> ## Content Index > Fetch the complete content index at: https://otherwise.ghost.io/llms.txt > Use this file to discover other available public pages before exploring further. # The Oxford Binary Trap - URL: https://otherwise.ghost.io/the-oxford-binary-trap/ - Published: 2026-09-05T05:50:17.000Z - Updated: 2026-09-05T05:50:17.000Z - Description: Why two sides are enough for argument—but not necessarily for knowledge - Author: Khaled Salih - Tags: Newsletter, #Migrated-1792244264359, #Import 2026-10-17 14:38 5 September 2026 --- # Note **Two Doors Are Not Enough** The Oxford Union asks people to choose. **Proposition or Opposition.** **Yes or No.** **One door or the other.** This series begins by refusing to accept that two-door architecture as the limit of possible inquiry. Its starting conjecture is simple: **binary debate can be excellent at producing opposition, but opposition is not the same as knowledge creation.** A bad question does not become better because brilliant people are assigned to opposite sides of it. And defeating one explanation does not make its designated rival true. The recent Oxford debate on whether *“Israel Never Truly Wanted Peace with Palestine”* provides an unusually demanding laboratory. But Israel is not ultimately the subject of this series. The deeper subject is what happens to human thinking when complicated reality is compressed into two choices. Drawing on Karl Popper, David Deutsch, Brett Hall, and my own work on error correction and the Backspace Key, I will ask a different question: ***What if the most important intellectual achievement is not choosing the correct door, but discovering that another door can be built?*** Not a compromise. Not centrism. Not a predetermined third answer. **The freedom to create a new conjecture when criticism shows that the existing menu is inadequate.** --- This series is an experiment in doing exactly that. I watched [the Oxford Union debate](https://www.youtube.com/watch?v=7j7OR-FEV4I&t=6096s&ref=otherwise.ghost.io) *“This House Believes Israel Never Truly Wanted Peace with Palestine”* twice. More than two hours each time. The second viewing was more painful than the first. Read the Introduction, **Article 1: “Why Two Doors Are Not Enough?”** [here](https://khaledsalih.substack.com/p/two-doors-are-not-enough). --- # Article 2 The Oxford Binary Trap *Why two sides are enough for argument—but not necessarily for knowledge* There is something profoundly civilized about the Oxford Union. Someone stands up and says something controversial. Someone else is invited to say that it is wrong. The proposition speaks. The opposition responds. Arguments are attacked in public rather than suppressed in private. A room full of people listens, judges, objects, applauds, heckles, changes its mind—or does not. For anyone influenced by Karl Popper and the idea of the Open Society, there is much to admire in this tradition. Criticism is permitted. Authority does not have the final word. No speaker is supposed to be protected from argument. And yet, while watching the recent Oxford Union debate on the motion \*“This House Believes Israel Never Truly Wanted Peace with Palestine,”\* I found myself becoming increasingly interested in something other than Israel or Palestine. I became interested in the debate itself. Not initially in who was winning. Not in which side had assembled the most devastating examples. Not even in which historical interpretation was correct. I began wondering about something more elementary: ***Why were these the choices?*** Before the first speaker had said a word, an enormous amount of intellectual work had already been done. A problem had been selected. It had been compressed into a proposition. The proposition had been formulated so that it could be affirmed or rejected. Speakers had been divided into Proposition and Opposition. And at the end, the audience would literally vote with its feet, leaving through one door or the other. Two sides. Two answers. Two doors. It is an excellent architecture for producing disagreement. I am no longer convinced that it is an adequate architecture for producing knowledge. ## Before the Debate Begins The problem is not that Oxford permits opposition. Quite the reverse. Opposition is one of the greatest inventions of civilization. The deeper problem is that the architecture decides what is to be opposed before criticism has begun. Consider the intellectual sequence. Reality is unimaginably complicated. From that reality, someone extracts a problem. The problem is converted into a sentence. That sentence becomes a motion. The motion becomes: **Yes or No.** Only then does criticism begin. **But** what if the mistake occurred earlier? **What** if the problem was badly formulated? **What** if an important concept in the motion is ambiguous? **What** if the causal mechanism does not correspond to either side? **What** if the evidence presented during the debate reveals that the original question should be abandoned? **What** if A is wrong, but Not-A is also a bad explanation? **What** if the best idea does not exist when the debate starts? The traditional debate format has surprisingly little institutional room for these possibilities. It begins too late. ## Popper Without the Next Conjecture Karl Popper taught that knowledge grows through conjectures and criticism. That is often summarized too crudely as: Make a claim. Try to falsify it. But Popper’s conception of knowledge growth is richer than a duel between two propositions. There is a problem. Someone proposes a tentative solution. Criticism exposes errors. The discovery of those errors creates a new problem situation. New conjectures become possible. And criticism begins again. There is no final intellectual authority. More importantly, there is no rule saying that if conjecture A fails, conjecture B must therefore be true. This matters enormously. Suppose the motion is: **Israel never truly wanted peace with Palestine.