Thinking differently about safety investigations

file000704919536Systems exist and function within diverse societal environments reflecting numerous societal constructs that have evolved over time in all societies. Societal constructs about justice, personal rights, behavioral norms, education, economics, governance, and others influence safety program and investigation perceptions, constructs and practices. Those societal constructs and their influences need to be incorporated by academics and practitioners into a new and broader “societal” socio-technical vision horizon for both safety programs and safety investigations if existing concerns, issues and conflicts are to be understood adequately and resolved.

For example, many societal desires are reflected in legal systems and their constructs. Legal systems exist in societies to provide an impartial and consistent process for binary resolution of issues involving deliberate or accidental harm or loss. Legal proceedings aim to satisfy several societal desires, including the determination of responsibility for causing harm, penalizing harmful behaviors, just compensation for harm or loss caused, deterring future harmful behaviors and bringing closure to victims’ concerns about incidents. Over many centuries, legal systems developed constructs needed to deal with all kinds of allegations involving acts or omissions that produced harm or loss covered by law. Legal proceedings also developed constructs for legal investigations to support parties’ arguments and legal decisions.

Safety domain programs and investigations evolved much more recently, to serve some similar societal concerns: the avoidance of harmful behaviors and harm, expressed as “prevention of recurrences”, and closure of concerns, expressed as desiring descriptions and explanations of what happened. As they evolved, safety programs and safety investigations used easily available legal investigation constructs to create a similar investigation process. Their use continued, even as safety investigation goals began to diverge from legal investigation goals. As far as I could determine, those constructs and the consequences of their use have not been investigated in detail.

To address this specific knowledge gap, such an investigation was undertaken. The goal of that work was to identify legal constructs incorporated into safety investigation processes, their effects on those processes and, if detrimental, potential changes to overcome any adverse effects. The work disclosed at least 7 related legal constructs with detrimental influences on safety investigations. The results suggest that further research into societal constructs’ influences on safety programs and safety investigations could be fruitful for identifying needed changes to both safety and societal constructs, as will be likely be required to resolve the criminalization of accident behavior issue.

A report of the inquiry can be accessed through the Resources page.


  1. William R. Corcoran, PhD, PE Reply


    Thanks ever so much for addressing this important area.

    The sword cuts both ways.

    The use of legal concepts in root cause analysis can enhance the insights from the investigation. However, the highjacking of legal terminology by safety investigators can baffle readers.

    It would be interesting to see some examples on both sides.

    1. L Benner Reply

      Bill, the examples of influences of legal constructs on safety investigations are in the referenced paper. It would be interesting to see some examples of insights provided by legal concepts in root cause analyses.

        1. L Benner Reply

          Re comment about aggravating factor.

          Doesn’t the Wiki cite about “circumstances attending a tort or crime which increase the guilt or enormity of a crime…” add another legal construct requiring subjective selection of evidence and conclusions – and eventual subjective determination of its validity by a jury or judge (or investigators?) This was addressed by 2.5, 3.5 and 4.6 in the paper, I believe. Isn’t its selection as a “aggravating factor” an analytical categorization decision rather than a description of objective reality? To counter an “aggravating factor” must not users determine who did what to create the factor before they can determine the relevance to their operations and whose behaviors to change and how to change them? Don’t safety researchers face this same disambiguation challenge when they look at archived accident factor data for trends for which they can fashion remediation proposals?

  2. William R. Corcoran, PhD, PE Reply


    The brief answer is, “Probably not.”

    Some legal constructs probably can be helpful when applied to safety investigations appropriately. Some legal constructs probably can be harmful when applied to safety investigations inappropriately.

    One of the intended results of some legal constructs is to assure that the wrong person is not punished while the right person goes free. Is this bad? I think not.

    The corresponding use in safety investigations would be to assure that the wrong conditions, behaviors, actions, or inactions are not addressed. Fixing the wrong thing can leave harmful conditions, behaviors, actions, or inactions in place to continue causing harm.

    If all of the legal constructs are purged from safety investigations, will we be throwing out the baby with the bath water?

    Let’s get specific.

    Which legal concepts are being used harmfully in safety investigations? Could they be used helpfully? What would it take?

    Let’s take them one at a time.

  3. William R. Corcoran, PhD, PE Reply

    What is the difference between a legal construct and a legal concept?

    Are legal constructs limited only to items constructed by the law, e.g., corporations, contracts, standards of care, standing, jurisdiction, and the like?

    Are legal concepts more inclusive? Is every legal construct a legal concept, but not the other way around?

    1. Ludwig Benner Reply

      In my view, a concept is an idea or perception about some subject or object whereas a construct is an entity consisting of conceptual elements brought together and refined over time. Thus constructs are more inclusive. Legal constructs have been adapted for safety, contracting and other activities.

  4. drbillcorcoran Reply


    Please do what you can to answer my questions here and in the comments in your Safety Differently posting or get others to do so.

    Let’s work one “construct/ concept” at a time.

    What is one construct/concept that you think is very harmful?

    What is the construct/concept?
    How is it defined legally?
    How widely is it used?
    What harm does it do?
    What are examples of the harmful use of it?
    What should investigators do about it?

    1. Ludwig Benner Reply

      I offer the following answers

      Let’s work one “construct/ concept” at a time.
      ​1.Causation models or accident perceptions, take your choice for first one.

