How to win an argument (Part 3)

In part 1 we had a look at four problems of reasoning, stemming from generalisations and implications:

1. Destroying the exception - all swans are white... until you find a black one.
2. Disregarding the exception - all surgeons are criminals cos they cut people up.
3. Mistaking the direction of the arrow (confirming the consequent) - clouds mean it must be raining.
4. Mistaking the meaning of the arrow (denying the antecedent) - no rain means there must be no clouds.

In part 2 we saw how distractions, intended or not, can come in many forms and begging the question is something journalists constantly get wrong:

1. Red Herrings - arguments that don’t address the issue.
2. Straw Men - arguments that address a superficially similar, but in fact different issue.
3. Loaded Questions - controversially complex questions that presume things that haven’t been proved.
4. Begging the Question - when the conclusion is assumed in the argument.

In this third part, we’ll look at laziness of thinking, particularly to do with cause and effect. This will lead us to a discussion of similar mistakes with words rather than reasoning.

In the next and final part, we’ll have a look at some real arguments and take them apart.

It’s worth noting at this stage that, while we’re looking at all these fallacies as ways to break down your opponent’s arguments, they should also be a guide for you to develop your own indestructible arguments.


-- Flying Penguins

Non sequitur is Latin for ‘doesn’t follow’. All logical fallacies can be described as non sequitur, but it is predominantly used to mean an argument where the conclusion doesn’t follow from the premises. This might be because the argument is invalid (we’ve been looking at ways in which this might be the case), or because it is not sound.

I should explain that difference:

Valid means:
If the premises are true, then the conclusion is true.

Invalid means:
If the premises are true, that doesn’t mean the conclusion is true (it could be true, but that would just be a coincidence).

Sound means:
A valid argument with true premises (and so a true conclusion).

Unsound means:
A valid argument with at least one false premise OR an invalid argument.

A valid and sound argument:
All men are mortal.
Ben is a man.
Therefore, Ben is mortal.

A valid but unsound argument:
All animals with wings can fly.
Penguins have wings.
Therefore, penguins can fly.

This is unsound because the first premise is false.
It is valid because, if the first premise was true, the conclusion would be as well.

An invalid (and so unsound) argument:
All men are mortal.
Jane is mortal.
Therefore, Jane is a man.

Can you spot why this argument is invalid? If you can, then you’re doing well. (Clue: Clouds.)

So the two ways an argument can be non sequitur (conclusion doesn’t follow from the premises) is either by being invalid or unsound.

There are lots of ways this can happen, but I want to look particularly at our laziness of thought when it comes to causation. Previously, we looked at mistakes that come from reasoning based on generalisations and the problem of inductive reasoning (going from specific instances to a universal generalisation). Inductive reasoning was not called that when Hume was discussing it. [Don’t know who Hume is? See Swans.] He talked about causal reasoning and was concerned with what it means for one thing to cause another. This is intricately linked with making generalisations (induction), but I won’t bore you with how (if you want to find out, check out this introduction - a little technical, but readable).


-- Farting Dogs

So what does it mean for one thing to cause another? Well, a major part of it must be that the effect comes after the cause and not the other way round. However, we need to be a bit more specific than that, because me turning on the TV just before the dog farts does not mean that the remote controls the dog. For one thing to cause another, it must always do so in the same conditions. This way, my pressing the power button on the remote causes the TV to turn on (if the batteries aren’t dead), but doesn’t cause the dog to let loose.

When one thing always follows another thing, they are said to be in ‘constant conjunction’. Hume was worried; he thought there was nothing more we could say about cause and effect. He was worried because, for example, day and night are in constant conjunction - but day does not cause night, nor the other way round. It’s these kinds of difficulties that make up our next batch of logical fallacies.

If two things always occur together, we feel like they must be linked in some way. Birds start singing in the morning, just before the sun rises - does birdsong cause the sun to rise? What about if your child develops autism after he’s been vaccinated - did the vaccination cause the autism? I’ve got a new flatmate and ever since she arrived the boiler has been playing up - does her presence cause the boiler’s problems?

Be careful! When we feel like there is a connection between things, we tend to misjudge how strong it is and can easily be swayed into believing the connection goes one way when it doesn’t. This mistake comes up a lot more often than you’d think.

Another way of making this mistake is to look at events that happen to occur at the same time and conclude that one causes the other. This is usually because there is a less obvious thing that causes them both. Let’s say I notice that, in the UK, people wear less clothing during the summer months. I want to know why. They also consume much more ice cream. I conclude that the increased consumption of ice cream is causing people to shed their winter-wear and jump into skimpy swimwear. Whoops.

