How to Write an Argument

Not open for further replies.


Est. Contributor
  1. Diaper Lover
  2. Diaperfur
  3. Little
  4. Carer
  5. Other
Are you arguing a point with someone? This guide will give the basics to writing an argument, with a view to being able to stand your ground, if not fight back, and even win. I have read threads on many diverse topics here on ADISC. Some are interesting and others are just entertaining, but many discussions just end up as shouting matches, repeating the same old rhetoric, and getting nowhere.

This guide will show you some of the perceptual and cognitive fallacies that you will need to avoid in order to present a sound, logical argument, and which you can identify and attack in the arguments of others. There is a whole lot more to learn than I present here, but if you can apply the principles I detail, you may find it that much easier to persuade someone of your arguments, and once you've mastered this, you can move onto more subtle ideas.

Step One: Know Your Fallacies

Arguably, this is the most important part of constructing a sound logical argument. If you know your fallacies, you can guard against them weakening your own arguments, and conversely, identify the flaws in your opponent's words.

So what is a fallacy? Well simply put, it's a failure in a person's reasoning that exposes their arguments to attack. Though there are a wide variety of possible fallacies, they tend to fall into four general types: logical fallacies, statistical fallacies, extrapolation fallacies, and personal fallacies.

Logical Fallacies

These boil down to a series of steps with an obvious flaw. Most of them are easily demonstrable by silly examples.

All birds are animals. All birds have wings. Aeroplanes have wings. Therefore aeroplanes are animals.

Only men have walked on the moon. I have not walked on the moon. Therefore I am not a man.

These are propositional fallacies. They happen when two statements independently are true, but a derived fact is not. They can easily be used against you if you're not careful. The classic one is with regards to evidence, "If I can see it for myself then it is true, I cannot see it, therefore it must be false."

As Pirates have decreased in number in the Mediterranean Sea the global temperature has increased. Therefore increasing the number of pirates solves global warming.

This is a known as the fallacy of correlation. Just because two things happen at the same time repeatedly, does not mean that they are connected; they could be sheer coincidence, or they could both be caused by the same factor. In the above example, technological progress has made piracy on the Mediterranean Sea more difficult, and that same progress has caused a vast increase in the demand for energy, which results in global warming. Alternatively, the direction of causation could be reversed – as the world gets warmer, pirates find boarding and looting ships too tiring!

Wood is less dense than water, so wood floats in water. Therefore if something floats in water it must have a lesser density. Therefore Wood is less dense than water.

This is known as circular logic. This is a 3-step example, but circular arguments can be as simple as rhetorical questions, or have many steps. Look out for them, as they hide in many philosophical arguments.

Statistical Fallacies

These tend to be down to a misuse of numbers, or human bias. All I can say is beware of numbers; they can be misleading. It is not for nothing that the phrase "lies, damned lies, and statistics" came to be.

With statistics, context is everything. You should be particularly careful about mixing percentages and absolute numbers. For example, a percentage increase is multiplicative when applied to an absolute number, but additive when applied to a percentage - i.e. a 2% increase from 2 is 2.04, which isn't much of a rise; but a 2% increase from 2% is 4%, which is double. Additionally, don't cite numbers if you cannot provide evidence for them; it is the first thing an alert opponent will challenge.

Sample bias refers to the tendency to use unrepresentative samples to illustrate a point about the general population. For example:

99% of the sample polled indicated that they would vote Obama in the next presidential election

This sounds very impressive, but it doesn't mention that the sample consisted entirely of people standing outside the Democrat headquarters. Alternatively:

Our newspaper's straw poll shows that 80% of the population support position X.

Again, looks impressive, but ignores the fact that the sample only consists of people who read that particular newspaper, and who cared enough about the issue to respond to the survey. Achieving a sample that is both representative and random is very, very difficult. That is why polling organizations exist.

Extrapolation Fallacies:

These are attacking an opposing stand by ridiculing it, or extending it to absurdity.

We can't legalize drinking for Teens, because it'll make them delinquents and they won't learn. If they don't learn, then they can't work, putting strain on an already stretched benefits system. The reduction in tax income and increase in spending will harm the economy until it collapses, soon after there will be anarchy and the end of civilization as we know it.

This is an example of a slippery slope argument. It constructs a causal chain from A to Z, via B to Y, where Z is an extreme and undesirable conclusion, and then asserts that A inevitably leads to Z, therefore A should be avoided. The causal chain may be logical, but to assert that each link inevitably leads to the next is extremely questionable. Avoid slippery slopes – they make your argument look weak. They can also be spread over pages of logical arguments, so are notorious in extended debates.

You said this ... which means that ... I attack that stance and imply an attack on what you said..

This is a straw man fallacy - taking arguments out of context, and belittling them to make them easier to attack. Essentially, constructing an argument that your opponent never made, and attacking that, rather than their original argument. Making such a caricature is enough to weaken your argument; attacking it will leave you in a losing position.

It's like this or it's like that...

This is the false dilemma. To say there are fewer options than there actually are is simplistic, and makes choice easier, but when the two options are limited to one end of a continuum, it can cause you to make a wrong decision. For example, an argument that "If we are to abandon oil as a fuel, either we can spend trillions turning the Sahara into a single PV solar panel, or we can spend trillions on nuclear reactors to do the same job," ignores all the other options.

