More text

Nothing says "I Love You, Dear" like screaming lower back pain!

Sometimes Wrong but rarely in doubt!

28 October 2009

You Can't Build the World out of Rubber Walls and Round Corners

One the things that I encounter frequently in reading various engineering publications is the tendency to look for increased safety for the public.  As an engineer I have an obligation to keep the public's safety paramount in performing my duties. The conundrum that I face is what is safe enough?  Engineering is almost always about trade-offs, trading safety for efficiency, one desirable physical property against another. Building a world out of round corners and rubber walls keeps you safe until the earth shakes or a stiff wind blows.  Designing products for perfect safety is impossible and eventually you have to say, 'Good e'nuff'.

That's why I have some trouble swallowing poor arguments that are started with the phrase 'Applying the precautionary principle...".  The Precautionary Principle as it is popularly applied is a plea for perfect safety which is an unachievable goal.  The need for evidence of harm or potential harm is dismissed by a person using the aforementioned phrase in a discussion.  The demonstrable logical fallacies of the Precautionary Principle also mean arguments shored up with its statement are essentially without merit.

I've had some difficulty dredging up a formal statement of the Precautionary Principle, the best I could find is as follows:
If an action might cause severe or irreversible harm then the burden of proof falls on those that advocate taking the action.
This is a slight paraphrase of the statement of the principle found on Wikipedia. Wikipedia is not the most reliable resource for factual information, but I feel it is fairly representative of how most people on the internet perceive a fact or principle. In this post I'm commenting on the popular statement and popular application of the precautionary principle rather than the formal statement of it (which I had trouble finding in a cursory internet search).

In and of itself this statement seems somewhat innocuous until you go back and look closely at the underlined word. A baseless accusation of possible harm due to an action under consideration can be made and the burden of proof lies at the feet of the proponents of the action rather than at the feet of the accuser.  This strikes me as somewhat nonsensical as it is the legal equivalent of the presumption of guilt. A good analogy would be to ban the consumption of solid food because someone may choke to death. Sorry all your food must be pureed lest get something lodged in your throat, it might happen after all!  As you can see from my analogy there is no evaluation of probability associated with the Precautionary Principle. The Precautionary Principle doesn't care if the odds of choking are one in a billion or approaching unity you now have to eat pureed food for the rest of your life.

Looking at the Precautionary Principle from a formal logic perspective we can demonstrate that it is a logical fallacy. Specifically the Precautionary Principle can be considered to be a straightforward Burden of Proof Fallacy .  Burden of  Proof fallacy occurs when the wrong side is required provide evidence.  In the case of the Precautionary Principle as popularly stated the burden of proof is being placed solely on the advocate rather than also requiring the opponent to provide evidence of potential harm.

In this article which purports to lament the misuse of the Precautionary Principle the authors spell out the Burden of Proof Fallacy fairly clearly as follows:
The objectors must bring forward evidence that stands up to scrutiny, but they do not have to prove that there are serious dangers. It is for the innovators to establish beyond reasonable doubt that what they are proposing is safe. The burden of proof is on them.
The quote above is conflating two conditions to mislead the reader into accepting the Precautionary Principle as a reasonable and logical moral. In fact there is a further formal logical fallacy in the above quote known as Special Pleading. The author pleads a special case for the objector's burden of proof rather than applying the same standard of proof to the objector and the innovator. There is no relevant difference between the objector and the innovator, other than the appeal to the fear of harm due to the innovation which is yet another fallacy.

Let's consider an example, DDT.  Rachel Carson's Silent Spring brought an accusation against DDT, specifically that DDT was responsible for the thinning of eggshells, particularly that of raptors such as owls and falcons. The thinning of the eggshells was purported to negatively impinge on the affected birds' reproductive success and DDT was eventually banned worldwide.  Carson's book made an accusation but it wasn't peer reviewed and it was not a scientific publication.  It simply created a groundswell of public support by engaging in a massive appeal to fear without providing scientific proof of its accusations.

Well let's also consider an accusation by one of Silent Spring's detractors, that the ban of DDT is responsible of millions of malarial deaths worldwide every year.  DDT was one of the world most effective anti-malarial strategies prior to the banning of DDT in the 1970's1 .

