Unit 22 · Reasoning and Argument

Read this small argument. All Chicago Public Schools are closed on Labor Day. Lane Tech is a Chicago Public School. So Lane Tech is closed on Labor Day. Notice what happened. You did not need to visit Lane Tech on Labor Day to be sure. Once you accept the two opening statements, the ending is forced. That is deductive reasoning: the conclusion is already contained in the premises, and the argument simply draws it out.

A premise is a statement the argument starts from, a reason offered in support. The conclusion is the statement the premises are meant to support. In the Lane Tech argument there are two premises, one general (all CPS schools are closed) and one specific (Lane Tech is a CPS school). The ancient Greek philosopher Aristotle called this three-part shape a syllogism, and he noticed that its power comes from its form, not its subject. Swap in any general rule, any member of the group, and the machine still runs.

Deduction is the reasoning of mathematics and of the law. A geometry proof moves from definitions and earlier results to a conclusion that must follow. A judge reasons from a law (anyone who does X is guilty of Y) plus a finding of fact (this person did X) to a verdict. In each case the argument does not gather new evidence at the end. It shows that the conclusion was already guaranteed by what came before.

The catch is that deduction is only as good as its starting points. If the first premise were false, if some CPS schools stayed open on Labor Day, the argument would still have a perfect shape and a worthless conclusion. So when you meet a deductive argument, you have two separate questions to ask. Does the conclusion really follow from the premises? And are the premises actually true? The next two lessons take those questions one at a time.

A nurse at Stroger Hospital notices that the last forty patients who came in with a certain rash had all eaten at the same food festival. She concludes that the festival is probably the source. She could be wrong. Patient forty-one might have caught the rash somewhere else, and the first forty might be a coincidence. But forty cases pointing one way make her conclusion strong. This is inductive reasoning: gathering particular observations and moving to a general claim that is probable, not guaranteed.

Almost everything you know about the world came to you by induction. The sun has risen every morning of your life, so you expect it tomorrow. Every raven anyone has recorded has been black, so you say ravens are black. Science works this way. A researcher runs a trial with two thousand people and concludes that a drug lowers blood pressure in people generally. The conclusion reaches past the evidence, to people who were never in the trial. That reach is the whole point, and also the risk.

Because induction never reaches certainty, we judge it by strength. A strong inductive argument has many cases, chosen fairly, that all point the same way. A weak one has few cases, or cases picked because they fit. Three friends who hate a new phone do not tell you much about the phone. A survey of three thousand owners, chosen at random, tells you a great deal. The difference is not the kind of reasoning. It is the size and fairness of the sample.

Deduction and induction work together. A detective reasons inductively from mud on a boot to a likely walk through the park, then deductively from a rule (the park gates lock at nine) to the time the suspect must have been there. When you read an argument, ask which move the writer is making. If it is a deduction, check the form and the premises. If it is an induction, check how many cases, how they were chosen, and how far past them the conclusion reaches.

Here is an argument with a perfect shape and a rotten start. All fish can fly. A trout is a fish. Therefore a trout can fly. Laugh, but look closely: if the first two sentences were true, the third would have to be. Logicians call an argument like this valid. Validity is about form only. It asks whether the conclusion follows from the premises, and it does not care whether the premises are true.

Now here is an argument with true sentences and a broken shape. All dogs are mammals. All cats are mammals. Therefore all dogs are cats. Every premise is true, and the conclusion is false, which proves the form is bad. Being a mammal does not make dogs cats any more than living in Illinois makes you a resident of Springfield. This argument is invalid, and no amount of true premises can rescue it.

An argument that is valid and has true premises is sound. Sound arguments are the goal, and they are rarer than they look. The trout argument fails on truth. The dogs-and-cats argument fails on form. Marisol’s crash argument from the chapter opening failed on form too: her true premise about crashes near home could not by itself support her conclusion about where driving is dangerous. A premise she never stated, that people drive equally everywhere, was doing the work, and that premise was false.

So checking a deductive argument is a two-step job. First, pretend the premises are true and ask whether the conclusion is forced. If not, stop: the argument is invalid. Second, if it is valid, test each premise against what you know. A good test for validity is to imagine a case where the premises hold and the conclusion fails. If you can build one, the form is broken. If you cannot, the form holds, and the only question left is whether the writer’s starting points are true.

Someone says: He went to Northwestern, so he must be smart. There is only one premise on the surface, and the conclusion does not follow from it alone. Something is missing. Fill it in and the argument becomes: Everyone who went to Northwestern is smart. He went to Northwestern. So he is smart. Now it is valid, and now you can see the weak spot. The first premise, the one nobody said out loud, is the one that is hard to defend. That silent premise is a hidden assumption.