Personal Fallacies:

The worst kind of logical fallacies are those used to attack evidence and people rather than arguments. They are not merely fallacious, but as an attack against other people, they amount to offensive insults in lieu of an actual argument.

A scientist argues X, a layman argues Y.

To reject the opinion of a layman on a scientific matter, just because he/she is a layman is fundamentally invalid. It is probable that the opinion of someone completely unqualified to express one is wrong, but you shouldn't just assume it. Conversely, just because someone is an expert in one field, does not meant that their arguments in another field should automatically be given merit. Attack the argument, not the person.

Drawing a comparison between yourself and an authority figure is also a big mistake. You have to justify your argument. Claiming that your qualifications make you immune from challenge or criticism is an automatic loser. In the same vein comparing the opposition to Nazis, Hitler or Stalin are considered argument losers in a big way. Just because Hitler advocated something, doesn't automatically make that specific thing wrong, even if many of the things he advocated were.

The evidence is bad because...

Sources are good if they can be independently corroborated (two people say the same thing each with their own proof) and have little bias. But are bad if they are conflicting, or have lots of bias. As a rule of thumb, pointing out bad evidence is only reasonable if you can counter it with two or more pieces of better evidence.

Step Two: Get your Evidence Right

If you're going to argue in posts, I would suggest the following as a way of delivering evidence: Pick out one or two sentences that provide the relevant ideas and allow others to look up the source and read the rest it makes you more credible. This is practical because you are providing the relevant bit from a source rather than just linking for others to look for it, and you're giving the context to show that you're using a source in a non-fallacious manner.

Example: "*points at Encyclopaedia Britannica* it says the universe is 12 billion years old" vs "under astronomy in The Encyclopaedia Britannica (An-Az) it says the universe is 12 billion years old"

(In fact, both of these arguments could be countered in that I might look up a more modern text with a better estimate of, say, 13.7 billion years, however I'd be far more willing to discuss it with someone who can point me to the page than someone who can point me to the right shelf in the library.)

So what is good evidence and what is not? Well good evidence has multiple sources of respectable quality. If you want to argue a point on psychology, one peer-reviewed paper is okay, but having a dozen such studies with the same results is better. Good evidence is also reasonably unbiased. If you're going to look at popular news, you should always ask what the writer/presenter has to gain from the story; papers and news networks will sensationalize stories to get sales/viewers, and many have a political agenda of varying degree. It's not a good idea to put all your faith in a single source - nobody is right about everything - but some evidence is better than no evidence.

Step Three: For the Sake of Balance...

Always try and think from your opponent's view, as it'll help you get to an ending position easier.

Some people like white chocolate as it's sweet and creamy however I understand why people might prefer the rich taste of dark chocolate. Dark chocolate is also a little bitter and hard, I'm not a fan of bitter tastes which is why I prefer white chocolate.

Here I had a list of facts about people's thoughts on chocolate and then I added a justification for which side of the line I fall.

You can do the same for any argument, if you take your time to think it through. Once you understand your opponent's position, picking the holes in the arguments becomes that much easier. You see where they used fallacies to their advantage, and where they slipped up and used them to their peril.

I like to think of making good arguments as if I were making Swiss Rolls; they start on a simple premise and were flat, then I mould them into the shape the maker wanted them to take (rolled it up), and then I deconstruct my own arguments from the perspective of my opponent (unrolled it), effectively picking holes in my own argument. Then I add some new information that let me put my argument back together (like the jam), and I roll the whole thing back up. I then take it apart again (unrolling), before finally adding the last missing piece (the cream), and roll it back together one last time. This leaves you in a very strong position as you've discussed both sides of the argument almost equally and come down on one side of the fence.

Step Four: Writing the Argument

Now you're ready to start writing. Once you have evidence, an outline plan of what to say, and a line of attack; you can put fingers to keyboard. As you write, I suggest you follow these tips:-

  • Avoid quote wars; if you're discussing a topic with a group of people, don't turn it into an argument with one person by quoting only them, debunking their work, then waiting for them to argue back. If you end up having a discussion that ignores everyone else's opinions because you and someone else are engaged in a tit-for-tat exchange, then I suggest you stop. Quote wars annoy everyone in a discussion and don't do much to progress a general discussion. If you can't stop quoting, at least move it out of threads and into PMs.
  • Don't repeat yourself. Find new things to say, new evidence, and if you need to change your perspective a little, do so. An argument is nothing if not a learning aid.
  • Remember to PEE in discussions; make a Point, provide some Evidence and then Explain why the evidence proves your point. It'll make the arguments simple to write and easier to follow, and once you've got it down, you know where fallacies usually turn up (in explanations). Point. Evidence. Explanation.
  • If things are becoming difficult to follow, dumb it down and KISS it better. If you can't be concise at least try not to waffle and cloud issues, but at all times think KISS: Keep It Simple, Stupid. Some arguments are unavoidably complex, but don't make an argument hard to follow just because you can.
  • Don't get emotional. If you're crying at, or angry with someone, then you've let them win. Take a step back and reread everything, then start your argument again. Ranting or whining won't help you at all.
Last edited by a moderator:
Not open for further replies.