Remember there is no requirement for proof if we apply the Precautionary Principle, the burden of proof in each case is on the people proposing action hence the lack of citations supporting either case.

Where does the Precautionary Principle lead us? In this case the Precautionary Principle argues for both action and inaction. On one hand we have bird's reproductive success and on the other we have millions of human deaths.  The Precautionary Principle because of how it's stated doesn't allow discrimination between the conflicting requirements, we must both stop using DDT to save eggshells and we must continue using DDT to prevent deaths from malaria. There's no middle ground and the Precautionary Principle provides no guidance on how to discriminate between risks. 

Risk management is the identification, assessment, and prioritization of risks followed by coordinated and economical application of resources to minimize, monitor, and control the probability and/or impact of unfortunate events2.
I consider the Precautionary Principle to be absolutely devoid of merit but without offering an alternative I'm just indulging myself in plaintively crying into my beer.  So consider the Risk Management as an alternative. Applying Risk Management to the DDT example, we assign values to the two identified risks and to the costs of mitigating the risks.  We can then decide between the lesser of two evils rather than remain stuck in the Mexican Standoff of the Precautionary Principle in the event of conflicting risks.  Part of the process of Risk Management must be the thorough and rigorous proof of both risk and beneficial outcome in order to correctly identify and assess the costs of risk and the costs of risk mitigation.

The last point I want to make is about the misuse of the Precautionary Principle in discourse between individuals.  In my experience entire well structured reasonable rational arguments are dismissed with a wave of the Precautionary Principle wand.  If you don't want to be rational about an issue then be open about the fact that emotion dictates how you form decisions and is an intrinsic part of the opinions you hold.  Don't get into a reasoned discourse and expect to use the Precautionary Principle to void arguments wholesale, it the equivalent of bringing a bouquet of flowers to a knife fight.

1 It is believed that [malaria] afflicts between 300 and 500 million every year, causing up to 2.7 million deaths, mainly among children under five years. Africa News, January 27, 1999]

2 Douglas Hubbard "The Failure of Risk Management: Why It's Broken and How to Fix It" pg. 46, John Wiley & Sons, 2009


  1. I think emotion, bias and personal preference/antipathy determine some part of everyone's world view.

    I further submit that the a priori assumption of rationalism as the best or the correct basis for making decisions or forming a world view remains an unproven underpinning of your final paragraph's thesis.

    I would be hard pressed to find anyone who is strictly rational in all aspects. So then, if an individual is not, what gives him or her the right to decide where rationalism should or should not apply and apply that standard to judging others?

    I agree that, in the absence of information, there can be no support for any statement nor plan of action where rationalism is your mechanism for decision making. Without information, no decision can be supported rationally.

    On the other hand, it seems to me that you are claiming special priviledge by putting rationalism out there as 'the' only correct way to come to a decision. I'd say this is an unproven thesis in general and in particular for any case you can cite. That thesis would need proven in any case of interest before I would agree you should be criticizing non-rational decision making approaches.

    Ultimately, by inclination and by profession, you are a rationalist. That could be considered a bias unto itself. There is nothing wrong with being a rationalist or even with rational decision making steps. But there might be other methods and views that are as valid and I think you'd need to debate the merits of method and approach in each particular case or win a much larger debate on rationalism in all cases as the correct or valid approach. Good luck with that!

    I'm confident, but lazy, so we'll call my last point here unsupported but I'll make it anyway: I'm fairly certain we can point to plenty of cases where non-rational decision making has lead to unpleasant consequences (perhaps predictable, perhaps not). I'm also fairly certain that strictly (allegedly) rational decision making has, on a fair few occasions, lead to unpleasant and perhaps unforseen consequences.

    Rationalism reminds me somewhat of democracy as seen by Winston Churchill: "Rationalism is the worst form of decision making, except for all the rest." (that is to say, imperfect, but perhaps the best of a bad lot... but you can't forget the imperfect part)

  2. My reply is too long so I'm taking it to email.


Polite and erudite comments by