Real arguments almost always leave a premise unspoken, and that is not dishonest. Spelling out every step would make speech unbearable. The trouble is that the hidden premise is often the one that carries the weight. An ad says: Nine out of ten dentists recommend this brush, so you should buy it. The hidden assumption is that what most dentists recommend is what you need. Maybe. A politician says: Crime rose after the new law, so the law caused it. The hidden assumption is that nothing else changed. Probably not.

To find a hidden assumption, write the stated premise and the conclusion on two lines and ask what sentence would have to sit between them to make the argument valid. Then ask whether that sentence is true. Marisol’s crash argument worked this way. Stated premise: most crashes happen near home. Conclusion: driving near home is more dangerous. The bridge sentence: people drive the same number of miles everywhere. False, so the bridge collapses.

This tool cuts both ways. When you write, your own arguments carry hidden assumptions, and a sharp reader will find them. Before you turn in an essay, do the two-line test on your main claim. If the bridge sentence is something your reader would accept without a fight, leave it silent. If it is the very thing a reader might doubt, you have found the paragraph you still need to write.

Suppose the claim is that later school start times help teenagers. Four writers support it four ways. The first tells the story of her cousin, who slept an hour more and raised his grades. That is an anecdote: one case, told vividly. It is powerful because it is human and specific. Its limit is that one case proves nothing about the rest. The cousin might have raised his grades for ten other reasons, and a different cousin might have done worse.

The second writer cites a study of nine thousand students across eight schools in which later start times were followed by higher attendance. That is a statistic: a number drawn from many cases. Its strength is breadth. Its limits are the questions it invites. Who was counted? How were the schools chosen? Did anything else change at the same time? A statistic answers one question precisely and can be quietly stretched to answer questions it never asked, as Marisol found.

The third quotes a sleep researcher from the University of Chicago who says teenage bodies are wired to fall asleep late. That is expert testimony. It matters because the expert has seen more evidence than you have. Its limit is that experts can be wrong, can speak outside their field, and can disagree with each other. A researcher on sleep is worth hearing about sleep, not about bus schedules or budgets. The fourth writer argues that we would never schedule a night-shift worker’s most important meeting at three in the morning, so why schedule a teenager’s hardest class at seven. That is an analogy: a comparison that says two situations are alike in the way that matters. It illuminates, but it proves only as much as the two cases truly share.

A strong argument uses more than one kind and matches each to its job. An anecdote opens a door; a statistic shows the door is not the only one; an expert explains what is behind it; an analogy helps a reader who has never walked through. When you evaluate someone else’s evidence, name the kind first. Then ask the limiting question that belongs to that kind. Does this one case stand for many? What exactly was counted? Is this expert in the right field? Are these two things alike where it counts?

A headline reads: Shark attacks double in a year. Alarming, until you learn the count went from two to four. Another reads: Ice cream sales and drownings rise together every summer. True, and ice cream does not cause drowning; hot weather sends people to both the freezer and the lake. A third shows a graph where a line shoots upward, until you see the vertical axis starts at 98 and ends at 100. Each number is real. Each presentation invites a conclusion the number does not support.

Three questions rescue you. First, what is the base rate, the ordinary background level? A doubling from two to four is not the same as a doubling from two thousand to four thousand, and a rate (crashes per mile, cases per thousand people) is almost always more honest than a raw count. Second, is this a correlation or a cause? Two things rising together is a correlation. Proving one causes the other takes more: ruling out a third factor, like hot weather, and often a controlled comparison. Third, what does the graph’s frame do? A truncated axis, a cherry-picked start year, or a scale that stretches one side can make a molehill look like a cliff.

Look at a small example. A school reports that students who attend tutoring score twelve points higher on average than those who do not. Does tutoring add twelve points? Maybe. But who chooses tutoring? Possibly the students who already care most about their grades. The comparison is between two groups that differed before tutoring began, so the twelve points may measure motivation as much as tutoring. A fairer test would assign students to tutoring at random and compare afterward. Until then, the number is a correlation with a story attached.

None of this means numbers are untrustworthy. It means numbers are precise, and precision cuts both ways. A statistic answers exactly the question that was asked of it, in exactly the group that was counted, over exactly the span that was measured. Your job as a reader is to recover that question. Your job as a writer is to state it, so that your reader never has to.

