Bedford and the Normalization of Deviance(rapp.org)
rapp.org
Bedford and the Normalization of Deviance
http://www.rapp.org/archives/2015/12/normalization-of-deviance/
10 comments
Well, out of context, in the abstract, tediously checking every single condition with a list may seem "stupid" to someone experienced in an activity.
In the context of flying, where a careless error can result in death, this tedious checking has a very logical explanation, one that I assume most pilots are aware of.
More broadly, I suspect that a lot of rules exist specifically because what seems obvious and natural in a situation gets you in trouble. But by that token, such rules can't help but appear "stupid" but also can't and shouldn't be avoided.
In the context of flying, where a careless error can result in death, this tedious checking has a very logical explanation, one that I assume most pilots are aware of.
More broadly, I suspect that a lot of rules exist specifically because what seems obvious and natural in a situation gets you in trouble. But by that token, such rules can't help but appear "stupid" but also can't and shouldn't be avoided.
I think it's also why there's real value in investing time to explain the rules and why they're important
and understanding that even over just 2 layers of bureaucracy its really easy to lose an understanding of the why behind a process
---------
Part of that is that its quicker to just rely on the institutional power
"I'm your boss, do it this way because I tell you to"
"My boss told me it needs to be this way, and I'm your boss, and the bs flows downstream, so do it this way"
uses fewer institutional resources, but that comes at the cost of no one actually knowing why things are being done the way they're being done,
and no one understands the costs of deviance from the processes being implimented
------
and in lots of instances, the benefits to deviance tend to be higher than 'we get to take off a little faster'
lots of times what we measure people on tends to be in conflict with the processes we want people to follow, so we even incentivize people inventing little shortcuts around the processes we want people to follow
and understanding that even over just 2 layers of bureaucracy its really easy to lose an understanding of the why behind a process
---------
Part of that is that its quicker to just rely on the institutional power
"I'm your boss, do it this way because I tell you to"
"My boss told me it needs to be this way, and I'm your boss, and the bs flows downstream, so do it this way"
uses fewer institutional resources, but that comes at the cost of no one actually knowing why things are being done the way they're being done,
and no one understands the costs of deviance from the processes being implimented
------
and in lots of instances, the benefits to deviance tend to be higher than 'we get to take off a little faster'
lots of times what we measure people on tends to be in conflict with the processes we want people to follow, so we even incentivize people inventing little shortcuts around the processes we want people to follow
Does it also mean you should try your hardest to obey stupid rules? Since you may not realize whether it's stupid or not until you have more experience than the person who made it?
If you think the world doesn't contain a large number of people whose attitude is that they are smart enough to decide what to do for themselves, and all the rules are made by stupid people, you have another think coming.
And blaming the rules for the behavior of those people is not the correct position.
(see also: why the only thing many HN users know about G.K. Chesterton is the fence thing)
And blaming the rules for the behavior of those people is not the correct position.
(see also: why the only thing many HN users know about G.K. Chesterton is the fence thing)
I'm sorry you read that into my comment, but that's not what I meant to imply. I simply meant that there's a balance to keep in mind when making rules.
Rules are organizational code. Be prepared to fix them when you find out they don’t work as well as you thought.
> Example: the boss in on board and we can’t sit here for several minutes running checklists;
If your boss does not notice that the aircraft stood idle for too short a time, and he doesn't fire you straight away, than the culture of safety did not permeate your company enough.
If your boss does not notice that the aircraft stood idle for too short a time, and he doesn't fire you straight away, than the culture of safety did not permeate your company enough.
Does it really matter if you know you operate in an environment where deviance is normalized? To me, it seems like realizing that you are in a cult. You aren't going to change the cult behavior because you aren't going to change human nature.
At the very least you can avoid joining the cult, and if enough people understand what's happening, they can do something about it.
Atul Gawande's appointment yesterday continues to echo around HN.
Corporate aviation is a very different place than Part 121.
It is cool to see the author point out that the Part 121 world has largely insulated itself from the "corrosion" of the normalization of deviance. (And that acknowledgement coming from people like us who chafe under the structure and rigidity of 121, but still give a shit about actual risk mitigation.)
The topic is interesting since the "dark ages" mentality of so much of corporate aviation is worth exploring as what stunts its growth, innovation, etc. (And I'd extend that to Part 135 too.)
It is cool to see the author point out that the Part 121 world has largely insulated itself from the "corrosion" of the normalization of deviance. (And that acknowledgement coming from people like us who chafe under the structure and rigidity of 121, but still give a shit about actual risk mitigation.)
The topic is interesting since the "dark ages" mentality of so much of corporate aviation is worth exploring as what stunts its growth, innovation, etc. (And I'd extend that to Part 135 too.)
One of the most important talks I've ever seen is Richard Cook's "Resilience In Complex Adaptive Systems"[1], in which he explains the fundamental abstract mechanism that causes the normalization of deviance.
