This paper reviews Gould's clock model for heterochronic processes and uses that model to develop simple matrix representations of growth and shape change. Matrix representations of growth and development provide a common formulation for all heterochronic processes. In particular, we show how neoteny can be diagnosed using such a matrix approach. The literature is rife with contradictory representations of how neoteny affects growth allometries and the timing of developmental events, and therefore of the role of neoteny in human evolution. Through the use of multivariate models, we explore these relationships and the internal consistency of opposing views. Gould's neoteny hypothesis for human evolution has been criticized for a number of reasons. Humans do not grow slowly. The slopes of our growth allometries show no common pattern of change vis-à-vis those of our closest relatives. Humans prolong rather than reduce rates of growth and development of body parts; the brain, for example, ceases growing later in humans than in apes, but during this prolonged period of early ontogeny, it grows at a rapid pace. This paper evaluates Gould's hypothesis and its critiques by focusing on particular questions. Does neoteny imply slow growth? Does it imply a unidirectional change in the rates of growth of traits? Under neoteny, should the brain cease growing in ancestor and descendant at the same age? Does prolongation of phases of growth and development confute neoteny? On the other hand, is paedomorphosis an inevitable consequence of prolonged growth and development? We show that, for all of these questions, the answer is no.