The case is considered in which the shape of a moving object is observed from a fixed point of reference along the line of motion. Three distinct recognition algorithms are introduced: for recognizing patterns per se, for recognizing patterns in symbolic context and for recognizing patterns in the context of their motion. The second of these algorithms requires two observations of the moving object; that is, it must be observed from two distinct points along the line of motion. There are implications that the introduction of context into pattern recognition may depend upon a degradation of the recognition mechanism. If a moving object is to be recognized independent of the speed with which it moves past a point of reference, then the shape of the object must be generalized to a number of segments over which the derivatives of the variables defining its shape are continuously contained in one of the four classes: {the negative reals}, {zero}, {the positive reals}, and {infinity}.