The Tricks in the Argument: A Plain Guide to Spotting Bad Reasoning in Politics

You don’t need a degree in logic to know when you’re being played. You just need names for a few moves — and the honesty to catch yourself using them too.

By Vivek Sanghi

Picture the last argument you saw about politics. Maybe it was a television panel where six people spoke at once and a seventh kept shouting the name of a state. Maybe it was your uncle at a wedding, or a message forwarded to your family group with the words “must read” and no source. Maybe it was you, at eleven at night, typing a reply you deleted.

Here is the strange thing about all of those. Very little of what upset you was a lie you could check. It was something slipperier — a way of arguing that felt like reasoning but wasn’t. A question dodged so smoothly you forgot it was asked. A real worry answered with an insult. A single frightening story stretched to cover a whole country.

These moves have names. Once you know the names, you cannot stop seeing them, and that changes how you watch the news, how you read a forward, and eventually — this is the real point — how you vote. This is a short field guide to the most common ones, in the plainest language I can manage, with made-up examples so that nobody reading it can guess which side I’m on.

One rule before all the others

Here is the catch, and I want it out in the open before we start, because everything else depends on it.

You are going to enjoy this. You will finish reading and go looking for these tricks — and you will find every single one of them, instantly, in the people you already disagree with. That feels wonderful. It is also the exact moment the whole thing turns useless.

The tricks are not a weapon for winning arguments against the other side. They are a mirror. Our minds are generous to their own team and merciless to the other, and that tilt is invisible from the inside — it always feels like you’re just being fair. So as you read, every time you recognise a move in them, pause and ask whether your own side pulled the same one last week and you nodded along. Do that even half the time and you’re already rarer than most of the people shouting on television.

Attacking the person instead of the point

The oldest move there is. Someone makes a claim; instead of answering the claim, you go after the person who made it. It works because it feels like a fair hit — and sometimes it even lands a real blow — but it leaves the actual question standing there, untouched.

Playing the man, not the ball (ad hominem)

An official says a new road came in on budget. A critic shoots back: “Of course you’d say that — you failed maths in school.” Notice what just happened. Whether the official failed maths has nothing to do with whether the road came in on budget. The numbers are either right or wrong on their own. Attacking the speaker is a way of changing the subject to a person, because a person is easier to dislike than a spreadsheet.

When it is not a trick

Here’s where people over-correct. It is not always foul play to bring up who is speaking. If someone selling you a medicine also owns the company, that’s worth knowing — not because it proves the medicine is bad, but because it tells you to check their claims harder. The difference is simple. Saying “don’t trust his numbers because he stands to profit, so let’s examine them closely” is fair. Saying “he stands to profit, therefore his numbers are wrong” is the trick. One raises your guard; the other pretends to have won.

“But what about them?” (whataboutery)

You point out that a leader broke a promise. The reply is instant: “And what about the other lot? They broke ten!” Maybe they did. But that is a different conversation. The other side’s ten broken promises do not un-break this one. Whataboutery is the national sport of political argument in India, and it is so common that it now passes for a real answer. The first wrong never gets addressed — it just gets buried under a second.

Poisoning the well

This one comes before the argument, not after. “You’re about to hear from Mr. Doe. Now, Mr. Doe has been wrong before, and honestly, who even invited him — but let’s hear it.” By the time Mr. Doe speaks, you’ve been quietly instructed to disbelieve him. Nothing he says has been answered; you’ve just been vaccinated against listening.

Quietly changing the question

The second family of tricks doesn’t attack anything. It moves things — the claim, the choice, the meaning of a word — while you’re looking away, so that the argument you end up having is not the one you started.

The bait and switch (the “motte and bailey”)

This is the most useful one to learn, because once you see it you’ll see it everywhere. A person makes a big, exciting, extreme claim — the bait. You push back. Instantly they retreat to a small, boring, obviously-true claim — the switch — and act as though that’s what they meant all along. Then, once you back off, the big claim quietly returns.

