Soon after writing this I've turned into a Hume fan. His views on being empirical (especially when offset with the so-called rationalists) are very sensible. His arguments regarding extension and other metaphysical objects, however, made me a bit angry. This is based on a short writing piece for a class I am in.
In the introduction of his Treatise, Hume declares his goals. He wishes to create a science of man based entirely on experience. He wishes to approach morality, logic, politics, and other such philosophical explorations as sciences. Ironically, not only does he barely defend his theory empirically, but he also leaves us without any meaningful takeaways. He fails to create a good basis for the rest sciences of man, because his system lacks what a good basis needs. A good basis should be easy to interpret, easy to falsify, and generate predictions of yet-unseen phenomena. At best, Hume’s system imparts us with useful language and negates the most unreasonable ruminations of his contemporaries. I will primarily compare his system to Newton’s.
Why was Newton’s system so good? His laws unify a wide gamut of phenomena into a short information-dense system with a few key properties. Firstly, it does not rely on esoteric knowledge. Anyone can understand Newton’s laws given time, since we are all endowed with both basic mathematical skills [1] and intuition about movement. This means that it is easy to interpret. While there are often doubts about what Hume really meant, there are no doubts regarding what Newton meant. Secondly, Newton’s system is easily falsifiable. Because there is a very tight range of interpretations, anyone can tell whether a law is incorrect or not given a physical observation. Truth cannot be contested once an experiment is made. Lastly, the system generates useful and unique, non-obvious predictions. Black holes and other objects not easily visible, were discovered in part due to Newton’s theory [2].
Hume’s system has no full strict interpretation. This is clear from the evident debate regarding his meaning. If Hume’s theory were more precise, the other problems would be easier to fix. With a strict interpretation, experiments are easier to devise and falsifiability would naturally arise from its the black and white nature. If we were to make his theory more scientific in this regard, there would be two clear paths to take. One is to reduce further, as the logicians and set theorists did to mathematics in the early twentieth century. Defining all of mathematics in terms of a formal system effectively takes the burden of knowledge since they need less context [3] to grasp its ideas. For historical reasons this door was not open to Hume. The second path involves the implementation of practice. This is what Hume should have taken. Many theories appear vague and loose in their interpretability. Consider psychology. Some of it may seem wishy washy, but much of a psychologist’s theory of the mind is implicit in his actions. The mind is loosely defined by the way it behaves [4]. In such fields, execution of theory is not yet separate from its expression [5]. A theory does not need to be fully explicit in language to be a coherent theory. In Zen Buddhism there is an understanding of the mind—a theory. However, it is kept implicit (partly since it is considered ineffable) in process [6]. Having practice mitigates problems of interpretation, since it more directly expresses the ideas than words may be able to do. In fact, this leads into an important point regarding theoretical arguments. If two arguments hold different positions in their language but behave the same in practice, it’s unlikely that they actually disagree. An atomist and a believer in infinite divisibility may in fact hold the same or similar views on extension due to their manipulation of their thoughts in the same way—it is not unlikely that the distinction is mostly formal. This would be especially true if the atomist believed that we would not find the “atoms” for a long time due to their small size. If we can’t break reality down to atoms, then we may as well be working with infinite divisibility. The disagreement may then be simply one of faith or lack of thereof in future discoveries [7]. Returning to Hume’s system, Hume need not have necessarily established an organization for the study of the mind, but giving ample instructions for its examination would have helped. As his instructions were highly limited, and he is not as precise as physicists are, we cannot tell if his model of the mind actually behaves differently from other models. It is clear from his explicit disapproval that he wishes to negate some of his contemporaries, but what he wishes to put forward is not. Thus, it cannot form a basis for the science of man.
This leads into my second point. Hume’s experiments are simply ineffective. Consider that which he outlines for the ink blot. There is no way to tell if the ink blot is truly the smallest possible. We can only walk forward or backwards in some smallest increment as our feet have a fixed length and our gate cannot be arbitrarily small. We would need infinite divisibility of walking to disprove infinite divisibility of the inkblot [8]. Moreover, our impression of the inkblot is too unstable to make any judgments regarding its simplicity. Hume, himself, argues in 1.2.4.7 that many delusions regarding the mind exist due to our inability to precisely detect and examine mental minima. How can he claim to have a reasonable answer based on experience if he too suffers from the same inability to experience accurately? I am confident that the tools to accurately examine the mind exist. Many Buddhist groups have developed methodologies to improve one's focus (and other such faculties) with the explicit goal of enabling an effective examination of the mind. Such tools are prerequisites to a theory that claims to explain that which is not visible to the layman. If Hume wishes to present a scientific theory of the mind backed up only by experience, he must have and/or provide the tools necessary to treat it scientifically, or cloister its meaning to only the parts of the mind accessible to the layman [9].