      What is one construct/concept that you think is very harmful?
      ​2. ​Cause probably does most long term harm for reasons I have documented in this and various previous papers​

      What is the construct/concept?
      ​3. Safety construct is that accidents are caused – the almost universal use of accident causation models.

      How is it defined legally?
      ​4. in several ways legally, but undefined technically for safety purposes. Need to separate safety needs from legal needs. Theyre different societal functions. ​

      How widely is it used?
      ​5. Very. Cause is almost universally employed by safety investigators and investigation methods, but not all methods.

      What harm does it do?
      ​6. That was covered in the paper, I hope. Quick short list – premature termination of investigations, selective problem definition and overlooked alternative opportunities, vulnerable to contention and conflict because it’s subjective decision, defies validation and replication, promotes linear thinking about accidents in safety field, too many variants of cause to choose from, evades contextual relationships, etc.

      What are examples of the harmful use of it?
      ​7. My favorite example is from aviation: crashes on approaches for years were allegedly caused buy pilot error until wind shear phenomenon was discovered by better investigations. I think saying operator error as the cause of the accident is a gross injustice to those involved, masks other risk raisers, is based on subjective hindsight, diminishes interest in making investigations more scientific, etc.

      What should investigators do about it?
      ​8. To overcome cause problem, two steps: 1) stop using the word cause and its der​ivitives, and substitute other words or phrases like what happened for what caused the accident, etc, and 2) start using input/output model for investigation data integration.

  5. Ludwig Benner Reply

    Highly recommended reading with a further example of misdirection and adverse consequences attributable to “cause” construct and why it should be purged from accident/incident investigations, safety research and the safety lexicon:

    “Trials and Errors: Why Science Is Failing Us” by Jonah Lehrer, Wired magazine, January 2012, accessible on line at

    Should have mentioned it in the paper.

  6. Ludwig Benner Reply

    I don’t use cause in my investigative activities, and only observe legal activities so I have no favorite legal definition. All the legal definitions I am aware of are variants of the underlying Aristotelian linear cause-effect philosophy, and are used in the legal community’s binary decision making process to determine innocence or guilt. That, I argue, should be left to the legal community exclusively.

    Which legal variant of cause do you think should be used in safety investigation reports, and why?

  7. Sheri Suckling Reply

    All too often legislation is enacted in response to political pressure from society, with legislators feeling pressure to be seen to act on an issue that has aroused public outrage. In the age of social media, this pressure is more strongly evident than ever before, especially as individuals feel both empowered and entitled to express an opinion.

    The ‘fast thinking’ that typically results from this pressure usually leads to ill-conceived and poorly constructed regulatory requirements, with little thought about upstream or downstream effects, including effectiveness, enforceability, and even whether society actually has the will to actually enforce these requirements when it comes to the crunch. Ultimately, any law is only as good as its effectiveness and application.

    Recent legislative changes in New Zealand – the Safety Reform Bill and Health and Safety at Work Act (HSWA) – are directly linked to findings of the investigation and enquiry into the Pike River mine disaster, in which 29 miners lost their lives. While numerous failings were identified, the investigation still fell short of understanding the human psychological factors behind those failings. Instead, they focused on the ‘symptoms’ and enacted legislation which included much larger penalties and mandated director accountabilities. This has largely resulted in a lot of fear (and fear-mongering), which has unfortunately directed more attention on avoiding those regulatory penalties (i.e., very large fines that cannot be insured against and jail sentences), added further distance and distrust between businesses and WorkSafe NZ, and emphasised the ‘covering your butt’ activities rather than actually understanding and doing things that make people safer at work! There is still very little understanding of the underlying psychological drivers, including societal norms, beliefs and archetypes, values, history, etc that shape people’s thinking and choices. Yet these are the factors that actually influence and shape behaviours.

    So far, it’s a bit like the scenario where one person who speaks English is trying to communicate with someone who speaks Swahili, The person who speaks Swahili doesn’t understand what the English-speaking person is saying, so the person speaking English just says it again louder….and louder….eventually shouting and becoming quite angry and defensive about not being understood, so they start shaking a finger at them and threatening them with punishment for not responding as desired. The person who only speaks Swahili is also becoming quite agitated and frustrated – they are doing the best they can with what they know; nothing is going to make them instantly understand or speak English. They feel trapped and fearful.

    As Einstein said, “The definition of insanity is to keep doing the same things and expecting a different result.” He also advised, “The problems we face cannot be solved at the same level of thinking we were at when we created them.” Clearly the present approach to safety is not working, and escalating the threat of punishment is only going to raise the background level of fear, mistrust and aggression, which will continue to escalate and divide people – at the very time when businesses most need agreement and alignment to run lean and efficient organisations!

    These times definitely call for ‘safety differently’.

  8. Joe Aiken Reply

    Sheri, I agree with much of your comments about the HSWA. I get the feeling that the substance has been degraded by the language used, much of which was taken from the existing Australian legislation , which itself was based on older UK & European legislation. In trying to explain the concept of a PCBU and the need for H&S Representatives (for PCBUs with more than 20 workers) etc., the idea that we should all be more aware of hazards & consequences and the risk of them happening appears to have got submerged.

Leave a Reply

Your email address will not be published. Required fields are marked *