Corrections are not a bad thing

I've been avoiding longer blog pieces of late in order to concentrate on CGI and work stuff. One I've had in draft for a while is a look at the general conditions under which people enjoy being proved wrong - and when they don't. Disproving old ideas is the foundation of scientific progress, but even there, it's not always fun if something you've worked hard on turns out to be junk. More on that in a future POTC post.

As far as science goes, it's pretty clear that reporting (in the mass media) has been in a pretty dismal state for some time :

Kovach and Rosenstiel call truth “a complicated and sometimes contradictory phenomenon.” It emerges as facts collect, and each new fact changes a society’s collective understanding of truth. The truth is a snowball rolling down a hill, and each new fact changes the snowball, making it bigger and more multifaceted. 

Journalists, however, have historically done a bad job of explaining to the public that each day’s news report is, by necessity, incomplete and provisional. CBS Evening News anchor Walter Cronkite’s famous sign-off, “And that’s the way it is,” sums up the attitude.

The evolving consensus on whether or not to wear masks in public is one example of a part of the coronavirus story that has changed quickly. As New York Times opinion writer Charlie Warzel pointed out, the official advice about wearing masks changed completely in the course of a month.

Again, the findings of science are evidenced-based and provisional, which is why stupid headlines like "mystery solved" or "we finally know..." are so infuriating - and damaging. Genuine proof is a rare thing indeed, not non-existent in any reasonable sense, but not something you should degrade by a cheap headline to get more readers. As scientific debate becomes more public, every minor disagreement between experts can be seen not as an essential, fundamental part of the process, but a petty spat between squabbling, tribalistic incompetents. This completely misunderstands how the scientific consensus is forged. Which is extremely dangerous when someone needs to make a decision based on limited evidence.

Long-developing, ongoing stories – such as the coronavirus today – are particularly difficult for journalists to convey in the day-by-day, article-by-article style which they and the public are used to. With many stories, from climate change to political campaigns to the coronavirus, journalists must rely on expert sources who collect and interpret data and tell the reporters what that data means. These sources change their interpretations, even just slightly, changing the long-term story as more information comes in. 

The problems of uncertainty are particularly troublesome in daily reporting and in the overheated atmosphere of social media, where today’s new development – or new partisan attack – is more important than the big picture... News organizations, of course, have developed systems for correcting themselves when they are provably wrong about something they have published. As my own research has shown, these systems, however, are something that newsrooms have historically entered into reluctantly....  most of the corrections The New York Times publishes are in the nature of a misspelled name rather than about putting facts in the wrong context or explaining how the common understanding of a situation has changed in subtle but important ways.

I would add that there's an even bigger problem for journalism than scientific reporting. At least scientists don't often care about the media reports, so eventually the process works its magic and quietly arrives at a solution behind the scenes. What I think may be in far worse shape is the regular political opinion column.

In principle, through a degree of light legislation and public ability to choose which outlet gets their money, society is supposed to hold journalists to account and produce a reasonable sort of commentary. In practise, this is bollocks. How often do you see a political commentator ever taking stock and reviewing what they got right and what they got wrong ? Hardly ever. It's not that they change their minds that's the problem, it's that they do so for the exclusive motivation to drive sales. If they were to say, "I was wrong about that because I didn't know about factor Y", that would be perfectly fine. But they never do. They never, ever stop to think about how they should form a conclusion, only which conclusions will sell papers. So they can freely say, "John Smith is the sexiest man ever" one day and "John Smith smells of manky bat guanao" the next.

(Just to clarify, I'm talking about predictive pieces more than analyses. There's much value in discussing unprovable notions like just how sexy John Smith is, nothing wrong with that. It's the "political party X is on the verge of disintegrating" pieces I'm concerned with here - the ones that do say something that can be analysed with some degree of objectivity.)

Take Boris Johnson. I've not forgotten his long, sad history of racism, idiocy, authoritarianism and infidelity. I have argued strongly that he should have been voted out of office for his monstrous inability to compromise. I have certainly not forgotten his disgusting attitude in Parliament when openly accused of enabling hate crimes. Nor should we forget his duplicity or egomania.

I could go on, but that's not my point today. My point is that - and I do want to openly echo Keir Starmer here - criticism has a purpose, which carries with it a host of implicit assumptions. It is not for the sake of enjoying someone else's suffering. It has an objective : in this case, to alter the dynamics of the system so that better decisions are taken in the future. The criticism in and of itself is utterly irrelevant. If Johnson's recent experience teaches him a much-needed degree of humility, then his reformed character should be welcomed with open arms. Attacking someone who reforms is only sending a signal to others that they should continue to hate you.

Not that I'm saying Boris Johnson is suddenly going to see the light. I'd bet heavily that he won't. But how many political commentators would ever say, "I was wrong about that" ? Few indeed. The result is we get a continuous stream of vitriol from all sides. Some continue to criticise the government even when mistakes where inevitable, some now applaud them - but few acknowledge their stance has changed in the slightest. The result is a total mess, a complete confusing mass about who's trustworthy, who's principled and who's just in it for the money, who's a diehard and who values evidence above ideology.

I should also mention a long-standing goal of this blog is to review the (sadly neglected) "predictions" category every once in a while. After this rant, I'd guess I'd better put my money where my mouth is pretty soon, so let's shoot for the end of next week. Possibly over on POTC depending on how long it turns out to be, but I'll put the short version here.

Anyway, to return to the theme of the linked article, it'd be great if newspapers (and TV stations) would do regular reviews of how they've done every month or so : what they've got right and what they've got wrong and why. Correcting typos of minor factual details are not enough - that's a way of burying the important results, as a form of bullshitting. Opinion pieces are valuable, but shouldn't be an excuse to say whatever crap will sell papers. Analysing what went wrong and why would help give commentators better perspective and make better predictions in the future. And while a good rhetoric-laden rant every once in a while is good for the soul, by and large I'd far rather have astute political commentary and satire than capricious ramblings about why a particular politician is nothing short of angelic / the devil incarnate / a big pile of poop.

Of course, it's not terribly likely that newspapers are going to start doing this themselves anytime soon. But perhaps it would be fun if someone external were to take up the challenge. Maybe not as a newspaper, but a blog or website - well organised, and with an emphasis on analysing why mistakes were made as well as successes - would be much more interesting than correcting petty typos. I bet a lot could be done just by using direct quotes alone. This is already quite a successful and popular approach to criticising politicians, so why not political commentators ? Not that I'm brave enough to go around mocking journalists myself, mind you...

Journalists are recognizing they're writing a rough draft of history -- and can't say definitively "that's the way it is"

On April 4, a Los Angeles Times story about the varying effects of the novel coronavirus contained a remarkable paragraph: "One thing to keep in mind before we continue: It is possible that the information you read below will be contradicted in the coming weeks or that gaps in knowledge today will soon be filled as scientists continue to study the virus."