I recently read The Duck that Won the Lottery by Julian Baggini. This contains 100 ‘bad’ arguments, i.e. arguments with a fundamental flaw. Whilst there’s no doubting all these arguments are flawed, some are still very persuasive if your reader doesn’t spot the fallacy at the heart of the argument.

This presents me with a moral dilemma – should I advocate the use of lies and deceit as a persuasive tool? Actually it isn’t much of a dilemma as these flawed arguments are used every day anyway, especially in politics and journalism. If you’re going to do it, you might as well understand what you are doing. This chapter pulls together a selection of interesting fallacious arguments from various sources and gives you some guidance on how to use them and, implicitly, how to spot them when used against you.

However, before going into detail, I want to include a couple of caveats:

1. Understand the risks. The problem with using a fallacious argument is that your reader may spot it. No one likes to be deceived so, if they do spot it, you will likely do more harm than good. Far better to use a properly reasoned argument if you possibly can.
2. Understand the consequences. Your primary objective in writing will be to persuade someone to do something. However, this will rarely be your only objective. For example, you will usually want to keep your reader happy and maintain a long-term relationship with them. Therefore, deceiving or threatening your reader may get you a short-term win but could seriously damage your future relationship.

So, these are last-resort tactics when normal persuasive methods won’t work. Use them sparingly and understand the risks and consequences. Take a deep breath, hold your nose and let’s delve into the murky depths of persuasion to investigate the dark arts.

ASSUME CAUSALITY

Sometimes you want to convince a reader that something works when it really doesn’t. This is seen regularly in the world of complementary medicine.

I want to sell a revolutionary new medical treatment that realigns your Chi by getting ants to carry tiny magnets across your body. I claim it can cure bad backs, depression and the common cold. At first glance, this seems an impossible task. However, with assumed causality, I’ll be selling the treatment in no time. The three main weapons I have at my disposal are:

1. It’s no coincidence. I can give people ant therapy and some of them will feel better afterwards. Most of this group will assume it is the ant therapy working, even though it is most likely just the attention, regression to the mean or placebo (see Ben Goldacre’s book Bad Science for more on this). But people like cause and effect so some will swear it worked. I now have my social proof and examples saying “the patient was ill, they had ant therapy and got better. It’s no coincidence”. In business, people use this all the time to claim credit for success. If you put in a little bit of input on projects once you know they will be a success, you can claim it’s no coincidence everything you worked on was successful.
2. There’s so much evidence. Repeat the above hundreds of times and you have ‘overwhelming proof’ ant therapy is the latest miracle cure. It doesn’t matter that it’s all circumstantial evidence; you keep bombarding the reader with enough of it and, sooner or later, they’ll start to think there must be something in it.
3. How else would you explain it? Of course, some unhelpful individual might start to question the way ant therapy works. In which case, you can always fall back on this argument. No matter how ridiculous your theory is, it will often be accepted if no one can come up with a better one. Challenge them to; when they can’t, you can claim this validates your theory.

Just to reiterate, these arguments have more holes than Swiss cheese but they’ve still been effective enough to help establish multi-million pound markets for some complementary therapies in the UK.

IT’S COMMON SENSE, STUPID

Everyone likes to think they have common sense or great intuition. It’s true everyone does possess common sense; it’s just that it isn’t always a good thing to have. If I won the lottery two weeks in a row with the same numbers, most people would assume I was cheating. However, if Sarah Jones in Redruth won one week with one set of numbers, then Steven Hughes in Aberdeen the next with another set, no one would bat an eyelid. However, both situations are equally likely (see ‘Abuse statistics’ below).

Yet common sense is still prized and calling someone’s intuition into question can be quite insulting. As such, people will tend to shy away from questioning something if you say it is common sense. After all, if they question your point, aren’t they showing they don’t have common sense?

So how do you use this in practice? Here’s an example:

We should pay an incentive bonus for staff in attendance on Mondays and Fridays. Obviously, this is when most staff will be tempted to ‘throw a sickie’ in order to take a long weekend. An attendance bonus will reduce absences and increase productivity.

