How does evolution work? It seems that evolution has determined many of the circumstances of the world we find ourselves in. Here are a few examples: (1) Cosmologists tell us that our present astronomical universe --- a conglomeration of planets, stars and clustered galaxies, together with dark stuff, all immersed in a sea of radiation and dark energy ---- evolved over 14 or so billions of years out of a hot, compact and quite uniform plasma of particles and fields ---- the early universe of some 14 billion years ago. (2) Biologists tell us that all the teeming plants and creatures that occupy the surface of this planet evolved over half a billion years or so from less numerous and simpler forms of life --- from “worms and worse”, as I’ve heard it put. (3) Political theorists tell us that modern representative democracy has evolved, from the simpler and harsher social systems of ancient times, to the complexities of the current presidential elections in the U.S. (4) Historians tell us that technology has evolved from fire and wheel to the complexities of modern life, like short circuits and traffic gridlocks. (5) And as you read this, silicon intelligence may be busy evolving a substitute for our ephemeral biological intellects. The details of how evolution (if this is indeed the right term to use) operates in each case are of course different. The purpose of this post is to ask folk who take an interest in the whys and wherefores of such matters (those who have a philosophical bent and some knowledge of physics?) if there could be a common factor that defines “evolution”. For a start I suggest that the existence in each case of some kind of self-promoting process might fit the bill. Since these are physics forums I focus on an example of physical evolution, that of our universe. Cosmologists seem to agree that ever since gravity began to transform density fluctuations in the almost homogeneous fluid of the primeval universe into the structures we know today, the second law of thermodynamics has ruled on a universe-wide scale. Consider the action of the second law as described by Penrose in his book The Road to Reality (Chapter 27) . Penrose presents his discussion of the second law in the context of what I will call the “landscape” of phase space, to borrow a term from fringe cosmology. Phase space is a well-known concept used in the discipline of statistical mechanics. It is an abstract multi-dimensional coordinate space in which the state of the system being considered is specified by a single point. As the system develops dynamically this representative point traces a trajectory through phase space. In the case of the universe it passes through a “landscape” that Penrose depicts as “boxes” of phase space with boundaries that are sensibly but rather subjectively chosen. Large boxes contain many states judged to be too similar to be distinguished. Small boxes contain only a few such states. It is implicitly assumed in Penrose’s persuasive graphic description of the action of the second law that the landscape of phase space remains distinct and unchanging as the universe ergodically explores its phase space; phase space is assumed to be an unchanging background, as it were, for the developing trajectory of the point that represents the state of the universe. It is in this context that one can assert with confidence that the chances of structures like stars and galaxies spontaneously assembling themselves without gravity are entirely negligible --- as are the chances of all the molecules of a gas in a macroscopic container spontaneously collecting together into one of its corners (or of all the atoms in a molecule of DNA spontaneously assembling into a double helix). How then did gravity then force the evolution of the complex and heterogeneous universe that astronomers now observe? Consider a modification of Penrose’s description of how the second law operates. Suppose that the boundaries of “boxes” that define the landscape of phase space, as envisaged by Penrose, are changed by the action of gravity as the universe’s representative point traces its trajectory through phase space. For example, suppose that the trajectory of the early, near- uniform universe passes through a box of states in which the universe has a single overdense patch. The attraction of gravity draws matter towards this patch, and in doing so moves or creates new box boundaries. I am here suggesting that gravity is an agent that can dynamically reshape the landscape of phase space while the universe evolves, and that this is possible because gravity stimulates self-perpetuating processes: i.e that gravity is an agent of auto-catalytic change. From a broader point of view, this is evident in many situations. For instance, in geomorphology gravity stimulates self perpetuating processes like the hydrodynamic dissection of a smooth slope by the formation of a drainage network of rivers and streams. In this process the phase space of fluid flow over the smooth slope is, I suggest, dynamically reshaped by the action of fluid flow itself, and therefore ultimately by gravity. And seems to me that it is indeed appropriate to talk of the evolution of the universe, because the principal agent of change that is involved is gravity, which stimulates self-perpetuating processes. In this respect physical evolution is like biological evolution, whose essence is the reproductive capability of DNA. The difference is that gravity is a phenomenon of the physical world determined by its geometry, whereas DNA is a physical structure made possible by the laws of nature. Is evolution in general characterised by processes that are self-perpetuating?