These causal misunderstandings (not ‘casual misunderstandings’ as I keep writing) can be very subtle. For example, we saw in part 1 how we can easily feel that rain and clouds have a deeper relationship than they actually do. When they are subtle, they are difficult to spot.


-- Non-existent Sofas

Another pair of mistakes that are related to this generalisation/induction misunderstanding are when we say something is true of the whole because it is true of some part, or vice versa.

“The university is seven hundred years old, therefore all the staff are seven hundred years old.”

“Every band member of Heaven’s Nails is brilliant, therefore Heaven’s Nails are brilliant.”

The first argument makes the mistake of saying that the whole has a property, so its parts must do too.
The second argument mistakes individual brilliance for band cohesion. It may be that they work really badly together, regardless of how brilliant they are individually.

Some more examples:

“This tiny piece of tin foil cannot be torn by hitting it with a hammer, therefore the spaceship I have made by wrapping large sheets of tin foil around a thin wooden frame is impervious to hammers.”

“Atoms are 99.9 per cent nothing. My sofa is made of atoms. Therefore, 99.9 per cent of my sofa doesn’t exist.”

Keep your eyes peeled...


-- End of Part 3

This time we’ve seen the most common of argument mistakes: non sequitur - what they claim simply doesn’t follow from what they have said. There were a few specific examples:

1. Two things happen in succession - this doesn’t mean the first caused the second.
2. Two things happen around the same time - this doesn’t mean they are linked.
3. The parts aren’t necessarily the same as the whole.
4. The whole isn’t necessarily the same as its parts.

Next time:
We’ll deconstruct some real arguments from Debatewise.org.

How to win an argument (Part 2)

In part 1 we had a look at four problems of reasoning, stemming from generalisations and implications:

1. Destroying the exception - all swans are white... until you find a black one.
2. Disregarding the exception - all surgeons are criminals cos they cut people up.
3. Mistaking the direction of the arrow (confirming the consequent) - clouds mean it must be raining.
4. Mistaking the meaning of the arrow (denying the antecedent) - no rain means there must be no clouds.

In this second part, we will be looking at some more ways to take down an argument; this time by identifying distraction techniques (even when your opponent wasn’t intending to distract).


-- Red Herrings

For some reason, the practice of strongly curing in brine and heavily smoking a kipper has lead to an idiom that means ‘a distraction’. This ‘red herring’ is also a logical fallacy and a debating tactic in which the respondent seeks to divert the opponent. When it’s not done on purpose, though, it’s called ignoratio elenchi. This roughly translates as ‘ignorance of refutation’, which in turn can be better expressed as ‘not understanding what response would correctly refute the argument’.

A ridiculous example might be:
Alice: “Monkeys are intelligent, because they can peel bananas.”
Bert: “You’re wrong, because whales live in the sea.”

Be careful of clever debaters who will employ subtle red herrings to avoid discussing difficulties in their arguments.


-- Scarecrows

Another way you might be duped by a devious debater is via a Straw Man. If you find that your opponent is attacking an argument that is superficially similar to yours, but not quite what you are saying, then they have built up a Straw Man and are attempting (maybe not on purpose) to follow this sort of reasoning:

(1) You present argument X.
(2) They, in what sounds like a refutation of your position, present argument Y that is superficially similar to your argument X, but isn’t the same.
(3) They show argument Y to be wrong in some way and so conclude that argument X is wrong in the same way.

So it looks like they’ve proved your argument wrong, when in fact they just made something up and proved that wrong instead.

Alice: “Monkeys are intelligent, because they can peel bananas.”
Bert: “Just because monkeys eat bananas doesn’t mean they’re intelligent. Bats, parrots, hamsters and mice all eat bananas - and they’re clearly not intelligent.”


-- Mistresses

Our final distraction technique is the loaded question. My favourite name for this fallacy (they all have many names) is the Fallacy of Many Questions. The famous complex question is “When did you stop beating your wife?” To answer this question, whether yes or no, is to implicitly admit to having a wife and having beaten her at some point.

There are three levels of complex question, two of them fallacious, but only one of them counts as a loaded question.

(1) Legitimately complex questions:
“Who is the Queen of the United Kingdom of Great Britain and Northern Ireland?”
To answer this question directly, you would be implicitly accepting that there was such a place as the UK and that it has a queen.

(2) Illegitimately complex questions:
“Who is the present King of France?”
To answer this question directly, you would be implicitly accepting that there was such a place as France and that it has a king - but it doesn’t. So this is a Fallacy of Many Questions.