According to the UK Office for National Statistics, weekday absence rates are pretty much equal, with Monday the least likely day for absences. The ‘common sense’ of Monday and Friday sick leave just doesn’t hold true across all industries, ages and regions. However, if you are a union lobbying for more pay for staff, an argument such as this may be very effective in increasing pay while seemingly benefiting the employers more.

Even if something isn’t common sense, claiming it is can be enough. Just don’t try to be too ambitious: although it’s common sense to use bicycle couriers in inner-city London, don’t claim they’re the obvious choice for international freight.

IT’S COMMON KNOWLEDGE

It’s amazing what you can get away with if you over-generalise. A quick search of this book reveals I have claimed something is ‘always’ the case at least five times but I doubt anyone has noticed (until now, obviously).

If you do not have the evidence to back up a claim, you can always try to generalise. The following phrases (or variants) are often used:

• “It is generally accepted that . . .”
• “The widely held/discredited view is . . .”
• “It is believed that . . .”
• “Always/never”
• “Everyone/no one”
• “The traditional view has been . . . However, it is now thought . . .”
• “The majority of people/businesses/managers . . .”

Of course, your reader may know you are completely wrong in your claim so be careful where you use generalisations. However, a few sprinkled here and there can get you out of a sticky situation without your reader noticing. In fact, it may even lead them to believe your claim as they subconsciously stash it away as fact.

ABUSE STATISTICS

The staple diet of journalists and politicians, dodgy statistics have contributed to health scares, wrongful convictions and even wars. Here are four of the most common ways of twisting numbers to suit your needs:

1. Ignore the base rate. I have software that can correctly spot a credit card fraud 90% of the time and calls a good transaction a fraud just 0.1% of the time (false positive rate). This seems pretty good on the surface. However, if the rate at which frauds occur (the base rate) is taken into account, is it still as good? If there is 1 fraud in every 1,000 transactions and I look at 100,000 examples, I should spot 90 of the 100 frauds. However, 1 in every 1,000 genuine transactions will be flagged up as a fraud when it isn’t. This means, of the frauds I detect, 90 will be genuine and about 100 will be false positives. This would probably be acceptable for the industry. However, if 1 fraud occurs in every 10,000 transactions, the situation gets worse. Now, I still have around 100 false positives but only 9 genuine frauds. This would be disastrous for the industry, as the money saved in preventing fraud would be swallowed up handling all the customer complaints from false positives. If you want poor statistical data to look good, ignore the base rate and just claim the percentages.
2. Compare best and worst cases. Let’s say you are comparing your product with that of a rival. On average, your product is worse. Never fear, simply compare the incomparable. For example, on a car, you could compare your predicted or laboratory fuel consumption with a rival’s actual average from real-world driving. Alternatively, you could compare their average to your extra-urban cycle, claiming your car could achieve ‘up to 30% higher fuel consumption’ than your rival. Note the use of the words ‘up to’. These have been responsible for more dodgy advertising claims than any other phrase. Without the details in place, your reader won’t be able to make an informed judgement and is more likely to simply accept your assertions.
3. Selection bias. This states that the results of any statistical analysis are heavily influenced by the source of your data. Take cosmetics for example. Almost every ad says something like “9 out of 10 women agree it gave them younger-looking skin”. So where do they get their numbers from? If it were me, I’d send free samples of the product to around 150 regular users of my brand and ask them to rate it. Not only do these people like my company already, they now feel obligated to be nice as I’ve given them free stuff. Secondly, I’d word the questionnaire to increase the likelihood of the user being positive, by using false dichotomies (see ‘And a few more’ later in this chapter). If I didn’t get the results I wanted, I would do another survey. Of course, I wouldn’t ever claim 100% satisfaction, as that might look contrived. Better to go with 77% or 86%. In fact, as an aside, a very unscientific search of Google shows most face creams to be between 75% and 90% effective.¹ The same approach can be applied to any other set of statistics – by cherry-picking your data and leaving out the detail, you can achieve the most persuasive numbers.
4. Find patterns in randomness. This links back to causality. If you can find a cluster of random events, you can claim they aren’t random but indicative of something else entirely. The book The Bible Code worked entirely on this principle. Worse, there have been several miscarriages of justice based almost solely on such dodgy statistics (in particular the case of Sally Clarke). In using this approach, you are asking the question “how likely is it to happen by chance?”, knowing the reader will say “very unlikely” in reply.
   To illustrate this point, I’m going to prove my staff are faking their sick days. Let’s say we have 15 people in the office. Bill and Sarah were both ‘off sick’ last Friday, but no one else was ill. As my staff are pretty healthy and only take one sick day a year on average, I’m sure they’re skiving off together. After all, what are the odds of that happening by chance?
   There are 220 working days a year. To prove my point, I can say there is a 1 in 220 chance Bill was off sick on Friday. The odds of Sarah being off sick at exactly the same time are 1 in 220 times 1 in 220, or 1 in 48,400. Time to start the disciplinary proceedings.
   Or is it? What I’ve done here is assume Bill and Sarah were both off on one particular day, which would indeed be unlikely. However, it could have been any two staff off together on any working day in the year. In fact, there are 105 different pairs of staff that could be off together on any given day. Although the likelihood of each pair being off on a particular day is small, there are so many possible times for a pair to be off, it soon becomes quite likely one combination will happen. In this case, the likelihood of a pair of colleagues being off together at some point in the year is 38%. Although the data are entirely random, I’ve spotted a pattern that seems unlikely so have assumed it couldn’t be chance. This is another example of the failure of common sense. However, even though I’d be wrong, most people would believe my reasoning and statistics and agree my staff were up to no good.
   In short, if you see a pattern in random data, you can claim it is significant by ignoring the fact it probably turned up by chance and asking the rhetorical question “what are the odds that this could happen by chance?” Throw in some dodgy statistics proving your point and you may just show the highly probable is almost impossible.

