A Treatise of Human Nature

David Hume was one of those philosophers I always thought I would read … someday. So I’ve had a copy of his Treatise sitting on the shelf for quite some time. Recently I picked it up just intending to read a few pages while I decided what I really wanted to read next. But somehow Hume’s thought immediately pulled me in because his basic premise was so simple and attractive. We have studied the natural world as a science, but this study is always conducted by human beings. So if we want to put our understanding of this world on a firm footing, we will have to back up and start with an underlying science of human nature. As the subtitle of the book indicates, Hume thought of this as, “An Attempt to Introduce the Experimental Method of Reasoning into Moral Subjects”. The goal is a sort of introspective empiricism that treats the mind as a collection of natural processes whose laws we can surely uncover if we carefully observe them. Bacon meets the brain.

In a way, this desire to back up and put our knowledge on firmer footing is not at all new to Hume. In fact, one might see almost all of philosophy in this light. Socrates was willing to question everything, and only knew that he knew nothing. Descartes famously resolved to doubt anything that could be doubted. Spinoza tried to demonstrate that all things derive from the concept of God in geometric order. What’s new with Hume is not the quest for a stable foundation of our logic, but the acceptance that this quest is bound to fail. Hume genius lies in the fact that he doesn’t actually provide us with something we can be ultimately certain of. That is, he converts what has always been considered an a priori question into an empirical one. Even the new science of human nature will inevitably be constructed by humans. This admission changes everything.

What Hume begins with as indubitable is merely that fact that there is experience. If there wasn’t some sort of experience, we wouldn’t be having this conversation at all. But he refuses to immediately betray the wildness of experience by attributing it all to a consistent and stable subject like Descartes’ thinking substance. Instead Hume takes for granted only what we actually discover when we take a moment to settle down and simply observe the mind in action – this place is a dump! All kinds of things float through experience. Sights and sounds, aches and pains, images of other people, ideas about what to have for lunch, sublime philosophical reflections, puerile humor – it’s all in there, jumbled up, constantly changing, seemingly with no particular order, and certainly not in our control. In other words, Hume begins with what happens to you in your first meditation. Mental fucking chaos – the real empirical starting point.

It’s clear that this starting point doesn’t satisfy us as a stable foundation on which to build a philosophy. And that’s the whole point. Hume explicitly tells us that there’s no sense in looking beyond experience for something more fundamental, because in any case this putative thing would have to be experienced to be of any use. Perhaps there’s something more under this experience, but it is senseless to speculate. We simply can’t reach the a priori philosophical certainty we always thought we wanted any more than Newton could understand the inner nature of matter. From an a priori perspective any thought is possible, and each one would count as a real experience. Like Newton, all we can do is observe how the mind behaves and try to notice any empirical patterns.

And that’s precisely what A Treatise of Human Nature attempts to do. Hume begins with patterns he finds in our understanding or reason (Book 1), explores the patterns in our individual passions or emotions (Book 2), and concludes with examining patterns in moral judgement or society (Book 3). ¹ Which is to say that the entire book is simply an empirical description of our various habits of mind. Hume isn’t trying to tell us how we should think, but examining how we do, in fact, think. He is not trying to provide us with the perfect starting point from which we can deduce all of the true thoughts and none of the false ones. He is beginning in the middle, with the observations made by an observer who may himself be fallible, since, after all, we are only capable of observing the mind with the mind.

Despite the clarity and simplicity of Hume’s overall vision, the Treatise can often be a strangely confusing book. First, we contend with the English of 1739.  While the spelling and diction are only a little different, the clause structure seems to have changed pretty substantially. Even though Hume writes informally, without academic jargon and only very infrequently citing the history of philosophy, I found I often had to read sentences several times before getting the gist. There are commas, where you least, expect them and some pretty weird; semicolon use. Second, the Treatise is fairly long, and the principles of its organization are not always obvious. Hume clearly foresaw that his new approach would fail to convince anyone if he didn’t also offer many arguments as to why the a priori metaphysical approach falls short of describing how we actually think. There are many interesting examples, but together they stretch out the main argument and dilute its line. Someone looking for a more digestible version of the same ideas is advised to read the Abstract, a twenty page summary of the main argument that Hume later published as a fictional review of his own book. Third, while Hume elucidates many clear principles to account for the way we think -- simple habits of mind like relating things we see as contiguous or resembling or related by cause and effect -- when he delves into specific empirical examples, the way he applies these principles can often appear a bit ad hoc.  

It may seem puzzling that this review hasn’t delved into the details of some of the claims Hume is most famous for, such as the idea that correlation is not causation, or that we can never know the external world. Partly, that’s just because it would end up being too long. Partly, it’s because I’ve already started in on some secondary literature which will give me a chance to address these topics elsewhere. But it’s also partly because these particular claims are not what’s really novel and important with Hume. Compared to other philosophers, I’ve read Hume quite casually so far, not necessarily getting concerned with understanding the details of every one of his arguments. Yet I found the book immensely valuable simply for its overall orientation to how we should do philosophy. So I wanted to emphasize this empirical shift as the most important contribution Hume can make to our thinking, without getting bogged down in the flash card details of what he thought of topic X.

1 Kant will attempt to solve Hume’s skepticism as best he can with his transcendental method, but it seems to me that the ding an sich was actually born with Hume, even though he didn’t bother to elaborate this fruitless concept. Also, it strikes me that what I understand of Kant’s three Critiques more or less follows the three books of Hume’s Treatise – Pure Reason, Practical Reason, Judgement.