It is not enough to be intelligent. It is not enough, even, to be emotionally intelligent. The rules of the game have changed: we also have to be socially intelligent. Among the cognoscenti is the nurse who can instantly comfort with a touch on the arm, the soldier who can be captured by insurgents and released with an apology, the diplomat who can defuse tensions with a well-placed word. And what happens to the socially dumb? They populate the ranks of unhappy spouses, inadequate parents, unfulfilled employees and loners.

So concludes Dr Daniel Goleman, the bestselling author who first coaxed the idea of emotional intelligence out of the academic wilderness and into public consciousness.

EI didn’t so much strike a chord as play a rousing symphony: it sold the comforting thought that there is something beyond general intelligence. It explained why supposedly clever people can rub others up the wrong way and why the pea-brained can become so popular. And it sold in its millions, allowing Goleman, who has a PhD in psychology from Harvard University, to give up his day job as a science reporter at The New York Times and to remodel the pretty, double-garaged New England home that he shares with his second wife, Tara Bennett- Goleman, two ageing horses and a copy of Emotional Intelligence in almost every publishable language.

But that was 11 years ago, and the onward march of neuroscience has given Goleman a new idea — and book — to sell. The central concept of Social Intelligence: The New Science of Relationships, which is published next week — and an extract of which appears in The Times tomorrow — is that we are “wired to connect”. Our brains are social tools, primed through evolution for promoting and guiding social interactions and relationships. Our sociable brains allow us to “infect” those around us with our emotions, and to “catch” the moods of others.

“Until now, all psychology, particularly in neuroscience, has been about one mind and one brain and one person,” Goleman explains, as he tries to pin down the “aha moment” that led to the writing of Social Intelligence. “Psychologists had begun to expand this to assess two people at the same time; the unit of measurement was becoming the dyad rather than the individual.

He expands on this further in the book: “My own model of emotional intelligence folded in social intelligence without making much of that fact . . . but, as I’ve come to see, simply lumping social intelligence within the emotional sort stunts fresh thinking about the human aptitude for relationship. The danger comes in fixating on what goes on inside us and ignoring what transpires as we interact. This myopia leaves the ‘social’ part out of intelligence.”

Goleman, 60, classifies human interactions, loosely, as nourishing or toxic; the socially intelligent will be party to very many more nourishing encounters than toxic ones (intriguing research suggests that successful marriages have a ratio, or “golden mean”, of at least five happy interactions to every rotten one; and no, he hasn’t calculated whether the golden mean is true of his own marriage). Both have a measurable effect on health: a pleasant life is a longer-lived one. Widowers die earlier than happily married men; those at the top of the management ladder suffer lower rates of heart disease than the minions on the drudgery-filled tiers beneath them.

So as well as telling us that SI plugs a gaping hole in the model of emotional intelligence, Goleman uses his new book to present a manifesto for a more compassionate, more socially interconnected, world. He implores us to reject the “inexorable technocreep” that results in so many of us conducting relationships by e-mail, which is faceless and voiceless and thus deprives the brain of vital social cues. And mirror neurons — the special brain cells that allow us to empathise and “catch” each other’s emotions (and which appear to be lacking or dysfunctional in those suffering from autism and psychopathy) — mean that human beings can wield great emotional power over others.

Goleman finds this troublesome: “Mirror neurons make us far more neurally connected than we ever knew; this creates a pathway for emotional contagion. If you really care about people, it gives a new spin to the term social responsibility: what emotional states are you creating in the people you’re with?”

Goleman, who has the composed aura of an Ivy League don, is not the first to expound the idea of social intelligence (EI was not his original idea either, but a revival of a 1990 paper by the psychologists John Mayer and Peter Salovey). Edward Thorndike, a Columbia University psychologist, beat him to it by more than 80 years, noting in a 1920 article for Harper’s Monthly Magazine that “the best mechanic in a factory may fail as a foreman for lack of social intelligence”. Thorndike, Golemnan notes, “was articulating something we all know”, ie, there is something about social adeptness that we know goes beyond professional competence, niceness or intelligence.

Thorndike’s observations were dismissed in psychology’s rush to embrace the fledgeling concept of IQ; social intelligence was judged to be general intelligence applied to social situations and, therefore, not a distinct intelligence in its own right. But the very modern science of social neuroscience — which involves peering into the brains of people during social encounters — now shows that our social behaviour is shaped as much by instinct as by rational thought. This contribution of the neural “low road” to our social behaviour — distinct from the cognitive “high road” associated with conscious mental effort — convinces Goleman that “the time is ripe for a revival of social intelligence on a par with its sister, the emotional type”.

Thorndike was also prescient in his observation that, while easy to spot, SI could not easily be measured. Goleman doesn’t think an SI quotient, akin to IQ or EQ, is around the corner, although some tests claim to measure it. “I don’t think there’s a good measure yet because the tests largely reflect what we know about social situations rather than how well we operate in social situations,” Goleman says.

“My argument is that it’s the latter that the tests should assess. And if you’re going to do that, you can’t just ask people questions about it, because that reflects their cognition. You also need to assess how well their automatic and unconscious circuitry operates, for example, the circuitry that tells me that your eyes are expressing interest and curiosity right now.”