David Hume

David Hume (7 May 1711 – 25 August 1776) was a Scottish philosopher, historian, and essayist; he is widely considered the most important philosopher to write in the English language. David Hume (1711-1776) was a Scottish historian and essayist.1 As one of the first important philosophers to write in English, Hume was a skeptic. Hume’s two largest contributions to the field of philosophy lie within his major philosophical works: A Treatise of Human Nature and Enquiries concerning Human Understanding. He is perhaps most well known for his interpretation of Aristotle’s causation in terms of matters of fact and relations of ideas, and for questioning the rationality behind induction.2 These are known as Hume’s Fork and The Problem of Induction respectively. These skeptical arguments posed a challenge to many great philosophical minds and continue to challenge philosophers today.

Historical Context

Hume is one of the most notable skeptics in all of history thanks to his skepticism of causation, and necessary connection. More specifically, within scientific change, Hume’s arguments are most notably used against infallibilism. Famously, Hume is often associated with the terms The Problem of Induction and Hume’s Fork. Hume creates most of his arguments for skepticism within his two texts A Treatise of Human Nature and Enquiries concerning Human Understanding. At the time, modern philosophers had strived away from Aristotle’s account of causation. In this rejection, they thought themselves revolutionary. Still, those same philosophers kept Aristotle’s distinction between knowledge and belief. Descartes and Malebranche, for instance, were certain of demonstrative scientific knowledge. Locke, on the other hand, was more interested in showing belief is rational enough to assert natural philosophy.2

Hume took a different approach. He divided that same distinction between scientific explanation and belief into relations of ideas and matters of fact. The distinction was still based on Aristotle's knowledge and belief, but it translated slightly differently. Similar to Aristotle’s distinction, Hume’s distinction agreed that all propositions could be exclusively divided into one category or the other. Different to Aristotle’s distinction however, it provided a different account for the two types of propositions. This distinction is commonly referred to as Hume’s Fork.3

Major Contributions

Hume’s Fork

In Hume’s entrance to the debate of causation, Hume translates the Aristotelean distinction between scientific knowledge and belief into his own terms. These are:

• Relations of ideas.
• Matters of fact.

Relations of ideas are ideas that are absolutely certain through either demonstration or purely through intuition. They are a priori, in that they are discoverable independent of experience. This categorization does not necessitate ideas to carry information dependent on the world and thus ideas falling into this category are independent of any existing thing. They are universal constants in that they hold true in all worlds. It should be noted, relations of ideas cannot provide any new information about the world. These types of propositions are simply a means used to help understand more complex ideas. They can be thought of as symbols or a series of simpler ideas describing a larger more complex idea. Common examples usually include geometry or math as formal sciences fall within this categorization. Examples of such statements include a 'square’s sides add up to 360 degrees' or '1 + 1 = 2'. Alternatively, a worded proposition may look something like 'when you run, you move your body,' or, 'all bachelors are unmarried'. Relations of ideas can never be denied as their denial would imply a contradiction in the very definition of the terms within the proposition.3

Matters of fact are the complete opposite of relations of ideas. Matters of fact are a posteriori statements and thus based on experience. They are, for this reason, not independent of worldly experience. Unlike relations of ideas, matters of fact do not hold true in all possible worlds. The contrary of matters of fact imply no contradiction and such statements cannot be established by demonstration. Matters of fact can show new information about the world but rely on the experience of the world. Examples of such statements include the sky is blue, or water is odourless, or all guitars have 6 frets. It should be noted that false statements can still be matters of fact. The level of coherence within false statements or contrary statements remains the same as within true statements despite being incorrect. In this sense, contrasting statements are, too, matters of facts.3

The reason behind this distinction was simple; it was to provide criteria by which to organize scientific statements. Through this distinction, all statements were categorized into either matters of fact or relations of ideas. This also ultimately meant that there was no type of idea which was certain and provided information about the world. In the case of matters of fact, propositions are reliant on senses and due to the fallibility of the senses, have no certainty. In the case of relations of ideas, propositions can be proven with absolute certainty through the use of other relations of ideas. Unfortunately, however, these statements cannot give any new information about the world. This distinction was often taken by the scientific community as a strike at Newton’s theory of motion. 4 Such a distinction has large consequences in the fields of science, religion, and even philosophy due to its prevention of certain real world statements. As an example this distinction would make useless the attempt to try to prove non physical entities as matters of fact.

1. α is a non-physical entity.
2. It has no observable effect on the world and its not made up of a physical thing.
3. α is a relation of ideas.
4. Relations of ideas are just assigned symbols helping to explain more complex symbols.
5. The statement ‘α exists’ proven or otherwise doesn’t say anything about the world; it is just a play on words.

Much akin to the reasoning the analytic/synthetic distinction uses, it is impossible, according to Hume, for a proposition not to fall within the distinction. In Hume’s eyes, such a proposition would be completely meaningless. It would simply not be a rational or reasonable endeavour.

