7/30/2019 Human associative memory

1/1

91

The book reflects the culmination of a fouryear collaborative relationship betweenmyself and Cordon Bower. When I arrived atStanford University as a graduate assistant toCordon Bower, there was an ongoingresearch program concerned withorganizational and imaginal factors invarious memory tasks. As we tried to becomeprecise, even quantitative, in fittingorganizational theory to free recall data, its

differences from associationistic models offree recall seemed to evaporate, frankly,because neither theory had been formulatedwith any real precision up to that time. Itbecame clear that what was needed was atheory of memory that was both general andprecise.

The outcome of rather intensiveruminations and discussions was the theoryfor human associative memory (HAM) thatwas proposed in the book. This was firstworked out in detail in a long dissertation

proposal in which I had several goals: Topresent an associative theory of sentencememory, to report evidence relevant to it, torelate the theory to the historical tradition ofassociationism, and to indicate how a fewstandard verbal learning phenomena mightbe interpreted in terms of this approach.

That document formed the basic outline forthe book. I wrote the language parser andquestion-answerer of HAM as LISP code.The proposal and computer simulation ledto a productive set of discussions andexperiments, many of which are scatteredthroughout the book.

In the spring of 1972 the collaborativewriting of the book began. Each day wasfilled with hours of fruitful discussionsfollowed by our individual writing efforts. Inthese discussions we came to adoptcharacteristic roles I as the proposer,interpreter, and defender of HAM, andCordon Bower as the critic, provider of moreproblems, and the demander of greatergenerality. However, like most fruitfulinterchanges, ours were freewheeling, andwe adopted various roles as the occasiondemanded. The discussions and writingturned out to be both personally and

intellectually the most gratifying moments ofour collaboration.Since that time HAM has been subject to

considerable experimentation anddiscussion. It has been found to be wrong onsome points and controversial on others. Iwas bothered by the fact that HAM did notsay much about how memories were used inthought. This led to the development of the

ACT model1 which has a production systemfor mental procedures which interfaced thisprocedural component with a memory

model derived from HAM. Cordon Bower wasbothered by the fact that HAM does notaddress the issues of how large bodies of textmaterial are processed, how largeknowledge structures are organized inmemory, and how these large structures areused to interpret text. This led him to aninterest in story comprehension, scripts andschemata, and reconstructive memoryprocesses.2

Still HAM is very much with us and hasimportant influences on our thinking about

these new problems. In HAM we find manyconcepts and analyses that we can use asfoundations for our new endeavors. Shortly,a revised edition of the book will appear.3

This edition is an attempt to focus on thoseaspects of the HAM theory that still seemrelevant for the 1980s.

The book described an associative theoryof human memory, embodied in a computersimulation that made a wide range ofpredictions about sentence memory andother verbal learning phenomena. Thetheory dealt successfully withorganizational phenomena thought todisconfirm associative theories. [TheScience Citation Index (SCI) and the

Social Sciences Citation IndexTM(SSCITM)indicate that this book has been cited over320 times since 1973.]

John R. Anderson

Department of Psychology

Carnegie-Mellon University

Pittsburgh, PA 15213

August 8, 1979

CC/NUMBER 52

DECEMBER 24-31, 1979This Weeks Citation ClassicAnderson J R & Bower G H. Human associative memory. Washington, DC: Winston &

Sons, 1973. 524 p. [Yale Univ., New Haven, CT and Stanford Univ., Stanford, CA]

1. Anderson J R. Language, memory, and thought. Hillsdale, NJ: Lawrence Erlbaum Associates, 1976.

546 p.

2. Bower G H, Black J B & Turner T J. Scripts in memory for text. Cog. Psychol.11:177-220, 1979.

3. Anderson J R & Bower G H. Human associative memory: revised edition. Hillsdale, NJ: LawrenceErlbaum Associates. In press, 1980.