Category Archives: Epistemology

A tweet-summary of David Hume’s “An Enquiry concerning Human Understanding”

After conducting a Twitter poll, I decided to re-read Hume’s Enquiry, one of my favorite works of philosophy. I posted my summaries of each section in a Twitter thread, which I’m reproducing here with some modifications.

Section 1: Of the different Species of Philosophy
In Section 1 Hume tells us he’s basically trying to construct a science of the human mind. He wants to do for the mind what Newton did for external bodies. This means not just describing the different parts of the mind, but actually explaining how the mind works by understanding its fundamental principles and capacities. This will clearly have consequences for the way we think about what kinds of knowledge we’re capable of.

Section 2: Of the Origin of Ideas
In Section 2 Hume draws a distinction between impressions and ideas. Impressions are things like sensations, desires and emotions. Ideas are dull copies of impressions we use when reflecting or thinking about them. The difference is here is the difference between (for example) actually having an experience of a red object, such as when looking at an apple, and later thinking about redness or red objects. Next, Hume says that while the human imagination is powerful, there are important constraints on what imagination is capable of. All complex ideas arise out of us “compounding, transposing, augmenting, or diminishing” simple ideas, which are in turn derived from impressions. Eg: we can think of a golden mountain by bringing together our ideas of “gold” and “mountain” which we were previously independently acquainted with.

Towards the end of the second section, Hume brings up a possible counter-example to this: the famous “missing shade of blue” thought experiment. All simple ideas are supposed to have their origins in corresponding impressions (and colors are supposed to be simple, not complex, ideas). But wait, we can line up various shades of blue progressing from darkest to lightest with a single space somewhere in the middle left blank instead of having a particular shade of blue. Isn’t it plausible that someone looking at this series would be able to imagine that shade of blue even if they haven’t had a direct impression of it? Hume acknowledges that this seems to be a difficulty for his account, but immediately sets it aside by saying that’s such a singular case that we don’t need to alter the general principle outlined here.

Section 3: Of the Association of Ideas
Section 3 is the shortest so far. In Section 3 Hume points out that our ideas seem to enter into various relations with each other, but that all of them can be boiled down to three fundamental ones:

– resemblance
– contiguity in time & space
– cause & effect

He thinks the last one is the most important, as we will soon see.

Section 4: Sceptical Doubts Concerning the Operations of the Understanding
Hume begins Section 4 by dividing human reasoning into two kinds: reasoning concerning relations of ideas, and reasons concerning matters of fact. Mathematics and logic belong in the former category, all forms of empirical investigation fall into the latter one. When we’re dealing with relations of ideas (math, logic) we can achieve certainty. When we’re dealing with matters of fact, the best we can hope for is probability. Hume then states that for all empirical reasoning that goes beyond immediate observation and memory, is based on cause and effect.

How do we know about cause and effect? They’re not things we can discover a priori. Causes and effects are distinct, so logic along will not get us from one to the other. So we must learn about causes and effects through experience. But there’s a problem. Our reasoning based on cause and effect proceeds like this: we recall that in the past we’ve observed that certain causes produce certain effects. But we go on to think similar causes will produce similar effects in the future. That is, we think the future will resemble the past. But why think this? What is the basis for thinking that because things have happened a certain way in the past, they will happen that way in the future?

It can’t be a logical deduction – “X has always happened in the past” doesn’t entail “X will happen in the future.” Hume points out there’s no contradiction in holding, for instance, that the sun, which up until now has always been observed to rise, will not rise tomorrow. OK, so it’s not something we know a priori. So maybe it’s something we infer based on experience? But hold on, we just said a short while ago that all reasoning concerning matters of fact presupposed the relation of cause and effect. But we also just showed that cause and effect presupposes that the future will resemble the past. So we appear to be reasoning in a circle. Not good!

Similarly, Hume says, you can’t say that because up until now, the future has resembled the past, it will do so in the future as well. That’s presupposing precisely what you set out to demonstrate – that the future will resemble the past. So where does that leave us? Hume’s conclusion in this section – a purely negative one – is that we don’t reason ourselves into believing that the future will resemble the past. He’s just shown that there is no possible route reason can take to do this. So if it’s not based on reason, why do we draw inductive inferences? That will be the subject of Section 5.

Section 5: Sceptical Solution of these Doubts
We were left with the question of what grounds our practice of using induction / inferences based on cause & effect, since it’s been argued that it can’t be reason. Hume is going to give us his answer to this question now. In Part 1 of Section 5, Hume asks to imagine that someone with all the normal faculties of reason is suddenly brought into the world – that is, he has no previous experience. In case like this he would observe regular successions of events but would not arrive at the idea of cause and effect. And so he would not be able to reason about anything beyond the senses and memory. After he’s been around for a while and gained more experience, he will begin to draw inferences: he will infer a certain object/event from the presence of another object/event because over time he’s observed constant conjunctions of objects/events.

And so we have our answer. What makes him – and by extension, us – do this is custom or habit not reason. In Hume’s own words, “All inferences from experience, therefore, are effects of custom, not of reasoning.” It’s simply a fact about us that we come to conclusions this way – every bit as unavoidable as feeling love or hatred.