** The word ***never*** makes this a very strong historical proposition. If one could establish even one genuine case in which an Israeli government, leader, movement, or substantial political constituency sincerely sought peace with Palestinians, the universal proposition would be in trouble. Several opposition speakers understood this perfectly. One explicitly argued that the opposition needed only a counterexample. Others pointed to Oslo, Camp David, the 2005 withdrawal from Gaza, Israeli peace movements, and Israeli leaders who pursued territorial compromise. But suppose they succeed. What has been established? Certainly not that Israeli policy has always promoted peace. Certainly not that settlement expansion is irrelevant. Certainly not that Palestinian grievances are invented. Certainly not that Palestinian political agency explains every failure. Certainly not that every Israeli offer was adequate. And certainly not that the conflict has therefore been explained. One proposition has been damaged. The problem remains. That is an important distinction: **Winning the debate is not the same thing as solving the problem.** A Popperian discussion should become more interesting when an original conjecture fails. A binary debate risks ending precisely there. One side wins. Everyone walks through a door. ## David Deutsch and the Missing Choice David Deutsch takes this much further. One of the deepest lessons I take from Chapter 13, “Choices,” of \*The Beginning of Infinity\* is that human beings are not confined to choosing among alternatives already available to them. Knowledge changes what can be done. People create possibilities. Nature did not originally present humanity with a menu saying: ***Walk.*** ***Run.*** ***Or fly.*** Flight became a human possibility because explanatory knowledge was created. The same principle applies, far less obviously, to politics. A political problem may appear to offer: dictatorship or chaos; assimilation or conflict; security or liberty; one state or two; submission or resistance; centralization or dissolution. But the existence of two familiar proposals does not demonstrate that reality contains only two possible solutions. They may simply be the solutions presently known. That is why a rational person should sometimes respond to: ***A or B?*** with: ***Why only A or B?*** And then with a harder question: ***What knowledge would have to be created for C to become possible?*** This is where the Oxford architecture begins to look epistemologically peculiar. Imagine that halfway through a debate a participant says: “I arrived here defending the proposition. The opposition has exposed an error in my argument. But I also think their explanation fails. The evidence presented tonight suggests a third possibility that neither side had formulated when we entered the chamber.” From the standpoint of knowledge creation, this might be the greatest achievement of the evening. The discussion has produced something new. **But** what happens institutionally? **Which** side does the speaker now belong to? **Which** door should the audience use? The best product of the debate may have no door. ## Israel Is Not One Thing The Oxford debate itself supplied evidence of the problem. Several opposition speakers challenged the very grammar of the motion. ***What does it mean to say that Israel wants something?*** Is Israel its government? Its electorate? Its prime minister? Its parliament? Its peace movement? Its settlers? Its soldiers? Its Arab citizens? Its protesters? Its institutions? Its political traditions? One speaker made the same objection regarding Palestine: does “Palestine” mean ordinary Palestinians, the Palestinian Authority, Hamas, Palestinian nationalism, or some combination of them? These are not pedantic objections. They reveal a level-of-analysis problem hidden inside the motion. An enormous collection of changing people, governments, parties, institutions and ideas has been compressed into a single intentional actor: **Israel.** The same compression repeatedly occurred in reverse. **A Palestinian** **decision** became evidence about Palestinians. **A Hamas** **atrocity** became evidence about Palestinian intentions. **A Netanyahu** **policy** became evidence about Israel. **A Zionist** **statement** became evidence about Zionism across generations. **An Arab** **rejection** became evidence about Arab political attitudes. This is where another question from my own work becomes indispensable: ***Where does the error go?*** A political decision is made. It fails. ***Where should criticism travel?*** Into the decision. Into its premises. Into the explanation behind it. Into the institution that produced it. Into the incentives that protected it. Into the leader responsible for it. Instead, political argument frequently sends the error elsewhere: Decision **→ people** **→ identity** **→ historical essence.** A government becomes a nation. A movement becomes a civilization. An atrocity becomes a people. Once this happens, explanation begins to disappear. ## Intelligence Is Not Rationality There is another problem hiding inside competitive debate. Oxford attracts intelligent people. That is not enough. One of the principles I increasingly use in my own analytical work is: Intelligence is not rationality. Intelligence can make a person extraordinarily capable of defending an answer. Rationality requires something more difficult: **the willingness to permit criticism to alter the answer.** Place an intelligent person on Proposition. Give that person ten minutes. Give them a motion to defend. ***What should one expect them to do?*** Find the strongest evidence. Construct the clearest argument. Anticipate objections. Neutralize counterexamples. Expose weaknesses in Opposition. Place an equally intelligent person on the other side and the same machinery activates in reverse. No dishonesty is required. The institutional incentives are sufficient. Intelligence becomes an ammunition factory. That may be excellent training in rhetoric. It may also be excellent training in understanding arguments one initially rejects. But it is not automatically rationality. The crucial question is: What is intelligence being optimized to accomplish? To protect an assigned answer? Or to discover where the answer is wrong? Those are different tasks. #When the Motion Becomes the Protected Object This produces a strange paradox. The Oxford debate is an institution dedicated to criticism. Yet something is comparatively protected from criticism: the architecture of the motion itself. Participants may criticize one another fiercely. They