Researchers at Harvard and the University of California in San Diego examined data from more than 12,000 participants in the Framingham Heart Study to reach scientific conclusions about this so-called cluster effect and published their findings in the New England Journal of Medicine.

They discovered that the stronger a social bond, the more likely it would influence an individual to give up the cigarette habit. Spouses were the most influential upon one another, followed by friends, coworkers and siblings. The likelihood that a smoker would quit after a spouse did, for example, was 67%, the study found. For friends, it was 36%; for coworkers in small firms, 34%; and for siblings 25%.

People who had more education were more subject to influence than those who had less. What the researchers found surprising was that members of groups tended to quit more or less simultaneously, and those who did not were marginalized.

James Fowler of UC San Diego, one of the leaders of the study, suggested this was a bad thing.

"Stigmatization is effective," he said. "But is it always advisable? Social pressure comes with a cost. We see here that we have stigmatized not only smoking but smokers too. Huddling in a corner with your smoking friends is going to make it that much harder for you to stop. And we also know from other research that social isolation is detrimental to your physical health on several other levels as well."

Hang on just a minute here. The thread of logic seems to be getting a little frayed. If people quit because of social pressures to do so, then it seems to me that those "huddling in a corner" are experiencing those pressures -- or they wouldn't be behaving as though they were ostracized, would they? That should make it more likely that they too would eventually quit, not less.

Fowler points out that thirty years ago, whether one smoked or not was socially irrelevant. "You could be central in your circle and be connected to lots of other people who were similarly central. You could be popular, in other words. By the 2000s, it had become highly relevant: If you smoked, you would, in some sense, be shunned," he says.

The difference from then to now, obviously, is the cultural shift in the social acceptability of smoking. Back in the 70s, smoking still had that James Dean, Humphrey Bogart, Breakfast at Tiffany's aura. There were still Marlboro Man billboards, ashtrays in coffee houses, and teachers' smoking lounges in schools. Nice people smoked. Elegant people smoked. It was cool.

After three decades of pummeling by antismoking forces, there's been a paradigm shift. Smoking is decidedly uncool. It's associated with illness, addiction, foul smells, discolored teeth, ignorance. Weak people smoke. Of course, smokers will continue to be more and more marginalized. That's as it should be.

Fowler seems to want to adopt a "hate the sin but love the sinner" attitude by suggesting that a little kindness toward smokers is in order. That, to be blunt, is horseshit. It would be inexcusable to release pressure on smoking -- or on smokers -- in the slightest.

Debating whether smoking kills now falls in the same category as arguing about the shape of the Earth. The reason smoking has precipitously declined in the US over the past three decades is precisely that people have been made to feel uncomfortable and unattractive doing it.

The worse smoking makes them feel in a social context, the more likely people will be to take the difficult steps necessary to break the addiction. That's a good thing. The suggestion that society should soften toward smokers -- to find a way to expunge their feelings of isolation and draw them back into the fold -- is counterproductive, to put it mildly.

I say let's keep pushing smokers into dark corners until they're ready to see the light.

Send your comments to Rita Jenkins.