Oct 13 2014

Anomaly Hunting and the Umbrella Man

This is not a new story, but it is worth repeating. At the moment that bullets were being fired into JFK’s motorcade, a man can be seen standing on the side of the road near the car holding an open black umbrella. It was a sunny day (although it had rained the night before) and no one else in Dallas was holding an umbrella.

This is exactly the kind of detail that sets a fire under conspiracy theorists. It is a genuine anomaly – something that sticks out like a sore thumb.

The event also defies our intuition about probability. Even if one could accept that somewhere on the streets of Dallas that morning one man decided to hold an open umbrella for some strange reason, what are the odds that this one man would be essentially standing right next to the president’s car when the bullets began to fly?

Our evolved tendency for pattern recognition and looking for significance in events screams that this anomaly must have a compelling explanation, and since it is associated with the assassination of a president, it must be a sinister one.

When you delve into the details of any complex historical event, however, anomalies such as this are certain to surface. People are quirky individual beings with rich and complex histories and motivations. People do strange things for strange reasons. There is no way to account for all possible thought processes firing around in the brains of every person involved in an event.

Often the actions of others seem unfathomable to us. Our instinct is to try to explain the behavior of others as resulting from mostly internal forces. We tend to underestimate the influence of external factors. This is called the fundamental attribution error.

We also tend to assume that the actions of others are deliberate and planned, rather than random or accidental.

The common assumption underlying all of these various instincts is that there is a specific purpose to events, and especially the actions of others. We further instinctively fear that this purpose is sinister, or may be working against our own interests in some way. In this way, we all have a little conspiracy theorist inside us.

I also find it interesting that these tendencies are often not inhibited by the many counter-examples that we encounter on a regular basis. The vast majority of the time, when I find the actions of another person puzzling, if I am able to simply ask them to account for their behavior, there is usually a non-sinister explanation. There were factors of which I was unaware. They knew (or at least believed) something I did not know, or were reacting to an external stimuli, or had a previous experience informing their current action. Sometimes they were just doing something whimsical for fun or to stave off boredom, or perhaps they were satisfying some minor curiosity.

In response to such experiences we should question our basic underlying assumptions of purpose and deliberateness. Instead, we tend to dismiss this data as quirky exceptions, and carry on with our assumptions intact.

Conspiracy theorists have essentially formalized the tendency to assume agency, deliberateness, and sinister motivations in the quirky details of events. Conspiracy theories are often an exercise in anomaly hunting. When anomalies, like the Umbrella Man, are inevitably found it is assumed that they are evidence for a conspiracy.

The assumption that anomalies must be significant rather than random is an error in the understanding of statistics, a form of innumeracy. It is also partly the lottery fallacy – which involves asking the wrong question. The name of the fallacy is based on the most common illustrative example. If John Smith wins the lottery our natural tendency is to consider what the odds are that John Smith won (usually hundreds of millions to one). However, the correct question is – what are the odds that anyone would have won, in which case the odds are close to one to one (at least over a few weeks).

The fallacy is in confusing a priori probability with posterior probability – once you know the outcome, asking for the odds of that particular outcome. This is perhaps more obvious when we consider the odds of someone winning the lottery twice. This occurs regularly, and when it does the press often reports the odds as being astronomical. They are usually also falsely considering the odds of one person winning on two successive individual lottery tickets. Further, they calculate the odds of John Smith winning twice, rather than the odds of anyone anywhere winning twice (the odds are actually quite good and match the observed rate).

So – conspiracy theorists tend to ask, what are the odds of a man standing with an open umbrella right next to the president when he was shot? Rather they should be asking – what are the odds of anything unusual occurring in any way associated with the JFK assassination?

There is another aspect to anomaly hunting and that is the use of open-ended criteria. What constitutes an anomaly? Well, anything you want to count as an anomaly. There are no specific criteria. In practice the criterion is – it seems weird to me. This then opens the door to confirmation bias. Seek and ye shall find.

What, then, is the explanation for the seemingly bizarre actions of the Umbrella Man? A nice documentary, recirculating on social media, has the answer. The man (Louie Steven Witt) was asked to come forward and explain his actions, and he did, before congress. The umbrella was a protest of Joseph Kennedy’s appeasement polices when he was Ambassador to the Court of St. James in 1938-39, with the umbrella being a reference to the umbrella often carried by Neville Chamberlain.

This is actually not as random as it may seem (and this is the one hole in the documentary’s treatment of the topic). An open umbrella was a common protest of appeasement policies. According to the historical society:

Umbrella protests first began in England after Chamberlain arrived home from the conference carrying his trademark accessory. Wherever Chamberlain traveled, the opposition party in Britain protested his appeasement at Munich by displaying umbrellas. Throughout the 1950s and 1960s, Americans on the far Right employed umbrellas to criticize leaders supposedly appeasing the enemies of the United States. Some politicians even refused to use them for that reason. Vice President Richard Nixon banned his own aides from carrying umbrellas when picking him up at the airport for fear of being photographed and charged as an appeaser.

That puts the umbrella protest into more context. In the early 1960’s I can see there still being people around who were angry at any attempts to appease Hitler and the Nazis prior to the start of WWII. I also wonder if Witt was protesting JFK in some way, but did not want to say so after he was assassinated and so blamed the protest on his father. Either way, the umbrella is not such a random detail after all.


The various aspects of anomaly hunting are critical to understand in order to avoid falling into this seductive mental trap. Poor intuition for statistics, logical fallacies, confirmation bias, and the use of open-ended criteria combine with the fundamental attribution error and the tendency to see patterns and significance everywhere to create the powerful impression that something (usually sinister) must be going on.

Our penchant for narrative then takes over. We love a good story, and the notion that some tiny clue in the form of an anomaly can reveal a vast unseen conspiracy is more compelling story telling than just random noise in the background of history. Unless, of course, your telling the story of how we fool ourselves. That, of course, is a story I like telling.

31 responses so far