Do NOT quote from this. Bits might be wrong
Introducing the annoying problem of information cascade
People who do their own research complain about Covid vaccines not being 100% safe or 100% effective.
Because this is plainly true they think it's case closed on their next assertion - that the entire vaccination programme and the research behind it was a scam, based on lies.
You can't fault the logic in this step.
One proposition necessarily follows from the other.
What invalidates the reasoning is that it flows from a false premise.
Because no one responsible for vaccine efficacy and safety seems to have ever stated that Covid vaccines are 100% safe and 100% effective.
Other people said it and then other people said it too, because those people had said it.
And now there are a number of people who sincerely believe that public health officials said it and were therefore lying. So now they keep saying that it was said.1
That's an information cascade. And people often build perfectly good arguments on top of them, not realising the fallacy of the original premise.
—
Most days I ensure I've walked more than 10,000 steps. I heard somewhere that this was a medically significant number.
It isn't. It's an arbitrary number that was chosen decades ago to market a Japanese pedometer. Then lots of people kept repeating it until 10,000 steps a day became a health benchmark.
Information cascades are an example of mimetic behaviour; people doing or believing things because they see others doing or thinking those things.
An information cascade is a process, rather than a thing that is either true or false, wise or foolish in itself.
You will often find one at the heart of any kind of narrative. (Half of the point of public relations is to start information cascade).
If you know anything about D-Day you're likely to believe that thousands of Americans died on Omaha Beach. You might even think and speak of that place as 'Bloody Omaha', while referring to Utah Beach and the others by their ordinary names.
In one episode of Angus Wallace's brilliant WW2 podcast he interviews a Normandy battlefields tour guide who describes reactions of fury from some visitors when he tells them the truth. These people have read popular books from some time ago which falsely reported the number of KIA (killed in action) on Omaha Beach and those books came to form the basis of a misleading narrative.2
Compared to Utah Beach, Omaha was much bloodier, but the KIA figure may not have reached 1,000 according to scholars of the landings.
Information cascades aren't necessarily bad, as such. There's nothing inherently bad about believing that many more Americans died on Omaha beach than the true figure. Goodness knows, enough of them died for it not to really matter in a moral sense that the figure has been overstated.
But those people who were influenced by the information cascade feel betrayed and don't like being told that they fell for an untruth.
There's nothing at all bad about me perambulating for more than 10,000 steps each day, just because the Yamasa Clock company thought it was a good number, rather than on the basis of robust medical research.
But they're inconvenient if reality matters to you - and deeply unhelpful in some cases. It must be quite stressful to go around thinking that all the official claims for Covid vaccine safety and efficacy are lies, especially with billions of people suckering for the jab at least once (I’ve had four of those bad boys and I’ll have another, if it seems advisable at any point, without stressing over it).
I personally know a couple of people who believe that the official line on Covid vaccines was that they are 100% safe and effective. It drives them nuts that most of us don't care about the figures they dig up about transmission rates from and between vaccinated people or the various and sometimes fatal or life-changing consequences of receiving one or more doses.
They aren't stupid, but they are often operating on the basis of a false claim about a claim that was perhaps never made about vaccine safety and efficacy. Sometimes various media did stupidly (albeit technically correctly) report some small study involving a particular cohort of vaccinated patients who didn't contract Covid after exposure and suffered no side effects from the vaccine. But journalists aren't public health officials. They're purveyors of stories.
We could also, at this point, go into the different meanings of 'safe', 'effective' and so on in a clinical research setting compared with everyday usage, but I'm frankly bored of Covid vaccine scepticism now and don't care what anybody thinks about any of it.
You do you. That's my policy.
The point is how information cascade causes confusion.
—
I recently spent weeks looking for the records of a WWII officer called ‘Lt. Alsauer’, who was reportedly in the cavalry unit whose history I'm researching.
After about a month I finally realised that we're talking about a ‘Lt. Al Sauermann’, probably from an infantry unit, who happened to be talking to a cavalry captain in a place called Quettehou on a particular day. The error comes from someone being tripped up by a line break in an old book and publishing the misinterpretation on their own massively popular website, which is now copied from endlessly by other websites.
It's interesting to realise how much plagiarism there is online. Time and again you see references to Capt. Brooks O. Norman ordering ‘Lt. Alsauer’ to go somewhere, when this probably didn't happen and ‘Lt. Alsauer’ didn’t exist. I only bothered chasing this one down because I’m a bit obsessive and because - having wasted so much time trying to surface records on a fake name - I was curious to understand how this particular information cascade started.3
My main current personal beef with information cascades is that they demand that you fact check everything so that the cascade stops with you.
Information cascades often start for very innocent and understandable reasons.