In Part 2 of Section 5 Hume analyzes the nature of belief. He begins by pointing out again that the human imagination can come up with various complex ideas by modifying and compounding simple ideas. What makes these “fictions” of the imagination different from beliefs? Hume thinks that beliefs have a different feeling attached to them. A feeling that “gives them more weight and influence; makes them appear of greater importance; inforces them in the mind; and renders them the governing principle of our actions.”

In other words, beliefs have a different feel to them than things we merely imagine, and this feeling in turn disposes us to act in certain ways. And this disposition is itself brought about by belief + the memories the belief triggers + custom, which is what makes us draw inductive inferences. So we can see that for Hume, custom or habit is of crucial importance, not something to be dismissed because it’s not a form of reason. Custom is “necessary to the subsistence of our species, and the regulation of our conduct, in every circumstance and occurrence of human life.”

Section 6: Of Probability
Section 6 is a short section on probability. Hume says straight away that “chance” is not something in the world. Rather, our ignorance of certain causes makes us form various degrees of belief, which is where probability has a role to play. Again, it’s experience + custom that’s doing all the work here. The more often we observe certain conjunctions of events – the more we see this sort of uniformity – the higher our degree of belief. And Hume takes this to be the essence of probability.

Section 7: Of the idea of necessary Connexion
After a brief introduction, Hume points out that all our ideas deas related to cause and effect, such as “power”, “force” and “necessary connection” are extremely obscure. So he wants to try to figure out what they are and where they come from. Hume reminds us about his theory of simple and complex ideas: simple ideas are copies of impressions, complex ideas arise out of modifying and compounding simple ideas. Do our ideas of power/necessity arise from experience? But we don’t actually perceive powers or necessary connections. We observe, for example, a billiard ball moving after being struck by another billiard ball, or that whenever there’s fire there’s heat. But our senses don’t actually grasp the causal connection between one billiard ball and the next, or between fire and heat. So there’s no external impression that generates the idea of power or necessary connection.

Maybe there’s an internal impression that does? Perhaps it could be argued that whenever we voluntarily move parts of our body, we feel this power, and that’s where the idea comes from. Hume argues that this isn’t actually what’s going on: we will to move a certain part of our body and this is followed by that part of our body moving. We don’t actually encounter any powers or necessary connections here, just a succession of acts: the willing and the moving. Besides, a close study of anatomy reveals there’s a whole bunch of stuff that happens between the willing and the final effect that we have no knowledge of, so it doesn’t make sense to say we’re actually aware of the connection between the willing and the acting. OK, we don’t encounter powers in cases when the mind tries to command the body. Maybe we do when we bring certain ideas into our mind or decide to turn our inner attention to certain ideas? But again, Hume argues, what happens here is similar to what happens in the case of moving parts of our bodies. We experience the desire / will to focus on a certain idea, and then we find ourselves focusing on certain idea. We still don’t encounter the supposed power that connects the first event and the second. So we’re still left with the fact that we only perceive conjunctions between objects/events, not connections.

Hume now points out that some philosophers, as a result of this, have adopted occasionalism. According to occasionalism, the powers are not in objects but in God. God wills all the conjunctions we observe. Hume’s response to this is that it doesn’t really solve the problem. It’s true that we don’t understand the power that determines how bodies operate on each other, but we’re equally ignorant of how a mind – even a supreme one – could operate on bodies. And the power of minds to act on bodies is also not something we have an impression of, as we’ve already shown. Having rejected these explanations we’ve considered so far, Human proposes his own.

We’ve shown that powers/connections are not things we observe – internally, or externally. What we observe are regular successions of events. But we also showed in Section 5 that as a result of custom, we always form certain expectations. Every time we see one billiard ball strike another, we see the second billiard ball move. And so the next time we see it, we expect it to move. That is, the connection between ideas is not something that is impressed on us by the world – it’s something felt by the mind. We project this felt connection on to the world. which is why we think there are necessary connections between objects/events. Hume then offers his own revisionary definitions of cause & effect:

– “An object, followed by another, and where all the objects, similar to the first, are followed by objects similar to the second”

– “An object followed by another, and whose appearance always conveys the thought to that other.”

Hume ends the section by worrying whether his chain of reasoning will be understood by the reader, but also says he doesn’t want to spend more time on it, lest he further confuse the reader. I say: Don’t worry, Hume! You did a really good job explaining what exactly the chaining of reasoning is, and I didn’t have much trouble understanding it.

Section 8: Of Liberty and Necessity
In section 8 Hume wades into the free will / determinism debate, or he calls it the question of liberty and necessity. Straight away Hume makes a bold claim: all the ink spilled over the supposed tension between liberty and necessity rests on a semantic mistake rather than some deep metaphysical principle. In his own words, “the whole controversy has hitherto turned merely upon words.” Liberty and necessity can be reconciled.

