Abandon naive realism all ye who enter here. Gregory, Director of the Harvard-Smithsonian Center for Astrophysics, describes the evolution of modern physics in terms of the language scientists have used in their theories, emphasizing that that language is essentially mathematical. In physics, there is no one-to-one correspondence between theoretical constructs and ""concrete reality."" Indeed, theories rise and fall according to their predictive value. As the author notes, ""The eight-fold way [the hypothesis invoking quarks as fundamental particles] was accepted, not because it provided a rationale for what was already known, but because it predicted something not known, the Omega-minus particle--and this prediction was supported by experiment."" Using such examples, Gregory traces how physicists have coined words and used the power of mathematics to build models of the universe that gain credence on the basis of experimental observations. The experiment may be the tail that wags the dog or, as in the case of Einstein's theories of relativity, merely bear out what his equations predicted. Current ""grand unified theories"" postulating ten-dimensional universes, superstrings, or supersymme-try suffer from observational impasses. Even the superconducting supercollider is too feeble to demonstrate the massive particles these theories predict. Thus, these new conversational modes of the language of physics may fail the test of utility and reduce physics to metaphysics. These heady ideas are neatly developed in chapters that trace the historical development of physics with Milan Kundera-like titles: ""The Imponderable Nature of Matter,"" ""The Ineffable Color of Quarks,"" ""The Unspeakable Power of Language."" Gregory also generously quotes Feynman, Bohr, and others with their caveats not to try to picture a quantum mechanical world. Instead, the reader can follow Gregory through the vocabulary of fields and forces, symmetries and symmetry-breakings that express how physicists now talk about the world.