Maths model tells if couple can go on ad infinitum

As the American Association for the Advancement of Science's annual meeting begins, we talk to the society's outgoing president and survey some hot topics

Valentine's Day may be near, but with the news that almost half of British marriages end in divorce, perhaps Cupid needs a scientific hand to guide his arrows, writes Stephen Phillips.

If so, John Gottman, a professor of psychology at the University of Washington, and his mathematician colleague James Murray may be just the people to help him. They have devised a model that, they claim, offers a 94 per cent success rate in predicting whether couples will stay together or split up after tying the knot.

The formula is built on dynamic non-linear models Murray used when he was professor of mathematical biology at Oxford University and had been used to model leopard's spots, and on empirical data on relationships that Gottman had amassed over a decade of tracking 700 Seattle couples.

They were invited to visit his "Love Lab" to discuss divisive topics such as sex and money. Researchers behind a one-way mirror studied every kind or cutting remark, leg-cross and frown exchanged.

"We were surprised how we could predict (marital fortunes) with high accuracy," Gottman says, adding that this prompted the need to build a scientific model to predict human behaviour.

Having studied maths to PhD level before switching to psychology, Gottman went for a mathematical formula.

Gottman and Murray scoured tapes of the couples' conversations, assigning positive or negative values to verbal and non-verbal cues, and created a "Dow Jones Industrial Average of couples' interaction".

The model was refined to account for a spouse's individual psyche and couples' "emotional inertia" or mutual responsiveness. Other variables included a "repair term", referring to people's willingness to patch things up if relations got strained.

The model, to be shown at the AAAS conference, has yielded interesting insights. "Positive emotions have great impact in courtship, but once couples get married, negative emotions are more influential," Gottman notes.

And countering prejudice about gay and lesbian couples, he says positive emotions are more enduring in these relationships because there are "fewer barriers to leaving" as well as a need "to continue renewing courtship".

Another finding is that after overcoming initial relationship problems, couples may have an easier ride in the long term because long-standing heterosexual unions tend to "mellow out on the negative emotions and value kindness more".

The findings also debunk the belief that biting one's tongue is the best policy for harmony. "Relationships where people object and try to repair are better off," Gottman says.

The researchers' book on their work met with predictable scepticism. "Most (psychologists) are maths-phobics, so we have to be very creative in how we communicate (it)," Gottman says.

Gottman and Murray will speak at "The science of love and marriage" session on February 14 at 11am.