A fallacy is a mistake in reasoning that looks like reasoning. The most common family is the fallacies of relevance: the reply is loud, but it does not touch the argument. Start with the ad hominem, Latin for to the person. A student proposes a later start time, and someone answers, You just want to sleep in. Maybe she does. Her motive says nothing about whether her evidence is good. Attacking the arguer instead of the argument is the oldest dodge there is.

Next is the straw man. The student says start times should move thirty minutes later. Her opponent replies, So you think school should start at noon and kids should learn nothing. He has built a flimsy version of her claim, knocked it down, and walked away as if he had beaten the real one. The test for a straw man is simple: would the original speaker recognize her own position in the reply? If not, the reply has changed the subject.

The red herring changes the subject openly. Asked about start times, a board member says, What we really need to talk about is the football field. It might be a fine topic. It is not this one. A close cousin is the appeal to emotion, which replaces evidence with feeling: a photo of a yawning child instead of attendance data. Feelings are not irrelevant to human life, but they are not evidence that a claim is true.

Finally the bandwagon: Every other district has done this, so we should too. Popularity is not proof. Plenty of districts have made mistakes together. The reverse is also a fallacy, sneering at an idea because it is common. When you meet any of these, do not simply shout the fallacy’s name. Say what was skipped. The evidence was never answered; the actual claim was never addressed; the subject was changed; the crowd was counted instead of the reasons.

Induction is judged by strength, so its fallacies are all versions of one error: leaping too far from too little. The hasty generalization is the plain form. Two rude cashiers at one store, and a customer declares the whole chain has terrible service. The sample is tiny and probably not typical. Stereotypes about groups of people are hasty generalizations with a long history and real damage behind them.

The post hoc fallacy, from the Latin post hoc ergo propter hoc, after this therefore because of this, confuses sequence with cause. A team changes its jerseys and wins the next game; the jerseys must be lucky. A city passes a law and crime falls the next year; the law worked. Maybe, but crime might have been falling already, or the economy improved, or the weather changed. That something came after is the weakest possible reason to believe it was caused. The correlation lesson already gave you the tool: look for the third factor.

The slippery slope predicts a chain of consequences without showing that any link actually holds. If we let students use phones at lunch, next they will use them in class, then during tests, then no one will learn anything. Each step is asserted, none is argued. Real slopes exist, and a writer may fairly warn about one, but only by showing why each step leads to the next. The fallacy is in skipping that work. The false analogy fails the analogy’s own test: the two cases are not alike in the way that matters. Running a school is like running a business, so let the customers decide. Students, though, are not customers who can be turned away, and a school’s product is not for sale.

A last one is the appeal to ignorance: no one has proved this supplement does not work, so it works. Absence of evidence against a claim is not evidence for it. The burden sits with the person making the claim. When you catch one of these leaps, the fix is the same each time. Ask how many cases, how they were chosen, what else might explain the result, and whether each step in the chain has been shown rather than assumed.

You are either with us or against us. That sentence has done a lot of work in history, and its work is to hide a third option. The false dilemma presents two choices as if they were the only ones. Either we cut the arts program or we raise taxes. Perhaps. Or perhaps there are grants, a smaller cut, a different program, a fee. The fix is a question: are these really the only two? Usually they are not, and the person who framed the choice knows it.

Begging the question is often misused to mean raising a question, but its real meaning is sharper. It is an argument whose premise already assumes the conclusion. This novel is the best of the year because no other book this year is as good. The reason restates the claim in different clothes. Circular arguments can run long enough that the circle is hard to see; a paragraph that says a policy is fair because it treats everyone justly has traveled nowhere.

Equivocation uses one word in two senses and hopes you will not notice the switch. The law says the government must protect our rights. I have a right to be angry about the new tax. So the government must protect my anger. The word right means a legal entitlement in the first sentence and a mere justification in the second. The argument looks valid only because one word is doing two jobs. A related trick redefines a group to dodge a counterexample: no real Chicagoan puts ketchup on a hot dog, and when you produce one who does, he is declared not a real Chicagoan. The claim has been made unfalsifiable.

Knowing the names of fallacies is useful, and it is also dangerous. A reader who has learned twenty labels can start slapping them on any argument he dislikes, which is its own kind of dodge. Use the catalog the way a mechanic uses a diagnostic list: to find exactly where the engine fails, and then to say so in plain words. The goal is never to win by naming. It is to see clearly what an argument has proved and what it has only claimed.

In the 1950s the British philosopher Stephen Toulmin noticed that real arguments, in courtrooms and newspapers and kitchens, rarely look like textbook syllogisms. He drew a different map, and it has been used in writing classes ever since. It starts with three parts. The claim is what you want the reader to accept. The grounds are the facts, data or observations you offer for it. The warrant is the general principle that makes the grounds count as support for the claim.

