Original Thesis Project

Overview

You are going to take a position on something that matters, build the strongest possible case for it, confront the strongest possible case against it, and defend your conclusion in front of people who will challenge you. This is not a school essay where you summarize what others have said and hedge toward a safe middle. It is an original work of argument — a claim that is genuinely yours, genuinely contestable, and genuinely defended.

The capacity to form an original position and hold it under pressure is rare, and it is the intellectual backbone of leadership. Anyone can have opinions. Far fewer can construct an argument that survives a serious adversary, change their mind gracefully when the argument fails, and tell the difference between the two. This project builds that capacity by forcing you through the full cycle: question, research, claim, defense, revision, and a live oral defense where you must answer for what you wrote.

This is also the most directly career-shaped thing in the whole pillar, and it is worth being clear-eyed about why. The work you will do at the frontier of any field — founding a company, advancing a science, shaping a policy, building a body of writing — is, structurally, exactly this: you assert something the world does not yet agree on, you assemble the case, and you defend it against people whose job is to find the flaw. A pitch to an investor is a thesis defense. A scientific paper surviving peer review is a thesis defense. A founder explaining why the incumbents are wrong is a thesis defense. By doing one now, deliberately and under real challenge, you are not completing a school assignment. You are rehearsing the central move of every consequential adult career, in a low-stakes setting designed to let you fail safely and learn from it. Treat the difficulty as the point. A thesis that was easy to defend was not worth defending.

The Deliverable

Two artifacts, judged by professional standards:

1. A written thesis of 4,000-7,000 words that states an original, contestable position on a complex issue and defends it with cited evidence. It must include a section that fairly presents and then answers the strongest opposing argument. It must be properly sourced with footnotes or endnotes and a bibliography.

2. A live oral defense of 45-60 minutes before a panel of at least two informed adults who have read the thesis in advance and will challenge it. You present for ten minutes, then answer questions. "Done" means you have defended the thesis on its merits — including conceding the points that should be conceded — and the panel agrees the position was held honestly.

The topic is yours, but it must clear three bars: it must be contestable (a reasonable, informed person could disagree), it must be complex (it does not collapse into a simple fact or a matter of taste), and it must be researchable with sources you can actually reach.

Materials & Tools

Project Phases

Phase 1: Plan — Find the Question and the Claim (Weeks 1-3)

A thesis project lives or dies on the question. Spend real time here.

Start with a domain you already care about — technology policy, education, a historical controversy, an ethical question raised by your own venture, a debate in a field you are building expertise in. Read widely and shallowly at first: get the lay of the land, find where the genuine disagreements are. You are hunting for a live controversy, not a settled question.

Then narrow ruthlessly. "Is social media bad for teenagers?" is too broad to defend in 7,000 words — every clause hides a separate debate. "The dominant policy response to teen social-media harm — age verification — will fail because it misdiagnoses the mechanism of harm" is a thesis. It is specific, it is contestable, and it is researchable.

Apply the contestability test before you commit: write down the single best argument a smart person would make against your claim. If you cannot find one, your claim is either obvious (not worth arguing) or vague (not yet a claim). If the opposing argument is strong and makes you slightly nervous, you have found your thesis.

End Phase 1 with: a one-sentence thesis statement, a paragraph explaining why it is contestable, and the single strongest objection written out in full. Bring these to your mentor and let them attack the claim. If it falls over in five minutes, it was not ready.

A note on scope, because this is where projects quietly die. The most common Phase 1 failure is not picking a bad topic — it is picking a topic so large that no honest defense fits in 7,000 words, then spending ten weeks producing a tour of a subject instead of an argument about it. The tell is that you cannot state, in one sentence, what specific claim a reader would have to accept or reject. "Artificial intelligence and the future of work" is a subject, not a thesis — there is nothing in it to be right or wrong about. "Within a decade, the labor-market disruption from AI will fall hardest not on low-wage work but on the mid-skill professional jobs currently considered safe, because those jobs are the most linguistically routinizable" is a thesis: narrow, falsifiable, and uncomfortable enough that a smart reader could push back hard. When in doubt, narrow further. A small claim defended airtight beats a large claim defended vaguely, every time, in this project and in life.

Phase 2: Build — Research and Draft (Weeks 4-10)

Milestone 1 — The evidence base (Weeks 4-6). Gather sources, prioritizing primary sources over commentary. A primary source is the original — the study itself, the law itself, the speech itself, the data itself — not someone's summary of it. Read your opponents' best sources, not just your own; you cannot answer an argument you have only heard secondhand. As you read, log every source with enough detail to cite it, and write a one-line note on what it actually establishes. Distinguish constantly between what a source demonstrates and what it merely asserts. A confident sentence in a credible outlet is still just an assertion until it is backed by evidence.

Milestone 2 — The argument architecture (Week 7). Before drafting prose, outline the logical structure. What does your thesis depend on? List the claims that must be true for your thesis to hold, and the evidence supporting each. If any load-bearing claim has no real support, you have found a hole — either fill it or revise the thesis. An argument is only as strong as its weakest necessary link, and finding that link now is far cheaper than having the panel find it later.

Milestone 3 — The first full draft (Weeks 8-10). Write the whole thing, including the section where you present the opposing argument at full strength — the steelman. This is the part most people get wrong: they build a weak version of the other side and knock it down, which fools no one and persuades no one. Construct the opposing argument so well that a thoughtful opponent would say "yes, that is exactly what I believe." Then answer it. If you cannot answer the strongest version, your thesis is wrong, and discovering that is a success, not a failure — it means you were doing honest work.

