Two retrograde indicators working in parallel, independent of one another: this is the essence of the biretrograde. Each hand travels its own arc, then snaps back to zero with a sudden, instantaneous movement. The name describes not an aesthetic, but a precise mechanical logic.
The retrograde principle itself is not new: a single hand that oscillates across a scale and returns at the end of travel has existed for centuries. The biretrograde complication, however, emerged from the need to manage two distinct pieces of information using this same principle, without the two mechanisms interfering with each other. Roger Dubuis and Jean-Marc Wiederrecht co-patented the solution in 1989, working on a system of cams, levers, clicks and springs capable of synchronizing two retrograde indicators within a perpetual calendar. The module appeared concretely in watches starting in the 1990s, with the day of the week on an arc to the left of the dial and the date on an arc to the right, both crescent-shaped.
From my workbench I have seen many examples of that generation of Roger Dubuis watches: the mechanics are dense, the module sits on top of the base caliber and the thickness is noticeable. It is not a complication that lends itself to easy miniaturization. The moment when the hand returns, technically called the fly-back of the retrograde, depends on the tension of the return spring and the geometry of the cam: if the spring is too weak, the return slows down and loses crisp definition; if it is too strong, wear concentrates on the release click.
The most common application is calendrical—day and date on semicircular arcs—but the principle extends further. In biretrograde chronographs the two indicators manage seconds and date, or seconds and minutes, on arcs that can range from 120° to 210° depending on the design. Gerald Genta pushed further, combining jumping hours, retrograde minutes and retrograde date on a single dial: three distinct movements that coexist, each with its own kinematics. In that case one speaks technically of three retrograde indicators, not two, but the conceptual framework remains the same.
One thing I have never seen documented with certainty is whether the 1989 Dubuis-Wiederrecht patent covers only the calendrical layout or also subsequent chronographic configurations. On the formal attribution of those variants, I prefer not to commit myself.