Graham Harman has a nice post up “On the Laziness of Comparing Object-Oriented Philosophy with Leibniz.” One of the points he brings up is that, even though he and Leibniz affirm windowless monads, monads are still determined by their relations in Leibniz , but their reality is non-relational in object-oriented philosophy.
Although I tend to follow Whitehead and Deleuze in thinking of monads as having windows, I also accept Harman’s Heideggerian conception of a non-relational reality, in which monads would have no windows. If we are going to continue describing entities as having windows, it would be important to add that contact through these windows is never direct but only happens indirectly… through a glass darkly. I am committed to affirming relational and non-relational aspects of monads, and I think that the image of windows with dark glass might help draw out the complex tension between relationality and non-relationality. Really, though, the window still implies too much accessibility, even if the window has bars on it or is wired with a bomb. Perhaps it would be better to say that a monad does not have windows at all but is a house of glass. The glass is dark, and what appears dimly in and through the glass is not comprised of reflections but of diffractions (not totally unlike the diffraction glasses that people wear for light shows, raves, fireworks, etc.).
Windowless and alluring, every object is a glass house of dark diffractions.