Unlike many spreading-activation theories, in which activation is assumed to be a binary variable, the present theory assumes that activation level is a real variable, with the activation level of node "j" at time "t" given by A("j, t"). Altough the theory is not incompatible with negative activation levels, only positive values occur in its present conception. The defining components of spreading activation are "spreading", "summation" and "decay".
- Dell (2004), a pag.132