Skip to content

← Thinking Clearly

What holds a claim up

Chapter 5 · Claims and reasons

In this chapter 4 sections
  1. What an argument is, and what it isn’t
  2. Finding the point
  3. Taking an argument apart
  4. Hearing the other side out

About two thousand years ago, somebody wrote a book about a king and a monk who argued. The king was Milinda, who ruled from a trading city called Sagala, probably today’s Sialkot in Pakistan. The monk was नागसेन (Nāgasena). After their first exchange, when the king asks whether they can talk again, the monk sets a condition. He will go on if the king discusses as a scholar, but not as a king.

What’s the difference? When scholars talk a matter over, the monk explains in the old English translation, one of them is shown to be wrong, he admits his mistake, distinctions are drawn, “and yet thereby they are not angered.” When a king discusses a matter and someone disagrees with him, he says, in effect, punish that fellow. The king agrees. He will talk as a scholar, he says, and the monk may speak freely.

That little exchange is a good definition of an argument. There are two ways to get someone to accept a claim. You can make them, by power, by shouting, by having five hundred supporters in the room. Or you can give them reasons, and let the reasons do the work, which means they can also fail. An argument is the second way. It’s a claim offered together with the reasons that are meant to hold it up, laid out so that anyone can check whether they do.

What an argument is, and what it isn’t

The book is The Questions of King Milinda, मिलिन्दपञ्ह (Milindapañha) in the Pali language it survives in, and it isn’t a record of a real conversation. Its translator, T. W. Rhys Davids, called it a historical romance with a teaching purpose. Milinda was almost certainly a real king, the Greek ruler Menander, who reigned in the second century BCE and whose coins have turned up in greater numbers than any other Greek king’s in the region, but nobody knows whether he ever met a monk called नागसेन. The book was written down some time between about 100 BCE and 200 CE. That makes it an odd sort of source, and an excellent sort of example, because its author cared about one thing above all: how a claim is held up.

Their best-known exchange goes like this. नागसेन has said that his name is only a name, a label people use, with no lasting self behind it. The king pounces. Is your hair नागसेन? No. Your nails, your teeth, your skin? No. All of them together? No. Something apart from them? No. Then, says the king, there is no नागसेन at all, and you are lying to me.

The monk answers with a question. How did Your Majesty come here, on foot or in a chariot? In a chariot. Then is the pole the chariot? The axle? The wheels, the frame, the yoke? All of them together? Something apart from them? The king has to say no each time. So, says नागसेन, “chariot” is a mere empty sound, and the king has told a lie. The five hundred Greeks who came with the king shout their applause and challenge him to get out of that if he can. He does, the only honest way. He didn’t lie, he says: “chariot” is the name in common use for the pole, the axle, the wheels and the rest put together. Exactly, says the monk. So is “नागसेन.”

Whether the monk was right about the self is a question for another book. What matters here is how he won. He didn’t say the king was foolish. He took the king’s own pattern of reasoning, ran it on the chariot the king had just arrived in, and showed that it led somewhere the king himself couldn’t accept. Reasons were met with reasons.

Not everything that sounds like reasoning is an argument, though. Four kinds of talk get mixed up all the time.

A story tells you what happened. “The king asked the monk his name, and the monk said it was only a name.” A story can contain arguments, as this one does, but telling it doesn’t argue for anything.

An opinion tells you what someone thinks. “Momo is the best food in the world.” Fine, but nothing holds it up. Add a reason (“because you can eat it anywhere in Nepal for very little money, and it comes in a dozen kinds”) and the opinion turns into an argument, a weak one perhaps, but one that can now be checked and answered.

An explanation tells you why something is true, when everyone already agrees that it is. “The milk went sour because it was left out in the sun.” Nobody is being asked to believe the milk is sour. They can smell it. The sentence is about the cause.