Take a small example. Claim: Ms. Okafor’s class should get the field trip money. Grounds: her class raised the most in the fundraiser. Warrant: the class that raises the most should get the money. Notice that the warrant is usually the hidden assumption from earlier in this chapter, brought into daylight and given a name. It is the bridge between fact and claim. If a reader accepts the warrant, the grounds carry the claim. If not, no amount of grounds will help.

Most disputes are really disputes about warrants. Nobody argues that Ms. Okafor’s class raised the most; the sign-in sheet settles that. What they argue is whether raising the most should decide who goes. Another parent might offer a different warrant: the class that has never gone should go first. Same grounds, different warrant, different claim. Finding the warrant tells you where the actual disagreement lives, which is where your essay needs to spend its time.

Try mapping your own draft. Underline the claim. Circle every sentence that offers grounds. Then write, in one sentence, the warrant that connects them, even if it does not appear in the draft. If you cannot write it, your grounds may not support your claim. If you can write it and it sounds like something your reader would dispute, you have found the sentence that must be argued, not assumed.

Toulmin’s map has three more parts, and they are what separate a strong argument from a bare one. Backing is support for the warrant itself. If the warrant is that the class that raises the most should get the trip, the backing might be that the fundraiser rules announced that in September. A warrant without backing is just an opinion about what should count; with backing, it becomes a rule the reader can check.

The qualifier states how strong the claim is. Words like probably, in most cases, unless, and under these conditions are not weakness. They are precision. A claim that a later start time will raise attendance in most large urban high schools is stronger than a claim that it will raise attendance everywhere, because the first can survive a counterexample and the second cannot. Readers trust writers who know the edges of their own claims. Overclaiming is the fastest way to lose them.

The rebuttal names the conditions under which the claim would not hold, or answers the best objection against it. Ms. Okafor’s class should get the trip unless the money was raised by a single donor rather than the students. A writer who states that exception before the reader thinks of it has taken the objection’s power away. This is the move students most often skip, and it is the one that most impresses a careful reader, because it shows the writer has argued against herself and survived.

Put together, the six parts form a full argument. Claim: the district should move start times thirty minutes later (qualifier: at its large high schools, for a two-year trial). Grounds: attendance rose at the two pilot schools; the sleep research on adolescents. Warrant: policies that raise attendance without added cost deserve a trial. Backing: the board’s own stated priority on attendance. Rebuttal: bus costs are the main objection, and the pilot schools covered them by staggering routes. You do not need to label the parts in your essay. You need to have them.

You now have every tool in this chapter. Here is the order to use them in. First, find the claim and write it in one sentence; if you cannot, the argument may not have one. Second, list the grounds and name the kind of each: anecdote, statistic, expert, analogy. Ask each kind its limiting question. Third, write the warrant, whether stated or hidden, and ask whether a reasonable reader would accept it and whether the writer offers any backing.

Fourth, decide what kind of reasoning links grounds to claim. If it is deductive, test the form by imagining true premises with a false conclusion. If it is inductive, count the cases and ask how they were chosen and how far the conclusion reaches beyond them. Fifth, scan for the fallacies: any attack on a person, any distorted version of an opposing view, any leap from sequence to cause, any forced pair of choices, any word that changes meaning. Sixth, look for a qualifier and a rebuttal. Their absence is itself information.

Here is the method on a short editorial. Claim: the city should ban plastic bags. Grounds: a photo of a lakefront covered in bags; a statistic that residents use hundreds of millions of bags a year; a report that a coastal city’s litter fell after a ban. Warrant, unstated: what reduced litter elsewhere will reduce it here. The photo is an anecdote, vivid but single. The statistic counts bags used, not bags littered. The coastal report is the real evidence, and the warrant it needs is an analogy: is a coastal city like Chicago in the way that matters? No qualifier, no rebuttal. Verdict: the claim may be right, but the editorial has established that bags are numerous, not that a ban here would work.

That last sentence is the goal. An evaluation does not end with thumbs up or thumbs down. It ends with a precise statement of what the argument proved and what it only asserted. Write that sentence for every argument you read this year, including your own. Marisol’s radio statistic proved that crashes cluster near home. It never proved that home was the dangerous place. The number was true. The conclusion was not. Knowing the difference is the skill.

Most people remember the Declaration as a list of stirring phrases. Read it instead as a piece of reasoning and it becomes something stricter. The second paragraph lays down the general premise. People are created equal and have rights they cannot sign away; governments exist to secure those rights; when a government instead destroys them, the people may alter or abolish it. Each clause is stated as a truth that needs no proof, which is exactly what a deductive argument’s first premise is supposed to be.