(3) Controversially complex questions:
“Does your mistress live by herself or with your other wife?”
To answer this question directly, you would be implicitly accepting that that you have both a mistress and a second wife. Not wise. This is both a Fallacy of Many Questions and a loaded question.

The loaded question can be used as a distraction device, but watch out for the more general fallacy when people accidentally use illegitimately complex questions (i.e. ones with false assumptions).


-- Journalists

According to Wikipedia, the loaded question is often mistaken for Begging the Question. I’m not sure why. I think it might be because very few people seem to know what begging the question actually is. Please excuse what is about to be a mini-rant.

I just went to guardian.co.uk and typed into the search box ‘begs the question’. There were 1394 results. Here are the first six (with my italics).
“..hours of dull activity that requires their personal intervention. It begs the question: why do they do it?” (22 Aug)
“...where it conflicts with the "principles of Islam" (which of course begs the question of how the principles of Islam are to be determined)” (5 Aug)
“...by Francis Veber's 1998 French comedy Le Diner de Cons, begs the question – what has happened to mainstream US comedy? “ (2 Aug)
“Even more than most titles based on popular pastimes, it begs the question "why not try the real thing?” (1 Aug)
“That of course begs the question about whether it is strictly necessary...” (28 July)
“Research has shown that boards comprising a variety of people from different backgrounds and experiences are more effective.
This also begs the question why the Bank of England's monetary policy committee, making key decisions about everyone's future, is all male?” (19 July)

I would put a lot of money on ninety nine per cent of “begs the question” on most UK newspaper websites being followed with what particular question is being begged. This is very, very, very wrong. What they mean is something like ‘implies the question’, ‘asks the question’, ‘raises the question’, ‘inspires the question’, or even ‘we should ask’.

If an argument begs the question, the question that is being begged is this - and ONLY this:

“How can you justify your conclusion, when you have already assumed its truth in your argument?”

A very simple example:
(Premise 1) Monkeys are intelligent.
(Premise 2) Monkeys can peel bananas.
(Conclusion) Therefore: Monkeys are intelligent.

This argument begs the question, because its conclusion has already been assumed in the argument. It usually isn’t so easy to spot.

One of the most famous examples of question begging is with inductive reasoning. Deductive reasoning goes from premises to conclusions mathematically and undoubtedly. Inductive reasoning is the reasoning we use in our everyday thinking. I was tempted to delete this whole section, as it goes off into a little history of philosophy, but I think it illustrates nicely some of the pitfalls arguments can fall into. Please feel free to skip it if you get bored, but that does mean you will go straight to the end of the article.


-- Swans

“Every time I have seen a swan, it was white; therefore, I can reasonably believe that all swans are white.”

Inductive reasoning is going from a finite number of observations to a general rule. (For you pedants out there, this is only one kind of induction, called enumerative induction - but it is the easiest to explain and there are lots of academic papers about how the problem of induction can be translated into these other kinds.)

The rotund Scottish philosopher David Hume was the first to note that this kind of reasoning is problematic. In fact, he concluded that it was entirely irrational. If we construct inductive reasoning in a deductive way, we’ll see why:

(1) So far, all swans have been white.
(2) Nature is uniform - i.e. the future will be much like the past; and ‘over there’ will be much like ‘over here’.
(C) Therefore: ALL swans are (and will be) white.

The problem is with (2) - it needs proving, we can’t just assume it. Hume identified that we couldn’t prove it deductively, because that would mean that we couldn’t conceive of it being any other way - but of course we can, we can imagine a black swan. So we have to try and prove it inductively.

So we might say something like “Nature has been uniform so far, so i can reasonably believe that it will continue to be.” But this argument, however you put it, begs the question. It assumes that nature is uniform in order to prove that nature is uniform. You can’t, concludes Hume, prove induction with an inductive argument. It begs the question. (For further discussion on this topic, particularly about the people who say you can prove induction with inductive arguments, go to the entry in the Stanford Encyclopedia of Philosophy.)

That is what ‘begs the question’ means - when an argument’s conclusion is assumed within the argument.


-- End of Part 2

So this time, we’ve seen:

1. Red Herrings - arguments that don’t address the issue.
2. Straw Men - arguments that address a superficially similar, but in fact different issue.
3. Loaded Questions - controversially complex questions that presume things that haven’t been proved.
4. Begging the Question - when the conclusion is assumed in the argument.

In Part 3, plain old laziness and a discussion of verbal fallacies.