When Hume talks of necessity he’s referring to what’s often called causal determinism: the idea that events are necessitated by prior causes. But remember, Hume says, we’ve shown that the idea of necessity is suspect – all we have are constant conjunctions we observe and our disposition to form expectations on the basis of our observations. Hume thinks this is important because he thinks what can tempt people to deny the necessity thesis is something like this: they think there are necessary connections in the physical world, but when it comes to their own mind, these necessary connections don’t exist. So we can secure liberty for the mind if we think it’s exempt from the causal necessities that physical objects are governed by. But once we ditch the idea of necessary connections entirely, this analogy doesn’t hold. The whole picture falls apart. There are no necessary connections out in the world or in the mind. What we have is constant conjunctions / uniformity – and this is true of both the mind (desiring, willing, acting) and the world (physical events).

Hume then argues that no one really denies the necessity thesis (or, if we’re being philosophically rigorous, the uniformity thesis) even when it comes to human minds, because we constantly make inferences / predictions about how humans will behave based on past observations. Hume spends a significant amount of Part 1 of this section showing how all our practical activity relies on the presumption of uniformity in human thought and action (even across very different cultures).

OK, so how can this necessity / uniformity be reconciled with liberty, which is what Hume promised us? I will share Hume’s own compatibilist definition of liberty: “By liberty, then, we can only mean a power of acting or not acting, according to the determinations of the will; that is, if we chuse to remain at rest, we may; if we chuse to move, we also may. Now this hypothetical liberty is universally allowed to belong to every one, who is not a prisoner and in chains.” In other words, freedom means being able to do what you want.

In Part 2 Hume wants to show that many of our ideas about reward and punishment, or praise and blame, require both necessity and liberty. What’s guiding Hume’s argument here is the thought that reward/punishment or praise/blame for an action only make sense if that action is a reflection of the person’s character. But if we grant this, we also grant that they require uniformity. That is, we hold that certain kinds of behavior regularly follow from certain kinds of mental states (beliefs, desires, intentions etc). If people’s actions did not follow from prior internal and external causes – that is, if actions did not follow this pattern of uniformity – then they are not a reflection of their character – they’re not a reflection of who they are – and so reward/punishment or praise/blame would make no sense. Similarly, Hume points out that our practices presuppose liberty, since it wouldn’t make sense to apply them when people acted because they were coerced by external violence, because once again, their actions don’t reflect their character.

Hume then brings up two difficulties for the reconciliation he has proposed between liberty and necessity. First, someone could argue that even though we may be free to act according to the determination of our will, the will itself is determined by a prior series of causes that presumably has its origin in God. If that’s the case, people cannot be held responsible for their wills, and all the responsibility lies with God. Here Hume responds by saying that it’s a mistake to think our moral ideas are born of reason alone – when we praise or blame we are motivated primarily by sentiment. It’s a mistake to focus on the question of ultimate responsibility since our beliefs about this won’t significantly alter the way our sentiments are formed. He doesn’t develop this much further here, and to fully understand where he’s going with this I assume you’d have to be familiar his sentimentalist theory of morality. The second objection is that if God is ultimately responsible for the chain of causes that created people’s wills (which in turned led to people engaging in harmful acts), then God must be imperfect or immoral. Hume doesn’t give us a response to this – he says that tackling this question regarding God is beyond the scope of philosophy.

Section 9: Of the Reason of Animals
Section 9 is a short but super underrated section on animals and their capabilities of reasoning. It’s remarkable how, in a pre-Darwin age, Hume was able to stress the continuity between humans and animals in many things – including the ability to reason. Hume says it’s clear that animals also reason in the sense that they are capable of learning from experiences, and this enables them to avoid things that cause them pain and seek things that they find enjoyable. In this sense they are similar to us – they experience various constant conjunctions, and their minds automatically form expectations on the basis of these experiences. He gives the example of a horse learning how high it can jump, and a dog learning to fear the whip.

Now, of course this doesn’t mean animals get by in the world with highly abstract reasoning or formal argumentation. But neither, Hume points out, do we – most of the time. A quote I really liked from this Section: “Animals, therefore, are not guided in these inferences by reasoning: Neither are children: Neither are the generality of mankind, in their ordinary actions and conclusions: Neither are philosophers themselves, who, in all the active parts of life, are, in the main, the same with the vulgar, and are governed by the same maxims.”

Similarly, Hume says, animals are often guided by instinct rather than by things they’ve learned through observation. But then again, so are we. So Hume once more stresses the continuity between humans and animals. I think this is a natural consequence of his stated aim: constructing an empirical/natural science of the mind.

Section 10: Of Miracles
Hume’s made plenty of bold claims throughout the book, and this section on miracles is no exception. In the very second paragraph of Part 1, Hume says “I flatter myself, that I have discovered an argument of a like nature, which, if just, will, with the wise and learned, be an everlasting check to all kinds of superstitious delusion, and consequently, will be useful as long as the world endures.” Strong words – let’s see if the argument lives up to them.

