The Four Horsemen

Researcher John Gottman of the University of Washington has made a career of studying marital interactions. He claims to be able to predict with 95% accuracy which couples will eventually divorce.

After an earlier study that gathered data from thousands of couples over 13 years, Gottman's research team studied 130 newlyweds intensively over a six-year period. The couples were invited to Gottman's laboratory, where they were hooked to instruments that measured variables such as heart rate, sweating, and muscle tension. The couples were recorded with a video camera while they had a conversation about a disagreement. In a frame-by-frame analysis, their facial expressions, movements, and comments were analyzed.

What sort of research did Gottman do? What were Gottman's "Four Horsemen"?

After correlating the data with marriage outcomes, Gottman found four factors that predicted divorce. He called these the Four Horsemen of the Apocalypse for marriage. They were:

1) criticism (telling the other person his or her faults)

2) defensiveness (reacting to certain subjects by denying responsibility, or refusing to discuss an issue the spouse regards as important)

3) contempt (making sarcastic or cutting remarks about the other person)

4) withdrawal (also called "stonewalling": showing no reaction, having a blank look, or ceasing to care)

Perhaps the most "corrosive" of the four, according to Gottman, is contempt, which he said should be "banned from marriages."

What did later analysis show?

Later analysis of his data provided a surprise for Gottman. While the Four Horsemen remained important, he found one factor that was the best predictor of all. This was a positive predictor, one that predicted long-term success rather than failure in marriage. Gottman found that marriages are likely to thrive when the man was willing to be influenced by his wife ("Want a successful marriage? Listen to your wife" CNN, February 20, 1998).

Why would the factor of being willing to listen to your wife be so highly predictive? Here it is tempting to fall into the trap of imposing a cause-effect interpretation on a correlation that might be due entirely to self-selection. After all, this is not random-assignment research. Gottman did not tell one randomly chosen group of men to respect their wives' opinions, while another randomly chosen group served as a control. The group is self-selected, so any variable that correlates with the variable of "willingness to be influenced by wife" could be the real cause of a more durable marriage. These could include higher income, more education, better social intelligence, more inclination to cooperate rather than compete, a more satisfying approach to name it. We have here a perfect opportunity to apply the critical thinking habits encouraged in Chapter 1. Any of these variables, or any combination of them, could be the real explanation for longer marriages.

Whatever the underlying reasons or combination of reasons for the correlation, the predictive information is still useful and interesting. If Gottman is correct, measuring a male's willingness to take advice or be influenced by his wife provides the most accurate way to make a forecast about the long-term success of a marriage.

Write to Dr. Dewey at

Don't see what you need? Psych Web has over 1,000 pages, so it may be elsewhere on the site. Do a site-specific Google search using the box below.

Custom Search

Copyright © 2007 Russ Dewey