The middle of the document supplies the second premise, and this one is not self-evident. It is a claim of fact: the present king has aimed at absolute tyranny over these states. A claim of fact needs evidence, and the Declaration announces that it will provide some. Let facts be submitted to a candid world, it says, and then it does, for more than two dozen separate charges. That sentence is the hinge of the whole document. It is where the argument stops asserting and starts proving.

The conclusion comes last, in the final paragraph, and it is stated as a consequence. Because the premises are true, these colonies are, and of right ought to be, free and independent states. Notice the careful wording. Not simply that they wish to be free, but that they have a right to be, a right that follows from the first premise applied to the facts of the second. The argument is a syllogism in the shape you learned last chapter: all governments that destroy rights may be abolished; this government has destroyed rights; so this government may be abolished.

Once you see the form, you can test it the way you test any deduction. Is it valid? If the premises hold, the conclusion does follow. Is it sound? That question splits in two. The first premise, that people have unalienable rights, was a philosophical claim drawn from writers like John Locke, and the Declaration does not argue for it; it declares it. The second premise, about the king’s conduct, is where the document does its evidentiary work. A reader in 1776, or today, who wanted to resist the conclusion would have to attack one of those two premises. That is the map of every argument that has ever been made about this text.

The long middle of the Declaration is the part students skip, and it is the part that does the work. Each charge begins the same way: He has. He has refused his assent to laws. He has dissolved representative houses. He has obstructed the administration of justice. He has made judges dependent on his will alone. He has kept standing armies among us in times of peace. He has plundered our seas, ravaged our coasts, burnt our towns. Read aloud, the repetition sounds like a drum, and that is deliberate. The device is called anaphora, repeating a phrase at the start of successive clauses.

But the charges are not only music. They are evidence, and they work by induction. No single charge proves a plan of tyranny. One refused law might be an accident, one dissolved assembly a misunderstanding. Stacked together, more than two dozen of them, all pointing one way, they support a general conclusion: this is not a series of accidents but a design. The Declaration says so explicitly, calling it a long train of abuses. That phrase is the inductive move, from many cases to a pattern.

Notice what the charges leave out. They name no dates, no places, no specific acts of Parliament. The taxes of the 1760s, the closing of Boston Harbor and the trials of soldiers who fired on colonists are all present, but only in general language a reader of 1776 could decode. Nine of the charges, about acts of Parliament, are grouped under one He has and begin instead with For. That was a choice. Specifics could be disputed one by one; general charges, delivered in a rhythm, accumulate weight. A modern reader evaluating the evidence should notice both things: the pattern is real and well documented, and the presentation has been shaped to make the pattern feel inevitable.

There is also a rhetorical shape underneath the logical one. The charges climb. They begin with legal and administrative complaints and end with violence: plundered seas, ravaged coasts, burnt towns. The order moves the reader from irritation to outrage, so that by the time the conclusion arrives, it feels less like a decision than a necessity. Evidence and rhetoric are working together, and a careful reader keeps them apart long enough to judge each on its own.

The rough draft of the Declaration is the closest thing the country has to a recording of an argument being made. Two edits matter most. The first is the smallest. Where the final text says self-evident, Jefferson had written that the truths were sacred and undeniable. The change, usually credited to Franklin, moved the premise from religion to logic. A sacred truth asks for belief. A self-evident one asks for nothing; it presents itself as the kind of statement no reasonable person could deny. That is a stronger position to argue from, and a riskier one, because a reader who does deny it has nowhere else to be sent.

The second edit is the largest. Jefferson’s draft contained a long paragraph accusing the king of waging cruel war against human nature by carrying on the slave trade, and of then inciting enslaved people to rise against the colonists. The paragraph is tangled. It blames the king for a trade that colonists themselves had profited from, and it seems angrier that enslaved people might be freed by the British than that they were enslaved at all. Congress removed it. Delegates from Georgia and South Carolina objected, and Jefferson later wrote that some northern delegates, whose merchants had carried enslaved people, were uneasy too.

The cut left a silence at the center of the argument. The first premise said all men are created equal. The document was written and signed by men who enslaved other men, and after the edit it did not mention slavery at all. From a logician’s view, that silence is a hidden assumption: that the word all in the first premise did not, for practical purposes, include everyone. For the next eighty-nine years, the fiercest American arguments would be about whether that hidden assumption could stand.