Imagine someone says, “This policy will destroy the country.” You ask how. They say, “Well, I just mean it’s not perfect and we should discuss it.” Of course we should discuss it — nobody disagrees with that. But that modest, safe claim is not what got everyone angry. The next day, the “destroy the country” version is back on the poster. They live in the big scary house when it’s safe, and run to the small sturdy one whenever they’re attacked.

The fake choice (false dilemma)

“Either you support this law, or you hate the poor.” “You’re either with us or against us.” The trick is to shrink a wide, complicated question down to two doors, both chosen by the person speaking, so that whichever you pick, you’ve accepted their frame. Real life almost always has a third door, and a fourth. You can support the goal and dislike this particular law. You can want the problem solved and doubt that this is the way.

Sliding the meaning of a word (equivocation)

A word gets used to mean one thing at the start of a sentence and a different thing by the end. “The government said it would bring reform. This is not reform, it’s just change — so they lied.” The word “reform” quietly shifted from “any improvement” to “the exact improvement I personally wanted” — one word doing two jobs in a single breath, so that a change of subject feels like a caught lie. Big words like freedom, development or justice are the easiest to slide this way, because they sound solid while meaning almost anything.

Winning by noise, not by reason

Drowning you in points (the “Gish gallop”)

This one wins by sheer speed. The speaker fires off fifteen claims in ninety seconds — some true, some false, some half-true, none explained. It doesn’t matter that most of them are weak, because you cannot possibly answer fifteen things at once. By the time you’ve carefully unpicked the first, the audience has moved on and the speaker looks like they had a mountain of evidence. They didn’t. They had a mountain of sentences.

“Look what will happen!” (appeal to consequences)

“If we admit the scheme failed, people will lose faith in the government, so we must say it succeeded.” Notice that this says nothing about whether the scheme actually failed. It only says that admitting it would be inconvenient. Whether something is true and whether it would be comfortable are two completely separate questions, and this move welds them together so you can’t tell them apart.

The louder assertion as a “rebuttal”

Watch for the moment when someone answers an argument simply by saying the opposite, but louder and with more confidence. No new reason is given. The volume and the certainty are meant to feel like a reply. On television this is nearly the whole format: two people saying “yes it is” and “no it isn’t” at rising volume until the break. A rebuttal gives you a reason to change your mind. A louder assertion just gives you a headache.

Bad evidence dressed up as good

The last family is the sneakiest, because it comes wearing the clothes of proper reasoning — facts, numbers, experts, examples. The trick is in how they’re used.

“An expert said so” — and when that’s fair (appeal to authority)

Quoting an expert is not automatically a trick. If a doctor tells you about a disease, that’s a good reason to listen — that’s what expertise is for. It becomes a trick in two ways. First, when the “expert” is an expert in the wrong thing — a famous actor’s views on economics carry no special weight just because he’s famous. Second, when one expert is quoted as if he settles a question that the people who actually study it still disagree about. The honest question is never “did an expert say it?” but “do the people who spend their lives on this actually agree, and is this person one of them?”

Picking only the facts that fit (cherry-picking)

Suppose prices held steady for eleven months of the year and jumped in one. A supporter can tell a completely true story — “prices were stable all year!” — by quietly dropping the twelfth month. A critic can tell the opposite, equally true story — “prices are shooting up!” — by quoting only that one month. Both are citing real figures; the lie is in the ones each of them left on the floor. This is the hardest trick to catch, because there’s nothing false to point at. The only defence is the nagging question: what am I not being shown?

“These two things happened together, so one caused the other” (correlation vs. causation)

Neighbourhoods with more coaching centres tend to post higher exam results — but that does not mean the coaching did it. It may be that better-off, harder-pushing families both hire tutors and raise high scorers, so a hidden third thing is driving both, and the tutoring is doing far less than it looks. In politics: “Ever since they came to power, prices have climbed — so they caused it.” Maybe. Or prices were already rising, or the whole world’s were, or something else entirely did the work. Two things moving together is a reason to ask why — never, by itself, an answer.

