The Principles of Life Tibor Ganti. (2003, Oxford University Press.) £70.00, $168.00, 220 pages. To review this book is in some ways analogous to reviewing Gregor Mendel’s Experiments in Plant Hybridization (1865) shortly after it had been rediscovered by Hugo de Vries in 1889. Both works contained dormant insights of fundamental importance. In 1971 Tibor Ganti, a chemical engineer living in Communist Hungary, wrote a book called Az elet principiuma (The Principles of Life). It passed largely unnoticed in the West. Now in translation with essays from his other books, The Principles of Life is an original, logical, and parsimonious framework for thinking about life. Moreover, it has inspired a research program to synthesize minimal life from scratch. His theories foreshadow and anticipate the cores of systems biology, evo-devo, and developmental systems theory by several decades. The book is the work of a genius, according to the late John Maynard Smith. As Ross Ashby applied cybernetics to the study of the brain, Ganti has applied cybernetic ideas with extreme clarity to the study of the fundamental organization of life. The book is beautifully written for the layman. It is philosophically rigorous and introduces chemical models of minimal cells that are easily understood by non-chemists. The OUP edition is critically commentated by Eors Szathmary, a theoretical biologist who studied under Ganti, and James Greisemer, an eminent philosopher of biology. I present an outline of Ganti’s theory below. Ganti characterizes a unit of life as a unit that cannot be decomposed without losing any of the following properties: A unit of life performs metabolism, it is inherently stable, it contains an informational subsystem, and it is regulated and controlled. The simplest extant units of life are cells. Viruses are excluded because they have no metabolism. A biosphere of viruses lacking cells for their replication would face disaster. Neither are clouds or candle flames units of life, for they lack a subset of components that self-referentially constrain the properties of other parts of themselves, that is, an informational subsystem. Everyone would agree that a mule is alive; therefore, reproduction and hereditary are unnecessary features of units of life, but are clearly necessary for the continuing existence of a biosphere composed of units of life. Units of life can be hierarchically organized. For example, a human is composed of cells that are themselves units of life. But the human itself is also an indivisible bounded unit of life, possessing a higher-order metabolic organization with specialized organs such as a gut and a brain that is a specialized informational control system. Nation states consisting of human units of life could also be considered to share the characteristics of life, since they arguably have a metabolism consisting of primary production, import and export, an informational control system consisting of government at its various levels, and a boundary that makes the country (to some extent) an indivisible unit. They are controlled and regulated dissipative structures. Contrast this with John Maynard Smith’s definition of units of evolution as units that multiply and have hereditary variation [8]. He goes on to define natural selection as an algorithmic process that occurs where there are units of evolution that have differential fitness. One can immediately see that units of life and units of evolution are only partially overlapping sets. Not all units of evolution are alive; for example, entities in a genetic algorithm, viruses, and RNA molecules replicating using a Q-beta replicase are all units of evolution but not units of life. And not all units of life are units of evolution; for example, mules are not. All existing units of evolution depend on units of life for their maintenance. Now we have the conceptual machinery to ask ourselves whether this has always been the case. But let us continue with Ganti’s argument for now.