Studying the draft teaches something about every argument. Premises are not found; they are chosen, and the choice is often a matter of what an audience will accept. The signers needed the votes of Georgia and South Carolina, so the argument was trimmed to keep them. A writer today faces the same pressure in miniature every time a paragraph is cut because a reader will not like it. The draft asks a hard question: when a cut keeps the audience but breaks the logic, what has the argument actually won?

The Constitution opens with a single sentence, fifty-two words long, and every part of it does a job. It begins with the actor: We the People of the United States. That phrase was itself an argument. An earlier draft listed the states by name; the final version, usually credited to Gouverneur Morris of the committee on style, made the people the author of the government rather than the states. It is a premise disguised as a subject: the authority for what follows comes from the people directly.

Then come six purposes, each introduced by a verb. To form a more perfect union. To establish justice. To insure domestic tranquility. To provide for the common defense. To promote the general welfare. To secure the blessings of liberty to ourselves and our posterity. Read them as reasons and they answer a question the Declaration had raised: if a people may abolish a government that fails them, what is a government for? The Preamble’s answer is this list, and the rest of the Constitution is the means to those ends.

The sentence ends with the action: do ordain and establish this Constitution for the United States of America. Grammatically, the whole Preamble is one subject, one verb and a long stack of reasons between them. Rhetorically, the delay is the point. A reader must pass through all six purposes before arriving at the act, so that the act arrives already justified. It is the same move as the Declaration’s list of charges, compressed into a sentence.

The Preamble has no legal force by itself; courts do not strike down laws for failing to insure domestic tranquility. Its power is as a statement of purpose, and purposes are what you test an argument against. When later Americans argued about what the Constitution meant, they often reached back to this sentence. Does a law promote the general welfare? Does it secure liberty to posterity? The Preamble is the warrant of the whole document: the principle that makes the rest count as a reasonable answer to the problem of government.

In the fall of 1787 the new Constitution went to the states for ratification, and in New York the opposition was fierce. Alexander Hamilton, James Madison and John Jay answered with eighty-five newspaper essays under the shared name Publius. Together they are the Federalist Papers, and they are among the most careful arguments ever written in American English, because their readers were undecided and hostile and had to be reasoned with, not commanded.

Madison’s No. 10 shows the method. It starts with a definition. A faction is a group of citizens, whether a majority or a minority, united by a shared passion or interest that runs against the rights of others or the good of the whole. Having defined the disease, Madison asks about cures. There are two ways to remove the causes of faction: destroy liberty, or give everyone the same opinions. The first is worse than the disease. The second is impossible, because people reason differently and own different amounts of property. So the causes cannot be removed. Only the effects can be controlled.

That leaves the question of how. A minority faction, Madison says, is checked by voting. The danger is a majority faction, which can win the vote and still oppress. A small direct democracy has no defense: the majority meets, feels one passion, and acts. A republic, where representatives decide, filters the passion through people who must deliberate. And a large republic does better than a small one, because it holds so many interests that no single one can easily become a majority. The conclusion, that the extended republic of the Constitution is the best protection against faction, arrives as the only door left open.

Look at the structure. Define the problem. Consider every cure. Eliminate the impossible ones. Show that what remains is what the Constitution provides. This is a form of deduction, sometimes called argument by elimination, and it is persuasive only if the list of options is complete. Anti-Federalists attacked exactly there, arguing that a large republic would grow distant from its people, an option Madison had treated as a virtue. Whether the argument is sound depends on whether his premises about human nature and distance are true. Reading No. 10 means testing them.

No. 51, generally credited to Madison, contains the most quoted sentence in the series: If men were angels, no government would be necessary. The line is a premise about human nature, and it is doing heavy work. If people cannot be trusted to be good, government is needed. But government is made of people, who also cannot be trusted. So the design must, in Madison’s words, oblige the government to control itself. The argument moves from an observation about people to a requirement for structure.

The structure is separation of powers with checks and balances. Each branch, legislative, executive and judicial, is given its own will and the means to resist the others. Madison states the mechanism in a phrase you should memorize as an example of compression: ambition must be made to counteract ambition. Rather than hoping officials will restrain themselves, the design makes their self-interest do the restraining. A president who wants power will resist a Congress that wants it, and the friction protects everyone else.

This is an argument from a pessimistic premise to a hopeful conclusion, and its logic is worth tracing. Premise one: people, including officials, pursue their own interests. Premise two: a government must be strong enough to govern but prevented from oppressing. Conclusion: build a government whose parts have opposed interests, so that each part’s pursuit of power blocks the others. The warrant, unstated, is that opposed forces produce balance rather than paralysis. Anti-Federalists doubted that warrant, and so have critics ever since, whenever branches deadlock.