Another reason his empiricism is questionable is that he seems to forget that absence of evidence is not evidence of absence. Hume does not summon anyone to corroborate his claims regarding the mind experimentally. He only explores his own mind. It is unlikely that he is exploring the mind with enough breadth to buttress his theory competently, given that he is using a sample size of one. Newton’s theories described all the planets visible to astronomers. It also described the mechanics of projectiles on Earth. If it had served to describe only a single planet, we’d laugh at it. Therefore, to try to find universal principles such as that time cannot exist without objects following one another or that extension cannot exist without colors, is unreasonable. They should not be universal or simply be better tested. Blind people could very possibly have a different view of extension from those who aren’t blind, and Hume would never know. By noting the absence of evidence of qualities such as colorless (or sensory-less) extension, Hume erroneously infers that we have evidence of their absence. As I’ve said this is made far worse by his lack of data [10].
Overall, Hume’s empirical theory is simply not empirically verified. Due to its imprecision it is also hard to falsify, which makes it possibly unverifiable. I believe he’s biting off more than he can chew. The mind is very, very hard to pin down, especially when compared to planetary motion.
[1] Some people may be genuinely unable to think mathematically, but these are rare. The main difficulties in math is that of notational complexity and problem solving, not some inability to grasp the ideas themselves once understood. (return)
[2] Black holes were first thought of after Einstein gave his theory of general relativity. However, many invisible bodies are found by observing motion which necessitates their existence. Some of the outer planets (and dwarf planets) of our solar system were found this way before telescopes were sufficiently potent to spot them. (return)
[3] By context I mean the knowledge which is necessary to interpret information that is encoded on some medium, such as writing. For example, instructions for the baking of a pie require some context about how to bake and how to read. There is interesting philosophical exploration to be had in whether information, and meaning, lie more in the encoded medium (i.e. words) or the context necessary to draw out the meaning, and whether generative processes (such as when someone who’s never cooked before, reads cooking instructions and learns to cook via trial and error based off the instructions) ought to be used as evidence that the encoded medium and/or context already has the meaning or not. (return)
[4] Behavior here includes our mental impressions and ideas, so this isn’t really behaviorism. (return)
[5] The unification of theory and practice is an interesting topic to explore. The line is not as clear cut in many situations as one may assume from the hard sciences where theory and practice have a very specific and strict separation. (return)
[6] Other branches of Buddhism write more formalistic theory, but the consensus is that the theory is not itself the aim. It is simply an aid. (return)
[7] Or, if the existence of atoms seems contradictory, it is likely simply a formal distinction. Things are what they are. “Atoms” are a simulacra (language is a simulacra) of reality, with the useful property of being isomorphic and thus having meaning. To say something “is” an atom is to express the mapping between reality and the simulacra. (return)
[8] As soon as we step to make the blot disappear, we need to step back partly in some limiting series to show that there isn’t actually a slightly smaller ink blot we can form between the distance at which it was invisible and that at which it was last visible. A natural algorithm would be some sort of infinite binary search. (return)
[9] Such as some of the ancient philosophers, or grandparents might when giving advice. Someone giving advice on how to be happy, often, does not need to explore the deepest inner workings of the mind. (return)
[10] The absence of data has additional deleterious side effects. It forces Hume to focus on examples which most of us have experienced. For example, anyone can think of white and black cubes and spheres. That means that his theory is especially good where a good theory is not necessary, and especially bad, where a good theory could provide the most value (i.e. for what we have not yet experienced). Effectively, he may just be lecturing birds on how to fly. Edge cases, such as the shade of blue, which the theory fails to explain, are some of the most importants parts of any theory. People do not use theories when they are not necessary. Thus, the theory should be most potent when it is necessary—that is, in edge cases. It would be better for Hume to have knowingly made a theory which was particularly good at explaining such cases while sacrificing the quotidian. (return)