[…] Every piece in the edition is supposed to look exactly the same. Any copies that are noticeably different from the others are excluded. The amount of variation tolerated in an edition varies from artist to artist, from piece to piece, from context to context. There is no “perfect” edition—every edition contains the very thing (variation, difference) that cancels its mode of being, that activates its mode of becoming. […]
I have, for some time now, been exploring the concept of the “variable” or “variant” edition. This is another one of those paradoxical ideas—an edition that foregrounds difference. I’m not sure where the idea of the variable edition comes from originally—I learned about it in one of my first printmaking classes, so it’s a thing that’s been “out there” for awhile. It’s not a new idea, but it is a powerful one, because it is an idea that by its very nature questions its own core concepts (the edition, repeatability, the multiple) and the discourse in which it exists (the limited edition, the restricted economy). The variable edition is the edition “under erasure.”
[Brief aside: Does the variable edition produce an object-in-multiple that questions the limited edition/restricted economy? Or does it produce an object-in-multiple that is perfect for the restricted economy because every copy is unique? Is the variable edition the ultimate limited edition? Is context/distribution relevant here?]
There are many different ways to physically produce a variable edition, but we can identify two main approaches, and those approaches are determined by the construction and use of the matrix or matrices that produce the edition. It can be said that the matrix is actually these two things: the object containing the information that re-produces the object-in-multiple, and the manner in which that information is actually used to carry out the production. Every matrix is information plus action.
[“Re-produces?” I love how slippery, muddy, and opaque the language gets out here. Like the grease and dirt caught under your fingernails after you’ve been working on a machine.]
The closed matrix: The closed matrix is a matrix that contains information corresponding to every single mark or feature that will appear on/in the finished object. Everything is there already. And, importantly, the matrix is used in such a way that all of that information is automatically present in the finished object. The etched copper plate, the litho stone, the polymer plate, lead type, the photographic negative, the mold—all of these “normal” matrices are closed matrices, when used in their “normal” way—to physically and directly make an object. But it is also important to note that a digital file, interpreted through “normal” reproductive software, is also a closed matrix, because it will produce (barring the unavoidable physical peculiarities of the “interpreting” interface, generally, the screen and/or printer) almost exactly the same object over and over again. All of the information is embedded in that file. Digital printmaking, technically speaking, is no more or less radical than “traditional” printmaking.
The open matrix: The open matrix is a matrix that contains just enough information to begin the action of making, and the relationship between the matrix and what is produced is often abstracted or removed. (A line on an etched copper plate physically and directly produces a line in that same shape. Instructions on how to draw a line will still produce a line, but the information is abstracted and only represents or describes the mode of production.) The action component of the matrix has chaotic or variant elements built right in. For example, an open matrix could be a set of instructions (the abstracted information), and those instructions could guide the re-production of a drawing (variation built in to the process). Or the information part of the matrix itself could be literal and closed (like a linoleum block) and the action component of the matrix could be open and chaotic (pouring or spraying solvent on the linoleum block after it has been inked).
Both kinds of matrices, open and closed, can be used to produce a variable edition, and the “two main approaches” to making a variable edition mentioned above stem from these types of matrices.
The first approach is the use of a closed matrix in a variety of ways within a single edition; for example, the same copper plate can be used to print a single edition in several different colors. This approach is, more or less, grouping different prints into the same edition because of the use of a common matrix.
The second approach is, simply, to utilize an open matrix, which, because of its active, chaotic elements, always produces a variable edition.
Every edition ever is variable to a certain degree. Any set of information translated or transposed to a new physical form inevitably incorporates a certain amount of variation. Such is the nature, the chaos, of the world. What matters though, is the threshold of perceptibility of that variation. If there are two prints on the table, and no one can see the differences between them, are they different at all?
These images are from the Wikipedia article on the Difference Engine, an early mechanical computer.