Hume begins laying out the argument by recalling that for all matters concerning experience, we rely on constant conjunctions / regularities / uniformities that we have observed. These regularities are confirmed to various degrees, so we’re always dealing with probabilities here. Some regularities are extremely well confirmed in that we’ve never observed anything contrary to them. Whenever we encounter a claim, we must measure its probability based on our experience of these regularities. In Hume’s words, “A wise man, therefore, proportions his belief to the evidence.”

Hume makes a point about testimony here, and it’s crucial to understand it, since it plays an important role in his argument. Why do we believe people’s testimony? As with anything else, says Hume, it’s because of our past experience that there is some sort of connection between testimony and truth. By “connection” of course I don’t mean a “necessary connection.” Rather, I’m referring to the correlation we’ve observed in the past between people saying X and X actually being the case. What Hume says is that we’ve observed in the past that people aren’t always mistaken or trying to mislead us when they report things, which is why we think we can rely on testimony. This is the only reason to think testimony has some value. If not for this experience of a correlation we would have no reason to take testimony seriously.

Hume then goes on to say we’ve also experienced that testimony is not always reliable, and there are many circumstances in which we would doubt someone’s reports. All of this lays the groundwork for the core of the argument. Hume now defines a miracle as a “violation of the laws of nature.” It’s worth stressing here that Hume’s conception of laws of nature is entirely epistemic. There’s nothing impossible in principle about violating the laws of nature. Laws of nature are simply our best confirmed regularities – the facts about the world we have the most confidence in. Hume: “It is no miracle that a man, seemingly in good health, should die on a sudden: because such a kind of death, though more unusual than any other, has yet been frequently observed to happen. But it is a miracle, that a dead man should come to life; because that has never been observed, in any age or country.”

The second point to note here, which Hume mentions earlier, is that the reason we’re usually given for believing a miracle is testimony – people report witnessing a miraculous event. Hume says that when this is combined with what he’s said so far, the conclusion that follows is “That no testimony is sufficient to establish a miracle, unless the testimony be of such a kind, that its falsehood would be more miraculous, than the fact, which it endeavours to establish[.]”

Part 1 ends here. In Part 2 of Section 10 Hume sets out to show why exactly testimony for a miracle doesn’t rise to the level of being more miraculous than the miracle it’s supposed to establish. Remember that at this stage we’re weighing two different probabilities. One the one hand, there’s the claim that a certain law of nature was not violated. Remember, a claim like this is one we have the highest amount of confidence in, given our experience of regularities. We’ve never seen an exception to something like “Dead men stay dead” ever. When it comes to matters of fact, probabilities don’t get much higher.

On the other hand, there’s the testimony from witnesses that the law was violated. Hume wants to show that we will have no reason to assign this testimony a higher probability than the alternative that the law of nature was violated. Some people have portrayed Hume as giving some sort of a priori argument against miracles here, but this is not borne out by the text, and goes completely against the spirit of Hume’s empiricism. The reasons against assigning the testimony in favor of miracles a higher probability are straightforwardly empirical, and Hume lists them:

1) Testimony is strongest when it comes from people who are well educated and have good reputations that they would not wish to tarnish. All the testimonies we have for miracles fail to meet these conditions.

2) It’s a fact about human psychology that we have a propensity to believe miraculous things, especially in matters concerning religion, because of a combination of fear and wishful thinking, among other things.

3) Most reports of miracles come from people who are “ignorant and barbarous.” That is, people who don’t know much about how the world works, and are especially prone to attributing effects to supernatural causes. Hume points out that people sometimes wonder why miracles and supernatural intervention seemed to be so abundant in ancient times but seem so scarce now. The answer is simple: people were more easily fooled into believing such things in the past than they are now.

Finally, there are various miracle claims across history purporting to be evidence for various religions, all of which are in contradiction with each other. Even the religious person, therefore, has to believe that the vast majority of miracle claims are false. But the kind of testimony that exists for different religions is of the same nature and quality, so there’s no reason to think that the miracles described in any specific religion are more likely to have occurred.

On the basis of all this, Hume says, none of the testimony we have for miracles so far is good enough to overturn our belief that a law of nature was not suspended, and it’s unlikely that there will be any such testimony in the future. But again, this doesn’t mean that such testimony will be in principle impossible to find, and Hume himself provides an example of something that could work as sufficient testimony for believing something extraordinary: If everyone in every country agreed that there was darkness for eight days starting on January 1, 1600, and there was absolutely no variation in the accounts about this among people of different nations, then it seems reasonable to believe this did really happen, on the basis of testimony.

Hume immediately contrasts this with a different example: On January 1, 1600 Queen Elizabeth died. A month later she came back to life and resumed her rule, as confirmed by many witnesses of the time in England. Hume thinks that in this case we shouldn’t believe that it actually happened, for precisely the kinds of reasons outlined above.

One final point Hume makes is that someone could respond by saying we can say that the probability of a miracle isn’t actually low because God can suspend the regularities we observe anytime he wants. Hume’s objection to this is that the only basis we have for learning about the nature of God’s actions is experience, and all our experience shows us that he doesn’t like to suspend regularities. We’ve observed mistaken testimony and fraudulent miracles far more often than we’ve observed God violating a law of nature, so even if you’re a theist, you should bet against the miracle actually having occurred.