Phase 3: Test & Refine

Hand the full draft to your mentor and at least one other informed reader, ideally someone inclined to disagree with you. Ask them for the harshest honest reading. Then do the hardest thing in the project: separate the criticism that exposes a real flaw from the criticism that is merely a different opinion, and respond to each appropriately. Real flaws get fixed. Differences of opinion get acknowledged and answered.

Revise for argument first, prose second. A beautiful sentence defending a broken claim is worthless. Check every citation — does the source actually say what you claimed it says? Misrepresenting a source, even accidentally, is the one error that destroys credibility instantly, and your panel will check.

Phase 4: Present — The Oral Defense

The oral defense is where the project becomes real. You will discover that defending a position out loud, in real time, against a person who has read your work and found its weak points, is a completely different skill from writing.

Prepare a ten-minute presentation: the question, your thesis, your core argument, and the strongest objection and your answer to it. Do not read it — know it well enough to speak it. Then open the floor. The panel's job is to challenge: to probe your evidence, to press the objection harder than your draft did, to ask what would change your mind.

Your job is not to win. Your job is to defend honestly. That means conceding the points that should be conceded — "you are right, that source is weaker than I treated it; here is how much of my argument survives without it" — and holding firm where you have the evidence. A defense where you concede nothing is a defense you are faking. A defense where you concede everything means you never had a thesis. The skill is knowing the difference in real time.

There is a specific mental move that separates people who defend well from people who crumble, and it is worth naming so you can practice it. When a challenge lands, your body will want to do one of two things: surrender ("you're right, I was wrong about everything") or dig in ("no, you just don't understand"). Both are panic responses, and both destroy your credibility. The trained move is a third thing — locate the challenge. Before responding to any hard question, silently ask yourself: is this attacking a load-bearing claim, or a decorative one? If a panelist demolishes a supporting example but your core argument has three other legs, you can concede the example warmly and instantly — "good catch, that one's weak" — and lose nothing, while looking like exactly the honest thinker you want to be. If they have found a genuine load-bearing flaw, the honest move is to say so and show how much of the thesis survives. Either way you stay calm, because you mapped your own argument's load-bearing structure back in Phase 2 and you know which beams matter. Defenders who have not done that mapping cannot tell a scratch from a mortal wound, so they treat every challenge as mortal, and they panic. Do the structural work early and the defense becomes almost relaxing: you are simply walking a prepared map of your own reasoning, out loud, with company.

One more piece of preparation that pays off enormously: before the defense, write down the three questions you most hope nobody asks. They are the soft spots you have been avoiding. Prepare honest answers to exactly those three. Almost without exception, a serious panel finds at least one of them — and the difference between a student who has rehearsed an honest answer to their own worst question and one who is hearing it for the first time in the room is the difference between a defense and a collapse.

Success Criteria

• The thesis is original and contestable — a reasonable, informed person could disagree, and the student can state the best version of that disagreement
• The written work is 4,000-7,000 words, properly sourced with footnotes and a bibliography, and free of misrepresented sources
• The thesis includes a genuine steelman of the opposing argument, followed by a substantive answer to it
• The student defended the thesis live for 45-60 minutes, conceding what should be conceded and holding what could be held
• The student can state, from memory, what evidence would have changed their conclusion

Common Pitfalls

• Choosing a topic that is not actually contestable. If everyone informed already agrees, there is nothing to defend. The discomfort of a strong opposing argument is the signal you picked well.
• The straw-man opponent. Building a weak version of the other side is the most common failure and the easiest for a panel to catch. Steelman or fail.
• Confusing volume of sources with strength of argument. Twenty weak sources do not equal one decisive one. A single well-read primary source, understood deeply, beats a pile of citations you skimmed.
• Treating concession as defeat. In a real defense, conceding a fair point makes you more credible, not less. Stubbornness on a losing point destroys your standing on the winning ones.
• Editing prose while the argument is still broken. Fix the logic and evidence first. Polishing sentences around a flawed claim is wasted effort.

Extensions

• Publish it. Submit a condensed version to a student journal, a magazine, an op-ed page, or your own public site. Defending a position in front of strangers who can respond is the next level beyond a friendly panel.
• Defend the opposite. As a separate exercise, write the strongest possible case for the position you argued against. The ability to argue both sides well is the mark of a genuinely independent mind — and it will make you far more honest about your own conclusions.
• Turn it into a debate. Find someone who genuinely holds the opposing view and stage a moderated, timed debate before an audience. Real-time argument against a real opponent is the hardest and most valuable version of this skill.
• Make it the foundation of a venture or a body of work. Many strong companies, nonprofits, and bodies of writing began as one person's defensible thesis about how the world actually works. If your thesis is right and others have not acted on it, that gap is an opportunity.
• Repeat the cycle annually. The single biggest return on this project is treating it as a recurring practice rather than a one-time event. Pick a new contestable question each year and run the full arc again. The second thesis is dramatically better than the first, because the bottleneck was never the topic — it was the unfamiliar discipline of building and defending an argument under challenge, and that discipline only deepens with reps. A person who has defended five original theses by the time they leave home is a different kind of thinker than one who has defended none.