On this episode of the Safety of Work podcast, we discuss which questions need to be asked, when trying to constructively resolve an argument about safety theory.
You may find that this conversation makes us look hypocritical, when discussing how to argue. We just want the best outcomes for all discussions regarding safety. Ultimately, this topic came out of some recent intense arguments in the safety field and we wanted to address how to constructively handle disagreements in person and online.
Topics:
Quotes:
“Unless an entire field is genuinely pseudoscience, it’s always very, very dangerous to dismiss an entire field.”
“A lot of the time that there is a disagreement with Safety II, it’s basically people saying we shouldn’t be throwing out everything to do with Safety I. What we actually need is some kind of middle-ground between Safety I and Safety II.”
“A lot of the time, when we argue, there’s an implicit assumption that we can only agree with one of the theories, because they’re somehow mutually exclusive.”
Resources:
David: You’re listening to the Safety of Work podcast episode 67. Today, we’re asking the question, what questions should you ask to constructively resolve an argument about safety theory? Let’s get started.
Hi, everybody. My name’s David Provan. I’m here with Drew Rae and we’re from the Safety Science Innovation Lab at Griffith University. Welcome to the Safety of Work podcast. In each episode, we ask an important question in relation to the safety of work or the work of safety, and we examine the evidence surrounding it.
For this week, we’re doing something a little bit different. Hopefully, we’re providing an answer to a question that will help you as a listener, as a consumer of safety research, or as a practitioner, to understand and take forward ideas in a constructive way, and contribute to the further development of those ideas.
Drew, what’s today’s question and why is it an important one that we answer on the podcast?
Drew: I fear that today’s question might be how hypocritical can Drew and David be when we talk about arguing on the Internet. Our question comes out of some thinking that we’ve been doing recently in response to some pretty vehement debates that tend to go on in social media about safety, and what can we as a community do to improve the quality of the conversations we have, so they take us towards a constructive place instead of disagreeing with each other.
As we did with the manifesto, these are as much may be reflections for ourselves as reflections for you, thinking about what would we like to be doing ourselves and what do we reckon that other people could do along with us, to try to be a bit more productive when we get into arguments on the Internet or elsewhere about the right way to think about safety.
There are a couple of premises that we’ve established through our discussions that we try to embody on the podcast. The first one is we think that most of the disagreements disappear once people understand what the disagreement actually is. Often, we argue for the sake of arguing and argue because other people disagree with us without being really clear about what we are standing for. We think, actually, if there is a genuine disagreement, then getting rid of some of that unnecessary arguing and focusing on what the genuine useful point of disagreement is, is what’s going to help us move forward.
David: The structure of the podcast today is we’re going to go through a series of questions that you can ask yourself or ask other people—in fact, openly ask people who you disagree with—to try to turn that disagreement about safety and process of safety into a constructive discussion that is formed by both theory and evidence.
Instead of saying I believe this and you believe that, therefore we can’t be friends anymore, what we’re trying to do in (at least) this episode is to say, that’s really interesting that you believe that. Now, let me understand why, how, and what is different from what I think is my truth, where are the gaps in both of our understanding, and how we can move forward.
You may never get everyone to agree on everything, but like you said, knowing exactly what we disagree about allows you to turn to the evidence, turn to the research, turn to your own practice, and try to learn something new.
Drew: Should we look into the questions?
David: Let’s do that.
Drew: The first one we’ve got here is just asking yourself, what is the original source? Like all questions, we don’t want people to turn this into a weapon. We don’t think the instant response to an argument is give me a citation, give me evidence you can point to. This is taking a very humble understanding that we all take short cuts; all of us. No one can read everything in its original form. A lot of what we know isn’t from the original sources, it’s from other people telling us about it. That’s fair enough. We need to understand that. We all believe lots of things about things that we’ve never actually read ourselves.
But if we’re going to move forward, it’s important that we sometimes read stuff in its original form. That matters more and more the more we want to take part in an argument. If you want to actually argue for a position, then it becomes more incumbent on you to start reading those original sources.
It’s good for everyone to consume that balanced diet of some summaries of stuff and some original sources. Summaries to know broadly what’s going on, original sources to gain depth on things that are relevant. Just like to find out what’s going on in the world, sometimes we skim our Twitter feed and sometimes we read long-form articles. We need to know broadly what’s going on and in-depth about the things that we care about, that we want to talk about.
David: I like the way, particularly, if you’ve got a position on something. Or at either end of the thing if you strongly agree or disagree with something, that’s an opportunity to go deep. And just make sure that when you know those positions there, that you’ve got a good understanding of the original ideas and what the theory or the approaches is based on.
