In our hearts, we want simple causation like X hits Y, causing Y to hit Z, with no exceptions or fudge-factors, in absolute accordance with the physical laws of nature.  In a murder mystery, the physical causation must make sense.  I recall the Ngaio Marsh mystery, A Man Lay Dead, where Inspector Alleyn surmised that the murder could only have happened if the victim, Mr. Rankin, was standing at the base of the bannister with his back to the bannister, fixing a cocktail from the tray just there, and the killer, Mr. Wilde, slid face-first down the bannister and somehow, at the bottom, grabbed a knife from a leather strip on the wall, and while still sliding, with his right hand, drove it home into the victim’s back, who then fell face forward striking the gong, the gong originally having sounded the beginning of the novel.  And there you have it.  That and a confession is how you convict a killer.

Mechanical causation is orderly.  The arrow of time moves forward and cause comes before effect.  When I throw a rock at my dog, the rock leaves my hand, flies in the air, hits dog, dog goes “yipe!”  That is, the rock follows its fixed trajectory in space without teleporting to another location, and the dog doesn’t go “yipe!” before the rock hits him.  Einstein explains why: the speed of light.  In Einstein’s system, light speed must be the maximum speed that anything can travel in the universe.  To teleport in space or go back in time is to exceed the speed limit.  For example, if the sun explodes, we on Earth have 8 minutes before the initial light waves fry us, because that’s how long it takes for light to reach Earth from the sun.  8 minutes to live? Ice cream!  Of course, we wouldn’t know that the sun exploded until the 8-minute mark because that’s when the initial light waves reach us.  Let’s pretend we have Bruce Willis on permanent standby.  We send Bruce racing to the sun immediately at the 8-minute mark.  Bruce, that lazy bum, can’t go faster than the speed of light, so if it took 8-minutes for the initial light to reach Earth from the sun, it’ll take Bruce, that fat slob, another 8-minutes to get to the sun.  Now 16-minutes have elapsed since the sun exploded.  Bruce is late by 16 minutes.  You had one job, Bruce, one freakin’ job, and you screwed it up.

Classical, mechanical causation works great for things in ordinary experience, things we can see with our eyes and touch with our hands.  Even better, it’s mathematical.  Newton and then Einstein created geometric systems for tracking things in space and time.  To track the flight of a cannonball, use Newton.  A GPS system uses Einstein’s general relativity to tell me where I am.  Both systems can accurately track the billiard balls on a pool table. 

Can we string out a mechanical chain of causation that goes from atom to molecule to billiard ball, and track their movements to the end of time?  Logically, yes.  Mechanist causation says that, writ large for all matter in the universe, if we have sufficient knowledge of all particles, we can predict the movement of everything in the future, forever, in a long, deterministic causal chain.  Don’t work though.  I can’t build a causal chain from an individual atom because the atom jumps around unpredictably.  Atoms have random movement, and that cuts a mechanical chain.  Randomness is everywhere, from jumpy atoms to organisms to complex systems.  In response, science has fled from the individual and its random behavior.  Scientists prefer the security of statistical probabilities.  To make predictions now, we look at aggregates of individuals and calculate probabilities based on the behavior of the aggregate.

The problem with mechanical causation is that its application is narrow.  It’s a delicate mental construct that works great for simple things like billiard balls bouncing around on a pool table, but it won’t support complexity.  It only works at one level of reality, in a thin cross-section, but the real world has levels upon levels of complexity piled on top of and interpenetrating one other.  Imagine that when the cue ball hits the 8-ball, I drill down to the quantum level, or I expand up to the geopolitical level of nations striving for dominance.  That’s the real world of complexity, and mechanical causation can’t handle the load.  I’d go crazy trying to put it all together. 

In brief, classical mechanics works in limited and controlled settings, and otherwise should stay in its lane.  Hence my theme: a causal theory only works when matched to the proper level of complexity.  There’s no one theory to rule them all.  Instead, we match the right tool to the job, the right causal theory to the right level of complexity.