Originally posted on March 10, 2017 @ 4:05 PM
Abduction in Risk and Safety
We all should know about induction and deduction but what do we know about abduction? Safety has a long tradition of deduction (not always that good), some reasonable history in induction (albeit, the wrong rite of passage to the wrong mentalitie) and virtually no one in safety speaks of abduction (the expectation and logic of expecting a surprise). When one is preoccupied with a discourse of controls, certainty, technology and objects, why would one accept the certainty of surprise? Why would one expect something alien to emerge from a system that has been created to manage risk?
Abductive reasoning (https://en.wikipedia.org/wiki/Abductive_reasoning) acknowledges the finitude of systems, controls and human perceptions. Abductive reasoning acknowledges mystery in the machine. It is why organisations, to use Dekker’s phase appear to ‘drift into failure’. In reality no organization drifts into failure, they never had ‘arrived’. The concept of completeness, perfection and infallibility is a concoction of the safety mindset, fixed on absolutes such as zero. This is why Weick’s work is so important, expecting the unexpected is an indicator of risk intelligence (Evans).
One of the worst mindsets that can get hold of the safety person is that everything is under control, everything is comfortable. This is the recipe for hubris, one of the greatest risks to safety. The moment all is comfortable, the moment we thing we have got that matrix colour down from red to green, we can stop thinking, stop interrogating the situation, all will go well. We need a dose of abductive reasoning. So, one makes a hypothesis but doesn’t ‘deduce’ or ‘infer’ that the outcome is predictable, one needs to explain why it is unpredictable.
The idea of abduction comes from Charles Pierce, one of the founders of semiotics (https://en.wikipedia.org/wiki/Charles_Sanders_Peirce ).
Recently I have learned that people are being seduced by a nonsense called ‘predictive analytics’. The idea is that one can predict what a human will do if one can collect enough data on the human, including video data (https://en.wikipedia.org/wiki/Predictive_analytics ). This has led to building sites overseas with hundred of cameras (big brother) on site collecting data in the naïve idea that behaviours are predictive and provide insight into what will happen next. Trouble is, humans are not machines and behaviourism is a nonsense. Humans are not the sum of inputs and outputs. Still people buy this crap amid promises to reduce injuries down to zero. Ah, the seduction of the absolute, lets buy it. As long as the discourse is safety we can rob any human of what we want.
Predictive analytics is from the same faith as ‘big data’ and ‘machine learning’ (sic, no machine can ‘learn’). This faith-belief in big data is behind the pain and suffering of thousands of Australians disturbed by the robo-debt debacle by Centrelink. How fantastic to attack the weak and vulnerable with fear, distress and cruelty in the name of ‘big-data ideology’ (http://www.smh.com.au/federal-politics/political-news/centrelinks-robodebt-creating-a-climate-of-fear-20170307-gut1z7.html ). We see in this debacle the ideology of predictive analytics. When will people begin to understand how humans make judgments and decisions? When will people begin to understand the social psychology of decision making? Instead, lets make some dollars out of promises in predictability and cover up the mess when we psychology injure tens of thousands of vulnerable Australians. Cool.
There is one thing for sure, humans cannot be predicted. Any ‘clap trap’ about ‘prediction’ is a religious belief in denial of fallibility, randomness and the reality of the world. You can read more about the religion here:
http://ehstoday.com/safety/advanced-and-predictive-analytics-safety-are-they-worth-investment
https://ohsonline.com/~/media/F706D1194701451FADFD66D80C1E06B4.pdf
For a moment, let’s apply some critical thinking, just interrogate the language and the power behind the language. The guarantee is to ‘predict’ behaviours (and hence the unconscious and thinking) and then only the claim to ‘reduce’ injuries. Huh? If you can predict what humans will do, then go for zero!!! Ah, the promise of god, is alive and well in the safety religion.
Do you have any thoughts? Please share them below