Why pilots (and humans in general) get a bad rap during accident investigations. In my earlier article on root cause analysis, I mentioned that every cause present in a given situation can be considered equally responsible for the occurrence of the failure in question. Many times, however, investigations end up ascribing the fault of the entire situation to a particular "root cause" that may or may not have happened without the presence of other contributing causes. When we identify the "root cause" as human error and fail to take actions against the system that set up that error, we are perpetuating the problem.
First, let's consider what we know about humans. We are generally intelligent, make rational choices, communicate well through language, sneeze sometimes, learn quickly, get sleepy, are omnivorous, need food several times per day, have excellent hand/eye coordination, don't have perfect aim, make mistakes in high stress situations, need several liters of water per day, become preoccupied with social concerns, etc., etc. While that list suggests that we as a species are highly capable, it also suggests that we are not machines that can perform a given action millions of times in a row without error.
So, we know for certain that humans will commit errors, it is only unknown at what time they will occur, and what form they will take. [ad name="Adsense Small Horz Banner"] Next, during accident investigations or other root cause investigations, the investigators are under tremendous pressure to explain why things happened the way the did. That is especially true if deaths were involved. The investigators are also under pressure to name a specific root cause and make suggestions to reduce the likelihood of that cause. Human error is nearly always present in every situation and so it is a likely target. Human will is also often the last line of defense against something bad happening, so it is the last point at which a different choice could have been made in any accident; the rest of the situation having been determined by physics.
However, given that we know today that humans will always commit errors, and that some situations put humans in a position where mistakes can be deadly (piloting an airplane, driving an automobile), it is a choice of system design that is the real cause of those accidents. We could trade an American-style automobile traffic pattern to one based on passenger trains on dedicated tracks and trade many small deadly errors for much smaller numbers of more deadly incidents. We could put computers in control of the trains with redundancy so that they always follow certain protocols that have been found to be safest. In any case, it is the choice of system design attributes that is the cause of the error and the cause of the failure or accident. The person is at fault only in cases of negligence, otherwise, the system is designed either to either control or allow a human's natural errors. Changes to the system design are the only solutions available, as we can't create a more reliable person (even excepting training, which has a limited and temporary effect).
So, let's remember that when (not if) a pilot or capitan or driver is found to be the cause of a deadly accident, the design of the system they were operating in is the real cause, as some time ago we as a society, or group of engineers thought it was an acceptable level of risk to have a person making that decision even though it was a given fact that errors were bound to occur.