Problem of Induction

The problem of induction stems from the reasoning behind causal inference. Given the previous distinction between relations of ideas and matters of fact, the connection within causal inference should fall under one of those categories. With this in mind, Hume found that causal inference cannot be a relation of ideas, and so must be a matter of fact.

The reason he came to this conclusion was that each cause is independent of its effect. The two are not associated with each other. For example, never having seen an anti-biotic, and without being told of its effects, would a person be able to establish its effect through sole means of reasoning and senses? Quite simply, no. Without experience, a person would have no understanding to what the purpose of the medicine is, or that it is medicine at all. And herein, Hume concludes, since a priori reasoning cannot be the source of connection between causes and effects, the inference must be a matter of fact.3

Describing causal inferences as a matter of fact provides leeway to use experience to determine the effects of a cause. Using the example from before, if that same person had tried anti-biotics in the past and had been cured, they begin to infer the reason for their getting better was the anti-biotic. Grasping this chain of events Hume attempts to formulate that person’s argument for their inference and states it as such:

1. In the past x has resulted in y.
2. Therefore, in the future x will result in y.

Here however, Hume notices a gap in logic. How does (1) infer (2)? Hume sees that demonstrative reasoning cannot fill this gap, because a complete contrast where x does not result in y implies no contradiction within the inference. Herein, it seems that there is an underlying assumption within the first premise: the future will be like the past. This assumption is more commonly known as the uniformity principle. Using such an assumption would fill the logical gap within the current argument, however before it can be used, it must be established that the principle is either intuitive or demonstrable. However, the only way to formulate an argument for the principle is to rely on the principle itself.3 As such the full argument would follow as such:

1. In the past, the future has been like the past.
2. Therefore, the future will be like the past.

Within this argument, the premise assumes the conclusion and, as such, the argument is circular. In this sense, the first example shows an irrational train of thought. It seems then, that Hume established there is no way that reason could be the connection between cause and effect. Thus, Hume sought another connection between cause and effect. He eventually recognized this connection to be custom or habit. This is more commonly known today as induction. As a person experiences something repeatedly, they grow to expect it to happen again. However, despite being an adequate connection, this solution forces the abandonment of reason within causal inference.3

Criticism

Historically, due to the threatening nature of Hume’s distinction between relations of ideas and matters of fact as well as his problem of induction there have been many critics of Hume. One of the most prominent critics to criticize Hume on the account of his distinction between the types of propositions was Immanuel Kant. Kant theorized that the world was interpreted through sensory and intellect and thus there must exist some sort of a priori synthetic proposition.4 The existence of such a proposition would of course result in a proposition that fit both categories of Hume’s distinction. Unfortunately, Kant’s a priori synthetic proposition was debunked with the arrival of probabilistic determinism.

As for criticisms on Hume's Problem of Induction, there are quite a few cases. One of the more notable cases was the critique Karl Popper had towards Hume, stating that induction is a myth.5 Popper argued science is created by conjecture and criticism rather than reference to the past, and that the main purpose of observations wasn’t to make inferences about the future but to refute present existing theories. Popper was committed to the idea that Hume was asking the wrong question: ‘how can theories that cannot be justified by induction, otherwise be justified?’ Popper, instead, preferred to look for a process by which to correct errors.6

Wesley Salmon responded to this criticism in Hume’s place, stating theories still need predictions to be tested. When Popperians have multiple theories, each sharing the same quantity of empirical content, Popperians would choose the theories which were better corroborated but lack any justification in this decision. The Popperians either make an inductive claim:

1. A theory was reliable in the past.
2. It will be reliable in the future having survived falsification.

Or they admit corroboration is not an indication of predictive power.7

Publications

Here are the works of Hume included in the bibliographic records of this encyclopedia:

• Hume (2008): Hume, David. (2008) Enquiry Concerning Human Understanding. Early Modern Texts. Retrieved from http://www.earlymoderntexts.com/authors/hume.
• Hume (2007): Hume, David. (2007) Dialogues Concerning Natural Religion. Early Modern Texts. Retrieved from http://www.earlymoderntexts.com/authors/hume.
• Hume (2000): Hume, David. (2000) A Treatise of Human Nature. Oxford University Press.
• Hume (1975): Hume, David. (1975) Enquiries Concerning Human Understanding and Concerning the Principles of Morals. Oxford University Press.

To add a bibliographic record by this author, enter the citation key below:

 

Citation keys normally include author names followed by the publication year in brackets. E.g. Aristotle (1984), Einstein, Podolsky, Rosen (1935), Musgrave and Pigden (2016), Kuhn (1970a), Lakatos and Musgrave (Eds.) (1970). If a record with that citation key already exists, you will be sent to a form to edit that page.