Justin SheehyDec 2

I’d go much farther than https://hachyderm.io/@nygren/115576621423960811 and replace “isn't necessarily useful” with “is actively harmful” though I sadly know of far too many people and places that still talk this way.

Quite simply: “root cause” is not a good term to use in incident analysis, use of that term should be completely removed, and better choices exist.

Erik Nygren :verified: (@[email protected])

Attached: 1 image This article and its headline is a great example of where "root cause" isn't necessarily useful in incident reviews, or even as something to focus on in hazard mitigations: https://www.cnn.com/2025/11/18/us/ntsb-dali-bridge-collapse While it might be tempting to place "misplaced wire label" as the "root cause" for why a container ship crashed into a bridge and killed people, I suspect that focusing on ensuring wire labels aren't misplaced isn't anywhere near the top of the list for preventing future catastrophes. The more interesting things are only alluded to (eg, if a system had no backup because it was only intended to be an infrequent backup itself but failed after running for months, where were the sensors and alerts to indicate that the system was in a hazardous state). #safety #incidentreviews

Hachyderm.io
Show threadJustin SheehyDec 2
Multiple brilliant folks such as (but not limited to) @allspaw and @gallego have written and spoken deeply and extensively on this over the years. I am occasionally discouraged by seeing that this is still at all controversial.
Show threadd@nny disc@ mc²Dec 2
@justinsheehy @allspaw @gallego i am a fan of @norootcause's work
Show threadJustin SheehyDec 2

@hipsterelectron @allspaw @gallego @norootcause

Absolutely! There is no shortage of folks who have thoughtfully approached this problem and he is one for sure. I just mentioned the first two folks who came to my mind at that moment. And for those truly wishing to go into this, I'd point at Dekker, Leveson, etc for the heavy parts.

Show threadd@nny disc@ mc²Dec 2
@justinsheehy @allspaw @gallego @norootcause i followed everyone you mentioned too just thought it might help to make the point to bystanders that some people even identify the rejection of the root cause hypothesis as central to their professional identity
Show threadd@nny disc@ mc²Dec 2
@justinsheehy @allspaw @gallego @norootcause i don't know so much how others conceptualize it but i find the rejection of root cause analysis itself to form an axiomatic foundation with sufficient explanatory power to extrapolate to further inferences. which is to say that it not only identifies a fallacy, but also that the proclivity to accept root cause analysis in itself indicates its use as a bandage for structural weaknesses which led to the incident in question. this is something i think even individual engineers can and should be able to grasp and make use of
Show threadWill GallegoDec 2

@hipsterelectron @justinsheehy @allspaw @norootcause

If you're at all interrested, the three of us (Allspaw, Lorin, and I) are in a community of like minded folks helping to promote these ideas - resilienceinsoftware.org

Show threadLorin Hochstein Dec 3
@gallego @hipsterelectron @justinsheehy @allspaw It's a great community! Come join us in pushing these ideas forward!
Show threadThomas Depierre
@norootcause @gallego @hipsterelectron @justinsheehy @allspaw can confirm
Dec 3 at 5:52amWeb
Show threadstudio craque 54 🏳️‍🌈Dec 3
@Di4na @norootcause @gallego @hipsterelectron @justinsheehy @allspaw Ahoy hoy!