The Myth of the Root Cause

20140721-073939-27579228.jpg

Advertisements

About stevenshorrock

I am a systems ergonomist/human factors specialist and work psychologist with a background in practice and research in safety-critical industries. My main interest is human and system behaviour in the context of safety-related organisations. I seek to enable improvement via a combination of systems thinking, design thinking and humanistic thinking. I am a Chartered Ergonomist and Human Factors Specialist with the CIEHF and a Chartered Psychologist with the British Psychological Society. I currently work as a human factors and safety specialist in air traffic control in Europe. I am also an Adjunct Senior Lecturer at the University of New South Wales, School of Aviation. I blog in a personal capacity. Views expressed here are mine and not those of any affiliated organisation, unless stated otherwise. You can find me on twitter at @stevenshorrock
Image | This entry was posted in Safety, systems thinking and tagged , . Bookmark the permalink.

11 Responses to The Myth of the Root Cause

  1. charles colbert says:

    There was no written text concerning the myth of root cause, no matter what link I clicked on.

    • That’s because a picture tells a thousand words…

      • charles colbert says:

        Unless you read the current safety thinkers, you’re not going to know this RCA is a myth. Thought it would be worth a write up, since it so fundamental and has been traditionally used for so long. Dr. Hollnagel’s article and now his new book have a section on this. Although, the quasi-scientific community may say this is hand waving by social scientists.

        It’s amazing what we all thought was true is unfounded in any stringent argument, i.e., lots of items to deconstruct by social constructionism. A quick perusal up on the website for safety consultants finds most of them based in the safety myths–say us.

  2. Bill Wilson says:

    Hmm. I read a lot of the current safety thinkers, as well as some of the classics… and I see this attitude of “no such thing as a root cause” as being a bit too eager to discard old for new, even when some of the old, established ideas still have some value. I contend that the concept of “root cause” is one of those still-valuable ideas, even for the analysis of complex systems. Maybe I have to discard traditional notions of cause-effect in some cases. Perhaps I need to change some of the details in my definition of “root cause”. It may be that I’ll have to use a different set of tools, methods — even mental models — to gain the insights needed to propose meaningful improvement actions. Any “root cause” (or causes) that I identify might be stated in completely different language and involve factors that are completely foreign to traditional root cause analysis.

    Even with all those modifications, the concept of a remote cause (set of causes) that may be fundamental to a problem, or a whole class of problems, and which identifies a point of leverage from which improvements (in reliability, resiliency, safety) can be driven, still has merit. Why throw out the baby with the bathwater?

    • The problem is first one of ontology. To understand causation in safety, it is necessary to understand causation more fundamentally (this is a good into: http://www.amazon.com/Causation-Very-Short-Introduction-Introductions/dp/019968443X). So one problem concerns the essential nature of causation, and on this ground not only does the idea of a root cause crumble, but the notion of a cause is also an ontological quagmire. On a more practical level, the notion leads down a path which ignores complexity and emergence, causal loops, and often even multiple causes, in favour of linear cause-effect sequences. It is of course an efficiency-thoroughness trade-off by the analyst, but hidden behind an illusion, that going ‘down and in’ will get you to the ‘real cause’, which is actually a social construction. The starting point should actually be to go up and out to how the system really works. Still, root cause analysis, even domino theory and loss control model, is much better than the identification of superficial primary/proximal causes. It is not a huge step, though, to use systems thinking methods that will give a richer picture. (So I guess I don’t really see a baby in the bathwater!)

      • Bill Wilson says:

        Thanks for the reference; I’ll check it out. I have researched causality fairly thoroughly over the years; one reference that I liked a lot was Judea Pearl’s book (http://www.amazon.com/Causality-Reasoning-Inference-Judea-Pearl/dp/052189560X). I also like the following exploration of causal models from the epidemiologist’s perspective (https://www.kent.ac.uk/secl/philosophy/jw/2008/cits/Parascandola%20Weed%2001%20-%20Causation%20in%20epidemiology.pdf).

        Now, I have seen this conjecture before, that the notion of root cause crumbles due to complexity, emergence, causal loops, etc etc. However, I’ve never seen this statement proven, or even convincingly justified in abstract… maybe if you only consider ancient ideas — domino theory? really? — then the notion of “root cause breakdown” might be valid. However, why assume that *all* root cause analysis practitioners are stuck in the 19th century?

        Just to clarify… I do consider causal loops, feedback mechanisms, and the potential for emergence whenever I see a system that appears to be on the verge of passing from “merely complicated” to “characterized by complex behaviour”. In order to do that, I often have to go up and out, in and down, left and right, etc. There is no single way to do this, there is no master toolset, there is no fixed liturgy that determines how I shall conduct an investigation. I’ll happily use any tool I can find that will give me additional insight into how a system operates. I am always after that “richer picture”.

        I just don’t see how or why the concept of root cause is considered to be diametrically opposed to systems thinking. Don’t make the mistake of taking the analogy of a “root” cause too seriously. In systems, “root” just means “fundamental”, “important”, “performance shaping”, “high leverage”.

      • Well it all rather depends by what you mean when you talk about root cause, but much is defined by the techniques people use in practice (and which can be sold to generate income). On your site you mention that “(RCA) it is probably the only way to find the core issues contributing to your toughest problems”. There is of course a long conversation there…but I think you probably use a variety of approaches to fit the situation and system. If you are working with a mechanical system, then the concept and methods associated with ‘root cause’ might be of particular value. In human and sociotechnical systems it is rather different, but may be have be used/adapted to understand interactions between system levers and any emergent patterns. Your stance appeears to be pragmatic, which is fine; then it is about understanding the trade-offs that one is making, how and why.

  3. Bill Wilson says:

    Perhaps we are in general agreement about most things; I suspect we are, given what I’ve read on your blog (excellent, by the way). The main difference may simply be that I continue to use the term “root cause”, while you do not (if you ever did).

    For me, the definition of “root cause” stopped being “original cause for a non-conformance, that, if removed, prevents recurrence” about 15 years ago. My current, preferred definition is the following.

    /* A problem’s root causes are the fundamental features of the involved system(s) which created unresolved contradictions, conflicts, or control issues that significantly increased the likelihood or magnified the impact of the adverse consequence(s). */

    Admittedly, I haven’t convinced too many people to buy into that sort of definition… but I’m trying. 🙂

    Thanks for the thoughtful conversation!

    • That is a useful adaptation which again is pragmatic, which is usually necessary to compromise and progress. I have a similar approach, e.g. for incident, where wholesale rejection lf existing approaches would not work practically, but an adaptation and compromise does. In the end, it is a case of choosing a mindset for a situation or need, but not getting taken hostage by it.

    • And thanks to you also Bill.

  4. Cliff Norman says:

    Thanks Bill for the reference to causal models from the epidemiologist’s perspective, very helpful! Cliff

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s