Recently, a colleague remarked to me, “Philosophically, of course, there is no such thing as absolute safety.” Could this be true? Is there a situation where a man could not possibly injure himself?
After considering this for a while, I thought that a man in a straight jacket in a padded cell would be pretty close to absolute safety. There is nothing in the room that could cause injury, and, with the straight-jacket, there is nothing the man could do himself to cause injury.
What is absolute safety? I suppose we could define absolute safety as the possibility of injury from all possible causes as being zero.
I thought my example of the padded cell came pretty close to considering all possible causes of injury. But my colleague replied: “Your man in the padded call could be struck by a falling meteor, among many other things from which his padded cell could not protect him.”
True enough. There are many unsafe things over which an individual has little or no control, meteors being one. There is simply no way, today, to avoid injury from a meteor. I suppose that some time in the future we may be able to detect incoming meteors and predict their paths. If so, we could board spaceships and dodge meteors in a cosmic game of dodge-ball. This alternative may not be absolute safety, but it is better than the certainty of being hit by a meteor.
Let’s further consider the problem of the meteor. To date, there have been few, if any, injuries due to being struck by meteors. Most of us conduct our lives on the basis that we do not indeed have absolute safety from meteors.
To do any productive living and thinking, we must put some bounds or limits on our lives and on our contemplations. Without those bounds, we would, for example, live in continual fear of being struck by a meteor. Our lives would be driven and consumed by this fear. I suppose that a few people are indeed consumed by such fears, in which cases those people are probably not contributing to our society. Rather, they are probably requiring full-time care from some of the rest of us. They may indeed take some degree of comfort in living in a padded room.
In practice, each one of us conducts most of our life as if we lived in absolute safety with regard to most hazards. The newborn child starts his life with the belief of absolute safety without any bounds. Parents provide that safety. As the child grows, parents teach the child about the hazards in his world. They gradually transfer the responsibility for his safety to him. Certainly, when we sleep, we consider ourselves as being in the state of absolute safety. As we grow, we put bounds on our lives, deciding what kinds of activities are not worth risking injury.
Ultimately, for many of the hazards of this world, each of us, individually, is responsible for prevention of injury to ourselves and to others. For example, when we drive a car, we must do so in a manner that does not injure us or others.
Since we can’t define all possible causes of injury, then the probability of injury cannot be zero, and we cannot have absolute safety. Fair enough. Full stop. Fin. End of discussion.
And THAT is the problem. If we define absolute safety in terms of all possible causes, then we have nothing left to talk about. There is nothing we can do to accomplish absolute safety. And if we can’t accomplish absolute safety, then is there any good reason to attempt any safety whatsoever? Since we are facing death at any instant due to a meteor, then why should we take care in crossing a street?
Well, we DO take care in crossing a street. For the most part, we DO act as if we are in a state of absolute safety. (There are exceptions where we voluntarily place ourselves in risky situations for the challenge. But, we accept such jeopardy only at given times and in given places. For example, sky-divers deliberately jump into risky situations, but they don’t spend their entire lives in such situations.) So, we put bounds on absolute safety.
And, when we talk about probability, we must put some bounds on the problem. We can discuss the probability of injury from a meteor. We cannot discuss the probability of injury without regard to its cause. Carrying this thought a bit further, if we can identify a particular cause of injury, then can we ever accomplish absolute safety for that particular cause? That is, for a particular cause of injury, can the probability of injury from that cause be zero?
Consider the Golden Gate Bridge. When you cross that bridge, do you consider the probability of its failure while you are on it? Maybe. But, if you do, it is a very abstract consideration. You can’t really imagine the failure of the bridge without also imagining some cause for that failure, such as failure of a cable. But the cable is comprised of individual wires. So, the cable can’t fail without also the failure of each and every individual wire. What would cause the failure of one or more individual wires?
With a bridge, we can declare that if the design conditions are not exceeded, we have absolute safety with regard to the performance of that bridge. In other words, the bridge will not fail given the particular traffic and weather conditions accounted for in the design.
A few years ago, the Golden Gate Bridge turned 50 years old. On that occasion, the authorities closed the bridge to automobile and truck traffic, and opened it for pedestrian traffic. During the midst of the festivities, the authorities suddenly realized that the bridge was being loaded far greater than any traffic load! The bridge design did not account for the tremendous pedestrian load. Under these conditions, the bridge COULD fail.
Consider the case of the “unsinkable” Titanic. The Titanic was unsinkable given the conditions specified by the designers. The designers did not consider the situation in which the Titanic was damaged, and it did sink. No probabilities involved. If the Titanic was subjected to the conditions specified by the designers, then the probability of sinking was zero. However, it was subjected to other conditions, so the probability of sinking was 100%.
Consider the infamous skywalk in the Kansas City Hyatt Regency Hotel. While under-designed, it nevertheless probably would not have collapsed. But, the original design was difficult if not impossible to assemble. So, the design was modified. And, the modified design failed.
In most cases, safety of products and many other things is provided by one or more safeguards. For example, one of the safeguards against electric shock is insulation. If we are dealing with solid insulation, then we know the electric strength of that insulation. Disregarding other deteriorating factors, if we never exceed the electric strength of that insulation, then the insulation will not break down. Therefore, we have absolute safety against electric shock – – provided the electric strength of that insulation is never exceeded.
Safeguards have bounds. No matter the safeguard, we can always subject it to a stimulus which will cause it to fail. For example, we can subject an insulation rated 3,000 volts to a 10,000 volts and it will fail. When the safeguard fails, it is no longer a safeguard, and we no longer have a safe situation. But, if we stay within the bounds or limits of the safeguards, then we have absolute safety.
Absolute safety from specified causes or hazards does indeed exist, provided that the conditions of the design of the safeguards are not exceeded, and provided that the safeguards are manufactured accordingly.
“To engineer is human, the role of failure in successful design,” by Henry Petroski, St. Martins Press, New York. ISBN 0-312-80680-9.
Richard Nute is a product safety consultant engaged in safety design, safety manufacturing, safety certification, safety standards, and forensic investigations.
There is no safeguard against Murphy’s Law.