Drew: Under that, I think there are a couple of more black and white principles we can try to follow. One of them is we shouldn’t accept descriptions about things coming from opponents of those things. That’s a really easy habit to get into because none of us likes reading stuff that we disagree with, but when we read commentary on stuff that’s constantly from people who agree with us, we get a very unfair and uncharitable interpretation of what the alternate point of view is. Particularly, if you want to join in the criticism of something, that’s when it’s really important to read the original thing that you’re criticizing.
David: So maybe don’t form your own view and argue if you’re on scientific management or behavior-based safety based on reading Sidney Dekker’s work.
Drew: Yeah, and in the other direction I have been tempted at times to put up a post on LinkedIn, Nine Signs That Someone Criticizing Safety II Has Never Read the Book. I’m wary of both sides, but I think in safety arguments very often both sides have not read the original sources that the other side is coming from.
Another useful thing is because we can’t read everything in-depth, keep track of people that you trust as reliable interpreters of original sources. When you do go and read the original (sometimes), keep track of who else on your side has given a fair summary and who has given an unfair summary. Learn to trust even people on your own side of an argument may not be fair interpreters, and some people are actually quite good at being balanced and fair as interpreters of sources.
Keep in mind that there may be more than one original source. Even just a single author can write multiple things, and it’s tempting to use their name as a shorthand for a whole body of ideas encompassed by them and other people. Be aware that the very original source of the idea may not be the most up-to-date version. The person who named an idea might not be the person who started that idea. This makes it harder to find the original sources, but it is something that we need to be careful of.
David: A few examples in this area, just for people to think about the original sources, an idea about resilience engineering, for example. It’s important to be clear when we’re talking about a whole field of study or a single idea. Sometimes, when we’re talking about a field of study we’re looking at sources that are more books where you get chapters written by different contributors or contributing a particular idea or a particular theory within that field of study.
You can get a flavor of the field by reading one or two of these edited books, but then you can’t really make this statement like I agree or disagree with resilience engineering. You’ve actually got to start talking about the individual ideas that sit within that particular field.
Drew: Yeah, unless an entire field is genuinely pseudo-science. It’s always very, very dangerous to dismiss an entire field since we tend to do that based on the worst representations of that field.
David: I think we do that a little bit with safety culture, don’t we?
Drew: I think that’s a fair accusation. One upside safety people might be familiar with is the way some people talk about evolutionary psychology. They pick the worst examples of things claiming to be evolutionary psychology, that even the reputable scholars in that field don’t even think belong to the field. As a result, they dismiss that entire field of study instead of listening to the best of what it has to say.
If you sort of different patterns to look for, sometimes when a single author is associated with a single piece of work, where it really is worth going back and reading that one—these are the classics in safety—if you’re interested in normal accidents, most scholarship by normal accidents is done by the original author, Charles Perrow. Most of what Charles Perrow has to say about it is in a single book called Normal Accidents. You want to talk about normal accidents, you read that book, you are ready to comment on it.
You get other authors who have very long careers, and over those careers they’ve got very focused research programs. Good examples of these are Nancy Leveson at MIT, and Tim Kelly originally of University of York. Those people do have solid chunks, but if you want to understand what they say you’ve got to update your knowledge with what they’ve written recently and get the most of the old version of that work.
Leveson has fairly recently put out a book, Engineering a Safer World, which is her most up-to-date all the ideas encapsulated in one place. But you’ll want to be careful in 10 years’ time. You can’t just go back and assume that Engineering a Safer World is the latest version of those ideas. Kelly went the other direction. Most of his ideas came out in his thesis, but then he steadily—paper by paper—built on those ideas. If you just go back to the thesis, you’re missing out on the best version.
There are other people that are harder to interpret because they’re in this constant dialogue with a community of other thinkers. The ideas in their work are bigger than any of the individual authors. You ought to be careful; are you talking about the conversation or are you talking about the single work?
A couple of authors like that are Hollnagel and Wyke. Hollnagel doesn’t own the idea of Safety-II. That draws on a long tradition and conversation of competing ideas in organizational theory. You can say, do you like the book Safety-I and Safety-II? Do you agree or disagree with what he says in the book? Even then, you might need to zoom in on the particular statement or particular chapter because he conflicts himself a bit. He doesn’t own that whole family of ideas in the same way that Leveson owns stamps or Kelly owns goal structuring notation.
And then we have these other authors who just dabble all over the place. Andrew Hopkins is a good example there. His journal papers are very different from his books. You can like both, you can like just the books and other papers, or you can like the papers but not the books. You got to take each idea on its merits because he has so many different ideas.
