Judith: In one of the presentations I attended of my father's-- I can't remember if it was in Linz, Austria or Santa Fe, New Mexico (USA)... but he was talking about the limits and the goals of science... and he said, "We all have to agree on two things: First, that there are consistent causal pathways in all phenomena in the natural world and Second, that our minds and senses are capable of perceiving and interpreting meaningful information about these causal pathways. If either of these two things are not true, then we can all go home."
HP: Of course I agree there are useful models of the natural world, but I would say that only the simplest, usually artificial, models are viewed as linear causal pathways. The real world is more like a network of forces all interacting at once in which no isolated causal path makes sense. As an exercise, try thinking of an event that can be fully explained by only one proximate cause or even a chain of causes.
Howard
Bob Ulanowicz has done lots of work on multiple influences, webs of causality, much like Howard is saying. Bob borrows from Popper and generalizes causes and forces - linked events of conditional probability equal to 1 (if A then B ever and always) - into "propensities" - linked events of conditional probability less than 1 (if A then sometimes B, C or D). He uses information theory in interaction networks (like ecosystem/trophic networks of matter/energy flux) to get at what Popper said is needed - "a calculus of conditional probabilities". Via this framework, other configurations beside objects can have "agency", such as autocatalytic cycles in networks. Cause, force and agency all become more non-local, distributed in space and in time, topological/relational and field-like than local and particle-like as in billiard ball A hits billiard ball B and causes it to move.
I can cite of Bob's some papers if anyone is interested. His own home page has all his papers listed also.