When one particular species of event has always … been conjoined with another,
we make no longer any scruple of foretelling one upon the appearance of the other,
and of employing that reasoning, which can alone assure us of any matter of fact or existence.
We then call the one object, Cause; the other, Effect.
- David Hume's (1748, §7) An Enquiry Concerning Human Understanding
Metaphysics of Causation v. Causal Epistemology
Metaphysics of Causation
X causes Y
S1: X causes Y.
S1 is an example of a causal claim
If S1 is true, then:
i) X and Y are causal relata
ii) There is a causal relationship between X and Y
The metaphysics of causation is concerned with:
i) The intrinsic nature of causal relationships
ii) The intrinsic nature of causal relata such as X and Y
Can causal relata include properties, objects, processes, omissions, facts, mental states, etc?
Questions about the intrinsic nature of causal relationships include:
Q1 (Connection): How do causally related and causally unrelated sequences differ?
Q2 (Direction): How do sequences related as cause to effect differ from other sequences (e.g. effect to cause, joint effects of a common cause, etc)?
Q3 (Selection): How do sequences involving causes differ from sequences involving mere background conditions?
Questions about the intrinsic nature of causal relata include:
Q1 (Immanence): Are causal relata immanent (i.e. spatiotemporal) or transcendent (i.e. non-spatiotemporal)?
Q2 (Individuation): How fine- or coarse-grained must causal relata be?
Q3 (Adicity): How many causal relata must there be?
Possible RESPONSES to Q3 (Adicity):
Causal Graph
Description
Causality is a dyadic relation with only primary relata X and Y: X causes Y (traditional)
Causality is a triadic relation (including an effectual difference) with only primary relata X, Y, and Y′: X causes Y rather than Y′ (van Fraassen, 1980)
Causality is a triadic relation (including a causal alternative) with only primary relata X, X′, and Y: X rather than X′ causes Y (Hitchcock, 1993)
Causality is a 4-adic relation (including a causal alternative and an effectual difference) with only primary relata X, X′, Y, and Y′: X rather than X′ causes Y rather than Y′ (Schaffer, 2005)
Causality is a triadic relation with primary relata X and Y and a secondary relatum M: X causes Y relative to M, where M is a suitable causal model for the situation (Pearl, 2000)
Causality is a triadic relation with primary relata X and Y and a secondary relatum N: X causes Y relative to N, where N is an ordered pair of natural outcomes relative to X and Y (Hart & Honoré,1985)
Name of Position
Description
Causal generalism
In Domain R:
A1 causes B1 at t1
A2 causes B2 at t2
A3 causes B3 at t3
⋮
In Domain S:
X1 causes Y1 at t1
X2 causes Y2 at t2
X3 causes Y3 at t3
⋮
Causal generalism is the traditional view according to which causal relationships are instances of a general causal regularity or law
∴ A (or X) causes B (or Y)
Causal pluralism is the view according to which the domains or systems in which causal relationships exist (viz. Domain R, Domain S, etc) are metaphysically diverse
A causes B in R
X causes Y in S
As the search for a unified theory of causation may be unlikely to succeed, causal pluralists will recommend a plurality of methodologies, causalities, etc
Causal singularism (Anscombe, 1971)
Causal singularism is the view according to which causal relationships exist in the singular between the event-token pairs:
A1-B1, A2-B2, A3-B3, X1-Y1, X2-Y2, X3-Y3
Causal singularists could remain agnostic about whether causal laws exist at type level (A-B, X-Y) and across domains (R, S)
Causal eliminativism is the view according to which there are no such things as causal relationships
Causal Epistemology
Causal metaphysics is concerned with the fundamental question of what causality (e.g. in terms of causal relata, causal relationships, etc) is
Causal epistemology is concerned with how causality can be inferred from the observational data
Causal epistemology is concerned with how we can recognize causes and learn about causal relationships
The metaphysics of causation has some implications on causal epistemology:
IMPLICATION 1: If causal pluralism is correct, then there may be a multiplicity of ways in which one can learn of causal relationships and a multiplicity of methods that are each suited to different domains
IMPLICATION 2: If process theories of causation (Salmon, 1994, Dowe, 2000) are correct, then causality involves the transmission of information and requires locality in space-time, making the use of spatiotemporal datasets (e.g. fMRI) more attractive