David: The first question we asked where we encounter a disagreement with our own ideas is what’s the original source, and checking the source around our own ideas and checking the sources around the opposing ideas.
The second question we need to ask which is obvious once we’ve located the source is what does the original source say? Drew, what would you be looking for here?
Drew: The important thing here is that all of the big ideas in safety are part of a conversation. This conversation has been going on since around the early 20th century. Obviously, safety existed before then and some of these ideas existed before then. But that’s when we formed societies and journals, and started having conversations about the ideas.
When you look at each work, you need to read it in the context of that conversation. Things you can ask are, what were the industrial conditions at the time? In particular, what recent accidents that happened that they might be directly or indirectly commenting on? What were the prevailing ideas, attitudes, and practices that they were commenting against the background that they assumed everyone knew or tended to be doing? Basically you’re like, what was the author responding to? They’re probably positioning themselves against trying to change a way of thinking. You have to understand what that way of thinking is that they’re trying to change.
David: I think the big one in the safety process is just the commentary that we make around (say) Frederick Taylor’s work on Scientific Management or Heinrich’s work in the 30s. I suppose we interpret those ideas or what the sources say in our current context in 2021 sitting inside our organizations. That’s not what they are responding to in their conversation, that’s not the circumstances that they faced. When we put ourselves in the shoes of the idea, context, and time then it gives us a better understanding of what was actually underneath what they were trying to propose.
Drew: One of the interesting things about both of them, particularly Heinrich because we’ve got the trace of older versions of his book, is that each time he wrote a new version he wasn’t just updating his ideas. He was responding to things that people had said in response to the earlier versions. How they’ve misinterpreted it, how they’ve applied it, so you got to see that on-going conversation rather than just take his work, plunk it in the modern day, and ask whether you agree or disagree.
David: Another example from our recent episode, which is Reason and Swiss cheese. We had almost a decade of conversation between Reason and other people around the Swiss Cheese Model.
Drew: Two particular recent examples are arguments that people have had and why we think this matters from current arguments. A lot of our readers might be aware that Nancy Leveson self-published recently a critique of Hollnagel’s Safety-I and Safety-II. It was called Safety-III: A Systems Approach to Safety and Resilience.
You can read that just on its own, but to understand where it’s coming from, you’ve got to know that Leveson and Hollnagel have for decades been part of the same general movement arguing for a more complex understanding of accident causation. They both call it a systems approach, they’re both responding in the same way to what they see as people over simplifying accident causation.
But then, in the book Safety-I and Safety-II, Hollnagel talked about Leveson’s work in a pretty negative way, essentially putting her work alongside the work that Leveson had been criticizing. You can read that as a dismissal of Leveson’s ideas and her life’s work by saying that Leveson is basically the same as the people she is criticizing; not a particularly fair thing to do. Leveson wrote a response which is a not unreasonable complaint when Hollnagel talks about the status quo. He’s lumping too many things together as Safety-I.
It’s not really appropriate or fair to read what Leveson said without understanding that conversation. Understanding that actually most of what Hollnagel says Leveson agrees with. It’s just what she disagrees with is not his proposal for the Safety-II. What she disagrees with is his characterization of what is essentially her work.
Your comment, David?
David: I’m just putting myself in the shoes of the listeners around this is starting to be a difficult proposition to understand these ideas because I got to find and read the source, and now that’s part of a broader conversation. I’ve got to think about what other sources might be out there before this particular piece is written.
It is hard and that’s why I think we’re probably a little bit frustrated when we see these really emotional, simplified disagreements. This is actually hard to go back to sources and understand the context, understand what’s come before because these people who are generally the sources are people that spend every day reading and thinking about these things.
Drew: It might sound like we’re saying, before you’re allowed to even take part in this debate there’s a heck of a lot of work you got to do. To a certain extent, that is exactly the conclusion to draw from this. It doesn’t mean that it’s an exclusive club, but we’ll talk a bit later on about how you can usefully join in the debate. If you want to argue at that level, then you do actually have to do the work to understand everything that’s been said so far. Otherwise, you’re that person who’s jumping into a conversation halfway, injecting their opinion without understanding what the conversation’s really about.
David: And I think knowing your network and having a network around. I suppose that’s a thing that LinkedIn is good for. It’s a network of sorts. An example we had I think on episode 17, I spoke to Carsten Busch about Heinrich, who’d done a thesis on Heinrich’s work. He had just written and published a book on Heinrich’s work, so I know from a trusted source. I know that if I’ve got a question about Heinrich, I can go to Carsten, for example, without having to start finding 1931 books and trying to understand that for myself.
