However, one of the questions that need to be asked when considering the ‘causal’ relevance of ‘unsafe conditions’ is this; do they materialise out of thin air? That is, do they create themselves?' Any thoughts from anyone other than Mr J? I don't think anyone with any knowledge of accidents and causation believes that unsafe conditions materialise out of thin air. Unsafe conditions are pre-existing conditions, each one may not be sufficient on its own to cause an accident, but linked to another unsafe condition or act may provide the chance concatenation required for an adverse event. In other words, most accidents do not just happen, there are latent and active pathogens which under certain prescribed conditions ie trigger event, will provide the conditions for an accident or disaster. Unless of course, you wish to rubbish the teachings of Turner, Reason, et al.

Hi Heather. I dont think (and sincerely hope) that anyone read it like that but apologies just in case. There is no reason for you to have a clue...I have'nt got one either! 14 years of trawling the relevant literature produced nothing; so we are in the same boat; i.e. there is nothing out there to support multiple causation theory and Ray might not be fully aware that Professor Reason’s Swiss cheese model is fully reliant on it. I'll do my best to answer the questions more fully by referencing the book but as I make clear in the response to the ASSE, they do not need the book to answer the questions that their causal philosophy supports.

Difford please note my posts at 12 and 18. I have found over the years that most operators (as has been said before) carry out their work after training and instruction. If then an accident occurs then the management system must be looked at besides any historical evidence as to how the system of work evolved. My example of the aircraft crash clearly demonstrates multiple causation where the desgner (also his supervisor and their management system) , servicers, and maintenance staff all failed to find the fault which downed the aircraft. There were 2 management systems at fault in this scenario. Thus as RayRapp asks do unsafe conditions just materialise? In my opinion "no they don't" and never will (save force majeure) - unsafe conditions will always have a human element (cause). Rich

Jonty - perhaps you'll get your answer from Difford perhaps you won't. Other factors in play causing an accident in unsafe conditions are not recognising that an unsafe condition is in existence. Back to the management system and possible lack of training for managers, supervisors and ultimately the workforce. It could be fatique, it could be a slip of memory, it could be mudane (not having your eye on the job), it could be traditional - we've always done it this way and never had an accident (old school thinking?). The list goes on. It keeps coming back to humans and their failings. Rich

Right on both counts Mr D.

Most would agree I think, conditions are man made. The problem arises when we look at unsafe conditions in the pure H&S sense as opposed to the pure causal sense. In the pure causal sense, a “condition” (the word unsafe is not necessary) is something that “is not a cause”. “...imagine a man walking along his local High Street. For reasons known only to himself, he walks into a lamp post...a lamp-post he has successfully negotiated hundreds of times before...not to mention all of the other lamp-posts that he has also avoided. So, what is the cause of this accident? True, “but-for” the lamp-post, that accident would not have happened. But, the question is not enquiring about things that would have prevented the collision, the question relates to what caused it! So, is the lamp-post the cause, or a mere condition?” Difford (2011:45). There will be occasions when we have to look further upstream (much further) in the causal chain but to assist Jonty’s post the above extract should suffice for now. (Heather, yes, the workshop fire would be a good one to extract but it covers seven pages and contains diagrams so hope you approve of this more basic example).

