28.6 Are Processes an Exception to Hume's Epistemic Principle?
David Hume famously made the following claim about our knowledge of causation in his Enquiry Concerning Human Understanding (Hume 1748):
Hume's Epistemic Principle. We cannot infer with certainty and on purely a priori grounds the existence of a causal connection between two events.
Hume offers, in defense of his principle, a set of thought-experiments that seem to show that we can always imagine any event's having a different result from the one we in fact observe. Hume assumes that if we can coherently imagine a claim p to be false, then we cannot know, on purely a priori grounds, that p is true.
Causal processes like locomotion (conceived of as continuous movement through space) may provide counterexamples to Hume's Epistemic Principle. Let A be the event of a material body's approaching closer and closer to location l as time draws closer and closer to time t. Let B be the event of that same body's being located at l at t. In this case, we can surely infer with certainty and on a priori grounds that A is the cause of B.
In response, Hume would have to ask us to imagine a world in which both A and B occur without being part of the same causal process of locomotion. We can imagine a world in which A occurs without B's occurring. Maybe the body approaches position l and vanishes exactly at time t, never actually arriving at l. This could involve the body's ceasing to exist at t, or it could involve the body's being instantaneously teleported to some remote location at t. Similarly, we can imagine B occurring without A (or anything remotely like A). We could imagine the body being created at time t in location l, or we could imagine its being instantaneously teleported from some remote location to l at the instant t.
However, Hume asks us to imagine both A and B occurring, yet without a causal connection between the two. In order to accomplish this, we would have to imagine the body's being teleported away at t, so that it never arrives at l by the ordinary process of locomotion. In addition, we would have to imagine the body's also being teleported away from this remote location to l in such a way that it arrives in location l precisely at time t. In order to do all this, we would have to imagine the body's being in two, widely separated place at time t: both in some remote location (as a result of the “first” teleportation) and in location l (as a result of the “second” teleportation). This is bad enough, but in addition we must imagine two simultaneous events of teleportation, one moving the body to some remote location r and a second moving the body from that remote location r back to l. However, it seems absurd to suppose that there could be two simultaneous actions, the second of which requires as a precondition the completion of the first action.
28.7 Conclusion: The Consequences of Causation
We saw in Chapter 26 that causation is needed in accounting for a wide range of phenomena—knowledge, semantics, modality, science, decision theory, agency—as well as being deeply embedded in our common sense view of the world. These facts strongly favor theories that treat causation and causal powers as fundamental features of the world, since all anti-realist and reductionist theory face the very heavy burden of providing an account of causation that is adequate for all of its applications.
In particular, these last three chapters (26–28) tip the balance in favor of Strong Powerism and against its rivals, especially Neo-Humeism (4.4T) and Nomism (4.4A.2). The latter two theories are inextricably tied to Causal Explanationism (27.1A) and share in its struggle to explain the real linkage between particular causes and effects (especially in cases of late preemption and trumping—Section 27.1.1) and the asymmetric direction of causation and time (Section 27.1.2), phenomena that Powerism, especially when combined with Temporal and Procedural Intervalism (19.1A and 19.2T) and Causal Connectionism, can readily accommodate. The only serious drawback for Causal Connectionism is the difficulty that that theory faces in explaining negative causation. However, as we have seen (in Section 27.2.2.2), a number of plausible are available to the Connectionists, including the selective reifying of some absences and the use of Totality Facts.
As this chapter has shown, the package consisting of Powerism, Procedural Intervalism, and Connectionism (which we might call ‘the Aristotelian package’) offers the only way of accommodating both discrete and continuous causation, both of which seem to play an indispensable role in our world.
Notes
1. We are assuming in this definition that causation is transitive. If it isn't, then we can stipulate instead that there be a finite chain of causal links from x to z and z to y, or vice versa.
2. In modern mathematical terminology, this ordering would be called ‘dense’ rather than ‘continuous’. We use the word ‘continuous’ here to mark a connection with Aristotle and Brentano.
29
Conclusion: The Four Packages
In this course of this book, a pattern has emerged. There are two quite coherent packages of answers to the issues that we've discussed: a neo-Humeist or Ludovician package (championed by David Lewis), and a neo-Aristotelian package. Ludovicians favor the following answers:
ludovician package
Truth supervenes on being.
Neo-Humeism about powers (the Humean mosaic as fundamental).
Nominalism—either Ostrich or Resemblance, with or without tropes.
Primitive identity and similarity relations.
Intentionalism, with a causal-representational theory of perception.
Mereological Universalism.
Modal Concretism, with Counterpart Theory.
Anti-Tensism (the B Theory).
Spatial and Temporal Pointillism and Infinitism.
Composition as Identity, and Compositional Equivalence.
Perdurantism, of either the worm or stage variety (or a hybrid of the two).
