Skeptical doubts concerning the operations of the understanding 

In this piece, David Hume is able to distinguish how objects of “human reason or enquiry” are either “Relations of Ideas” or “Matters of Fact” (Hume, 1). ‘Relations of ideas’ can be found through “operation of thought” and proved by demonstration without knowing if anything is tangible. This category constitutes math, especially geometry, algebra, arithmetics, and other sciences discussing logic that can be compared. The latter are discoverable through experience, and by proving so Hume dismisses the fact that it can be achieved through deductive or “a priori” reasoning (Hume, 3). This is because it is never possible to find something to be true (acting in a certain way) before it has happened. It is only possible to know the apple will fall downwards by experiencing it. It can also never imply a contradiction, and if it did it would cease to be something distinct to the human understanding. This is to say that any instance of which we have no experience will never fall within our scientific reasoning/logic.

There is a connection in this last sentence to what Barbara Forrest says with regards to Methodological Naturalism (i.e. all theories are tested with respect to real-world events through the scientific method) (Forrest, 4).  Forrest believes that scientific knowledge/method will never be able to “procedurally” (not logically) prove the existence of the supernatural (Forrest, 4).  Although both authors discuss the philosophy of science at various levels, they have some similarities with respect to how causality plays a role in methodological naturalism or ‘Matters of Fact’. According to Hume, ‘Matters of Fact’ are most firmly grounded in the Cause and Effect relationship but Hume makes it clear that an effect can never be found in the cause and therefore the first notion of the effect is completely arbitrary (Hume, 4). As previously stated, cause and effect can only be discovered through experience (Hume, 4) and can never be “mandated by a priori” (Forrest, 8). While a cause can be a natural instance or a supernatural one, the set of effects arising from them can only be of the same instance (Forrest, 6).  Throughout Forrest’s piece, we can see that methodological naturalism deals with the existence and “causal efficacy” and as seen by both authors there is a “begging the question” (Forrest, 11) or circular logic involved (Hume, 9). We use the scientific method to prove the validity of the scientific method. On a broader scale, both authors agree that the scientific method has flaws of its own, but it is still the best (only) method through which we can order information from the natural world and bring ideas to fruition.

The basis for the problem of induction is the notion of causation. Induction is concerned with how things act when they go “beyond the present testaments of our senses or the records of our memory” (Hume, 1). Meanwhile, causation as previously stated is the foundation of all “reasonings concerning matter of fact” (Hume, 2). A cause can have multiple different effects and these effects can never be found in the cause. No one and no object can ever discover through their/its senses the causes that have produced the effects, or the effects that will appear from the cause, there is no reasoning involved in this process, it can only be experienced (Hume, 3). So, the foundation of all cause-and-effect relationships is experience and not an ‘a priori’ or deduction (Hume, 4).

But from a collection of experiences, one can use inductive reasoning and predict what a future effect with a similar cause can be perceived like. Although this effect may be based on a probability or outside our domain of thought, we inherently assume that the future will continue to act in some way like the past. But we also know that the future does not necessarily resemble the past. There is a problem here, our experimental knowledge/observation cannot be justified. We use experimental observation to attain information about the effects arising from a cause and then project it on similar future causes to predict the effects again while simultaneously assuming that the future will conform to the past, thereby doubting the very principle through which conclusions are derived. Both, demonstrative reasoning and probable reasoning, are inadequate and we have ended where we started, thereby, “taking that for granted, which is the very point in question”, a question of existence (Hume, 9). This is the whole problem of induction and Hume aptly describes it. An example that makes it clear, is the example of a kid who burns his hand while touching fire and realizing that he must not touch it again. But he does not know if the next fire he touches could freeze his hand instead. If he uses the causal relationship to understand that fire burns, that means he uses experience, and this cannot be justified due to the problem of induction.   

Although not explicit through the work of Hume, I think induction implies a probability, there is a probability that such an event will occur, but it can also be true that some other event occurs instead. So, cause and effect cannot be reasoned but from the collection of experiences, it can be predicted. Therefore, a more accurate prediction is one that has considered the probability of the occurrence. I think we do have to learn from experience/induction because not only is it the best we can do but it is the only method we have to explore the world around us. Since the results obtained from induction are practical, they should not be disregarded. But we should still acknowledge the fact that induction does not always imply our ability to prove anything and keep in mind that it may be incorrect. Hume comes around to this point when he says, “we shall at least, by this means, be sensible of our ignorance, if we do not augment our knowledge “(Hume, 11). This to me sounds like the best solution.