Drew: That’s a good tip. Another thing is consider the overall intent of the work rather than taking it too literally. As an academic myself, I really hate saying you shouldn’t take academics literally. The fact is, there is a whole tradition of scholarship that links together rhetoric and metaphor, along with little acclaims. A lot of safety comes out of that tradition. The result is that you really do need to do a bit of reading between the lines and say, yeah they said that, but is that really the point of what they were saying?
A good example that’s definitely the case with Heinrich and his discussion of ratios. It is literally true that Heinrich did give these precise ratios and did make claims that these precise ratios were true, but that was never the main point that he was trying to make. The main point he was trying to make was this relationship between accidents have causes that go back in time, and we’ve got to choose the right point in time that is not so far back that it’s meaningless, and not so close to the accident that the accident already happened.
What a lot of people do is they pick on those precise ratios—medical studies—testing whether those were the correct ratios or not, rather than engaging with do I agree or disagree with Hollnagel’s overall claim, which is the best point to intervene is agree that there are systemic causes but focus on the behavior because that’s the best point to try to interfere with the accident sequence.
David: I think the other literal example is Safety-I and Safety-II, so people who (I suppose) disagree with Safety-II. A lot of the times that there’s a distream of Safety-II is basically people saying we shouldn’t be throwing out everything to do with Safety-I. What we actually need is some middle ground between Safety-I and Safety-II.
That’s exactly what Hollnagel has been saying the whole time, and even throughout the whole book with a few exceptions. He says actually Safety-I is just not the whole picture. Do this other stuff as well, and maybe here are a few things that you can let go off a little bit. You can see in some of these arguments his idea that, well I don’t agree with Safety-II because I believe some of the things in Safety-I are good.
Drew: And it’s annoying that you can actually go through Hollnagel’s work and you can find sentences or paragraphs where he does clearly say throw out Safety-I. But then, that just makes him a confusing writer; that’s not his overall position. Look for the overall position that someone’s giving.
Along with that is remember that almost all the big ideas in safety are really a lot more nuanced than either their critics or their advocates suggest. Because of the way we tend to promote and spread ideas in safety, and perhaps the borderline between research and consultancy, then a lot of people take rhetorical positions that are much stronger than their true intellectual positions.
Usually, that’s because they don’t like the status quo and they’re trying to push it in one direction. Because they’re trying to push, they come off much more strongly than probably there are more reasoned ideas are. Don’t get sucked into responding to that heavy push, rather looking for what their nuance of the position is.
David: I think one example of that is even the last three weeks where we recorded the three episodes on Just Culture. Even if just the rhetorical distream is never blame anyone, never punish anyone, never hold anyone accountable, that simple sound bite is nothing about what is in the nuanced discussion about what different retributive and restorative Just Cultures actually look like in reality of organizations, societal expectations, and all of those other things that we spoke about over the last three weeks.
Drew: That’s a good example, David. I’m going to throw in a mea culpa on our own safety clutter paper. I’ve given talks at conferences that are basically rants against safety clutter, like advocating for decluttering because it’s taking strong positions like that that make for interesting talks rather than putting up lots of detailed academic description.
You could be forgiven for thinking that I believe we should get rid of all safety bureaucracy and that’s what my work is about. Whereas, you read the text of the paper, most of it isn’t about decluttering at all. It’s a careful description of the organizational mechanisms that create clutter and an argument that there’s no point in decluttering unless we try to first understand these mechanisms and try to subtly disarm the mechanisms rather than focusing on the clutter which is more of a symptom. That’s a good example of where I think I have publicly taken positions that are much stronger than our academic work actually suggests or supports.
We move on to the next one?
David: Yes, I think of the third question. You’re in a debate, you’ve gone and found a source, you’ve done some reading around it, you’ve worked out what it says, then you’ve got to basically take what the broad ideas are saying and ask the third question, which is what does this theory say about how accidents are caused? Ultimately, that’s one important thing that we’re trying to understand. We got the idea and now, how can we think about accident causation? What does the theory say about accident causation?
Drew: The thing we need to give is a brief disclaimer, that it’s not absolutely essential that all theories on safety are about how accidents are caused. Most theories in safety have a large element—either explicitly or implicitly—of saying this is how accidents are caused, therefore this how we stop accidents. Then they think that’s furthest to the exception (I think) might be HRO theory, which some people interpret as the theory for how to prevent accidents, but when you read it, it often doesn’t make that claim.
David: When I thought about HRO theory, I think what the theory is doing is describing, for example, a set of characteristics that researchers found to be common in organizations that seemingly operated these complex high-risk operations reliably over time, so Robertson and colleagues thought that these things were contributing to that. It doesn’t tell you anything about how accidents might happen in HROs or in any other organization that’s not an HRO. It’s just theorizing some common organizational characteristics of seemingly reliable high-risk organizations.
