Abstract
Modal Worlds is at once modest and ambitious. Our universe harbors a zoo of structures from the tiny, such as quarks and leptons, to the cosmological, such as the cosmic microwave background radiation. The sheer number and diversity of structures that pop up in scientific research fuels a mitotic explosion of scientific fields, from neuroparasitology and nutrigenomics to exometeorology and cliodynamics, each focused on some newly encountered species in the zoo. Against this trend of narrowing focus, the ambition of Modal Worlds is to embrace and understand the entire zoo. Its point of departure is the insight that the world has “regularities” that sometimes become correlated to create “modes”: Modal Worlds is a new approach to Reality. The regularities are simple and easy to understand. The key is that the observation of one property allows the inference of other properties of the object under view. These are the “modal models” we seek. The aim is to show that the world seen this way is principled. The result is not one integrated edifice, but rather a web of evolving correlations. (p. 8) Modes are defined by their regularities. But the regularities we are invoking, namely patterns, categorizations, ribbons of an unspecified nature, etc., do not have obvious scientific metrics. Thus it is not clear how one can build quantitative models for modes. (p. 26) Modal descriptions of the world about us lie in a domain that is semi-independent of Science. Hence we have two domains of study: Modes and Scientific inquiry. (p. 79)
This survey has its intended effect. One comes away impressed with the wide variety of regularities and modes, questioning whether any one theory could possibly capture it all. This question may take several forms, and it is of interest to ask which is advocated by Modal Worlds.
One might, for instance, question the generality of physics, namely, the claim that all events that fall under the laws of any science are also physical events that obey the laws of physics. Modal Worlds does not explicitly deny the generality of physics. Or one might question the unity of science, namely, the claim that laws and natural kinds of the special sciences must, in principle, be reducible to laws and natural kinds of physics. Modal Worlds does seem to deny the unity of science. Or one might question the generality of science, namely, the claim that all human knowledge is within the purview of science. Modal Worlds flatly denies the generality of science. These denials are highly consequential. Forcing the reader to come to grips with these denials, their evidence, and their consequences is the goal of Modal Worlds. Vision researchers of all stripes—from Bayes to embodied to deep learning—will be challenged to think deeply about the provenance and interpretation of the manifold structures and regularities of the visual world.
Modal Worlds has a thought-provoking introduction by Jan Koenderink—a first-rank vision scientist whose influential research includes the perception of color, shape, motion, and depth. Koenderink clarifies the issues at play in Modal Worlds and helpfully puts them in the context of related work in philosophy, physics, and ethology, with special (and appropriate) emphasis on the work of von Uexküll. He hopes, contrary to the claim of Modal Worlds, that a formal theory of modes is possible, and sketches the outlines of such a theory. This is exactly the kind of reaction that Modal Worlds is designed to evoke.
Near the end of his introduction, Koenderink writes: “Whitman’s book is inspiring. It launched my thoughts on various trains of thought. It is much like the kind of matter Whit would feed his students in an attempt to set them off on new tracks” (p. O). I couldn’t agree more. Whit was my graduate advisor at MIT from 1979 to 1983 and a dear friend thereafter. He did inspire us. He set our thoughts off on new tracks. He did it without trying. It was just who he was: endlessly inquisitive about the vistas ahead. And it was contagious. So is Modal Worlds.