How to Plan an Essay: From a Question to an Argument

The first task is to work out what is actually being asked, because marks are awarded for answering the question set rather than the adjacent one. Underline the instruction word and take it seriously: discuss, assess, evaluate, compare and account for demand different things, and to what extent is an invitation to argue a degree rather than to pick a side. Identify the limits, since a question about the causes of a war in a particular decade is not a question about the war. Note any assumption smuggled into the wording, because a question phrased as why did a policy fail can often be answered best by disputing that it did. A useful test is to write the question out in your own words and check that the rewrite still means the same thing; if it does not, the misreading has been caught before three days were spent on it.

An essay needs a single contestable claim that the whole piece supports, and the word contestable is doing the work: if no informed person could disagree, it is a summary rather than an argument. This is the thesis, and it should be expressible in one sentence that could be written on the back of a hand. Getting to it usually means reading first and deciding afterwards, so the sequence is to gather material, notice what you actually think, and then state it, rather than picking a position and hunting for support. A good test is whether the opposite claim could be argued by a competent person. Another is whether the thesis answers the question directly: if the question asks how important something was and the thesis says it was important, the essay has a topic and no argument, whereas saying it mattered less than a rival factor and only in a particular phase gives every paragraph a job.

The plan is a list of paragraphs, each with one job, in an order that a reader could not rearrange without loss. Write it as sentences rather than headings, because a heading hides whether you have anything to say:

• •One sentence per paragraph stating the point that paragraph will prove, not the topic it will cover
• •Under each, the evidence that will support it, named specifically, so gaps are visible now rather than at midnight
• •An order with a reason, usually from the strongest supporting line to the most qualified, or chronological only if the argument is genuinely about change over time
• •At least one paragraph engaging the strongest opposing case honestly, since an essay that never meets a counterargument reads as though it has not looked
• •A note of what each paragraph adds to the thesis, which catches the paragraph that is interesting and irrelevant

Introductions are usually too long and too general, opening with a sentence about the dawn of human history that could precede any essay on any subject. An introduction needs three things: what the question is really about, what you will argue, and the route you will take, in that order, and it can be written last. Conclusions should not summarise, since the reader has just read it; they should say what follows from the argument, what it implies, and where its limits are. Topic sentences are the most valuable single habit, because an essay whose first sentences, read alone in sequence, lay out a coherent argument is nearly always a good essay, and this is a check you can run on your own plan in two minutes. Evidence needs to be worked rather than displayed, so every quotation or statistic should be followed by a sentence saying what it shows, and the paragraph that ends on a quotation has stopped one sentence early.

Planning scales with the format. For a timed exam essay, five minutes on a rough thesis and four or five paragraph points is enough, and it is the single highest-return five minutes available, since it prevents the answer that drifts and the one that runs out of time halfway through the best material. For a coursework essay of a few thousand words, the reading, note-taking and planning should take most of the elapsed time and the drafting comparatively little. Work out the word budget in advance by dividing the count between the paragraphs, which exposes immediately whether the plan has nine points for a piece with room for four. And leave time between the draft and the edit, ideally a night, because the errors that matter, an unsupported claim, a paragraph in the wrong place, an argument that quietly changes halfway, are invisible while the text is still warm.

Take the question: to what extent was the assassination of Archduke Franz Ferdinand the cause of the First World War? The instruction, to what extent, asks for a judgement of degree, and the wording invites you to test the idea that one event caused the war. A summary would narrate June to August 1914. An argument might say: the assassination was the trigger rather than the cause, because it became a general war only through alliance commitments, military timetables and great-power rivalries already in place. Now every paragraph has a job.

The skeleton follows. Paragraph one: the assassination gave Austria-Hungary a reason to crush Serbia, which explains why the crisis began where it did. Paragraph two: Germany's backing of Austria turned a Balkan quarrel into a continental risk. Paragraph three: Russian mobilisation and German war plans, which depended on striking France first through Belgium, made the crisis hard to stop once it started. Paragraph four meets the objection that without the shots in Sarajevo there might have been no war at all. The conclusion then weighs trigger against cause rather than repeating the four points.

A practical way to get from a pile of reading to a skeleton is to break your notes apart. Put each fact, quotation or idea on its own card, sticky note or line in a document, with its source, and spread them out. Then sort them into piles according to the claim each one could support, not by which book it came from or which topic it belongs to. Each pile that forms is a candidate paragraph, and it earns a sentence stating its point. Cards that fit no pile, however interesting, are usually irrelevant to the question.

The piles then tell you things about your plan that are hard to see in a notebook. A pile with a single card is a weak point that needs more reading or should be cut. A very large pile is probably two paragraphs pretending to be one. Cards that cut against your thesis belong together, and they become the paragraph that meets the opposing case. Most usefully, the act of shuffling breaks the order in which you happened to read, which is the order weak essays tend to follow, and forces the material into the order the argument needs.

Questions that ask you to compare two things invite a particular planning mistake. The easy structure is the block plan: everything about the first subject, then everything about the second, with a closing paragraph noticing similarities. It feels natural because it follows your notes, but it produces two separate summaries joined by a hinge, and the comparison, which is what the question asked for, gets squeezed into the last few lines. The stronger structure is point by point: choose the criteria on which the two can be compared, and give each paragraph one criterion, discussing both subjects within it so that every paragraph makes a comparative claim.

Suppose the question is: compare the causes of the French Revolution of 1789 and the Russian Revolution of February 1917. A point-by-point skeleton might run like this. Paragraph one: in both cases the state was in financial crisis because of costly wars, though Russia's war was still being fought. Paragraph two: food shortages and high bread prices turned discontent into crowds on the streets of Paris and Petrograd. Paragraph three: the loyalty of the army and the monarch's handling of the crisis differed in ways that shaped how quickly each regime fell. The thesis then says which shared cause mattered most, and where the two revolutions part company.

The difference between a summary and a good essay is made in planning: read the question closely enough to catch its instruction word and its hidden assumptions, then find one contestable claim that a competent person could dispute, and state it in a sentence. Build a skeleton of paragraphs, each expressed as the point it proves rather than the topic it covers, with the evidence named and one paragraph meeting the strongest objection. Write introductions last, make conclusions say what follows, and check that the topic sentences alone make the argument.