Causal Explanationism, including the Counterfactual Theory of causal explanation and the Mark Transmission theory of processes.
What unifies these 12 theses are two fundamental commitments: to Ockham's Razor and to the sufficiency of modern physics. Ludovicians put little weight on common sense beliefs, especially when they are embedded in ethical and legal practices, and they do not rely heavily on the “manifest image of the world” (to use Wilfred Sellars's phrase): that is, they do not read too much into how things appear to us. In addition, Ludovicians share the proclivity of much of modern science to explain things in a bottom-up fashion, privileging the very small and very short-lived over the large and enduring.
ARISTOTELIAN PACKAGE Aristotelians favor positions on this issues that are typically inconsistent to those of the Ludovicians:
Atomic or Spectral Truthmaker Theory, with the One Truthmaker per Fundamental Property principle.
Strong Powerism.
Constituent Realism, with universals as parts of their instances.
Individuation by bare particulars or by a structure-relational (amphibian) form of Bundle Realism.
Direct Realism, with Perceptual Dualism or Meinongian direct realism.
Organicism (with emergent biological and social powers).
Modal Abstractionism, with an Aristotelian theory of possibility, actuality defined in terms of metaphysical fundamentality, and Transworld Identity.
Tensism (the A Theory), usually in a modal form (Changing Possibility or Changing Actuality, or both).
Moderate Anti-Pointillism and Temporal Intervalism (with dependent boundaries), along with Finitism.
Aristotelian Compositional Reductionism.
Classical Endurantism.
Causal Connectionism.
Like Ludovicians, Aristotelians are interested in discovering elegant and powerful theories in metaphysics—theories that can be the basis of real metaphysical explanations. Consequently, they share with Ludovicians a commitment to Ockham's Razor. However, Aristotelians seek to explain a broader and richer body of data, a body that includes insights into our selves and even into the nature of the world of matter that go beyond the formal conclusions of theoretical physics. In a
ddition, Aristotelians rely more heavily on our semantic intuition about what could possibly be the truthmakers for familiar kinds of claims, including those involving tense and modality. They are, therefore, less willing to embrace radically revisionary semantic theories, even when these theories assign the right truth-values.
We leave it as an exercise for the reader to trace all of the interconnections and dependencies among the positions in each package. We are not claiming anything as strong as that these are the only two logically possible positions. In fact, there are at least two alternatives, each deviating from both the Ludovician and Aristotelian package in several ways.
First, there is a more micro-physicalist approach that deviates from the Aristotelian package by taking physical simples to be the only fundamental entities and which embraces Pointillism, Instantism, and Anti-Tensism, as represented by David Armstrong or John Heil. This view differs from Ludovicians by embracing either universals or trope theory, along with a Constituent Ontology, and by relying on Atomic Truthmaker Theory. We will call this group the ‘Fortibracchians’, translating Armstrongs's name into Latin (‘fortis’ for ‘strong’, and ‘bracchium’ for ‘arm’).
Second, there is a more Platonic approach that deviates from the Aristotelian package by rejecting the truthmaker principle altogether and eschews nearly all talk of grounding and fundamentality. This sort of philosopher makes little or no appeal to Ockham's Razor or ontological economy. They do not think of metaphysics on the model of a theoretical science, like physics, astronomy, or chemistry. Rather, the metaphysician's job is almost entirely a priori, to be performed from the philosopher's armchair. We don't associate this package with any one paradigmatic philosopher, as we do the other three (Aristotle, David Lewis, and David Armstrong). Instead, it is a kind of composite image, which a number of philosophers resemble to some degree, such as Roderick Chisholm, Trenton Merricks, Alvin Plantinga, George Bealer, or Peter van Inwagen. We will call the position, for want of a better term, ‘Quietism’, in the sense that its proponents do not aim to make revolutionary discoveries of a theoretical nature but rather to organize, synthesize, and harmonize the elements of reality that appear clearly to us in our most reflective moments.
In Table 29.1, we've listed the thesis most likely to be adopted by each package on each of the issues considered in this book. In most cases, the association is quite clear, but there are several that are more debatable. We've listed the Aristotelian package as including Meinongianism, ontological vagueness, and Tensism on the passage of time. It is certainly possible to be an Aristotelian and to adopt the opposite positions on these theses, but we think that Aristotelians should be somewhat more open to them than those committed to the other three packages, and so associating them with the Aristotelian package helps to further highlight the differences between the four.