An argument tries to get you to accept something you might not accept yet. “The milk has gone sour, because it smells sour and it’s full of lumps.” Here the sourness is the thing being shown, and the smell and the lumps are offered as reasons to believe it.

The two milk sentences use the same word, “because”, and that’s the trouble. Words like “because”, “so” and “therefore” turn up in explanations and arguments alike, and sometimes the only way to tell which you’re reading is to ask what the speaker is trying to do. Are they telling you why something you both accept is so, or trying to get you to accept it? The difference matters, because an explanation that’s wrong about the cause can still be right about the fact, while an argument whose reasons fail leaves its claim with nothing holding it up.

Finding the point

Every argument has a point, the claim it exists to make you accept. It’s called the conclusion, which is a slightly misleading word, because the conclusion doesn’t have to come at the end. It often comes first, sometimes in the middle, and now and then it’s never said at all, left for you to work out.

There’s a simple trick for finding it, taught in British critical-thinking courses for years and called, without much imagination, the therefore test. Take a sentence you suspect is the point, and put “therefore” in front of it. If it reads naturally as something the rest leads to, you’ve probably found it. If it sounds silly, you haven’t.

Here it is at work on something said at a tap in Bhaktapur. “Don’t drink from that tap. The water comes straight from the old stone spout, and the spout tested dirty last month.” Put “therefore” before each sentence in turn. “Therefore, the water comes straight from the old stone spout” makes no sense: nothing before it leads there. “Therefore, the spout tested dirty last month” is just as odd. But “therefore, don’t drink from that tap” fits perfectly. That’s the point, even though it was said first, and the other two sentences are the reasons.

The test isn’t magic. It works best on short, tidy passages, and real talk is rarely tidy. But it builds a habit that pays off everywhere: before you decide whether you agree with something, find out exactly what it’s asking you to accept. A surprising amount of disagreement is two people arguing about different conclusions without noticing.

Taking an argument apart

The best way to see the parts of an argument working together is to take apart a good one. The Greek philosopher Aristotle, who lived from 384 to 322 BCE, wrote a book called On the Heavens, and in it he set out to show that the Earth is round. It holds up very well.

His conclusion is that the Earth is a sphere. His first reason comes from eclipses of the Moon. The Moon itself shows every kind of shape over a month, a thin curve, a half, nearly full, he points out, “but in eclipses the outline is always curved.” Since an eclipse of the Moon happens when the Earth gets in the way of the Sun’s light, the edge of that shadow is the shape of the Earth. And the only shape that always throws a round shadow, whichever way it faces, is a ball.

The full Moon with a dark, curved shadow across its lower left edge.
Figure 5.1 The Moon in partial eclipse on 7 August 2017, seen from Kuwait. The dark part is the Earth’s shadow, and its edge is curved. Photo: Irvin calicut, CC BY-SA 4.0

His second reason comes from the stars. Travel a little to the north or south and the stars overhead change. He gives an example: some stars seen in Egypt and near Cyprus can’t be seen at all in lands farther north. On a flat Earth everyone would see the same sky. On a curved one, moving along the curve tilts your view of the sky.

Those two reasons are independent: each one stands on its own. If somebody found a flaw in the eclipse reasoning, the stars would still hold the conclusion up, and the other way round. Reasons like that are worth having in pairs, because an argument resting on one reason falls if that reason does. Other arguments use reasons that only work together, like two legs of a ladder, and those are weaker in a different way: knock one out and the whole thing goes down.

The example about Egypt and Cyprus does a different job from the reasons. It doesn’t add a new reason. It gives you one real case of the second reason, so you can picture it and check it. Examples are how an argument gets its feet on the ground.

Aristotle also does something he didn’t have to do. He brings in the other side. Some people, he says, think the Earth is flat like a drum, and they have a reason: when the Sun rises or sets, the edge where the Earth hides it looks straight, not curved. That’s a fair observation. Aristotle doesn’t pretend it isn’t made. He answers it. These people, he says, leave out of account how far away the Sun is and how big the Earth is. A small piece of a very large circle, seen from a distance, looks straight.

