This paper describes an architecture for symbolic processing from the viewpoint of an experienced computer engineer rather than of an academic. Moreover, it is a description of a LISP machine from Symbolics, Inc. The paper is worth reading if one is interested in (1) a design example for symbolic processing, (2) the general flavor and difficulty of designing architecture for symbolic processing, and (3) the design philosophy and history of products from Symbolics, Inc. The paper is strong at presenting an “architecture” as a three-layer specification: the system, the instruction set processor, and the processor architecture; the change of each layer will not impact the capability of the layers above. This is the fundamental philosophy advocated by the paper, and it has been very well organized in this straw man description of machines from Symbolics, Inc. The paper unfortunately does not give a characterization of symbolic processing nor does it mention other related architectures for symbolic processing. In this regard, the reference list is incomplete.