My first publication appeared in 1956 when I still was a graduate student, while my most recent one came last year (2014), a decade after I retired from my position as a psychology professor and director of an Institute of Cognitive Science. That is a long time, and psychology has changed a great deal during that period, and so has my work. Nevertheless, much of my career has focused on the same problem: to study how people read and comprehend texts, both in the psychological laboratory and in educational settings. Today, the study of text and discourse comprehension, text memory, and learning from text is a flourishing branch of experimental psychology and cognitive science, and I had a small part in making this happen.
I came to text and discourse from the study of memory. Memory research was one of the big achievements of the new cognitive psychology. Indeed, it is not too much to say that the modern conception of memory was formulated in the 1960s. Of course, there has been significant progress since then, but the basic framework for our understanding of memory was worked out at that time. This achievement was made possible because the new research on memory adopted a principle that goes back to the earliest days of experimental psychology. Namely, the laboratory studies of memory avoided dealing with meaningful material, by adopting nonsense syllables in the tradition of Ebbinghaus, or by working with word lists. This was a brilliant move in that, by disregarding the complexities of meaning, much could be learned about basic memory processes. It was, however, no easy task to eliminate meaning. The subjects in our experiments found ways to introduce meaning even into lists of nonsense syllables, and meaning was always a major factor in how subjects remembered word lists. In trying to understand memory, we had to deal with meaning, whether we wanted to or not. Once I realized that, I decided to leave nonsense syllables and word lists to others and focus on meaning directly – that is, on how we understand and remember texts.
The Representation of Meaning in Memory
At some point, memory research comes down to this: you give your subjects a set of items – nonsense syllables, words, even sentences – to study and then count how many they have remembered.