Kinograffiti continued: the analog and the digital

Posted by Rhizome | Sun May 25th 1997 1 a.m.


+ + +


If today we find this sense of digital media in any way familiar, not to
mention appealing, it is perhaps because we have come to regard its
artifacts as ephemeral and consistent with a sense of passing time, a
momentariness to which we find difficulty holding on. To put this into
perspective, it is important to place this sense of digital media in
relation to other media -- those we now tend to call "analog".

It is generally acknowledged that digital media differ from the analog
variety. Nonetheless, both are created for the purpose of storing
information or data. Data, in any case, requires a structure or
substrate in which it can be stored at any one time. At any one moment
it is a resource in reserve for a later time. Analog storage is usually
characterized as being formed by means of an impression left by a
marking tool upon a receiving surface. In this category we have such
familiar forms as hieroglyphs, easel paintings, LP records, video tape,
and so forth. Digital storage, on the other hand, is characterized by
there being a substrate that is a logical matrix -- rather than a
physical surface subject to direct impression and manipulation. The
logical matrix is an architecture of cells or slots, each with its
unique address, within which a number, a symbol, rather than an
impression is stored. In the case of analog storage we might say that
the impression comprising the stored information resembles the marking
tool. That is, there is what is called an "indexical" relation, one of
spatio-temporal contiguity, between the tool with which one writes and
the mark or impression made. In the case of the digital device, the
symbolic encoding of the information effectively bears no visible or
tactile similarity to the encoding tool. In addition, the fact that each
slot of a digital matrix is uniquely addressable makes possible a type
of direct access to its symbolic content, known as "random access".
Analog information, being mapped onto a physical surface with variable
impressions, rather than a uniform matrix whose slots are filled with
symbols from a known domain, tends to reinforce a linear mode of access.
We tend more to run our eyes and hands, reading (metaphorically and
literally), over a surface containing analog information.

Much has already been said about this. I will only touch on one issue
which is important to my work. An operation that digital storage and
computation make possible is a sort of universal translation of
information from one physical manifestation into others seemingly quite
different ones. Thus, my voice, once recorded digitally, may be
converted into light on a display, or, through a motor, the motion of a
robot, or any number of other sensible phenomena. Not unlike the
symbolic system known as money, digital representation enables an
economy. Money enables exchange of objects independently of their size
or material. The structure of money, its tangible exchangeability, makes
it seem as if goods are more similar materially, and, though distant,
attainable, than they may actually be. So with digital computation, the
specificity of one form of experience, voice, for example, is dissolved
as one experience is translated into another, voice into picture. This
operation seems to correspond with a sense that things and ideas never
stand alone, but somehow relate to each other, and that there may even
be ways of connecting seemingly different or opposite concepts (such as
political positions).

Despite this alchemy, I don't think that this operation by itself
disturbs our senses of identity in contemporary culture. This is because
what I have been describing has been a conservative operation. In other
words, in stating that a form is translatable, through digital
computation, into other forms, I have taken for granted that our
original voice recording will have remained preserved. Untouched. Those
of you familiar with today's multimedia tools such as Adobe's Photoshop
are also familiar with the "Undo" facility. Identity is preserved by way
of duplication. One changes a "working copy" of an image rather than the
saved version.

On the other hand, I do think that we experience apprehension when
operations are found to involve a destructive capacity. In terms of the
digital matrix, it is possible to change the symbolic information
contained in one of its slots with another symbol without retaining any
history of such an exchange. In terms of something we do everyday, like
communicate through E-mail, it is possible to erase or overwrite one's
correspondence without the hope of retrieving it. No doubt many of us
here have tried in vain to rescue from oblivion a file mistakenly erased
with some "unerasing" tool . As it was pointed out in one of the
questions at my talk, the Internet today actually duplicates our
information so that if we lose one version, another version may be
archived and retrieved from another part of the system. This is true.
But it is also the case that this archival system of backups is
organized to discriminate, and I use this word in several senses. What
is archivable is such because it is programmed or determined as such.
Furthermore, any future reading of such an archive will be
technologically determined in a way radically different from the
archeological procedures we use on collections of letters and books,
even when these are written in yet undeciphered scripts. We know we must
rely on computer mechanism for future retrieval rather than rely on our
unaugmented senses. This particular aspect of the digital archive is one
with which we might feel uncomfortable, specially as it suggests a
rethinking of the nature of memory.

While conventional forms of correspondence can be lost or damaged, such
occurrences often leave behind some recognizable traces, for example, a
fragment of a note, an ink blur in the place where water washed over
writing, rendering it illegible, though not unreadable. Our foundations
of historical research lie in identifying and extrapolating from such
traces. Erasure in the realm of digital representation need not retain
this dialectical concept of trace. Properly speaking, there is no
erasure of digital information. There seems to be only the possibility
of copying it or of copying over it. But it is perhaps not even copying
that occurs. Perhaps a better description would be to say that symbols
are "placed" into unique addresses. To "erase" in this domain is to
"place" a different symbol into a slot in a matrix where previously a
different one had been "placed". We might add that symbols are encoded
into (a) place, into an address. Each address frames an entryway into
the symbol located there.

But what does this have to do with how we experience digital
representations? Well, quite a bit and very little. Most of these
processes for which I have been trying to find a proper articulation
occur at sizes and speeds whose scales are quite divorced from our sense
of experience. In the course of experiencing digital representation most
of us will not be concerned with the status of each bit, let alone
remember it. It is here that the "experience" of digital representation
needs to be explored further.

As I noted before, digital computation enables a sort of universal form
of exchange between other media. This function is certainly made
possible by the placement operation and addressable structure I
described. But those operations, while imaginable, are in themselves
unreachable to our senses. More importantly, they bear in themselves
little resemblance to how we are accustomed to crafting new objects with
tools and material. However, we are able to sense their effects. Insofar
as these operations work "invisibly," behind the scenes, it does indeed
seem that magic is being performed when, for example, sound information
is used to make a picture. It is as if the medium through which the
change is enabled has no form in itself, and is pure operation. Again,
as in the case of money, there is an enabling of exchange without the
enabling medium seeming to have material consequence on the value of the
transaction. The digital operation, like a monetary transaction, seems
to provide a function while exacting neither a profit nor a loss in the
process. It is in this respect that the digital medium often seems
immaterial. It seems to be pure function unhampered by form.

Again, while we have all probably felt this diminished materiality with
respect to either monetary or digital processing, we have also, no
doubt, seen the other side of the coin. In the case of money we are
perhaps most often aware of its effect on goods exchange when we note
the value of one monetary unit in relation to another. Dollar versus
Yen, for example, changes daily. Inflation too is a form of distortion
of the symbolic value of money -- it renders its materiality visible. It
is not that we see money for what it is, paper or metal, but that we
realize that it is a particular material to which an arbitrary symbolic
value has been assigned. Rather than seeing its truth, we are reminded
of its un-truth. With digital processing, when viewing an image of a
digitized artifact such as a photograph, we may note aberrant "symbols"
placed into parts of the digital image by the process of encoding the
analog version. These artifacts of encoding and data compression attest
to some kind of work that has been performed on the image. That work
marks the digitized images by literally changing its component numeric

What I am saying is that digital representation is not immaterial just
because it places such artifacts into our digitized images, but because
experiences of digital representation cannot be separated from the
material forms through which they is made sensible. That is, although
with computers we have the potential to "translate" one representation
into another, provided we have these representations stored digitally,
we cannot separate this "translation" process from an apparatus through
which it is made sensible: the apparent immateriality of digital media
is problematized by the tools and interfaces which make them manifest
and usable.
Your Reply