Historically, most scientists have taken the disciplined approach of not even considering their work as concerning causation. Causation was thought to be a question for priests and philosophers. That's much of the reason behind the scientific method being carefully arranged to only use measures to predict measures, offering purely descriptive rather than explanatory results. How you predicted things is all that could be explained. Knowing how things actually happen didn't seem to matter that much anyway.
So... if it doesn't really matter, why does it matter? Because scientists have started trying to cross that carefully constructed boundary, but tending to do it by resorting to a metaphysics of mathematical rules, i.e. making a mess of it. Studying that mess is important. If the history of science teaches anything it is that logical contradictions are fruitful to explore, and the metaphysics of modern science is full of them. Nature is described as an interaction of measures, but measures absolutely never physically interact, only things do. Nature is described as random, but that usually refers to distributed classes of events, not any demonstration that any particular event is. When you carefully study the course of individual events, everything you can actually find is discretely continuous. What we need to explore are the tell tale signs that might show us what these contradictions can tell us about ourselves and our world. .
It may be taking some liberty to characterize physical science as representing two primary views of causation, a mechanical and a mathematical one. Accepting this momentarily, though, allows characterizing a third view, local evolution. Physical processes can be understood as developing right where they are found, evolving locally as they occur, rather than being determined from afar.
Looking closely at how individual events develop seems to make this third view ever more necessary. There is simply nothing else but local interaction present, anywhere, and local events on every observable level take shape rapidly, smoothly and intricately. When a finger touches a key on a keyboard, the interaction seems to begin in the fields of the least sub-atomic parts which then communicate through their respective universes by way of great chains of super organized atomic structures (i.e. body parts). The responses delivered by the least parts in contact develop from interactions taking place throughout the whole hierarchy of systems they are part of. It's not rules, particularly not our own rules, that are operating, but highly complex locally developing interactions. One of the tell tale signs of how it works is the usual presence of growth curves (smooth non-linear transitions) connecting different trends.
Yes it appears strange, but other scientific ideas that first appeared strange but logically necessary include the discovery of the solar system, the evolution of species and relativity. Concerted investigation led to first reinforcing their apparent necessity and then to confirming evidence. One of the tell tale signs is that nature is densely populated by individually unique transient connecting processes (mysteriously appearing vanishing exponential like transitions). They have been given remarkably little attention. They seem of little importance, being of such variety and irregularity, hard to study, and soon gone anyway. They're just what's happening, that's all!
Despite its magnificent power, the statistical / mechanical view has long been found quite useless for explaining the unique individual character of particular events. A central purpose of the statistical method is that it allows that question to be avoided entirely. If studying the local origins of particular events, however, represents a 'better' view for considering causation it certainly could not avoid the need to explain the presence of predictable mechanisms. I think we can watch them develop and learn how it happens.
First, though, one needs to learn how to 'see' again. That might seem a little pompous, but there is clearly a kind of retraining of perception required, a matter of learning to look at physical events with the idea of documenting the actual steps of their occurrence. That means using the eye of a medical artist, paleontologist, geologist or historian in basic physics.