11 Hypotheses about Unobservables

  • Hume’s theory of causality

    • Hume argued that psychological habits shape our beliefs about causation. Causal ascriptions are inevitably the product of correlations that have been observed in the past.
    • Any underlying causal powers responsible for C causing E must involve a spatial “causal chain” connecting C and E.
    • But we have to accept the fallibility of inferences that depend on correlations alone.
    • Hume’s view is that limits on human knowledge leave us without a way to know if there is any underlying necessity between events A and B. So, we treat repeated correlations between events A and B, where A consistently occurs before B and where A and B are spatially contiguous as if they are causally related. 休谟的观点是，人类知识的局限使我们无法知道事件A和B之间是否存在任何潜在的必然性。因此，我们把事件A和B之间的重复相关性，即A始终在B之前发生且A和B在空间上连续的情况，视为因果关系。
    • Hume’s theory of causality is that we must use correlations as evidence for causality, but beliefs based on this habitual assumption are always defeasible. 休谟的因果理论是，我们必须把相关性作为因果关系的证据，但基于这种习惯性假设的信念总是可以被推翻的。

  • Hume's issues with the limits of our possible knowledge

    • (a) There must be “secret powers” that are causally responsible for observed patterns in the physical world.
      • That Hume accepts claim (a) frankly surprises me. It is a strong metaphysical assumption.
    • (b) Because we can have no direct experience of these unobserved or unobservable causes, even on Hume’s empiricist theory of causation, we can’t know anything about their properties.
      • If we accept claim (b), it seems that we are left looking for, at best, very limited ways to justify claim (a). 如果我们接受(b)的观点，那么我们顶多只能寻找非常有限的方式来证明(a)
        • 1) The events we observe are just “brute facts” for which there is no underlying causal explanation.
        • 2) appeal to common sense analogies to justify (a)
    • Remember that Locke seems to suggest that there are epistemic pathways to knowledge of his version of “secret powers,” i.e. what he identified as the “primary qualities” of physical objects.  While accepting Locke’s hypothesis of primary or secondary qualities, Hume did not accept that there were ways that we could justify our beliefs about them as true.

  • The strategy for justifying beliefs about unobserved causes

    • If E can be explained by C, then that is evidence that C causes E.
    • An explanatory relationship – especially if that relationship involves a deduction of E from C – is evidence for a causal relationship between C and E.
      • basis for one of the most widely used epistemological strategies in scientific inquiry.

  • Hypothetico-Deductive Method:

    • The general form of inference involves hypothesizing an unobserved cause (H) for an observed effect (E), demonstrating that H would explain E, and using this explanation as evidence for H's truth.
    • Discovering unobserved causes involves conjecturing them based on the requirement that they must account for observed effects.
    • The example of Neptune’s discovery illustrates this approach: Le Verrier’s mathematical predictions led to Galle’s observational confirmation, though there is debate about who truly "discovered" Neptune.
    • "Mechanical philosophers" justified their theories of corpuscular structures and motions - Their conjectures about "secret powers" were constrained by metaphysical presuppositions: they were corpuscles of extended matter with different sizes, shapes, motions, and resulting collisions with one another.
    • The logic used for justifying these hypotheses was similar to Le Verrier’s: explaining observable phenomena by deducing them from conjectured causal powers.
    • Hume's analysis and the history of science show that this method is fallible, as scientific theories are often revised or replaced with new explanations that contradict earlier ones.
    • The "hypothetico-deductive" method involves proposing hypotheses about unobserved causes and deducing observable consequences to provide explanations for the phenomena of interest.
      • “hypothetico” part: a major feature of this approach to scientific inquiry is offering conjectured hypotheses about what Hume called the “secret (or at least unobserved) powers” of the unobservable world.
      • “deductive” part: we have provided an explanation of a phenomenon when we can deduce a description of that phenomenon from one or more explanatory assumptions. The assumptions are what explain the phenomenon. Put another way, this view of explanation is that it is a deductive inference.

  • 17th and 18th Century Scientific Theories:

    • Scientists like Descartes, Boyle, Hooke, and Grimaldi used mechanical philosophy to explain light and color as interactions of corpuscles (particles).
    • Different theories (e.g., Descartes' particle rotation vs. Huygens' wave theory) explained optical phenomena, illustrating the method's inconsistency and potential unreliability.

  • Underdetermination Problem:

    • Multiple conflicting theories can explain the same set of observations.
    • The history of science shows theories being replaced as new evidence emerges (e.g., Newtonian vs. Einsteinian explanations for Mercury's orbit).

  • Contemporary Examples:

    • The hypothetico-deductive method persists in modern science, such as astrophysics, where hypotheses about dark matter attempt to explain observed galactic behaviors.