WIDEN OR NARROW DEFINITIONS

Wouldn’t it be nice to claim you were the best at something? However, not everyone can be the best. If you’re not, why not change the meaning of ‘best’? You can do this by widening or narrowing the definitions you use.

For example, perhaps you want to claim you have more customers than any other firm in the same industry. If you don’t, try changing the definition of ‘customer’. You could widen it to include quotes issued and not just work completed. Alternatively you could narrow the definition to exclude numbers that don’t help your cause. Your new definition of ‘customer’ could be only those you have a long-term relationship with or who spend over a certain amount. Again, this trick is used by journalists to create sensational headlines. A city can be called the most violent in Britain on the basis of all crime, all violent crime, all gun crime, all murders, all deaths or even all gun-related murders. By tailoring your definition, you can legitimately make bold headline claims that draw your reader in.

FLATTERY WILL GET YOU EVERYWHERE

As a general rule, people have a higher opinion of themselves than is actually warranted. For example:

• between 70% and 90% of people think they are better drivers than average
• 87% of MBA students in a study rated their academic ability above average
• people of lower skill are more likely to over-estimate their ability.

This information is far more important than reality – as far as our reader is concerned, this is reality. Two particular biases make the situation worse. First, confirmation bias (see Chapter 3) ensures your reader will ignore anything that disproves their ability/intelligence. Secondly, attributional bias will reinforce their belief in their own superiority. Attributional bias means we attribute our successes to skill and our failures to bad luck. Likewise, we attribute others’ success to luck and their failures to their lack of ability. Along with others, these biases ensure the majority of us have an overly inflated opinion of ourselves.

Rather than battle this, we can improve our chances of persuading a reader by agreeing with how great they are. Consider using some or all of the following forms of flattery:

• We trust you to make the right decision.
• We remember exactly who you are.
• We wouldn’t exist without you.
• We have special options for people of your ability.
• We need your help because you’re the best.
• As you clearly understand this, we’ll just give you the outline.