For weeks I was baffled that the award of a particular French honour to the unit I’m researching was being misreported all over the internet.
Everyone says it was awarded to the 4th Cavalry Reconnaissance Squadron.
But the citation describes an engagement that was verifiably fought by the 24th Cavalry Reconnaissance Squadron. I have several original records of the engagement.
Try telling anybody this, though, when they've seen it in black and white - the 4th Cavalry Reconnaissance Squadron. It's right there, on the record, in an old document.
But ... do you see the unit listed below the 4th Cavalry Reconnaissance Squadron?
The 0th CML? (that would be the zero'th Chemical Decontamination Company, which obviously didn't exist)
The left hand side of the page has been cut off in the scanning of the original document. The 4th Cav Rcn Sq (Meczd) is missing a 2 in front.
This seems to be the source of the original error. The reason why so many people report incorrectly that the 4th Cavalry Reconnaissance Squadron received a Croix de Guerre for their actions in the battle for Bourg de Lestre (in which six men from the 24th Squadron died).
That document is the source of an information cascade, because someone who didn't know the details of the engagement meriting that medal made an honest mistake - and everyone else is now copying them.
It's funny how gaining a little knowledge in a field leads to noticing false claims.
I've read a few times that the German commander occupying the pretty French fishing port of Barfleur pulled out before the Americans arrived (Troop C of the 24th Cav Recon, to be precise, baby) because he was so fond of the place and didn't want to see it damaged.
I'm sceptical. Because I now know that the Germans were ordered back from this area to defend Cherbourg on exactly the night that they pulled out of Barfleur.
He might very well have loved Barfleur, but he vos only following orders.
Someone started an information cascade with that story. Nice story though it is, I'm no longer at all sure that Barfleur was spared combat damage thanks to a public spirited German officer. Who, incidentally, left so many landmines around the place that a significant number of locals died long after the Germans left.
Good 'sensemaking' seems to be a worthwhile pursuit, but it's often really hard.
So much for my war stories ... this stuff affects everyone who is interested in the issues of today. If we can't all agree on the basic facts, we've seen from the toilet of Twitter/X, Facebook and elsewhere what the arguments look like.
And yet it seems risky to trust anything other than original source documents, because you cannot know whether someone has been mistaken when you read their later version of something.
Throw in awareness of the Gell-Mann Amnesia Effect and it becomes very difficult to trust anything you read or hear.
Like all problems of its type, this one is much easier to describe and explain than it is to solve.
Endnote
So, a Rarely Certain milestone is approaching. Thanks to recommendations from bigger names it's about to pass the 900 subscribers (mostly free) mark.
I try not to make this a numbers game, but no one’s pretending it won't feel good to reach 1,000. So, if you liked this, please consider contributing to the metrics by signing up.
And because subscribing to things soon gets costly, there's always the option of showing support through Buymeacoffee. That gets you pro-rata paid subscriber benefits, too.
These thoughts were inspired by a thread on the often useful Metabunk website, in which one user makes the point that the original claim of 100% efficacy and safety may never have been made. Trawl through it here.
You can hear that episode here https://ww2podcast.com/ww2-podcast/allies/219-d-day-tourism/
UPDATE [29-03-24] we are talking about 1st. Lt. Albert C. Sauerman, a Silver Star medal recipient who was indeed under the command of Capt. Brooks Norman, in Troop A of the 24th Cavalry Reconnaissance Squadron. Glad we cleared that up. Albeit a week later.
It seems so weirdly coincidental that you talked about death tolls of D-Day; I was just last week talking to my 14-year-old daughter about war (she'd asked some questions about the US Civil War), and it led to looking up the beach death tolls. I was really surprised to see how low they were! Not because a couple of thousand deaths is small, but because of the mythology built up around them, they make you FEEL like it must have been tens of thousands of soldiers. Sort of like the number of unarmed black people shot by police. No one ever quotes numbers, but the STORY that's told about it makes it feel like the number must be high. That's not the same as an information cascade, but it's a related way in which narratives can mislead us without any one of the individual stories within the narrative being incorrect.
Regarding information cascade, I agree with everything you said. It seems to me that, while the behaviors that drive the cascades are fairly built into human social behavior, our contemporary information ecosystem demands these kinds of information efficiencies far more than prior generations where one could conceivably read the world's significant news in a few hours, say, or even earlier, where one could conceivable read a pretty fair percentage of books in one's language. Do these problems continue to get worse? Does AI help us start to handle them more effectively, or does it merely mirror our already-overstressed (perhaps overwhelmed) mechanisms?
So I forwarded this to my elderly anti vax conspiracy theory mother, it seems she read the first two sentences, quoted them back to me and then wrote me a 1000 word essay on why vaccines are not safe. QED!