Whenever I look back on the banality of my recently defunct career, I can’t resist the urge to abstract myself from the story. Who is this guy who wasted so much time on so much trivia? That can’t be me! It says here in the script that I was an intellectual behemoth who offered the human race a new rôle in the universe. So why was I in a meeting trying to decide whether the results of flushing a lavatory in a stationary train should be classified as ‘Special Waste’ or ‘Controlled Waste’? Spirit of Christmases Fucked, why are you showing me this?
Actually, I think I can answer my own question. It’s just a timely reminder that ‘self-identified’ is the postmodern term for ‘deluded’.* Fantasy and fate have a habit of avoiding each other’s company, but am I bitter? Just a tad.
I could carry on in this vein, entertaining you with the wistful angst of an aging Caucasian male, but I also have a serious point to make. First, however, I must admit that not all of my career was as soul-sapping as the Great Train Toiletry Debate. Towards the end of my working life, for instance, I became involved in the honourable trade of safety-systems engineering, specialising in the functional safety analysis of computer-based systems.
To master the subject, l had to tackle a weighty tome called ‘IEC 61508—Functional Safety of Electrical/Electronic/Programmable Electronic Safety-related Systems.’ This is a standard that details, with almost talmudic opacity, the development lifecycle to be followed by safety engineers in the design and construction of computer-controlled, safety-critical systems. In particular, it tells you how to identify necessary safety functions, together with the standards of reliability required to achieve specified levels of safety. The reliability targets are quantified as Safety Integrity Levels [SILs], ranging from SIL 1 to SIL 4 (with SIL 1 being the least failure-proof). For each level, IEC 61508 explains the techniques that are both prescribed and proscribed for achieving the safety target.
It was all good stuff. But there was one thing that troubled me.
You see, impressive though IEC 61508 is, I was always struck by the complete lack of evidence that adherence to its instructions would result in the various levels of safety promised. SILs are specified in terms of failure rates, with SIL 4 (as a case in point) representing ‘failures upon demand’ that may be as few as 1 in 105. For such high-integrity systems one could be waiting millennia for confirmation that the required reliability had been achieved. Even when historical checks were practicable, nobody seemed to be conducting them to see if systems developed under IEC 61508 were doing what it said on the tin. It all looked very suspect to me, so when I had the privilege of working with a safety consultant of considerable standing in the field, I posed this question of evidential support to him. This was his response:
“Yes, John, I too was troubled by that. So I asked one of IEC 61508’s authors to outline the process they’d used to decide how to correlate the approved methods and techniques to the SILs. The story he told me was that he met up with two of his colleagues at a pavement café in Bruges, where they determined themselves to thrash it out. Basically, they just shouted things out and wrote them down, and as the strong Belgian lager flowed, the job just got easier and easier.”**
And that’s how the world’s safety engineering community established the scripture we used to ensure the functional integrity of everything from automatic braking systems to nuclear power stations. I am not saying that the evidence for all computer-controlled system safety is spurious, but you need to know that a lot of the evidence cited in support of a system’s safety case will often amount to little more than an auditor’s confirmation of IEC 61508 compliance. Consequently, I advise that one should treat any claims for the achievement of specific safety levels with just a small wheelie bin of salt. Furthermore, knowing now the ease by which drunken whimsy can take the reins of power, perhaps you should be on the lookout for other signs of bogus authority. Is the IPCC immune from the same Wizard of Oz syndrome? Did anyone check Moses’ breath when he came down from Mount Sinai?
To answer the former question it’s useful to appreciate that climatology is just a branch of systems safety analysis. It is an attempt to understand and, thereby, predict the future behaviour of a system (the climate system) that has a posited failure mode (anthropogenic global warming) with the potential to cause serious and irreversible harm (CAGW). A safety case has been developed by an authoritative body (the IPCC) and safety case reports (Assessment Reports) are regularly produced to communicate said case to the policy makers. There are even proposed interventions that have been deemed necessary (though not necessarily adequate) to reduce the risks to acceptable levels.
When I look at climatology, I see nothing I haven’t seen many times before in the field of safety engineering. There are decisions made with incomplete evidence. There are uncertainties that cannot be reduced (at least within the timescales required). And, above all, there is an over-reliance on appeal to authority. (All too often in safety management, it is not the truth that matters but the extent to which the experts can agree).
But there is one major difference: Safety engineers know what they are doing, and they know they are not doing science.
Meanwhile, back at IPCC headquarters, lead authors are developing standards for the communication of uncertainty, and are confidently using their cherished terminology to classify the likelihoods that their various proclamations may be true. Would it be churlish of me to enquire who, exactly, was on the team that dreamt up these classifications, and how many lagers were consumed in the process? ■
* And don’t think you can teach me anything about delusion–I bloody well invented the word.
** This story may be slightly hyped in the re-telling but please do not mistake it for a conspiracy theory. It’s just a little-known insight, which I’ve chosen to share with you, dear reader. You would actually need a conspiracist mindset not to believe me.