I have, at last, managed to find the time to read both the linked documents and remain puzzled as to the objective of the rebuttal of the ASSE article. It seems to me that the ASSE article is warning against placing too much specific reliance, if any at all, upon Heinrich’s work. Particularly not quoting it as fact and drawing conclusions that were never part of his work. That seems to me to be sound advice. Difford appears to be challenging the validity of multiple cause theory by discrediting the ASSE article but I have not been able to determine what Difford is actually saying. One could draw a picture in which multiple cause is the weapon to blame management and Heinrich is the weapon to blame the injured person. I don’t think either is an inevitable outcome but accept that they could be used that way. But does misuse discredit anything other than those who misuse? Consider if you will for a moment that Heinrichs statement ‘unsafe acts by WORKERS are the principal causes of occupational accidents’ is updated to read ‘unsafe acts by PEOPLE are the principal cause of occupational accidents’. Does it then have more or less acceptability or validity in 21stC society? Is this the basis of the discussion? Or is it that multiple cause theory is the key myth that needs to be debunked rather than Heinrich or that both are false? Whichever it is I would be interested to read what is being proposed as a more proper approach to occupational accident investigation in the 21stC. It may be that those are outlined elsewhere in documents beyond the scope of this forum? It seems to me a universal truth that humans make mistakes and thus, in a day to day sense, that statement needs no scientific support to justify it. It is also self-evident that understanding why and how we fail is a useful skill for reviewing any organisational failure. However, it is probably true that day to day many, if not most, will lack an adequate understanding of those mechanisms to use them consistently and to full effect. So we have less than perfect people using less than perfect models. How much should we be concerned at that state of affairs? Isn’t that a pretty good description of life? I am sure many of us recognise that the various models and research on multiple cause theory are not perfect science. However, we also recognise that the use of such models or theories has dramatically improved the general quality of investigation. Therefore, to simply discredit it without providing any alternatives would be a significantly retrospective move in regard to workplace safety. Quo Vadis? Finally, I am really puzzled by the lamp post example given. What does this demonstrate exactly? Did a human make a mistake? Yes, therefore an error on the part of the injured person caused the event. That supports the 98% principle given in the topic. Does it mean this is the only error that caused the event? No. Do I need to know why the mistake(s) occurred? Yes. Is the lamp post relevant? Yes. Does it really matter whether it is a condition or cause? No. Is the fact that the person has successfully completed the task many times before a reason to discount anything in my investigation? No, just the opposite. P48

Pete The point of the lamp-post example (I think!) is to demonstrate the difference between a "cause" and a "condition". In this example, the lamp-post (or the existence of it) is a condition and therefore not considered to be a cause of the accident. I have no problem with this example. Where I struggle a bit is if the "condition" is more transitory. How about this example: Man hurrying, floor is wet, man slips. Is the wetness of the floor a condition or a cause? Let's be clear here that with your "safety head" on you obviously look both at the behaviour of the man (why was he rushing?) and the fact that the floor was wet (and then look further to find out why!). However with your "accident causation" head on, what was the cause of the accident?

Pete. Thank you very much indeed sir for taking the time to read not only the linked articles but to produce such a thoughtful and objective post here. Refreshing. Its not so much about discrediting the ASSE article as asking the ASSE and Fred (Manuele) to provide the evidence for multiple causation theory (I have laid out the case against MCT fully elsewhere but that is immaterial, nobody should need that in order to address my response). Petersen provided nothing in support of the theory (surprising as that might be) and that point has never been challenged until now. If my post does nothing more than highlight that fact, my job is done. Critically perhaps, Petersen also rejected Heinrich’s common cause hypothesis. However, that, in effect, challenges the belief of the masses that PEOPLE cause accidents and has resulted in a relentless attack on management and the system. Even if management or a system could be THE CAUSE of an accident, it simply cannot be so on an exceptionless basis; i.e. it cannot always be the case. Unfortunately, there is an exceptionless philosophy under pinning MCT and it is wrong according to science and logic. Yes Pete, MCT is the myth that must be debunked as far as I am concerned. Heinrich was not blaming people; that is a safety myth and has largely closed the doors on the social sciences assisting industry in general. That said, and whilst I find nothing but support for Heinrich’s central premise, there is nothing that science or logic can support in relation to his so-called ‘domino theory’ and I reject that also. As for the matter of causation in its purest sense, the models out there assist neither that nor investigation. What the current models do is reduce the investigation to a post-accident audit or inspection. Consequently, numerous management failures, regulatory breaches and system non-conformances (conditions) get identified. However, and whilst these must be attended to, they do not equate to cause; that is another myth that needs to be debunked.

As regards the central statement in the book, the following arises in the linked document... "Unpredictable and uncontrollable consequences of natural phenomena aside, human behaviour, suitably defined, will be the underlying cause of any accident. [that statement is supported, inter alia with this] The behaviour of a man who has been found to be the underlying cause of an accident is attributable only to him...” (Difford 2011:196). That statement does not place causal responsibility on anyone in particular. What it should be highlighting is that the underlying cause of any accident cannot be a non-entity such as an organisation. Similarly, a system cannot be the underlying cause of an accident since (as Rich alludes to in one of his posts) only a man could have specified it (or, of course, permitted its under specification). Good investigations inform and improve our systems, they do not blame.