A big number with no “out of how many?” (no denominator)

“Ten thousand people were affected!” sounds like a catastrophe or a triumph, depending on who’s talking. But ten thousand out of what? Out of twenty thousand, it’s half of them. Out of a hundred and forty crore, it’s a rounding error. A number with no denominator — no “out of how many,” no “compared to last year,” no “compared to everywhere else” — is not information yet. It’s a mood dressed as a statistic.

One vivid story standing in for everything (anecdote as evidence)

A single, heartbreaking or infuriating story can do more to change your mind than a page of statistics — which is exactly why it’s used. One family’s disaster is real and it matters, but it does not tell you whether the thing that happened to them happens to one family in ten or one in ten lakh. We also only ever hear the stories of the people still around to tell them. If a hundred businesses tried a scheme and one succeeded, you’ll be shown the one, on stage, smiling — and never the ninety-nine who quietly closed.

How these tricks travel in India

Every trick above is universal — people have been dodging questions since there were questions. But a few of our own habits make some of them thrive, and it’s worth naming the local machinery.

Start with the television debate itself. We tend to blame the guests, but look at the design. Eight boxes on one screen, a countdown to the ad break, an anchor rewarded for heat rather than light, and a mute button used like a weapon. The format does not merely allow the tricks — it requires them. Nobody can make a careful, slow point in the nine seconds before someone talks over them, so the careful points don’t get made. What survives is the interruption, the louder assertion, the fifteen-claims-in-a-breath. The panel is not a place where reasoning happens badly. It is a machine built, whether anyone intends it or not, to produce the bad kind.

Then the forward. A message arrives in the family group. It has no author, no date, no source — just confidence, and often a photograph that may be from another country or another decade. It feels true, and here’s the quiet reason why: you’ve now seen the same claim from three different relatives, and your mind treats “I’ve heard this a lot” as if it meant “this is well established.” It doesn’t. A thing repeated ten thousand times by people who all copied it from each other is still just one unsourced claim, wearing ten thousand coats.

Notice the “sources say” construction, which does a lot of work in our news. “Sources say the decision has been taken.” Which sources? Saying what, exactly, and how would they know? An unnamed source can’t be weighed, can’t be checked, and can’t be wrong in a way anyone will remember next week. Sometimes the source is real and genuinely needs protecting. But the phrase has become a way to give a rumour the costume of a fact, and the honest reader keeps a small amount of suspicion in reserve for anything that won’t say who’s talking.

Watch, too, how an argument can change shape between languages. The same claim is often made softer in English for one audience and sharper in a regional language for another — or the reverse — on the bet that the two audiences won’t compare notes. When you can, it’s worth asking what the same leader or channel said about the same thing to a different set of listeners. The gap between the two versions is frequently more revealing than either one alone.

So what do you do about it?

Spotting the tricks is most of the battle, but people always ask the same fair question: fine, I can see it — now what do I say? Here it matters not to become the thing you just learned to spot. Shouting “that’s a logical fallacy!” at your uncle is itself a kind of showing off, and it never once in history changed a mind. Three plainer moves work better.

First, name the move, quietly, mostly to yourself. Half the power of these tricks is that they work in the dark. Just thinking “ah — that was a fake choice” loosens its grip on you, whether or not you say a word out loud.

Second, if you do reply, say their point back in its strongest form first. “So what you’re really worried about is X — have I got that right?” This does two things at once. It stops you attacking a version of their view that nobody actually holds, and it makes the other person feel heard, which is the only state of mind in which anybody has ever reconsidered anything. If their argument was strong, you’ll have to deal with it honestly. If it was one of the tricks, restating it plainly often makes the trick fall apart on its own, in the open, without you having to be clever.

Third, ask the one question the format skipped. Out of how many? Compared to what? Which version do you actually mean? Who exactly is the source? These aren’t gotchas. They’re the ordinary questions a careful person asks, and asking them calmly, without the sneer, is the single most useful political act available to a private citizen.

None of this is about winning arguments at dinner. It’s about the quieter thing that happens over years — the slow work of not being managed. Every election, an enormous amount of effort and money goes into moving you with exactly these moves, on the reasonable bet that you won’t notice. So notice — and insist the arguments be real ones, from every side including your own, before you hand any of them your vote.