Hume's Argument Against Reason in Conclusions of Cause & Effect

           In the Enquiry Concerning Human Understanding, David Hume sets out to show that all knowledge can be derived from experience. Hume takes a particular look at the experience of cause and effect and our knowledge of the relation between the two. It is in the fourth section of the Enquiry that he presents what one commentator (Peter Millican) describes as Hume's “most celebrated argument” concerning the relation of cause and effects. It is obvious that we experience things of which we call causes and other things which effects, but Hume is interested in how and why one can conclude anything at all from such an experience. Hume answers in part two of the fourth section, that conclusions from an experience of seemingly cause and effect, cannot be founded by reasoning. He therefore takes it as his aim to explain and defend this answer throughout the fourth section and the rest of the text. I will therefore argue in this paper that Hume's argument for his solution functions as a viable defense for his solution.              

          Prior to Hume's discussion of cause and effect he makes an important distinction between two kinds of objects of reason. That is to say a distinction between the kind of things that one can reason about. This distinction will aid us in our understanding of Hume's discussion of cause and effect. One kind of object of reason is what Hume calls relation of ideas. By ideas Hume means the less lively and less forcible perceptions of the mind. Ideas are faint copies of another kind of perception of the mind known as impressions which are lively perceptions (such as hearing, seeing, feeling) and are forced upon our minds. Relation of ideas Hume means, “every affirmation, which is either intuitively or demonstratively certain”(4.1) Things that are intuitively certain are things that are known through a direct intellectual grasp, such as our knowledge that 2+2=4. Things that are demonstratively certain are things which can be proven by a sequence of intuitive steps, such as a proof of Pythagoras’s Theorem (that the square of the hypotenuse of a triangle is equal to the square of its two other sides). Relation of ideas are thus objects of reasons which can be known a priori. By a priori I do not mean to imply that there is any innate knowledge in people (recall that ideas are derive from impressions which are derived from experience), but that only that it can be justified without experience.

       The second kind of objects of reason Hume calls matters of fact. Matters of fact are in contrast, a posteriori, that is they can only be learned through experience. They concern ideas in relation to the actual world. For example, it is a matter of fact that words are appearing on the screen of my computer after I press buttons on my keyboard. And it is not a contradiction to suppose that a matter of fact such as this could be otherwise. For it could be the case that I press the buttons on my keyboard and birds start to fly out my computer screen. For this reason Hume thinks that “every matter of fact is still possible” (4.2) since every outcome which we conceive by the mind would have the same “facility and distinctness” (4.2). Therefore, matters of fact cannot be demonstratively or intuitively certain.

                 Now we have sufficient understanding of Hume's distinction of objects of reason to proceed to unravel his enquiry of cause and effect. Hume decides to investigate a little deeper into the nature of the evidence that gives us an assurance of matters of fact. Matters of fact he states are found on relations of cause and effects. Recall, I found that it is a matter of fact that words appear on my computer screen after I pressed keys on my keyboard. And still as I continue to type this paper I am making an on-going inference that there is a connexion between my pressing the keys and words appearing on the screen. But on grounds is this inference made? Hume answers that the connexion between cause and effect is not intuitive and there is a required medium (grounds) from which to make the inference. As this is as Hume says a “new” (4.17) question we should not trust our own “penetration” (4.17) of the question and instead lay out all the different kinds of reasonings and show that none of them make a viable medium for inferences of cause and effect.

                 Hume divides all reasoning into two kinds, demonstrative reasoning and moral reasoning. Demonstrative reasoning is the reasoning concerning relation of ideas. Arguments from this kind of reasoning involve intuitive steps such as, if you recall, my earlier example of Pythagoras’s theorem. Moral reasoning concerns matters of fact and involves uncertain inductive steps. I am using moral reasoning to infer at this moment that words will continue to appear on my screen after I press the keys because that has been the case several times in the past. Therefore, if Hume is to show that our conclusions from experience of cause and effect cannot be derived from reason he will have to show that that these kinds of reasoning are not plausible mediums from which to infer the connexion between cause and effect.

                   Hume first refutes the possibility that the grounds for the inference of cause and effect can be founded on demonstrative reasoning. Demonstrative reasoning if you recall are always certain. That is, a denial of it would imply a contradiction. 2+2=4 can be not be conceived to be any other way or else it would be a contradiction. But as Hume points out, “it implies no contradiction that the course of nature may change” (4.18). For there is no contradiction in supposing that birds may fly out of computer screen after I press on the keys. Such conceptions would seem to oppose the nature of demonstrative reasoning. Another much simpler refutation to this kind of reasoning is that demonstrative reasoning requires no experience, but we need experience in order to infer an effects from its cause. For a person that has never seen a gun could not infer without experience that a bullet will be fired after the trigger is pulled.

                 Hume then continues to investigate whether moral reasoning could be a plausible medium for the inference. We have said that matters of fact are founded on the relations of cause and effects, and that we know that relation from experience. In our ending conclusion we want to be able to say that the future will resemble the past. But in moral reasoning this means, the future will resemble the past, because future has always resemble the past. But this doesn't tell us why the future must still resemble the past. The evidence itself supposes the conclusion which generates a circular argument.

It must follow then that our conclusions from experiences of cause and effects (namely that there is a connexion between the two) cannot be founded on reason. This negative argument (one that derives it conclusion through the negation of the other options) is one that is not only plausible but also perfectly viable if we are to accept Hume's empiricist philosophy. For in order to refute Hume, one would either have to show how relation of ideas and matters of fact do not cover the all objects of reasoning, or prove that there is an object-less reasoning which both exists and is viable solution for making inferences of cause and effect. Otherwise we must accept as Hume does in the fifth section of the Enquiry that we only draw conclusions of cause and effect out of a habit of seeing the events constantly conjoined.