Interact is a computer program that displays verbal descriptions of what people might do in a given situation, of how they might respond emotionally to events, and of how they might attribute qualities or new identities to themselves and other interactants in order to account for unexpected happenings.

Interact achieves its results by employing multivariate non-linear equations that describe how events create impressions, by implementing a cybernetic model that represents people as maintaining cultural meanings through their actions and interpretations, and by incorporating large dictionaries that index cultural meanings. The program's predictions can be, and have been, tested in experimental and field studies, and results largely support the validity of Interact simulations.

Interact first and foremost is a research tool for examining the implications of Affect Control Theory. While the theory is simple - people try to have experiences that confirm their basic sentiments - detailed application of the theory is complicated by computations and data processing, and Interact is required to keep analyses from getting bogged down.

Read a primer on how to work with Interact.

Reminiscences of David Heise

The first version of Interact was built in 1972. It was just an experiment to see if the basic ideas of Affect Control Theory could predict actual behaviors in social interaction. I didn't have much hope frankly. At that time the best theories worked just with numbers. Nobody tried to predict actual behaviors that might occur in social situations.

Electronic hand calculators had come on the market a couple of years before. I bought one and worked through equations for a social situation involving two people. It took days, and I wasn't sure my calculations were correct. I translated numerical results into words by scanning lists visually, and I wasn't certain if I was selecting the right words to fit the numbers. Results seemed promising, but were they right? I'd never know unless I wrote a computer program to do the calculations and to search the dictionaries automatically.

So I wrote the program. After a few months' work, the program seemed to be harboring its last bugs. I set up an analysis: two enemies, Mac and Bob; Bob insults Mac; what will Mac do? In those days you set up an analysis, submitted it to a university computer, and came back hours later to find out what happened. I went out to dinner with my wife Elsa and stopped at the computer center on the way home to pick up the printed output, fully expecting another bug.

Hey, no bug this time! It ran! Quick, find the results page. There. Mac hits Bob!

That first Interact outcome seemed the best behavior given the limited choices in that version of the program. I was encouraged. In the next couple of years I assembled dictionaries of  identities and behaviors, got better equations to use in calculations, rewrote the program so it was easier to run. I prepared a report, Understanding Events, which was published jointly by Cambridge University Press and the American Sociological Association.

In the late 1970s the National Institute of Mental Health provided financial help. Lynn Smith-Lovin, Chris Averett, Beverly Wiggins, Bernadette Smith, and others joined in working on the theory. Bigger dictionaries and better equations were assembled. Experiments tested the theory. The theory expanded to cover behavior, labeling, attribution, emotion, and settings. Neil MacKinnon began work that has culminated in a set of  Canadian dictionaries and the first cross-national replication of the equations.

Then microcomputers appeared, and I translated the program from PL/1 to a combination of Basic and 6502 assembler language. However, back then microcomputer technology changed almost yearly, and by the mid-eighties, I'd translated the program back to PL/1 to run on a CP/M personal computer. Then I translated it to Pascal to run on MS-DOS systems. The MS-DOS program was marketed by Wm. C. Brown Publishers, Software of Dubuque, Iowa in 1988 (with distribution being discontinued in 1993). In the late 1980s I also created Attitude, a program that allowed respondents to rate concepts on a graphic rating scale displayed on the computer screen. Attitude provided a more efficient and more precise means of gathering data on affective meanings than paper-and-pencil questionnaires.

In order to develop a graphical user interface for Interact, I transferred my work to a Macintosh computer and developed a version of Interact using HyperCard and and HyperTalk during the mid-1990s. The Macintosh program was the first to display facial expressions besides listing words describing emotions.

Development of the World Wide Web in the 1990s raised new opportunities. I realized that publishing the program on the Web would make it easier to distribute and would reduce the cost to students and researchers to zero. People anywhere in the world could have instant access to the software. So I re-wrote Interact in Java and published a Java version on the Web in 1997.

In the 1980s and 1990s Neil MacKinnon, Herman Smith, and Andreas Schneider were building international data sets that nucleated into Project Magellan - Affect Control Theory's cross-cultural data-gathering enterprise. In 2001 I provided infrastructure for Magellan by creating Surveyor, a data-gathering program that uses graphic rating scales in indigenous languages at computers around the world and that archives the data via the internet. I also internationalized Interact in 2001, allowing it to be translated readily into any indigenous language. 

Milestones in the development of Interact are indicated in the following list.

Time Chart

Technical Description of the Java Program

JavaDoc