invasion. (Morrison, 1988: 257).
Let the bridge over the sciences and the humanites break under this jubilee and let jubilee machines rememorate their borders.
1. My use of emergent histories (in spite of the difficulties that can be found with the term) comes from the work of Homi Bhabha: "The postcolonial passage through modernity produces a form of retroaction: the past as projective"(Bhabha, 1992: 56); see also Sara Suleri: "If both [colonizer and colonized] are identifiable as victims of traumatic change, then the idiom of trauma itself requires reformulation that can provide a language for the slippage of trauma from apocalypse into narrative"(Suleri, 1992: 5); and Paul Gilroy: "Defenders and critics of modernity seem to be equally unconcerned that the history and expressive culture of the African diaspora, the practice of racial slavery, or the narratives of European imperial conquest may require all simple periodisations of the modern and the postmodern to be drastically rethought"(Gilroy, 1993: 42).
2. Von Neumann computer architecture separates memory from processing and is built according to nineteenth-century notions of localizing brain function: X marks the spot for memory, speech etc. For a classic discussion of memory in computing see von Neumann, 1958, especially pp. 36-37; see also Smith, 1989: 277-303; for a critical discussion of early computers designed to match stimuli with fixed memory stores (the model used by Simon in his essay) see Rosenfeld, 1992.
3. My comments here draw on an extensive literature of historiographical critique, see especially Dirks (1990), de Certeau (1988), and Spivak (1991).
4. My discussion here is based on the following texts: Lévi-Strauss (1966), Brotherston (1985), Derrida (1976), and Jacobs (1993).
5. From an interview with Gibson in McCaffery (1991). See Biddick (1993) for a detailed discussion of the models of memory used by Gibson in the trilogy and a critical inquiry to rememoration in cyberspace.