Drew: I don’t want to be too nit-picky. Remember what we said about reading work in context, that they wrote these HRO papers in direct response to Charles Perrow’s work claiming that accidents were inevitable in complex organizations. In a sense, once you look at the whole conversation it’s quite a reasonable reading to say that they’re reading this as a solution to the problem. That they think that accidents are not caused the way Perrow says that they’re caused, and that therefore accidents are prevented the way HROs prevent them.
The interesting thing is that once you dig into what the theory says, almost all of the theories agree on the basics and they agree far more than they disagree. Pretty much, everyone agrees that there is a large causal field leading to the accident. In other words, lots of causes and those causes interact with each other in complex ways.
There is no actual author who really says accidents are caused in really simple ways. What they’re trying to do is they’re trying to take this understanding that they are complex and saying what is a relatively simple way to think about this, that is going to help us prevent accidents? They all acknowledge the complexity and then artificially try to impose some simplicity. When they do that, often what they’re doing is they’re trying to establish categories of causes, and they try to place particular significance on some causes. You look at a different theory and basically if you read every theory as, yes accidents are complex, but let’s focus on this.
David: Yes, and I think that’s where we see the emergence of the significant causes. It might be behavior-based safety where we’re placing a significance on the individual behavior as part of the cause. In safety culture, it might be placing significance on the role of leadership.
In some areas like engineering, we might be placing particular emphasis on goal conflicts or some of these other things. It’s all drawn from the same causal field. I think each of these theories would recognize each of those other causes, just not place the same significance on them as these specific theories do.
Drew: There are a couple of things that we can take out of that. The big one is that if you read everyone as saying that, it makes it much easier to agree instead of disagreeing if you read behavior-based safety saying some of what contributes to accidents is human behavior. Yes, everyone should agree with that. They then say, we think therefore we should focus on this human behavior. That bit you might disagree. You might say, yes I agree that given the matter, but I think, generally, we should focus a bit less on human behavior.
Or you might talk about picking a circumstance. When we talk about driving, we can have a conversation. How much of driving do you think is about design, rules, and processes, and how much do you think it’s about human behavior? What about working it out? How much is it organizational decisions? How much is equipment? How much is human behavior? Once we agree that we all agree, we can focus on the disagreements much more productively.
The other one is that when you’ve got theories that disagree about where to focus, you don’t need to decide who’s right. It doesn’t hurt you that someone else is focusing on human behavior even though you want to focus on system design. In fact, in your organization you might want to draw from both of those ideas. You might want to be doing something about design; you might want to be doing something about behavior. There’s nothing fundamentally incompatible or wrong with doing that.
David, do you want to say anything about any of the particular theories? We’ve sort of mentioned behavior, implicitly mentioned system safety by talking about engineering processes.
David: One of the things that I’d probably say—probably if I think about some of these questions that we’ve asked about what is beyond the sources, what does it say about accident causation—is probably the more recent theories that say less and less about that, so even beyond HRO into RE, Safety Differently, Safety-II, human organizational performance, and some of these ideas. It actually becomes very hard to understand what these theories say about how accidents are caused other than they all just say the accidents and emergent property of the complexity of the system.
Drew: That is fair and I think partly that’s a product of the fact that we’ve hit the limits on being too reductive. We have learned that we can achieve improvements in safety by focusing just on human behavior. We’ve learned that we can cause improvements on safety by focusing just on human design. But we also discovered that we can do both of those things and then a major accident still happens. We need to then stop narrowing the focus and try to come up with more explanations of complexity, and be friendly enough when we try to explain complexity, our explanations get complex.
David: Are we going to question number four?
Drew: Okay. This is where it starts to get useful. So far we’ve talked about what the theory says, but the takeaway from that is all about just whether we agree or disagree is not particularly useful. At some point we got to get to the point of asking, what does the theory say that we should do?
A lot of the time when we argued, there’s an implicit assumption that we can only agree with one of the theories because they’re somehow mutually exclusive. Once we get down to the point of what does it say that we should do, even quite fundamentally different theories can still tell us to do the same or two different things that we can do both of, or we can do both at different times, or we can focus on one or focus on the other.
And then about your thoughts. My personal theory is that it’s because people try to justify the new ideas by saying the old ideas are insufficient, inadequate, or bad therefore you should listen to my new idea. That’s why we get into this habit of thinking it’s got to be one or it’s got to be the other.
