Asteroids (and comets) are remnants of the original building blocks which formed the planets. As such, they are recorders of the conditions which existed in the solar nebula approximately 4.6 billion ago. Since that time, however, these minor planets have been pummeled by a continuous rain of impacting asteroids and comets, which produces craters and, occasionally, catastrophic disruption events. These collisions are responsible for the creation of most of the sub-100 km bodies now observed in the main asteroid belt. Therefore, to understand how asteroids have evolved over time, we need to understand the physics of hyper-velocity collisions and how these events have modified the shapes and internal structures of the surviving bodies.
Our knowledge of asteroids has grown substantially over the last decade. For example, meteorite samples and spectroscopic techniques tell us that asteroids are composed of non-volatile rocky and metallic materials. Detailed information on the composition and orbital distribution of asteroids is given by the paper by J. Remo (this issue). Unfortunately, no observational measurement can probe an asteroid's interior, making it impossible to say whether small asteroids are (i) solid fragments from homogenous parent bodies, (ii) a solid fragments from differentiated parent bodies, or (iii) piles of rubble produced by collisions. Despite this problem, models (i) and (ii) have been favored by planetary scientists for many years, probably because they match our preconceived notions about what ``rocks in space'' should be like and they explain many asteroidal surface features observed by ground-based spectroscopy. As I will present in the next section, however, model (iii), the ``rubble pile'' model, appears to provide a better match than (i) or (ii) to observational constraints and new results from numerical impact simulations.