'Where I struggle a bit is if the "condition" is more transitory. How about this example: Man hurrying, floor is wet, man slips. Is the wetness of the floor a condition or a cause?' Heather, the wet floor surely is a condition...and partial cause, the other cause being the man hurrying. Accidents are caused by unsafe conditions and/or unsafe behaviours. In the above case it is clearly both. That said, someone not hurrying may have slipped. Therefore the primary defect must be the wet floor. In accident investigations I prefer to use the terms - immediate cause and underlying causes. There are often many underlying causal factors of differing degrees of influence. I dislike the term 'root' cause because with some accidents a single root cause is not easy to define, hence the term can be misleading.

I guess it will depend on how the water got their in the first place. For example, if it was due to a cleaner washing the floor, then the cleaner should have provided signage to warn the unwary. If it was due to some mishap, it could have gone unnoticed or there may or may not have been signage or a mop available. Tell me Difford, does your accident aetiology accept that some accidents just happen, no one was to blame and by implication they are not preventable?

I should have said notwithstanding hindsight bias, which has the effect of making everything crystal clear after the event, nor the courts interpretation of the facts, except Lady Hale's assertion that "some things are nobody's fault" - Hatton v Sutherland [2002] 2 All ER 1 (CA). Blame was an unfortunate choice of word, although 'fault' could just have easily been used - semantics. The fact is sometimes things do just happen, albeit usually only attributable to minor incidents. It appears the world we now live in there must be a cause and corrective action for everything, as opposed to [expletive deleted] just happens.

Not familiar with either case and I suspect they are not UK case law. In the UK the courts rarely accept the plea of 'not reasonably foreseeable' in the case of fatalities, R v HTM [2006] EWCA Crim 1156 being the exception. Difford, I must say you are very handy with the cut and past...LOL!

Ray. Re Cut n paste. I know my work inside out and have an electronic copy. Simples.

Agreed - so why do so many in the profession support a philosophy which says that organisations, not people, cause ALL accidents?

So there are competing theories on accident causation. Your point is...? Am I the only practitioner out in the real world who gives this about as much importance per se as Saint Thomas Aquinas' findings (thanks A Kurdziel post 22). Whilst it is perhaps comforting to know that there are people who seek to expand the intellectual underpinning of our profession I'm bound to ask myself this question: In the last few years, how many times have I felt the need to resort to causation theory to understand the circumstances of an accident such that I can draw proper and justified conclusions that both will withstand scrutiny and, vitally, will point me towards improvements? You guessed right! As there seems to be some debate about the availability and analysis of evidence, could I perhaps also question the assertion above "Multiple causation theory is the basis of the causal philosophy that is currently held by most safety practitioners". Your evidence for this is...?

Johnv. I agree with your mantra but fail to see how anything in this thread might encourage anyone to question its validity. That said, you also say that "there are so many factors implicated in accident causation" and that the "whole chain of factors is so unpredictable". With respects, that seems totally contradictory and, in any event, I disagree with the latter points entirely in the context that you seem to provide them. However, if what you say is true, what is it exactly that your remedial interventions are trying to prevent...the accident you just had, or the one that is yet to come? If it is “the one that is yet to come”, could you clarify how you do that when you have already said that, for you, the whole chain of factors is so unpredictable?

Jonty. I suspect that the majority of us don't really believe particularly strongly in one theoretical mechanism for causation or another. I personally don't stop to think "now where are the management failings?" or "now hang on, have a found a cause or a condition" when I'm investigating an accident and I am willing to bet most other practitioners on here don't either! "Management" is a pretty all-embracing word anyway and taken to its logical conclusion, most managers that I have met are more or less human, so in saying the failures are "management failures" one is indeed saying that they are "human failures" surely. As for your last sentence. Are you seriously suggesting that peoples' lives depend on whether we define something as "causing" an accident or "associated" with it? I can assure you that if I was investigating the case of the man slipping on the wet floor I wouldn't be worrying about semantics like that!

All accidents may be preventable in theory - in practice it is impossible. Like Peter, I also take the view that some accidents cannot be prevented, regardless of how much money is thrown at preventing them. If only senior managers would take the same view! Zero accidents mantra is my no.1 pet hate.