Table 29.1 Positions of the Four Packages
Ludovician Aristotelian Fortibracchian Quietist
Truth Truth Supervenes on Being (2.1A.1A.1T) Classical (2.1T) or Spectral Truthmakers (2.1A.1T.1) Atomic Truthmaker (2.1T.4) No truthmaking (2.1A.1A.1A)
Undefinable Grounding? Yes (3.1T.1T) Yes (3.1T.1T) Yes (3.1T.1T) Definable? (3.1T.1A)
Powers Neo-Humeism (4.4T) Strong Powerism (4.4A.3) Strong Nomism (4.4A.2) Hypotheticalism (4.1T)
Universals Extreme Resemblance Nominalism (8.1T.4.1T) UP Realism (7.1T.1T) UP Realism (7.1T.1T), or Moderate Nominalism (8.1T.4.1A) Ostrich Nominalism (7.1A.1A)
Particulars Constituent Ontology (9.1T) Constituent Ontology (9.1T)
Relations Reductive Separatism (10.1A.1A.1A) Constituent Connectionism (10.1A.1T.1T.1T.1A) Constituent Connectionism (10.1A.1T.1T.1T.1A) Ostrich Separatism (10.1A.1A.1T)
Non-Existence Possibilism (12.1A.1T) Meinongianism (12.1A.1A), or Possibilism (12.1A.1T) Possibilism (12.1A.1T) Actualism (12.1T)
Vagueness Multiple Meaning Theory (12.2A.1T) Ontological Vagueness? (12.2A.1A)
Perception Intentionalism (13.3A.1A.1A.2) Perceptual Dualism (13.3A.1A.1T) or Meinongian Direct Realism (13.3A.1A.1A.1) Intentionalism (13.3A.1A.1A.2) Perceptual Dualism (13.3A.1A.1T)
Possibility Concretism (14.1T.1T) Aristotelian Modality (15.2T.7) Combinatorialism (15.3T) Magical Abstractionism (15.1T.1T)
Actuality Modal Indexicalism (14.2A) Modal Anti-Indexicalism (15.2T) Modal Anti-Indexicalism (15.2T) Modal Anti-Indexicalism (15.2T)
De Re Modality Counterpart Theory (16.1A.1) Transworld Identity (16.1T) Transworld Identity (16.1T) Transworld Identity (16.1T)
Space and Bodies Spatial Monism (17.1T.1A.1A) Aristotelian Relationism (17.1A.1A) Body-Space Dualism? (17.1T.1A.1T)
Regions and Boundaries Spatial Pointillism (18.1T) Volume-Boundary Dualism (18.1A.2A)
Bodies Infinitism (18.4A) Material Anti-Pointillism (18.3A), Finitism (18.4T)
Instants Instantism (19.1T) Moderate Intervalism (19.1A.1A), Temporal Finitism (19.3T)
Processes Procedural Instantism (19.2T) Procedural Intervalism (19.2A)
Time's Passage Reductive Anti-Tensism (21.2A.1A) Tensism (20.2T)? Reductive Anti-Tensism (21.2A.1A) Tensism (20.2T)?
Future/Past Eternalism (20.2A.2T) Indeterminacy of Future Contingents (20.3T) Eternalism (20.2A.2T)
Compositional Free Lunch? Compositional Equivalence (22.1A.1T.1) Universal Bottom-Up Priority (22.1T.1T)
Heaps Fundamental Heapism (22.2T) Anti-Heapism (22.2A) Anti-Heapism (22.2A)
Point-Parts Priority of Spatial Point-Parts (22.3T) Non-Priority of Spatial Point-Parts (22.3A)
Special Composition Question Mereological Universalism (22.8T.1A.2) Homogeneous Continua Plus Organisms (22.8T.1A.10), Organicism (22.6T), Anti-Artifactualism (22.5A) Priority Atomism (22.7T), Universal Atomism (22.4T) Organicism (22.6T)
General Composition Question Composition as Identity (23.1T.1.1) Aristotelian Compositional Reductionism (23.1A.1T.1A.1) Arbitrary Compositional Reductionism (23.1A.1T.1T)
Persistence Classical Perdurantism (24.1T.1T.1A.1T) Classical Endurantism (24.1T.1T.1T.1T) Classical Endurantism (24.1T.1T.1T.1T) Classical Endurantism (24.1T.1T.1T.1T)
Constancy Ramsey-Sider- Lewis Perdurantism (24.3T.3) Mereological Inconstancy (25.1A) Mereological Constancy (25.1T) Mereological Inconstancy (25.1A)
Motion At/At Theory (24.5A.1T) Motion Intervalism (24.5A.1A)
Causation Causal Explanationism (27.1A) Existence of Uncaused Causes (26.1T.1A.1T) Principled Causation (26.1T.1A.2T), Causal Connectionism (27.1T) Causal Explanationism (27.1A)
Table 29.2 Principles Appealed to by the Four Packages
Ludovician Aristotelian Fortibracchian
PMeth 1 Ockham's Razor 18.1T, 24.1T.1T.1A.1T 17.1A.1A