Some of these are more than just bare-faced flattery. Special options can cost more, for example. My nephew was told he was too good for the cheaper group guitar lessons and so needed more expensive one-on-one tuition. No parent would wish to deny their child that sort of opportunity if they could afford it, so the guitar tutor makes more money. I’m sure my nephew really does need extra tuition but, if I were a hard-up guitar tutor with low moral standards, I’d do something similar.

The only thing to watch with flattery is that you don’t go too far. Words like sycophantic, toadying and brown-nosing are not compliments. Also, flattery works best if your reader has a high ego. Those with low self-esteem can be made to feel good but you run the risk of their viewing your efforts more suspiciously. As always, know your reader.

AND A FEW MORE . . .

I don’t have space to include individual sections for every technique so I’ve summarised some more here. Beyond this, have a look for lists of fallacious arguments on the web. Just remember, most are not effective and, if you possibly can, you should use a properly reasoned argument.

• False dichotomy – only give the reader two options, despite there being many more. Comparing a complete package with nothing at all, despite there being options for only part of the package.
• You’re a Nazi if you disagree – link the alternative to your view to something unpleasant. “Enforced redundancies are akin to corporate euthanasia. We strongly oppose any such Final Solution.”
• Naturalistic fallacy – if it’s natural it must be good. “Belladonna and Nightshade smoothie – 100% natural ingredients!”
• Bribes and threats – direct threats are not something you want to put in writing but veiled bribes and threats are used all the time in business. These appeal directly to the powerful greed and fear emotions. Not good for long-term relations though. “Acecorp shares its experiences of sub-contractors with our partners and peers across the industry” – subtext: keep us happy or you won’t work again.
• Perfect/imperfect solution – if there is a rival solution or option, try to find weaknesses. If it is perfect, claim it is too perfect to be credible. If it’s not, highlight those omissions.
• Prove a negative – you can’t prove something doesn’t exist. You can claim your argument must be true if no one can prove it isn’t. Has been used for millennia by people to ‘prove’ the existence of God. “You can’t prove my company can’t deliver, therefore, you must assume it can.”
• Beg the question – this means avoiding the question. By carefully wording your statement, you can prove something is true because it is true, thus avoiding the question ‘why is it true?’ “Our integrity cannot be questioned, as only a company with the highest moral standards can grow as big as we are.” This example begs the question ‘can a company with low moral standards grow big?’ Essentially, we end up saying ‘we must have high integrity because we’re big, and we’re big because we have high integrity’.

SUMMARY

Wherever possible, use a well-crafted argument backed up with sound data. However, as a last resort, you may be able to carefully apply one or more of the following:

• Assume causality – it’s no coincidence, there’s so much evidence, how else could you explain it?
• It’s common sense, stupid – even though intuition is often wrong, make people believe you by claiming your argument is common sense.
• It’s common knowledge – everyone knows this is true so you should believe it too.
• Abuse statistics – ignore the base rate, compare extremes, bias your selection and find patterns in randomness. These will all produce impressive but misleading statistics.
• Widen or narrow definitions – tweak the meaning of a term until it suits your needs.
• Flattery – most people have high opinions of themselves so make them feel special, especially if they make the decision you want.
• And more . . .
  • False dichotomy
  • You’re a Nazi
  • Naturalistic fallacy
  • Bribes and threats
  • Perfect/imperfect solution
  • Prove a negative
  • Beg the question.

THE END

And that’s it! Or rather it isn’t. This book has merely scratched the surface of persuasive writing. More detail on all this content can be found on the web and in a range of books, so get reading.

In the Appendices, there’s a summary of the rules, some alternatives to supersized words, some useful tools and a worked example of the seven-step concise writing process.

Finally, if you take away nothing else from this book, remember these two points:

Reader Response equals Result

and

If following a rule makes your writing less persuasive, break it.

1 I spotted this in my proofread. The creams aren't shown to be effective; the stats just show 77% of women in a contrived sample agree with a statement. I've done exactly what the ads want me to do, which is assume anecdotal survey results equals proof of effectiveness.