ABSTRACT
In this chapter we will develop our theory regarding the encoding of sentences into memory, provide an explicit mathematical (probabilistic) model of the process, and report several experimental tests of this explicit model. Memory for sentences has been one of the favored ways to study their mental representation. The general working premise has been that contingencies in recall of sentential elements reflect proximities of the elements in the underlying mnemonic representation of the sentence. We shall not review here the research of others using this strategy, but will refer the interested reader to reviews by Fillenbaum (1971) or Wanner (1968).
