The Investigation Process Research Resource Site
A Pro Bono site with hundreds of resources for Investigation Investigators
Home Page Site Guidance FAQs Old News Site inputs Forums
. . . .. . . . . . . . last updated 8/8/09


to advance the
State-of-the-Art of
investigations, through
investigation process

Research Resources:

Search site for::

Launched Aug 26 1996.


Source: ISASI forum September 1992. published by the International Society of Air Safety Investigators, Sterling, VA USA


Since this was written, I have become increasingly aware that the "cause or probably cause of the accident" concept is a product of England's common law that has been adopted in many other countries, but is not universal. I would welcome inputs about the "cause concept" in other nations from readers of this document. A survey of ISASI papers shows significant variations in perceptions of cause among these investigators. There has been a lively exchange of mails about this topic on the web, which I have captured and will edit when time permits, unless I have a volunteer to do the editing. See also investigation definition '

In December 1999, a study of NTSB recommended action in this area.

L. Benner

Guest Editorial
Ludi Benner

Recent exchanges on the subject of "Causes" suggest to me that three major issues face ISASI:

1. Determination of "Cause" in investigations reflects a legacy which the legal community imposed on investigators;

2. Determination of "Cause" as presently practiced is demonstrably indefensible from a technical and scientific point of view; and

3. As investigators, should we make an effort, or not, to change society's demand for us to determine "Cause" (read blame and fault) in our investigations?

At the outset, I believe that our work should not ignore other traditions and legacies. Where it serves us, we should respect those traditions and legacies. But we should be their master, not the other way around. Societies' legal system, which historically has been the driving tradition and force behind "Cause" determination in accident investigations, is designed to provide a mechanism to resolve disputes. It is based on the weighing of evidence, and fact is determined by the weight of the evidence developed. Investigators have a different mission: prevention of future accidents. In my eyes, our mission is to develop an understanding of what happened and why it happened, to predict the effects of the findings on future activities, and propose controls if those effects are unacceptable. We have different processes for determining the technical truth of our findings, related more to scientific inquiry than to legal processes. I have discussed this with prominent trial attorneys who confirm that this is one of the major differences in the respective roles.

I also believe strongly that if traditions and legacies detract from the practice of our profession, we must challenge them, and ultimately change them in the work we do within our profession. The need to change the tradition of determining the accident "Cause" demands action. In my professional practice, I have observed first hand the harm that flows from determining "Cause" in investigations. I have seen how the search for "Cause" corrupts and prematurely aborts promising investigations. I have seen judgments of "Cause" result in technically unjustifiable blame, shame and dishonor in the form of guilt feelings, damaged reputations, impaired job security, and even unjust punishment - to individuals who were victims of a rush to judgment by investigators or investigation authorities. I have seen misdirection of prevention efforts because of arbitrary or prejudicial selection of a "Cause" in an investigation. I have seen perpetuation of myths and misconceptions about safety problems in accidents because traditional "Causes" like "unsafe acts" or "human error" have been cited as the "Cause" of accidents for too long. I have seen "Cause" determination frustrate understanding of the risk decisionmaking process faced by employees and managers in all kinds of situations. I have seen workers blamed for "causing" what happened in impossible work environments. I have seen "Cause" statements that are impossible to address with corrective actions. I have seen determination of "Cause" align one investigator against another, undermining the efforts of both to improve safety. I have seen the quest for "Cause" blind managers to opportunities to improve operational performance. I have seen government investigators trample individual rights with their determination of "Cause". I have seen good investigators and investigations being manipulated to arrive at opinions of cause that cannot be reconciled with the nature of occurrences. I have seen experienced investigators so blinded by the tradition of determining "Cause", and the aura of expertise it accompanies, that they are unwilling to listen to, let alone try to learn new investigation ideas and technology. I have seen experienced, thoughtful investigation managers demeaned and ostracized by traditionalists for having the audacity to question the determination of "Cause" and what it accomplishes. I have seen the eagerness of investigators with true potential for valuable contributions to safety handcuffed by environments emphasizing determination of "Cause", fault and blame rather than understanding, compassion and progress. I have experienced the frustration of making the facts of an accident fit a list of predetermined "Causes" on an accident form. And I have been subjected to pressures in the past to be silent on these issues in my published works.

With this experience, and the experience of my research into better investigation methods, I made a personal decision that in my performance as an investigator I cannot and will not knowingly compromise demonstrably valid principles in work products I produce for my clients. I take very seriously the ISASI Code of Ethics including Section 4 about LOGIC and 45 which cautions us against "value judgments based on personal experiences which may influence the analysis." Participation in the drawing of a technically unsupportable conclusion like "probable cause" or "root cause" of an accident, by whatever name the exercise of drawing these kinds of conclusions is called, is in my view a breach of the ISASI Code of Ethics by which I feel bound.

I have acted consistently in accordance with this position since I understood the nature and consequences of this issue. For over 10 years, my aim has been to develop new and blameless descriptions and explanations of accidents of all kinds in my investigations and research. I refuse to propose an accident "Cause". I successfully deliver technically defensible, validated and quality-checked descriptions and explanations of accidents, without attributions of "Cause". They illuminate the role of people in these complex occurrences. My problem statements and recommendations flow demonstrably from the accident description and explanation. I have experienced no adverse repercussions for my clients, harmonious investigations, gratifying new safety insights and improvements, and personal feelings of accomplishment about my investigation results.

A second result of my approach has been that others' demands for me to compromise my professional integrity by demanding a determination of "Cause" fade away. I cannot accept that there is nothing we investigators can do, or that the problem of Cause is so big or so deeply rooted that we should just make the best of it and live with it as best we can. If we in the investigation profession take a stand to bring about change within our membership, society's view will also change. Subordination and eventual abandonment of "Cause" determination by investigators can be accomplished. Three key steps are required:

First, I think we in ISASI must adopt the position that an accident is a process for investigation and prevention purposes, and that accident processes must be thoroughly understood before they can be properly controlled.

Second, we must present complete, validated descriptions of what happened that also demonstrate why the occurrence happened as it did. That must be the basis for all uses of our work.

Third, each of us needs to respond to those who demand "Cause" by demonstrating the technically irreconcilable nature of that demand with comprehensive descriptions of what happened. We can enhance that effort by substituting demonstrably robust, valid descriptions of the entire accident process for subjective judgments of "Causes".

It is noteworthy that in discussing this issue at length with a prominent and very successful litigator, he said these actions would pose no problems for him in litigation. A sample of one is not proof, but maybe this change will be easier to accomplish than many investigators now think possible. It seems worth trying.