[1] https://www.youtube.com/watch?v=PGLYEDpNu60
[1] https://www.youtube.com/watch?v=PGLYEDpNu60
Related article by Richard Cook: http://web.mit.edu/2.75/resources/random/How%20Complex%20Sys...
that was interesting however at the end he says all their papers are at ctlab.org, which is now a parked domain...
Nice startup idea in the article for anyone interested:
>A better model would be that of the FOQA program, where information from flight data recorders is downloaded and analyzed periodically in a no-hazard environment. The pilots, the company, and the FAA each get something valuable. It’s less stick, more carrot. I would also add that this sort of program is in keeping with the Fed’s recent emphasis on compliance over enforcement action.
>A better model would be that of the FOQA program, where information from flight data recorders is downloaded and analyzed periodically in a no-hazard environment. The pilots, the company, and the FAA each get something valuable. It’s less stick, more carrot. I would also add that this sort of program is in keeping with the Fed’s recent emphasis on compliance over enforcement action.
What I don’t understand is why the gust lock mechanism is gated on not be ingengaged vs being disengaged. The report goes into detail about having the lock in an intermediate state, but if you actually made it so that the throttle was gated at X% unless the lock was completely disengaged (eg mechanically release the throttle lever only when the lock was latched into the disengaged position)
IS there some mechanical/aircraft-related reason that such a lock isn’t possible?
IS there some mechanical/aircraft-related reason that such a lock isn’t possible?
Well, think about it this way: airplanes don't have differential gear boxes that transmit power to the drive train in contact to the ground. In other words, power on the ground is not delivered by the wheels.
This means the engines that deliver the thrust in the air, are also doing the same work on the ground. The traction is delivered to the air, and the rolling is all relative to to the thrust.
The manner of control for the power of the thrust isn't based on throttle alone. There are a handful of factors that tie together, to generate force.
Just like stepping on the gas, when a car is in neutral, you can spin up the turbines or propellers, without directing the air in such a way that moves the aircraft. The pitch of the blades matters. Also, with multiple engines, thrust must be synchronized, or the effect is a turn, and one engine could push the plane in circles, if controlled improperly.
So, it's not like leaving the emergency brake engaged in a car. There are other variables at work.
I think it's also a factor, that often enough, with complex machines created to solve for a complex problem domain, leaving configurable items as independent as possible prevents more footgun scenarios than it creates. When quick thinking is needed in an emergency, to deal with real world forces in play, systems are best designed so that in a proverbial sense, you can quickly throw yourself at a problem, when stopping and thinking about a strategy might take too long.
This is not to say that stopping and thinking carefully about strategy should be skipped, but that when trouble finds you, despite best efforts, and seconds count, having the ability to skip checklists and improvise in the heat of the moment does represent a desireable pathway for saving throws.
This means the engines that deliver the thrust in the air, are also doing the same work on the ground. The traction is delivered to the air, and the rolling is all relative to to the thrust.
The manner of control for the power of the thrust isn't based on throttle alone. There are a handful of factors that tie together, to generate force.
Just like stepping on the gas, when a car is in neutral, you can spin up the turbines or propellers, without directing the air in such a way that moves the aircraft. The pitch of the blades matters. Also, with multiple engines, thrust must be synchronized, or the effect is a turn, and one engine could push the plane in circles, if controlled improperly.
So, it's not like leaving the emergency brake engaged in a car. There are other variables at work.
I think it's also a factor, that often enough, with complex machines created to solve for a complex problem domain, leaving configurable items as independent as possible prevents more footgun scenarios than it creates. When quick thinking is needed in an emergency, to deal with real world forces in play, systems are best designed so that in a proverbial sense, you can quickly throw yourself at a problem, when stopping and thinking about a strategy might take too long.
This is not to say that stopping and thinking carefully about strategy should be skipped, but that when trouble finds you, despite best efforts, and seconds count, having the ability to skip checklists and improvise in the heat of the moment does represent a desireable pathway for saving throws.
This is interesting in that here "the normalization of deviance" describes situations where behaviors that can be measured as objectively bad still wind-up normalized through implicit group consensus.
That such deviance still happens suggests strongly that deviance around things that are less clear and less likely to result in harm to the professional would be even more likely to occur. The simplest example is anyone who faces the public and is expected to give full service. It's easy for that to breakdown when it isn't convenient. One can see this today in everyone from bureaucrats to police officers.
That such deviance still happens suggests strongly that deviance around things that are less clear and less likely to result in harm to the professional would be even more likely to occur. The simplest example is anyone who faces the public and is expected to give full service. It's easy for that to breakdown when it isn't convenient. One can see this today in everyone from bureaucrats to police officers.
> First and foremost is the attitude that rules are stupid and/or inefficient.
This is a great reason why to try your hardest to avoid implementing stupid rules. When some rules are stupid and become ignored, other rules lose their power.