It’s interesting how everything Hume’s said so far can be seen as a direct attack on Christianity, and indeed that was how it was received in his time. I’m sure Hume didn’t make things easier for himself with the way he ended this section: “So that, upon the whole, we may conclude, that the Christian Religion not only was at first attended with miracles, but even at this day cannot be believed by any reasonable person without one. Mere reason is insufficient to convince us of its veracity: And whoever is moved by Faith to assent to it, is conscious of a continued miracle in his own person, which subverts all the principles of his understanding, and gives him a determination to believe what is most contrary to custom and experience.” Oof.

Section 11: Of a particular Providence and of a future State
In Section 11 Hume tackles philosophy of religion, although his treatment of the subject is pretty short here. But you can see the germs of the criticisms that would be more fully developed in his later work, the Dialogues Concerning Natural Religion. Amusingly, Hume places most of the criticisms of theistic arguments in the voice of a friend, who’s also playing the part of an Epicurean arguing against the Athenian gods rather than the Christian one. It seems pretty clear Hume wanted to distance himself from being seen as an outright atheist.

Hume’s main target here is the argument from design. It can be stated in a lot of ways, but the general idea is that given the extraordinary order and structure of the universe, we infer that its cause be some powerful intelligence(s) behind it. We’re reasoning by analogy here: if we came across some complex artifact we would infer that it was designed by an intelligence. The same applies to the universe. Hume’s main criticism here is that from any effect you can only infer proportionate causes – that is, you can only infer causes that have the particular capabilities that the effect requires. Anything more is just a shot in the dark. So if we want to infer that the universe has a designer, all we can say is that the designer has the ability to create the universe. We don’t have any grounds for inferring that the designer has attributes like omnipotence or benevolence or anything approaching perfection. This is especially true given all the imperfections we see in the world.

Hume goes on to say that if there’s no reason to think the gods are just, there’s no reason to think there will be a proper distribution of justice in the world either – that is, that good deeds will be rewarded and bad deeds will be punished. At this stage Hume considers an objection. Suppose we see a half-finished building with various tools and materials around. Wouldn’t it be reasonable to assume from this, not just that there was a designer, but that the designer had not yet completed the project? Why can we not reason in an analogous fashion, on observing the universe, that while rewards and punishments may have not yet been meted out in a just matter, they will in the end, as events run their final course? Hume points out the relevant disanalogy here: in the case of the half-finished building we’re justified in drawing that kind of inference because we’ve observed humans and how they behave in the past. The same is not true of the gods – we have no experience of them that tells us what their behavior is like or what their intentions are.

A final point Hume makes about inferring causes from effects is that we can only infer like causes from like effects. If we’re dealing with a singular effect that has no parallel in our experience, there’s not much we can say about its cause. This is what Hume thinks about inferring things about the cause of the universe as a whole. When we’re looking at specific objects within the universe, we can infer their causes based on constant conjunctions we’ve seen before with similar objects. But the universe as a whole is something singular; there’s no similar kind of object we’ve observed in the past to come to conclusions about its cause.

Section 12: Of the academical or sceptical Philosophy
In Section 12, the final section, Hume goes over the different varieties of skepticism. First there’s Cartesian skepticism, which begins by universally doubting the veracity of all our faculties, and seeks to arrive certain foundations from which knowledge can be built up. Hume thinks this is hopeless: once you entertain Cartesian doubt, you’re never going to reason your way into knowledge. It can’t be beaten. Another form of skepticism is more modest and more specific: We have reason to doubt our senses because we know they can be deceived (hallucinations, illusions). Then there’s skepticism about the external world. Each of us is naturally inclined to believe that there exists a world out there, independent of our minds. But we know that all the mind has access to is impressions. How do we know these impressions are caused by, and resemble, objects in the world? To do this we’d have to compare the impressions to the objects. But there’s no way to do this – there’s no way to get outside the impressions and at the objects themselves, independent of our impressions, because all the mind has access to is impressions. If all we have are impressions, there’s no proof that they were caused by or resemble objects in the world.

After this Hume brings up another variety of skepticism, which has its origin in the distinction Locke drew between primary and secondary qualities. According to this view, which Hume takes as being universally acknowledged by philosophers, the “sensible qualities of objects” such as color or hardness don’t exist in the objects themselves – they exist as impressions in the mind, produced by objects. The primary qualities of objects are things like solidity and extension. Now, rationalist philosophers like Descartes thought these things could be grasped as intrinsic qualities of the objects by the mind (as opposed to the senses). As an empiricist, Hume obviously doesn’t think this is true. Hume things ideas like extension are also acquired through sensible qualities perceived by the senses.

