Hume, D. (2006). An Inquiry Concerning Human Understanding, Section IV, Part II. Urbana, Illinois: Project Gutenberg. (Original work published in 1748)
A standard philosophical view is that knowledge beyond sense perception can be acquired by way of inference.
This discussion on knowledge beyond perception depends on assuming that at least some of our beliefs about external objects are based on our perceptual experience.
“None but a fool or a mad man will ever pretend to dispute the authority of experience, or to reject that great guide to human life.”
Hume is not denying the importance of inductive inferences on the basis of past experiences, but there is a problem of the philosophical justification of induction.
the problem of induction = what rational basis do we have for believing that the causal relations we’ve experienced in the past will persist into the future? we cannot justify it deductively and we cannot justify it inductively.
Foster-proposed-solution-to-the-problem-of-induction noted that observed regularities are evidence of natural necessity, and the inference to laws embodying natural necessity justifies conclusions that past regularities will persist in the future.
In contrast, for Hume, observed regularities brings us to expect that future cases will be like past cases but they don’t justify these expectations. Hume’s view is because of his concept of causal necessity that necessity is just a projection of our subjective feelings of expectation onto the world.
Limits of perceptual knowledge
- Predictive beliefs = it cannot be justified on perceptual experience along as it is a belief about events that haven’t occurred yet
- generalizations = these claims have implications regarding not only observed cases but also cases that have not been observed
Both predictive beliefs and generalizations tend to be made on the basis of extrapolation from observed cases.
Induction as causal reasoning
Hume argues that our inductive reasoning is grounded in our idea of Causality. When we see one event regularly followed by another, we form the belief that the first caused the second and we can then make predictions.
on what basis do we make all our causal inferences and predictions about the world? Hume’s answer is past experience.
In Humean-theory-of-causation, the metaphysical claim that necessary-connection between cause and effect is not real but psychological leads to his epistemological view that our confidence in induction is not rationally justified, but rather a product of habit formed by repeated experiences of conjoined events.
Hume notes that experience provides us with an answer to how we go about making causal inferences, but it is more complicated to explain exactly how our limited experience provides justification for these kinds of inferences. i.e. what kind of reasoning would allow us to draw conclusions about unobserved cases on the basis of observed cases?
“As to past Experience, it can be allowed to give direct and certain information of those precise objects only, and that precise period of time, which fell under its cognizance: but why this experience should be extended to future times, and to other objects, which for aught we know, may be only in appearance similar; this is the main question on which I would insist.”
Principle of the uniformity of nature
Hume notes that in making an inference from observed cases to unobserved cases, there seems to be a missing premise called the principle of the uniformity of nature1 = premise that nature is uniform in the sense that regularities that have occurred in the past will persist in the future, or as Hume puts it, “the future will be conformable to the past”.
This makes an inductive argument into a deductive one. Example:
- Object of type X have always been observed to be accompanied by effects of type Y
- principle of the uniformity of nature = the future will be conformable to the past Conclusion: Therefore, in the future, any object of type X will be accompanied by effects of type Y
The principle of the uniformity of nature acts as bridge between the observed cases to the unobserved.
Why believe in the principle?
Hume notes:
- the principle is not a tautology (necessary truth)
- it is not an a priori
Hume concludes that the principle, if justified at all, would have to be justified based on past experience. And would need an argument like:
- In the past, some regularities have been observed as persisting over some stretch of time Conclusion: Therefore, past regularities will persist into the future
However, by arguing from past experiences, it forms an inductive argument to defend the principle of the uniformity of nature.
Problem of circularity = in order to justify inductive arguments, the principle of the uniformity of nature is required to bridge the gap between the premises about that we have observed to the conclusion about cases that we haven’t observed. However, in order to justify the principle itself and inductive argument is needed. This makes the whole approach viciously circular / it seems to “beg the question”
i.e. that the attempt to justify inductive inference already relies on inductive inference
causal powers as evidence
from the regular connection that we’ve observed between fire and smoke, can we not infer that fire has a causal power to produce smoke, generally?
Even if we infer that objects have causal powers from past experience, our belief that those powers will continue to operate in the same way in the future already presupposes what we’re trying to prove — that the future will resemble the past.
Hume’s positive view of Induction
Hume critiques that inductive inferences are not rationally justified, but he believes that we do and we have to make inductive inferences.
“… But still he finds himself determined to draw [this inference] … There is some other principle which determines him to form such a conclusion. This principle is custom or habit.” - Section V, Part I, An enquiry concerning human understanding
For Hume, like his idea of necessary-connection, our inclination to make inductive inferences is due to the habit bred in us by repeated experiences of past regularities.
For Hume, inductive inferences provide no rational basis for beliefs beyond experience, even though we’re psychologically determined to form those beliefs on the basis of evidence that we have.
Footnotes
-
Hume does not use this name particularly, but this name is used in literature of induction ↩