That move has a name in critical-thinking courses, a counter-argument, and a good argument is stronger for including one. It shows the writer has looked at the strongest thing the other side can say and still has an answer. Leaving it out doesn’t make it go away. It only means a reader who already knows it will trust you less.

Try it

Mapping Aristotle’s argument

Aristotle’s case for a round Earth, retold in plain words, its sentences in the order a writer might put them. Tap each one to put it on the map.

The other sideSunrise edgelooks straightIts answerA big curvelooks straightReasonEclipse shadowalways curvedReasonStars changenorth to southExampleEgypt’s starsnot seen northConclusionThe Earthis a sphere

Tap a sentence of the passage to put it on the map.

Laid out as a map, the argument shows something the passage hides. The sentences came in one order, with the conclusion last and the other side first, but the argument stands in quite a different shape: two reasons holding up the conclusion, an example holding up one of the reasons, and the drum view set to one side with its answer pointing at it. The answer doesn’t prove the Earth is round. It only knocks away the drum’s support. Seeing what leans on what is most of the work of judging an argument, because a claim is only as strong as whatever is underneath it.

Aristotle didn’t stop at the shape. Since a short trip north or south changes the stars so quickly, he went on, the Earth must be not only round but “of no great size.” Here a conclusion becomes a reason for another conclusion, a step in a chain. From that he went one step further. If the Earth is that small, he said, the people who think the land around the Pillars of Hercules, the Strait of Gibraltar at the western end of the Mediterranean, is joined by one ocean to India in the east shouldn’t be dismissed too quickly. And he passes on the evidence they offered: elephants are found at both ends.

The elephants are where a map earns its keep. Put them on it, and they’re clearly holding up the claim that the two ends of the world are close, and they hold it up very poorly. Finding the same animal at both ends says nothing about how wide the ocean between them is. The conclusion at the top of the map was right, and the eclipse and the stars still convince anyone who checks them. The chain hanging off the bottom was shakier. Drawing an argument out is how you see that its parts can be of very different strength.

Hearing the other side out

Counter-arguments can be misused in a way worth knowing about, because it happens all the time on a phone.

In 1859 Charles Darwin published On the Origin of Species. Near its end he called the whole book “one long argument,” and he meant it: hundreds of pages of reasons, examples and evidence for a single conclusion, that living things change over generations by natural selection. Like Aristotle, he brought in the other side. He gave a whole chapter to the difficulties of his own theory, and in that last chapter he wrote that he had tried to give the objections “their full force.”

The most famous of those difficulties was the eye. In the first edition he wrote that to suppose the eye, with its focusing and its control of light, could have been formed by natural selection “seems, I freely confess, absurd in the highest possible degree.” Then, in the very next sentence, came his answer: “Yet reason tells me” that if a series of eyes from simple to complex can be found, each one useful to its owner, and if eyes vary and the variations are passed on, then the difficulty, however hard to imagine, “can hardly be considered real.”

That first sentence has been quoted on its own many times, as if Darwin had admitted his idea was absurd. It’s the other side of the argument, set up so he could answer it, presented as his conclusion. The therefore test catches it at once. Put “therefore” in front of “it seems absurd that the eye was formed by natural selection,” and it doesn’t fit what comes before or after. Put it in front of “the difficulty can hardly be considered real,” and it does.

So when a single sentence is passed around as proof of what someone believes, it’s worth asking which part of their argument it came from: the point, a reason, an example, or the view they were about to knock down.

In the end all of this, finding the point, sorting the reasons from the examples, noticing which way an answer faces, is what नागसेन asked for when he set his condition. Talking as a scholar means laying out your reasons so that someone else can see what holds your claim up, and being willing to have it shown that something doesn’t. A map shows what an argument says. Every argument also leans on things it never says out loud, and those are often where it gives way.

Found a mistake? Please tell us.