Hume also says this “grasping by the mind” doesn’t make sense because the process of abstraction involved in it doesn’t really make sense – we can’t conceive of an extended object that is devoid of sensible qualities. Hume challenges to conceive of “a tangible or visible extension, which is neither hard nor soft, black nor white” and concludes that it’s impossible. In my opinion the argument here is not sufficiently developed. Hume is far too brief here, and it’s hard to evaluate what exactly the chain of reasoning is, although I’m sure you could reconstruct a sympathetic version of the argument if you tried. Anyway, Hume’s conclusion is that primary qualities are every bit “in the mind” as secondary qualities, and once we accept that, there’s really nothing we can say about the nature of the external world: “Bereave matter of all its intelligible qualities, both primary and secondary, you in a manner annihilate it, and leave only a certain unknown, inexplicable something, as the cause of our perceptions; a notion so imperfect, that no sceptic will think it worth while to contend against it.”

Moving on, Hume hints at some paradoxes related to space & time which can also lead to skepticism, before eventually distinguishing between two kinds of skepticism: Pyrrhonian skepticism, and mitigated or academic skepticism. Pyrrhonian skepticism is an “excessive” form of skepticism about all our faculties and abilities to reason. If we were capable of taking Pyrrhonian skepticism seriously, we would be paralyzed and unable to act. Hume rejects this form of skepticism. The point of mitigated skepticism is to restrict our areas of inquiry to fields where we may actually arrive at answers. What are these fields? The same fields Hume listed in Section 4: relations of ideas and matters of fact. The former includes mathematics and logic, the latter includes everyday empirical investigations. What Hume wants to rule out is reasoning about deep metaphysical questions that are beyond the purview of empirical investigation.

Further, Hume insists a priori reasoning (which can grant us certainty) is only applicable to relations of ideas (that is, mathematics and logic), it cannot demonstrate metaphysical conclusions. And for empirical reasoning, we must work with the cautious, modest, systematic application of custom, which can lead us to probable conclusions. This is what we’re left with once we adopt mitigated skepticism. And so we come to the dramatic conclusion of the Enquiry, which Hume ends by declaring: “When we run over libraries, persuaded of these principles, what havoc must we make? If we take in our hand any volume; of divinity or school metaphysics, for instance; let us ask, Does it contain any abstract reasoning concerning quantity or number? No. Does it contain any experimental reasoning concerning matter of fact and existence? No. Commit it then to the flames: For it can contain nothing but sophistry and illusion.”

Final thoughts:
I’m more convinced than ever that Hume is rightly considered one of the greatest early modern philosophers, and the Enquiry is rightly considered one of the greatest early modern works of philosophy. It isn’t perfect – there are reasonable objections you could raise at various stages, but that doesn’t change the fact that it’s a fantastic piece of philosophy. Also, Hume is a great writer. He manages to be clear and profound. I honestly think someone with next to no background in philosophy could pick up the Enquiry, read it, and grasp most (if not all) of it.

My favorite sections: Section 4 (induction), Section 7 (causation / necessary connections), Section 8 (free will & determinism), Section 10 (miracles).

Measures and Metrics

It was a day at the races for two young boys in Gimli, Manitoba. As crowds gathered to watch cars flick back and forth on the track, they took their bikes a few hundred metres away to re-enact the races on a derelict runway. Unbeknownst to them, a giant silently bore down upon them. Air Canada’s flight 143 had run out of fuel and was attempting to glide to a safe landing. We can’t imagine their terror. We can’t imagine what it was like to see their short lives flash before their eyes as they turned to see the jet nearly upon them. They stood still in shock for a moment then raced down the runway as fast as they could. They made it. Barely.

This crash was no tragedy. But it could have been. Hundreds of people could have been injured or killed. And for what? For a conversion error. The ground crew thought that they were adding fuel in kilograms but were actually adding pounds. Hence only about half the fuel was added than intended. That is, the Gimli Glider was a casualty of Canada’s metrication process. There were others, though much less severe. And yet, metrication is worth it. The metric system is the best measurement system we have available to us.

This is a weird claim. All measurement systems are arbitrary. Rulers and thermometers were not the gifts of Heaven like those law tables carved in stone. Humans have always used what was available to them to keep track of weights and measures. Humans have used stones and feet and thumbs and brines and cups and so on. The metric system is just one more in this long list of arbitrary measures.

But arbitrary things can also be more and less apt to accomplish what it was designed for. For metrics, there are at least three primary criteria we use to evaluate aptness: ease of use, ease of conversion, and its ability to represent real difference.

Most metrics throughout history were designed with ease of use in mind. Weights and measures needed to be compared to some standard, and that standard had to be readily available. The cubit is the gold standard in ease of use: what is more available to a carpenter than her own forearm? If she wishes to measure out a board, she need only count how many of her forearms it is long, and she has some measure. And since her forearm does not change in size, everything she measures with it will be bound to the same standard.

Now, the problem with a cubit is that it isn’t easily convertible into other measures. How many thumbs go into a cubit? How many feet? It isn’t obvious, and the carpenter must determine that for herself. If another carpenter were to use her measurements, he might find that his cubits are of a different size, and the ratios of his feet to his cubits might be different altogether. People solved this long ago by standardising their previously variable measures into formal units. The foot no longer corresponded to the carpenter’s own foot but to some arbitrary standard foot, say the king’s. And so too with thumbs and all the rest. And once this is established, the ratios between different measures remain static. There are twelve thumbs in a foot. There are sixteen ounces in a pound. There are two pints in a quart. And so on. This may slightly reduce the ease of use for these measures, but it vastly improves the overall utility of the measurement system.

