Last week, PLOS ONE published a piece by a physicist, David Robert Grimes, which deals with the likelihood of conspiracies being kept secret over time.

Said article has been talked about a fair bit on the news over the last few days, and I've been asked by a great deal of people either to comment on it, or to outright debunk it. Given that I don't like commenting on something I haven't read, I held off saying much about it until such time I found time to not just read it, but also carefully pick through it. That time has past; I have read the piece and I am ready to give my opinion on it. This is the short version of my critique; the original version ballooned out to almost 4000 words, and I probably should turn that into an academic article of some sort. If you want the tl;dr version of my opinion, then skip to the end. Otherwise, let me edutain you.

Grimes' argument

Grimes' argument centres around the viability of conspiratorial activity, and how that weighs upon considerations of the rationality of belief in associated conspiracy theories. The structure of Grimes' argument is to:

1. Establish a baseline for the viability of known conspiracies, and then
2. Use that to assess a set of putative conspiracy theories.

There are problems with both of these parts of Grimes' argument, so let us go through them one by one. Namely:

1. The examples he uses to generate his model don't fit his model,
2. The model is insensitive to the vagaries of conspiratorial activity, and
3. He has a weird definition of what counts as a leak.

This short article only focuses on the first part of Grimes' article, establishing the baseline of his model. I have thoughts on the second part of his analyis, but by showing up problems for the viability of the model, I can show that no matter what we think of those purported conspiracies here-and-now, the Grimes' model is not up the task of analysing them.

Known conspiracies

Grimes admits conspiracies occur. It is interesting, then, to see what his chosen examples of conspiratorial activity are. They turn out to be:

1. The NSA's Mass Surveillance programme,
2. The Tuskegee syphilis experiment, and
3. The FBI forensic scandal.

Three examples really isn't very many, but Grimes uses them to estimate the parameters of his model about the viability of conspiracies over time.

1. Conspirators are generally dedicated to keeping their activity secret or concealed.
2. Leaks from a conspirator expose conspiracies and render them redundant.

However, his chosen examples do not fit these two assumptions, which is why he can't use them to generate a baseline for the viability of conspiratorial activity.

Take the Tuskegee syphilis experiment. If this cover-up of the unethical treatment of African-American men was exposed by conspirators, then Grimes is stretching the idea of a leak too far. Information about the experiment was published in medical journals; the cover-up, so to speak, was that the patients were not told about the experiment. The leaks – the various academic publications about the experimental results – did not end the experiments. Peter Buxton, who is often labeled as the whistleblower in this case, came to know about the experiments because of his job with the United States Public Health Service. He leaked the information to the press because his worries were not taken seriously by management. He was not a conspirator but, rather, a worried public official who learnt about the experiments without being involved in running them.

Then there is the case of Dr. Frederic Whitehurst and the FBI forensic scandal. This, like the Tuskegee syphilis experiment, is a case where someone who was not a conspirator blew the lid on suspicious practices; Whitehurst was someone who, whilst checking and rechecking work, discovered that FBI management was not willing to discuss issues with that work openly. ((It's not even clear that this is a case of a conspiracy; what might be taken to be a cover-up could have just been the FBI just not listening to expert advice.))

What this says is that it's not clear two of the three examples of known conspiratorial activity fit Grimes' fit his model, which is a problem for his generation of the estimates that drive his predictions about putative claims of conspiracy.

Who are the conspirators?

A bigger problem still is that Grimes does not distinguish between conspirators and whistleblowers.

Take the NSA mass surveillance conspiracy theory, for example. Grimes would have us believe that Snowden was a conspirator who leaked information about what it was the NSA were up to, and this lead to the exposure of PRISM and related surveillance programmes. That certainly is one version of that story, but it's important to note that Snowden's narrative is that of being a whistleblower: Snowden – at least in his recounting – learnt what the NSA were up to, played the role of a good contractor, and was eventually granted access to the files he needed to show the world what it was the NSA were really doing. Snowden, on this version of the story is not so much a conspirator, but a whistleblower who went undercover at the NSA. As such, the Snowden revelations do not fit with the assumptions that drives Grimes' model.

Grimes does not distinguish between kinds of conspirators, let alone conspirators and whistleblowers. Conspirators are not homogenous; a conspiracy can look big, yet only feature a small number of people who know the full extent or aim of the conspiracy. Some members of the conspiracy will be lackeys, goons or even unwitting conspirators. It's possible to be involved in a conspiracy without realising you are conspiring. Not everyone in the NSA need necessarily know that the data they are collecting and processing has been illegally obtained, and FBI agents who are using forensic evidence to secure convictions may not have been informed by senior personnel that the evidence they are using is of dubious merit.

What is a leak, anyway?

Grimes asserts in his paper that a leak from a conspirator will expose a conspiracy and render it redundant. But 'leak' is an ambiguous term; some leaks are accidental, some are purposeful, and some 'leaks' don't even come from conspirators at all.

Take, for example, accidental leaks. These are cases a conspirator fails to cover up some facet of their conspiratorial activity. So, for example, in the Watergate Affair, it was surely a mistake or accident that Alexander Butterfield revealed the existence of the White House tapes. This was an accidental leak, one which – in part – helped expose the cover-up of who knew about the break-in of the Democrat National Headquarters in Washington, D.C.

Contrast this with a purposeful leak, where, say, a disgruntled member of the conspiracy leaks information about the conspiracy in order to expose it. Such leaks are, for example, common when it comes to leadership coups, and a variety of political conspiracies, where there is an advantage to some former conspirator to out what has been happening, in order to either change sides or escape blame. ((Indeed, sometimes these leaks presage the fact that the conspiracy is on the verge of collapse, and the leak is a pre-emptive strike by certain conspirators to save face.))

Whatever the case, Grimes is committed to the idea that conspirators leak, and that such leaks expose and make redundant the conspiracy. But it is not clear that conspirators leaking is the only way in which conspiracies get exposed. Admittedly, Grimes is not expressly committed to the idea only conspirators leak. But, given that his model seems insufficiently sensitive to how leaking works, let alone the difference between leaks from conspirators and external whistleblowers, this is a problem.

Indeed, none of the examples Grimes uses, as outlined above, fit with the known conspiracies he uses to derive his model, let alone give him any justification to then show that a number of purported conspiracy theories are irrational to believe. As such, Grimes does not do enough to warrant his own theory, let alone allow him to then extend that analysis to purported conspiracies now.

tl;dr: Grimes’s article doesn’t just misrepresent the nature of conspiratorial activity, let alone how they are exposed, his article seems to mostly exist to justify our suspicion of ‘crazy’ conspiracy theories.