Whereas, that rhetoric is not actually a logical argument. If it’s worth doing something, then it’s worth doing it for its own sake, not just because what you’re currently doing is bad. You could tell learning teams without criticizing any other form of activity by saying, whatever else you’re doing, whether it’s good, bad, it’s helping, it’s not helping, here are the benefits you’ll get from learning teams; here is the evidence. But we don’t. We tend to say, your current accident investigation processes are failing you. Here’s what’s bad about them. Therefore, do learning teams.
David: I think that’s right. The other thing that you mentioned there when you said what should we do about these things not being mutually exclusive because you’re trying to string together a management safety effort in your organization. You might go, for example, scientific management in 1911 says to understand work to the Nth degree so that you can best possibly set the operation up to be reliable, successful, profitable, and safe.
Critical Steps, for example, a book about to come out now in 2021 is basically going to tell you the same thing—understand work and identify the really critical steps and risk important actions in that work activity, and make sure you get really dependable human performance and role performance around those critical steps to manage the risk.
If I thought about (say) Safety Differently or learning teams, I say go and understand work-as-done by talking to the people involved. You identify some really important behavioral controls. Then you actually go, actually now I need—even hesitant a little bit, behavior-based safety program—a way of creating reliable human behavior in certain situations in my organization.
If we suspend our judgment on the practical application of these ideas and the things we’ve seen in organizations, and stick with just the ideas themselves, you actually need to have a holistic safety management effort in organization. I personally believe that you have to actually string some of these ideas together.
Drew: I think there’s an important subtlety here. Just because two ideas end up suggesting the same action doesn’t mean that there’s nothing new in the new idea. It doesn’t mean that it’s not a contribution to our overall understanding. It doesn’t mean it’s not going to help us do that idea better.
It’s a pretty poor criticism of a new idea to say, yeah but we’re already doing that, because what the idea suggests we do is not the end of the idea. What it means is that we don’t let the academics have that argument, that there’s no need for practitioners to buy into one theory or the other.
If three different theories all suggest you should do something that looks like learning teams—even if you sometimes call it one thing, sometimes call it another thing; it still looks like a learning team—then you don’t have to say I’m doing this because I believe in Safety-II, or because I believe in 1960’s total quality management, or because I believe in 1930’s make sure you understand the work before you try to manage it.
They all say do the activity, so it’s quite okay to just do the activity. Have the argument if the argument leads you to different conclusions not just because it leads you to a different level of being right about the theory.
The other thing that’s related to this is that a lot of the people who present their own theories are critical of other theories, not because they’re mutually exclusive just because of the research traditions they’re in.
A good example of that is there are a number of different criticisms on safety culture and safety climate. None of those come from people who have to get rid of safety culture and safety climate in order to sell their own ideas. In fact, their ideas might be quite compatible with also doing safety culture and safety climate activities. It’s just that because of the research traditions they come from and the type of scholarship they do leads them to criticize this idea.
As far as practitioners are concerned, those sorts of criticisms aren’t particularly interesting or useful. Just because you might love Dekker’s ideas for safety management doesn’t mean that when Dekker criticizes safety culture and safety climate you’re obliged to agree with that criticism in order to be consistent in your own thinking. Just because you hate Dekker doesn’t mean you’re obliged to love safety culture and safety climate just because you disagree with him.
David: I think the other thing that I’ve also fallen to the trap a bit is to know when you are critical of an idea or whether you’re critical of the way that ideas are being translated into practice in organizations in your experience. Many times what we see in organizations, whether we talk about these things, even safety culture, or behavior safety, or even Safety Differently or something, we see what one organization does and we then make (I suppose) a judgment of the idea itself rather than the application of that idea, which is typically always not what the person came up with the idea intended.
Drew: I think that works in a few different ways. The interesting ones is I think it works backwards. I think from an early theoretical point of view, zero harm deserves massive criticism. But from an application point of view, no one applies it, actually, as it’s originally specified. It is applied in so many different ways. They’re just saying I hate zero harm because it doesn’t make sense. It is not a fair criticism of the thousand different things out there that are called zero harm that are done in different ways. Likewise, you’ve seen safety cases done once badly is not a good reason to say I hate safety cases.
It would be fair to say, for example, that yes this idea is good in practice, but I’ve never seen it done well. At that point, you can start leading the practice with the idea and saying if there are no good examples of it, then maybe the original idea isn’t quite as useful as people thought it was.
David: In a bit of a negative spiral here talking about criticisms of other ideas and criticisms of what we see in practice. I think the next bit’s really important is we’ve got to stop focusing so much on the criticisms of what people say we shouldn’t do and what people say is wrong in other people’s ideas because we get in this endless cycle of negative debate. Instead, what we need to be doing is trying to focus on the useful suggestions that people have not the things that they say are not useful.
