A few weeks ago, when the Atlantic article came out saying that Donald Trump had called people in the military "losers" and "suckers", I got into it on-line with someone about evidence. They didn't approve of anonymous sources. So I posted the video of Trump calling McCain a loser and "not a war hero":
And the Howard Stern interview, where Trump called STDs his "personal Vietnam. I feel like a great and very brave soldier."
And said that I thought we could perhaps extrapolate future behavior from prior statements. Now I wouldn't have minded an argument on freedom of speech - I'm always ready to defend that one - or even the validity of judging someone by their past behavior. BUT the reply I got was that "the videos are just circumstantial evidence." So I blew a gasket. Because of course no, they're not.
BTW - one of the great examples in literature of historical arguments (and how much analysis, deduction, argument, and debate they require) is not Dan Brown, but Tom Stoppard's 1993 play Arcadia, which deserves far more productions (and at least one movie! Please!) than it has gotten. Alternating between the mid-1800s and the present, the present scenes show dueling historians arguing over the following primary sources: mid-1800s "game books" (i.e., hunting records at a country house), a diary (by a young girl), and a number of letters and notes tucked into a poetry book which itself was heavily underlined. All of which seem to indicate that Lord Byron killed minor poet Ezra Chater in a duel over Chater's wife at this country house where they were all guests at the same time, and after which Mr. Chater disappeared, and Lord Byron fled to the Continent for two years. Solving what belonged to whom (including who did all that underlining) is a masterclass in historical deduction and detection.