Because when you study complex systems and how they fail from famous examples (Chernobyl, Challenger, Air France Flight 447, etc. etc.) to (for me) personal experience addressing software systems downtimes... what you find is single causes are quite rare. What you do usually find is a dozen independent things, all of which need to be fixed, any of which if it were fixed would have prevented the failure.
Attributing "cause" to any of them is an exercise in preference of importance. Do you pick the last one that happened? Do you decide which one had the biggest impact? Do you pick the one that was the easiest/hardest to find? It's a silly exercise to pick one. You will find that the failure was caused by a concert of things working together, the only reason to pick one is a personal preference for there to be one thing to pick.
A common pattern is the "straw that broke the camel's back" in a relatively small aberration or mistake which pushes a system from a working state into a phase change to a non-working one revealing all sorts of structural deficiencies. Do you say that last straw was the "cause"? Only if you're particularly short sighted.
I've experienced this personally many times. No we didn't have 3 days of downtime because there was a 5% performance regression in the last release, we had 3 days of downtime because systems A through G were constantly on the edge of collapse and nobody noticed until all hell broke loose.
That's true in engineered systems. It happens because when you make many copies of a thing and use them for a while, all the simple failure modes get worked out.
But societies aren't like that. The world didn't make N copies of the bronze age Mediterranean. If, as Drews suggests, the collapse was a near-deterministic consequence of the overshadowing of expensive chariot-based centralized armies by armored sword bearing infantry forces, then it would have occurred the first time this technology came out (and it became widely known that previously protected rich cities were ripe for the sacking by Joe Conan and friends.)
Who introduced the armored sword-bearing infantry? It's not exactly cheap. If it was an existing civilization, would they not have simply taken over the region, like the Macedonians, Romans, and assorted others? If it were a migration of people, as the Egyptian writings indicate, why were they migrating?
Why were the Egyptians and Babylonians able to survive, albeit with considerable losses, while others such as the Mycenaeans didn't? Why weren't the existing civilizations able to adapt and adopt?
The Egyptians adopted the new fighting style; it's how they defeated the "Sea Peoples" attack in the Nile delta (there are famous depictions of this battle in Egypt).
Babylonia was soon afterward conquered by militaristic Assyria; Assyria also adopted the new approach and, later in the early iron age (as the neo-Assyrian empire), became the dominant superpower in the Fertile Crescent.
The others? The Mycenaeans were closest to the part of Europe where the new sword technology had evolved and matured.
I suggest you read both Cline's book (there's a new edition out) and Drews', and compare and contrast their takes on the subject.
Attributing "cause" to any of them is an exercise in preference of importance. Do you pick the last one that happened? Do you decide which one had the biggest impact? Do you pick the one that was the easiest/hardest to find? It's a silly exercise to pick one. You will find that the failure was caused by a concert of things working together, the only reason to pick one is a personal preference for there to be one thing to pick.
A common pattern is the "straw that broke the camel's back" in a relatively small aberration or mistake which pushes a system from a working state into a phase change to a non-working one revealing all sorts of structural deficiencies. Do you say that last straw was the "cause"? Only if you're particularly short sighted.
I've experienced this personally many times. No we didn't have 3 days of downtime because there was a 5% performance regression in the last release, we had 3 days of downtime because systems A through G were constantly on the edge of collapse and nobody noticed until all hell broke loose.