The example here is Safety-I and Safety-II. Even though by its very title it looks like Safety-I is in opposition from Safety-II and most of the way that Hollnagel talks about his ideas are do this instead of this, or new versus old, or so on, but then a lot of people don’t necessary would object to Hollnagel’s core ideas—understand work-as-done, understand why things go right as well as why they go wrong, and work to try to support work going right more of the time.
I think people would generally say that it’s very rational, hard-to-disagree-with point of view, but a lot of people who are against Safety-II are more against when Hollnagel talks about the criticisms of all of the other ideas that are out there or of Safety-I, that then blind them to his ideas about Safety-II.
Drew: I think that’s well said, David. The next one is what does the theory say that we should not do and why. When we’ve emphasized the importance of looking at what we should do, sometimes there is theoretical work which is critical reflection on theory and practice.
The main thing for practitioners to understand is that most of this work isn’t actually talking to practitioners. Mostly when academics criticize other academics, their audience are those other academics which we’re having debates among ourselves. That’s going to be true of any safety theory or practice, that there’ll be a lively academic debate around it. Practitioners need to be careful about what this debate does and doesn’t mean for practice.
David: I think you’re exactly right, letting the debate play out and being informed about the debate rather than necessarily using that debate to take some particular course of action in your own organization. It should spark some curiosity because if we’re seeing a lively academic debate about something, then it should mean to us as practitioners that if we think we understand this then we probably got more to learn about it.
The example of zero harm before. Although it’s not really currently, there is some level of academic debate around it. If we’re passionate one way or the other as practitioners, we should really be curious about why we’re so passionate one way or the other when it doesn’t seem to be academically resolved.
Drew: Remember that we were talking earlier on about the level of reading to be ready to join into a debate, and how we weren’t necessarily suggesting that every practitioner should do that. But there are ways to be part of a debate without taking sides in it. One of the things that I think is useful is to focus on the evidence that’s being produced, not purely the arguments.
Academic debates are going to play out in one of two ways. Either someone’s going to actually win the argument and there’s going to be some consensus of ideas, or someone’s going to do research, get empirical results, and come up with evidence to help resolve the debate and therefore to guide practice.
It’s important to know about finished debates. It’s more just interesting to follow along with the debates that are going on now. What you can do, though, when you know that there’s a debate that’s the time to start being a little more humble about practice and a little bit more curious about your own way of doing things.
You mentioned zero harm as a good example of this. If you know that there’s a scholarly debate about zero harm, you don’t have to take sides. That’s definitely time to stop getting dogmatic with zero harm is or isn’t a good thing. You should be going out there and saying everyone has to have a zero harm program. We demand you have zero harm. Or saying that people who do zero harm are wrong, evil, and don’t know about safety. The fact that there is an active debate is a good reason to be less certain in our own opinions.
The other one that I’m interested in your opinion is coming from an academic point of view of this. I don’t think it’s helpful for practitioners or consultants to evangelize particular practice or ideas on behalf of the academics. I think this is one of the reasons that we get into very non-constructive debates. The original sources of their ideas sometimes they’re intemperance, sometimes they put a bit too strongly, but they also almost always have more nuanced positions.
Other people start arguing on their behalf, so they’re much more likely to be ideological about it, much more likely to be dogmatic, much more likely to be exclusive of people who disagree and to start misrepresenting the other position.
I really have to ask you, what is the advantage of joining in a debate on the side of someone else, unless you have genuinely new arguments or new evidence to present?
David: I suppose there’s no underlying human desire for association that probably is driving some of that joining. We’ve spoken before but we’ve never actually recorded a podcast on the role of gurus in safety, and I think some of the academic people—if you think about Dekker, Hollnagel, or Leveson—have communities around them. They have a status within that community. They have respect and sometimes admiration from practitioners.
It’s probably not that different to someone cheering for their sports team on social media, someone cheering for their favorite ideas or favorite academics. I think that might just be a bit of a human condition to evangelize the practices and the people that you agree and align with to get a greater support for your own ideas (I suppose) or the things you believe.
Drew: That is fair. I think there are two separate things there. One of them is a desire to spread and share ideas, which is definitely worthwhile. If you have never heard of Safety-I and Safety-II, what’s going to make you pick up the book in the first place? It’s hearing someone talk about the ideas in a positive way and saying, hey this a good book. You should go and read it. Possibly even someone giving you a potted summary of the ideas here. It’s not just a good book, but it’s a book that says these sorts of things. You might want to read it.