Notice how No. 10 and No. 51 fit together. No. 10 controls the effects of faction among citizens by enlarging the republic. No. 51 controls the effects of ambition among officials by dividing power. Both rest on the same premise: you cannot make people better, so you must arrange them so that their flaws cancel. Whether you accept the Constitution’s design depends heavily on whether you accept that premise. That is the sentence a reader of the Federalist should be arguing with.

In July 1848, about three hundred people gathered in a chapel in Seneca Falls, New York, for the first convention in the United States devoted to women’s rights. Elizabeth Cady Stanton drafted the meeting’s central document, and she made a decision that was itself an argument. Rather than invent a new form, she took the Declaration of Independence and rewrote it. The result, the Declaration of Sentiments, follows the original almost line for line.

The first premise is where the change is made. The 1776 text says all men are created equal. The 1848 text says all men and women are created equal. Everything else in the opening follows: rights that cannot be taken away, governments that exist to secure them, the right to demand a new arrangement when the old one fails. Then comes the list of charges, and where Jefferson wrote He has about the king, Stanton wrote He has about man in general: he has never permitted her to vote, he has made her, if married, civilly dead in the eye of the law, he has denied her a thorough education.

The form is the argument. By keeping the syllogism and changing two words in the premise, Stanton forced her readers into a corner. Anyone who accepted the Declaration of 1776, which in 1848 was nearly everyone, and who rejected the Declaration of 1848 had to explain what made the two cases different. Either the first premise did not really mean all people, in which case the founders’ argument had a hidden exception, or it did, in which case the conclusion about women followed. This is argument by parallel: matching a new case to an accepted one so exactly that denying it means denying the original.

The tactic had costs. Some readers found the parallel too neat, and the demand for the vote was controversial even inside the convention; Frederick Douglass, who attended, spoke in its favor when others hesitated. But the document’s logic outlived the room. For seventy-two years, until the Nineteenth Amendment in 1920, arguments for women’s suffrage returned to Stanton’s move: hold the country to its own first premise.

Frederick Douglass escaped from slavery in Maryland in 1838 and became the most famous Black orator and writer in the country. On July 5, 1852, he was invited to speak in Rochester, New York, at a celebration of Independence Day. He began by praising the founders honestly. They were brave, he said, and their Declaration was a great document, and its principles were right. Then he turned. The holiday, he told his white audience, was theirs, not his. To the enslaved American, the Fourth of July revealed more than any other day the injustice he lived under.

The structure of the speech is the structure of this unit. Douglass accepts the Declaration’s first premise: all people have rights no one may take. He then supplies grounds: the laws of slaveholding states treat the enslaved as men, punishing them for crimes, which proves the slaveholders themselves know they are men. If they are men, the premise applies to them. The conclusion follows by strict deduction: the nation’s practice contradicts its founding argument. Douglass refuses to argue the premise, saying the point is too plain to need proof, and spends his force on the contradiction.

He also refuses a false dilemma. Some abolitionists, following William Lloyd Garrison, held that the Constitution was a proslavery document and should be rejected. Douglass had once agreed; by 1852 he argued the opposite, reading the Constitution’s text and finding no word of slavery in it. The Constitution, he said, was a glorious liberty document when read plainly. This mattered strategically. If the founding documents were on the side of freedom, then abolition was not a rebellion against the country’s principles but their fulfillment.

The rhetoric is fierce, full of irony and direct address, and students sometimes take the fierceness for the whole. It is not. Under the heat is a syllogism the audience already accepted, applied to a case they had excused. That is what makes the speech an argument rather than a lament. When you write in dialogue with a founding text, Douglass’s method is the model: grant the premise, prove the case, and let the reader feel the conclusion close.

Abraham Lincoln practiced law in Springfield for more than twenty years, and his speeches show it. He argued the way a good lawyer does: pick the premise the other side has already granted, and build from it. His chosen premise was the Declaration’s. In June 1858, accepting the Republican nomination for the Senate in Springfield, he warned that a house divided against itself cannot stand, and that the nation could not endure permanently half slave and half free. That is a deductive claim: a divided house must eventually become all one thing or all the other.

That fall he debated Stephen Douglas seven times across Illinois, from Ottawa to Alton. Douglas argued that the founders had never meant to include Black Americans in the phrase all men. Lincoln answered that the founders had stated a principle they had not yet lived up to, a standard set for the future. The disagreement was about the first premise’s meaning, exactly where the chapter opening said every argument about the Declaration would be fought. Lincoln lost the Senate race and won the argument’s future.