Finally, the ultimate purpose of any measurement system is to record real differences in the world. We want to know whether two boards are the same size or different. We want to know whether our apple-only diet is helping us lose weight. We want to know whether it is warmer or cooler outside than it was yesterday. And this is the hardest part about any measurement system: some of those differences are more important to us than others, and our measurement system must recognise this.

So given all this, why metrication? There are of course practical questions. Metric is the international standard; US manufacturers lose millions of dollars every year because they have to convert quantities into metric for export, a phenomenon that does not plague manufacturers elsewhere. This also puts American students behind their peers from other countries. Metric is the official measurement system of science, and so students in American schools must not only learn the imperial system, but also the metric system. However, these pragmatic questions say nothing about metrication. The US could without too much difficulty exert its influence to motivate other countries to switch back to imperial. If we want to justify metrication, we need to do this internal to a theory of measurement alone.

So why is metric the gold standard? In no uncertain terms, it unambiguously meets the measurement criteria better than any other measurement system. Metric is no more difficult to use than any other measurement alternative. Both imperial and metric require standardised instruments, and both are equally available. But metric makes conversion easier than any other measurement system. Imperial maintains constant ratios between different units, but these ratios are not uniform. There are twelve inches in a foot, three feet in a yard, and 1280 yards in a mile. A person needs to remember each one of those ratios, or at least look them up in some rule book, before one can complete even the simplest conversions. Metric dispenses with this difficulty. The conversion ratios in metric are not only constant, but also uniform. Every unit is distinguished from the others as multiples of ten. And when in doubt, the conversion ratio is included in the name of the unit. A kilometre is a kilo-metre, or a thousand metres. Easy.

Perhaps the most impressive benefit is the metric system’s ability to represent real difference. The metric system employs a greater diversity of units, in part due to its uniform conversion ratios, than any other measurement system. Objects as small as atoms and molecules or as large as stars and galaxies can be easily and coherently represented in metric units without ambiguity. And while the very small and very large do not affect ordinary people very often, the metric system does a better job with ordinary measures as well. The best example of this is temperature. The Fahrenheit scale is pinned to three different benchmarks: ice and salt brine, the freezing point of pure water, and human body temperature. These correspond to 0, 32, and 96 degrees respectively. These choices are arbitrary, of course. But they are also meaningless. There is no firm relationship between these standards that ties their temperatures nicely to a single scale. And moreover, they tell us nothing about what the temperature of a thing is like. The Celsius system is very different. It is pinned only to the phase changes of water. 00 is freezing. 1000 is boiling. This too is arbitrary, but it does tell us quite a lot about the world we live in. Water is fundamental to Earthly life, and its properties determine life’s conditions. It is plain that negative temperatures represent a different kind of weather than positive temperatures. Negative temperatures are associated with snow and ice. Positive with rain and warmth. Fahrenheit never gives us that clean divide. It could be snowing at 300 but raining at 350. Those five degrees are far more important than the scale lets on.

This is the reason the world has adopted the metric system. This is worth all the troubles metrication caused. The metric system is completely and unambiguously the superior measurement system. Now we need only for the most powerful country on Earth to catch up.

 

Anti-Anti-Skepticism

We live in an era of anti-skepticism. Skeptics are the enemy. They must be destroyed and overcome. This is what drove Descartes and Leibniz, Locke and Hobbes, and even Hume. They sought to naturalise the world, to conquer it, knowing its features with certainty. But this new tradition of anti-skepticism is very different. Philosophers today do not claim to certainly know. They in no way claim to refute the skeptic. The older eras held knowledge to be something lofty, fragile, godly. Knowledge was something not for man but for gods, for elites, for those deserving of special authority. In this new era, the anti-skeptics have lowered knowledge. It is not lofty: it is mundane. It is not godly: it is human. They have not refuted the skeptic: they have made her irrelevant. And the consequence is dire.

I begin by motivating contemporary anti-skepticism from its roots in Moore and Wittgenstein. Then I show how this tradition fails to overcome skepticism.

Moore begins with knowledge. Of course he knows that idealism is false. Of course he knows that he has hands and was once smaller than he currently is. What he doesn’t know is how to correctly analyse any sentence saying so. This looks dumb. Surely Moore cannot possibly know that he has hands: what if he is dreaming? What if he is hallucinating? What about Descartes’ evil demon or any of Hume’s or Kant’s skeptical challenges? But Moore is not troubled by these suggestions: these are only competing analyses of what it means to know. And these others are inferior. Why? As Moore suggests in “A Defense of Common Sense,” the skeptical analysis commits skeptics to far more than Moore’s analysis. Moore is committed to the truth of only a small set of basic claims—that there is a world, that he is an element of that world, that he was once younger and smaller than he currently is, and so on. The skeptic, on the other hand, is committed to these basic claims plus their skeptical claims. As Moore notes, the skeptic does not stop herself from going about in the world as a normal human being. She still wakes in the morning, commutes to work, constructs arguments for skepticism, and presents them to an audience of other philosophers. She demonstrates that she knows these things through her actions. Only she does not claim to know them: she both knows them and commits herself to not knowing them. This is extravagant. This is impossible. The consistent skeptic sits demure like Johannes Climacus, never stirring, never speaking, never thinking.

