I investigated many of these theories, often by consulting the sources I’d cultivated. Kennedy was right that the 2004 election had been rife with irregularities and efforts at disenfranchising voters, particularly in Ohio, where a partisan secretary of state, Kenneth Blackwell, had overseen several divisive voting measures and obstacles. But pretty much every expert I talked to said it was unlikely that any of the issues were big enough to have undone Bush’s win. An investigation by the Democratic National Committee which looked at precinct level voting counts found that the data “does not suggest the occurrence of widespread fraud that systematically misallocated votes from Kerry to Bush.”

And so: I wrote a point-by-point debunking of Kennedy’s breathless claims. Then Kennedy wrote a rebuttal to my rebuttal, which I, again, rebutted.

For a week or two this dust-up took over my life. Salon, a generally liberal-leaning publication, was deluged by letters from readers angry that I was defending Bush’s win. Thankfully, my editors supported me, and I remember coming away from the episode feeling bruised but journalistically vindicated: A man with a famous political name was wrong on the internet, and, armed with the facts, I had stepped in to correct the record.

Looking back, though, I cringe. The other day I went back and listened to a debate I had with Kennedy on public radio’s “The Brian Lehrer Show.” Lehrer opened the program by asking Kennedy for his big-picture case. But whether Kennedy is talking about vaccines, elections or other out-there topics (he told Rogan he was “aware” that he could be assassinated by the American government), he tends to present his theories in a particular way. He starts with a few sprinkles of truth — Ohio’s vote was run by a partisan official, some vaccines have serious side effects — and then swirls them up with enough exaggerations, omissions and leaps of logic to create a veritable McFlurry of doubt.

Such was his effort when we met on “Lehrer’’: Kennedy offered an assortment of claims about the election that, in big and small ways, were unsubstantiated. So when Lehrer turned to me, I felt I had no choice but to start out by correcting Kennedy’s misstatements. I did so pretty handily, but because I had to point to sources and tease out the nuances Kennedy had elided, I couldn’t help but sound like the boring, persnickety nerd stuck in the weeds. After a few rounds of this back-and-forth, I can’t imagine that much of anything had been clarified for the audience. Instead, the impression was one of earnest complexity: One side says X, the other says Y, but whoever is right, it sure seems like this is a debate we should be having.