PMeth 1.1 2.1A.1A.1T, 4.4T 2.1T, 4.4A.3 2.1T.4
PMeth 1.2 18.1T, 22.2T 7.1T.1T, 18.4T, 19.3T, 22.2A 7.1T.1T
PMeth 1.3 4.4T 4.4A.3
PMeth 1.4 8.1T.4.1T, 17.1T.1A.1A 7.1T.1T 7.1T.1T, 15.3T
PMeth 1.4.1 8.1T.4.1T 7.1T.1T 7.1T.1T
PMeth 2 Scientific Realism 18.4A, 20.2A.2T 20.2A.2T
PMeth 2.1 4.4A.3 4.4A.2
PMeth 2.2 4.4A.3 4.4A.2
PMeth 3 Structuralism 18.4T
PMeth 4 Redundancy 22.2A
PEpist 1 Imagination Guide to Possibility 4.4T, 19.1T, 20.2A.2T, 22.1T 4.4A.3, 18.1A.2A, 19.1A.1A 4.4A.2, 20.2A.2T
PEpist 2 Common Sense 20.2T
PEpist 4.1 Sensory Error Minimization 13.3A.1T, 13.3A.1A.1T
PTruth 1 One Truthmaker per Fundamental Property 7.1T.1T, 18.1A.2A 7.1T.1T
PMeta 1 Test for Grounding 4.4T 4.4A.3 4.4A.2
PMeta 2 Intrinsic Powers 4.4A.3. 19.3T, 20.2T
PMeta 5.2 Patchwork Principle 18.4T, 19.1A.1A, 20.2T, 24.1T.1T.1T.1T
PNatPhil 1 Possible Gunk 22.2T, 22.6T.1A.2 22.8T.1A.10
PNatPhil 2 Motion and Substantial Change Independent 22.2T 18.1A.2A 22.4T
PNatPhil 3 Continuous Motion 19.3T 22.4T
Correspondence Theory of Truth (2.2) 2.1A.1A.1T 2.1T 2.1T.4
No Metaphysical Cheaters (2.2) 2.1A.1A.1T 2.1T 2.1T.4
&
nbsp; Principle of Indifference (5.2.3) 4.4A.3 4.4A.2
Priority of the Actual (15.1.2) 15.2T.6 15.2T.6
Acquaintance Model (15.1.2) 14.1T.1T 14.1T.1A.1 15.3T
Physicalism (22.6.2) 22.2T 22.7T
Connected Occupation (22.4.2) 22.1T
Table 29.3 Special Principles Appealed to by Aristotelians, with Associated Theses
Principle of Indifference (5.2.3) 4.4A.3
Phenomenal Exportation (13.3.1) 13.3A.1T. 13.3A.1A.1T
Pure Physicality of Physical Things (13.3.2) 13.3A.1T
Physics Carves Nature at Joints (13.3.2) 13.3A.1T
Branch (15.3) 15.2T.7
Infinite Points in Space (18.3.1) 18.4T
Symmetry (18.3.3) 18.1A.2A
Unlimited Divisibility (18.3.3) 18.1A.2A
Maximum Velocity (19.1.2) 19.3T
Impotence of Identity (19.1.2) 19.3T
Spatial Occupancy (22.4.1) 22.2A
Intrinsic Composition (22.5.1) 22.4A
Reasonable Reliability (25.2.6) 24.1T.1T.1T.1T
Cogito ergo sum (22.6.1, 25.1) 22.6T
No radically discontinuous motion (24.3.2) 24.5A.1A
Ludovicians are centrally committed to Neo-Humeism. Consequently, they reject universals, which would fit most naturally with Nomism. Ludovicians could be Ostrich Nominalists, but given their commitment to economy and given the resources of modal concretism, extreme resemblance nominalism seems the best fit. The Humean mosaic is a bottom-up picture, so Priority of Point-Parts follows, as does Instantism and Pointillism. Neo-Humeism is also consistent with gunk, so heaps should be as fundamental as atoms. The mosaic picture also leads most naturally to Anti-Tensism and to Ramsey-Lewis-Sider perdurantism.
Fortibracchians' core commitments are to Nomism and against Extreme Nominalism. These commitments fit well with a Constituent Ontology. On the question of properties, some Fortibracchians (like Armstrong himself) embrace universals only, but others adopt tropes instead of universals or a four-category ontology (with both universals and tropes). Given the importance of the immediate inclusion of universals and tropes in particulars for Constituent Ontologists, and given the Fortibracchian commitment to bottom-up grounding, gunk is a problem. The best metaphysical mereology for Fortibracchians is, therefore, Priority Atomism. Since the actual laws of nature, as formulated by scientists, refer to persisting particles, some form of endurantism is best. Why Anti-Tensism? Because science has no need of a metaphysically privileged present. Given the available resources of universals and laws of nature, Combinatorialism would be the best Fortibracchian account of modality.
The Atlas of Reality Page 101
