A biologist says: 'Natural selection (a population-level process) caused the evolution of antibiotic resistance (changes in molecular DNA).' A strict bottom-up physicalist would respond:
AThis is a genuine case of downward causation, since selection operates at the population level
BThis is acceptable shorthand for complex upward chains of physical causation at the molecular level
CSelection is not a cause at all — it is merely an observation about which organisms reproduced
DThis statement violates the conservation of energy
A strict bottom-up physicalist holds that the real causal story is told entirely at the micro-level: specific chemical reactions, molecular interactions, and physical processes. 'Natural selection caused X' is a convenient high-level description, not a genuine higher-level cause. Whether this is mere description or genuine causation is exactly what the debate over downward causation is about — the physicalist says the former.
Question 2 Multiple Choice
The 'causal closure of the physical' principle creates a problem for downward causation because:
AIt implies that higher-level properties do not exist
BIt implies that every physical effect has a sufficient physical cause, leaving no causal gap for an additional higher-level cause to fill
CIt implies that causation must always proceed forward in time
DIt is incompatible with the existence of emergent properties at any level
Causal closure says: for every physical event E, there is a sufficient physical cause of E. If my intention (a higher-level property) causes neuron N to fire, and prior neural activity (a micro-level cause) already sufficiently caused N to fire, then we appear to have overdetermination — two independently sufficient causes. This is the problem. The response that macro and micro descriptions are just different descriptions of the same process avoids overdetermination by denying they are two separate causes.
Question 3 True / False
Multiple realizability supports the claim that higher-level descriptions can do genuine causal work that is not fully captured by any particular physical realization.
TTrue
FFalse
Answer: True
If the same mental type (e.g., pain) can be realized by different physical substrates (mammalian neurons, octopus neurons, or hypothetically silicon), then no single physical description captures what pain causally is. The higher-level description (pain → withdrawal) predicts and explains behavior in a way that any particular physical description tied to one substrate would not generalize across. This suggests the higher-level description is doing real explanatory and potentially causal work.
Question 4 True / False
Downward causation is straightforwardly impractical if the physical causal closure principle is true.
TTrue
FFalse
Answer: False
Some philosophers argue that macro-level and micro-level descriptions are simply different descriptions of the same underlying causal process — the intention just *is* the neural pattern, described at a different grain of analysis. On this view, there is no overdetermination and no violation of causal closure, because the higher-level cause and the micro-level cause are identical, not competing. Whether this identity claim is satisfying or just redefines the problem is contested, but it shows that causal closure does not straightforwardly rule out all forms of downward causation.
Question 5 Short Answer
How can a defender of genuine downward causation respond to the charge that it violates physical causal closure, and what is the strongest version of this response?
Think about your answer, then reveal below.
Model answer: One response is the identity strategy: the higher-level cause (an intention) just is the neural pattern described at a different level of abstraction — not a second cause but the same cause redescribed. This avoids overdetermination while preserving the claim that the higher-level description matters. A stronger response appeals to genuine emergence: some higher-level properties have causal powers their individual components lack (like water's surface tension), so the whole system has causal efficacy that cannot be reduced to the sum of the parts' interactions. The emergentist accepts that causal closure must be revised or that emergence adds new causal structure to the world.
The debate turns on whether 'macro-level cause' and 'micro-level cause' pick out the same event or different events. If they pick out the same event, there is no overdetermination and no conflict with causal closure. If they pick out distinct events, then either causal closure is false or one description is mere redescription. The question of whether genuine downward causation is possible is ultimately a question about whether higher-level organization adds anything causally real to the world.