applied to fantasized violence against women (for a glaring example of this, see Auerbach 1982). Rather than congratulate ourselves, we must bear in mind the disparity between the reality of women's historical power and the size of the shackles historically placed upon it.
Like the pornographic model, the cross-sex fantasy model offers no exit from gender hierarchy. The female is still the site of violence, no matter what the location of the subject. Even if the magician and the lady change places, he is still taking her place.
Proprium humani ingenii est odisse quem laeseris.
[It is proper to human nature to hate one whom you have hurt.]
-- Tacitus Agricola 42.3
We need a political model that will both describe the magician's act and suggest a way to end it. Let me postulate that the problem is not gender but hierarchy: within hierarchy, violence is a right, and the control of violence diminishes liberty. An anarchic system is thus a precondition for the deletion of the pornographic. Though escape from hierarchy has seemed impossible, I would postulate that there are some "open" discourses that permit it: theory, mathematics, nonrepresentational art, music. Other, "closed" systems--humor, fantasy, narratives, film, and representational art--all interrelated, form the bars around hierarchy.
The structure of these closed discourses is political, and they have four main characteristics: (1) They contain a cue that says any item is untrue, creating what I call the "Archie Bunker fallacy" ("It's just a joke!"). Ovid actually asserted this in his poems from exile (e.g., Tristia 2.491-96). (2) Content follows function and is not arbitrary. (3) The relation between each item and reality depends on the status of the users; these discourses maintain the status quo. (4) Historically, though perhaps not necessarily, the hierarchy has been gendered. The position at the bottom, so often a woman's, has never been pleasant; something in it "exposes the meatiness of human flesh" ( Carter 1978: 140; see Kappeler 1986: 63-81; Rabinowitz, Parker, Brown in this volume).
Where does this leave us? On the one hand, history weighs heavy, and closed discourse is more comfortable than open. Revolutionary discourse is intrinsically unamusing. How ephemeral, how dry this essay is compared with Ovid's poetry! And insofar as it amuses, it fails. On the other hand, when we see problems of discourse as systemic, we can gauge our task. The female can no longer be by definition the site of violence--nothing can. What happens if we say, as Kappeler does (221), "Art will have to go"? Maybe there is something else. Meanwhile we must use what exists to show what is wrong.
How can women read? And why should we read Ovid? How badly do we need this history? I borrow an answer from Toni Morrison. We're stuck with Philomela; she's