The idea is this: language is a practical thing. It has consequences. It does not matter whether we are dreaming, for we do not act like we are dreaming. If a debate is predicated on something so ethereal, something so benign, then it is not predicated on a correct understanding of language. Language is substantive and effective. We all know this: we all use language for some purpose. Our claims bear in some way on our lives. And what skeptics mean is just this: that when we say such and such, we could be wrong. This is trivially true, but the skeptic overstates her case. And for Moore, there is no sense in which we could ever be meaningfully wrong about whether there is a world or whether we were once smaller than we are now. There is a subset of propositions (Moorean propositions) that we cannot consistently and meaningfully deny.

Moore did not succeed in this argument. But, says Wittgenstein, he is not wrong. He has rather failed to carry his insight to its logical conclusion. Moore remains vulnerable to skeptical attack by those who do not understand what it means to be certain, to know, and to doubt. Wittgenstein does. He will finish what Moore began.

For Wittgenstein, knowledge is located within a discourse: one knows such and such only if competent others recognise her as knowing such and such. This means that the claim must be recognised as true, but also that the reasons she gives for her claim are recognised as appropriate. This is not to reduce knowledge or even truth to something communal or legislative. A community cannot agree to collectively upend their knowledge system by changing what they believe and accept as valid reasoning. We are bound to a language in which we have no say. Our language gives us the concepts and thoughts to which our perception of the world conforms. In Kantian terms, language is the source of the categories of experience. Some of these are grounded in our form of life. In this Wittgenstein agrees with Moore. Others are grounded in particular grammars. But in neither case is the result a kind of relativism. It is rather a kind of contextualism: language merely shapes the world; it does not construct it.

The skeptic here has no room to move. Doubt can only occur concurrently with certainty. That is, where one aspect of a discourse is brought into question, it is questioned on account of other aspects of that discourse. When Copernicus doubted that the Sun revolved around the Earth, he did so on grounds acceptable to him, namely the simplicity of the heliocentric model and God’s preference for simplicity. Without these certainties, it is incoherent to doubt geocentrism. Doubt itself must be justified.  The skeptic does not do this. She holds an apple in her hand and asks “Is this apple an apple, or is it something else?” The question is senseless. Of course the apple is an apple. If it were not, it would not be an apple. The skeptic’s only reply undermines her claim: “You miss my point, Wittgenstein. I mean to say that we all know what an apple is, but cannot with certainty suppose that this object I hold in my hand is an instance of the general concept. And so also with any instance of any general concept.” This may be a legitimate and sensible question. Surely someone once pointed to a whale and asked whether that fish is a fish only to be answered in the negative. Surely it is never obvious when any object is an instance of a concept. But skepticism this isn’t. The relevant doubt is localised and predicated on a kind of certainty: that one knows what an apple is. And in general, if one is comfortable identifying the object of doubt with a name that is under dispute, the doubt is easily assuaged. Of course the apple is an apple: we call it an apple.

Whether a doubt is sensible and appropriate always depends on the context it is present. Some contexts place a greater demand on one’s knowledge of a claim than others. And some others will open up the possibility of doubt from some claims and not others. This is the skeptic’s greatest error. They always wish to doubt that which is never open to doubt. They wish to doubt those claims that are predicated on our form of life. They wish to doubt Moorean propositions. It’s curious that the skeptic claims to doubt the external world, yet stands upon the stage to address an audience. It really makes one think.

But in truth, the skeptic is well within her rights to do this. The most pressing questions in philosophy are not about apples or astronomy. They are about who we are and how we ought to live. There is nothing more fundamental to our form of life than not knowing who we are or how to live. And answering those questions requires inquiring into our history and our constitution. Not everyone who so inquires is a skeptic. The skeptic is only one who has so inquired and found herself at a loss. She is one who feels uneasy about who she is and how she ought to live. This is not absurd. This is not incoherent. It strikes me that this is the most natural, the most human feeling possible. For who among us truly knows how to live?

Contemporary anti-skeptics don’t satisfy these questions. They rule them out from the outset, and they themselves know this. Both Moore and Wittgenstein famously stood in an odd relationship to philosophy’s most pressing questions. For Moore, “the good is good, and that is the end of the matter.” (PE, §6) Wittgenstein agrees. Value is indefinable. It is something nonsensical: though it is everywhere present to us, it cannot reside in the world of facts. Both Moore and Wittgenstein surrender to a kind of bland mysticism. And what good is mysticism in answering our deepest questions? What good is anti-skepticism if it cannot actually overcome our skepticism? Moore and Wittgenstein offer no authority. They offer no guidance. Anti-skepticism is useless.