may challenge evidence. They may attack historical interpretations. They may dispute motives. They may question moral assumptions. But the evening still has to end with a decision about the proposition that existed before any of that criticism took place. During the Israel debate, one proposition speaker even declared, in effect, that the subject was not really a debate but a statement of fact. That is almost a perfect Popperian warning sign. A conjecture has ceased to present itself as conjectural. But the deeper lesson is not merely about that speaker. The architecture encourages certainty. A speaker has not been invited primarily to discover whether the proposition survives. The speaker has been invited to make it survive. The motion becomes temporarily protected from Backspace. #The Backspace Key I use Backspace as a metaphor for error correction. Most people imagine correction occurring after an answer has been given: Answer → criticism → correction. But David Deutsch’s argument about choices suggests something deeper. Sometimes the error is not the answer. Sometimes the error is the menu. Sometimes it is the question. Sometimes it is the premise that made the question seem inevitable. A genuine Backspace architecture must therefore permit several levels of correction: Backspace the answer. Backspace the alternative. Backspace the framing. Backspace the premise. And, when necessary: Backspace the question itself. A discussion that permits criticism of answers but not reconstruction of the problem possesses only part of the error-correction machinery. #Brett Hall and the Means of Correction Brett Hall has formulated a principle that I regard as indispensable: Do not destroy the means of error correction. Oxford deserves considerable credit under this principle. It preserves public criticism. It preserves opposition. It protects the possibility of hearing an argument one dislikes. These are civilization-level achievements. The right response is therefore not to abolish adversarial debate. It is to complete it. An error-correcting discussion needs at least four freedoms: the freedom to criticize; the freedom to revise; the freedom to abandon one’s original position; and the freedom to create an alternative that was not available at the beginning. Traditional binary debate strongly protects the first. It provides much weaker institutional recognition of the remaining three. That is the defect. ## Two Bad Explanations Do Not Make One Good Choice The Israel debate makes the consequences unusually visible. The Proposition developed a broad explanatory architecture around Zionism, expansion, settlements, occupation and the obstruction of Palestinian statehood. The Opposition developed another around Jewish self-determination, repeated Israeli peace initiatives, Palestinian rejectionism, terrorism and the refusal to accept Jewish sovereignty. Each side contained important arguments. Each also risked becoming a story capable of absorbing almost everything. A negotiation could become proof of Israeli peace-seeking—or proof of Israeli deception. A withdrawal could become proof of compromise—or proof of strategic repositioning. Palestinian violence could become proof of rejectionism—or evidence of what occupation produces. Settlement expansion could become proof that peace was never intended—or something separable from other Israeli peace efforts. When explanations become sufficiently flexible, almost every new fact can be recruited as confirmation. That is precisely when David Deutsch’s demand for good explanations becomes useful. A good explanation should be hard to vary while still explaining what it claims to explain. The purpose is not to find the most rhetorically adaptable story. It is to find the explanation that survives criticism without arbitrary adjustment. And sometimes the answer will be that both narratives have identified mechanisms that operate simultaneously. Israeli actions may change Palestinian choices. Palestinian actions may change Israeli choices. Violence may increase fear. Fear may strengthen political hardening. Territorial expansion may reduce belief in negotiation. The collapse of negotiation may strengthen maximalists. Maximalism on one side may supply evidence to maximalists on the other. None of this means responsibility is equal. Causal interaction is not moral equivalence. It means that reality may contain a feedback mechanism for which neither Oxford door is adequate. ## Build the Third Door This leads me to a conjecture. Binary adversarial debate is a valuable technology of criticism but an incomplete architecture for knowledge creation when its binary outcome is mistaken for resolution of the underlying problem. The solution is not less criticism. It is more Backspace. Keep Proposition. Keep Opposition. Let them attack one another as vigorously as ever. But do not stop there. After adversarial criticism, add another stage. Ask: What did Proposition get wrong? What did Opposition get wrong? What survived? Was the original problem correctly framed? Which hidden assumptions have now become visible? What alternatives were missing? Has anyone created a better conjecture? And allow participants to say the sentence that competitive politics and debating too often treat as defeat: I was wrong. That sentence should be counted as an intellectual achievement. Then ask another Deutschian question: What do we understand now that we did not understand when the debate began? If the answer is merely: “our side won,” little knowledge may have been created. But if the answer is: “we entered with A and B and leave with a better problem and a conjecture C that nobody possessed beforehand,” then criticism has done what criticism is for. It has created the conditions for knowledge growth. Oxford has spent two centuries defending the importance of disagreement. That is worth preserving. But disagreement itself is not the destination. The purpose of criticism is not to make two sides permanently better at opposing one another. It is to discover errors. To change explanations. To create possibilities. To make yesterday’s impossible option conceivable. And sometimes to discover that both doors were built in the wrong wall. Oxford gives criticism two doors. Knowledge growth may require building a third.