At Gettysburg in November 1863, in about two hundred seventy words, he compressed the whole case. The nation was conceived in liberty and dedicated to the proposition that all men are created equal. Notice the word proposition. Jefferson’s self-evident truth had become a claim to be tested, and the war was the test. The speech ends by turning the audience’s grief into a premise for action: the dead gave their lives so that the proposition might hold, and the living must finish the work. The logic is a chain, and every link is a single sentence.

His Second Inaugural, in March 1865, weeks before his death, is the hardest of the three, because it argues against his own side’s certainty. Both sides prayed to the same God, he observed, and neither’s prayers had been fully answered; the war might be a judgment on the whole nation for slavery, not on the South alone. Then the famous close: with malice toward none, with charity for all. The passage moves from a premise about shared guilt to a conclusion about how to make peace, and the antithesis of malice and charity fixes the conclusion in memory. Lincoln’s speeches endure because their rhetoric is the shape of their reasoning, not a decoration on top of it.

By now you have seen three ways to argue with a founding text. The first is to grant its premise and demand the conclusion, as Stanton and Douglass did. The second is to contest a definition, as Douglas did with all men and Lincoln did back. The third is to reject the text outright, as Garrison did with the Constitution, calling it a pact with slavery. Each move has a different relationship to the original, and each requires different evidence.

The twentieth century kept all three alive. In 1963, on the steps of the Lincoln Memorial, Martin Luther King Jr. described the Declaration and Constitution as a promissory note that every American was to inherit, and said the nation had defaulted on it for Black citizens. That is Douglass’s move updated: accept the premise, prove the contradiction, demand payment. King’s Letter from Birmingham Jail, that same year, argued with a different set of texts, the statements of moderate clergy, by granting their premise that law deserves respect and then distinguishing just laws from unjust ones. A student writing today should study these as arguments, not only as history.

Courts argue with the founding texts constantly, and their moves are the same three. A judge who reads the Preamble’s general welfare broadly is contesting a definition. A dissent that says the majority has abandoned the framers’ design is granting a premise and charging a contradiction. Even ordinary editorials do it: a column that opens with We the People and asks whether a law serves the people is running the Preamble as a warrant. Once you know the three moves, you will see them everywhere, and you will notice when a writer performs one badly, quoting a founder as decoration instead of as a premise that does work.

That last failure is the one to avoid in your own writing. Quoting the Declaration proves nothing by itself; the founders are not an authority whose words settle a question, and treating them so is an appeal to authority outside its field. What the founding texts offer is premises that most of your readers already accept. The work of the essay is to show that your conclusion follows from those premises, or that your opponent’s does not. Douglass did not win by quoting Jefferson. He won by showing that Jefferson’s own sentence, applied honestly, condemned the thing Jefferson’s country did.

Here is the assignment this chapter has been preparing you for: take a foundational text and write an argument that engages it. Not a report about it, and not an essay that quotes it once for flavor. An argument that uses one of its premises, or contests one, to reach a conclusion about a question that matters now. The method has five steps, and they are the Toulmin parts wearing new clothes.

Step one: state the text’s premise in your own words, fairly, so that a defender of the text would nod. If you are using the Preamble, say what promoting the general welfare means in ordinary terms. If you are using Federalist 10, say what Madison means by faction, using his definition, not a cartoon of it. A straw man of a founding text is the most common failure in this kind of essay. Step two: state your own claim, with a qualifier that marks its edges. Step three: supply grounds from the present, evidence about the actual question, using at least two of the four kinds and asking each its limiting question before you use it.

Step four is the hinge. Write the warrant that connects the founder’s premise and your grounds to your claim, and then decide whether to state it. If it is something your reader will grant, keep it quiet. If it is where the disagreement lives, argue it directly, with backing. Step five: find the strongest reply someone could make using the same text, and answer it. Douglas answered Lincoln with the same Declaration; Garrison answered Douglass with the same Constitution. A foundational text can nearly always be read against you, and a serious essay shows it knows that.

Look at a short model. Claim: Chicago should restore its ward-level participatory budgeting programs, at least in wards that have used them before. Premise from the text: the Preamble makes the people the author of government and its purposes the measure of it. Grounds: reports from wards that ran the programs, a survey of participants, a public-finance expert on cost. Warrant: a process that lets the people direct spending toward their own stated purposes serves the Preamble’s design. Rebuttal: Madison would warn that direct decisions invite faction, and the answer is that the budgets are small, the choices are local, and elected officials still hold the final vote. Write yours the same way. Grant the premise. Prove the case. Let the reader feel the conclusion close.