I think that sort of thing is definitely positive, but I think there’s a line that’s crossed between sharing ideas and evangelizing those ideas. I realized as I say that, that I don’t quite know where that line is. I just do know that it’s something that I try to be conscious of myself and maybe it’s more not something to police in other people but to be just constantly aware of ourselves.
What’s my purpose in joining in here? Is it because I think there is knowledge that needs to be shared? Or is it because I think there is ignorance that needs to be fought? If we could recognize that temptation to fight, argue, and evangelize in ourselves, and maybe reign it back in in favor of explaining, sharing, yes and-ing rather than no but-ing.
David: I think all of our business or those who have seen some things on LinkedIn in this space will know for themselves where that line is between sharing and evangelizing just from the stuff they’ve read. It might be hard to describe that line, but I’m sure we’ll see it all the time.
Drew: Yes and I expect to see it easier in other people that we see it in ourselves. Maybe that’s what I’m saying is let’s try to apply the same standards of that to ourselves that we get annoyed at in other people.
I think it leads into the next one which is questions six: What knowledge or evidence would move the debate forward and how can you contribute to that?
David: I think this is a great last question because at the end of the day the difference in views and the disagreement is a constructive thing. It describes uncertainty, maybe in the safety science evidence, it describes confusion in the ideas themselves. The disagreement should be seen as a very positive thing, and it will remain positive if people ask themselves this question about what’s going to move it forward and how can I contribute to that.
We did write the manifesto for reality-based safety science. I think we spoke about it on episode 20 of the podcast. Do you want to just talk about that in relation to this question about moving forward?
Drew: One of the ideas that we were trying to push through that paper was that we sometimes get confused about what academics bring to the table versus what practitioners bring to the table, undervaluing both of them and sometimes reversing their roles. Academics deal in broad theories and generalizations, so principles of what is generally true. Practitioners deal in local knowledge, deep understanding of the particular circumstances at a particular place in time, and what is working there. Sharing that knowledge between each other is very, very helpful.
Practitioner knowledge very seldom is able to prove or disprove broad theories, but what it can do is it can take any theory and advance it forward. The types of things a practitioner can easily contribute is, I’m trying to apply Safety-II more in my workplace. I tried this and this is what happened. Just sharing that knowledge is really helpful because it takes a broad theory and it adds something to it. It adds a practice that you have tried and it adds feedback on what you’ve tried.
Or, I’ve tried out two different things and this worked better than that. Or going further and you’re deliberately building in collecting reliable information on those practices, and collecting that information, reporting back, we did a trial in our workplace. We had one site do it this way, one site did it this way, this the data we collected, this what it showed.
Those activities might be broadly informed by the theories, but the selection and application of safety practice has to be done by the practitioner locally. The information is collected by the practitioner and is shared back to the community. That doesn’t have to be necessarily positive or negative. It could be, I tried doing this and I couldn’t work out how to do it. I had other people thought about this. What if they tried?
David: That sounds like a very constructive discussion to have online, more so than some of the debates which have been the inspiration for this episode. For practical takeaways, when you see a debate unfolding in safety, in a competition of ideas, personal attacks, or whatever else it might might look like, you find yourself in it or your or you’re watching it, there are six questions we’ve gone through in this episode to ask yourself to either think about your contribution to that debate or to try to make sense of the debate that you’re seeing go on.
Question number one is, what is the original source? Do I know the original source? Do I think that either side is representing the original source? What does the original source say and how close is what the original source says to the details of the actual debate that I’m seeing? What do these theories say about how accidents occur? Are we having an accident causation discussion or are we having a discussion about the ideas? What does it say that we should do? What does it say that we shouldn’t do? And why?
I suppose whether we’re in it or observing it, to reflect on that final question which is, what’s the specific point of disagreement here and therefore what knowledge or evidence would actually move this forward, and can I play a role in that? That sequence of questions and checks might actually help us be constructive in our discussions around safety rather than just personally upsetting.
Drew: I’m a little bit wary about throwing this episode out and then asking people to respond to it on LinkedIn. Things they would like to know are, what do you find helpful to talk about and read about in the space of safety ideas? Think of times when things have not been constructive. Think of times when you actually have come away from something thinking that you have genuinely learned something or you feel you’ve genuinely contributed something. In particular, is there anything that is missing from this? Any tips you have for a positive engagement in that space between safety theory and safety practice?
David: Our intention in this episode is consistent with the intention of the whole podcast, is to have a constructive, rational, non-judgmental discussion about what works and what might not work to improve the safety of work, and we would like to see more of that discussion happening. We like having another podcast and we like to see more of that discussion happening in the broader safety community.
Drew: Respond to this episode by flaming us on LinkedIn or send any comments, questions, or ideas for future episodes to feedback@safetyofwork.com.