From First Idea to Finished Essay

When a Draft Changes the Question

Imagine drafting an essay about a library’s evening opening hours. Your first argument is straightforward: staying open later would help more people use the building. Then you examine the evidence. Attendance records show visits, but do not explain who could not attend. A reader points out that the last bus leaves before the proposed closing time. Another asks whether staffing the extra hours would reduce services earlier in the day.

This is an invented example, but the writing problem is concrete. You can keep polishing the original argument, or let those questions change it. A more useful essay might ask which combination of hours, transport, and staffing would improve access—and which evidence would be needed to answer.

The revision begins before a sentence is shortened. It begins when the writer allows the investigation to alter the question.

That is the purpose of the AuthZ writing process. We develop ideas, test their support, explain them to a reader, and return to earlier work when something does not hold. The checks should fit the purpose and complexity of the piece.

Where the Process Came From

We developed the process by bringing together approaches we valued in other writers and adapting them into a working editorial method. Five influences became central to that development: Yuval Noah Harari, Malcolm Gladwell, Steven Johnson, Nassim Nicholas Taleb, and David Graeber.

From our reading of Harari, we took an interest in connecting a present question to a wider historical context. Gladwell offered a model for using a concrete story to bring a larger question into view. Johnson encouraged us to look for useful connections between disciplines. Taleb prompted us to test an argument against uncertainty, error, and failure. Graeber encouraged us to question familiar explanations and examine the evidence they leave out.

These are the lessons we chose to adapt, rather than complete accounts of those writers’ work. They did not design or endorse the AuthZ process. The framework is our synthesis, and a borrowed approach must still earn its place in the particular essay: a story needs evidence, a comparison needs limits, and a challenge to conventional wisdom needs more than an unexpected conclusion.

We then combined that framework with a practical content-quality checklist: checking paragraph purpose, logical connections, specificity, sentence rhythm, and readability. These checks direct an editor’s attention to passages that need work.

Give Different Problems Different Attention

A passage can read smoothly while making an unsupported claim. A source can be accurately cited while doing little to support the paragraph beside it. A careful argument can still leave its reader wondering why it matters.

These problems call for different kinds of attention. Exploring a possibility gives us material to investigate. Testing an argument asks what the evidence can establish. Editing makes the reasoning easier to follow. Verification checks whether the source, wording, locator, and claim agree.

One person may do all of this work. It helps to make separate passes: first ask whether the argument holds, then whether a reader can follow it.

AI can help suggest questions or organize an outline. Any reference it generates still needs to be located and read. Before using it, we check that the source exists and supports the claim.

Nine Stages, With Room to Return

Our writing process has nine stages. Each produces something the writer can examine before deciding whether to continue or return to an earlier question.

1. Define the project

Name the audience, purpose, form, and limits. An essay for library users has different obligations from a budget proposal for a council. Write a brief that explains what readers should understand or be able to do after reading.

Sample brief: A short essay for library users explaining what the board should investigate before extending evening hours. Help readers distinguish a promising proposal from an established benefit.

2. Develop the question

Turn a broad subject into something that can be investigated. “Library access” names a topic. “Who could use the library if it closed two hours later?” asks a question. Record the assumptions inside that question, including whether closing time is the main obstacle.

Working question: Would later hours improve access for people who cannot currently attend? The starting assumption is that lack of time, rather than transport or another constraint, keeps them away.

3. Map the evidence

List the claims you expect to make and the evidence each would require. Distinguish existing attendance records from evidence about unmet demand. Note what a source establishes, what you infer, and what remains unknown. Look for material that could change your initial view.

Sample evidence note: In our fictional case, suppose the library closes at 8 p.m. and the last bus leaves its stop at 8:15 p.m. A proposal would move closing time to 10 p.m. The timetable establishes the scheduled departure, but tells us neither how many prospective visitors rely on that bus nor whether another way home is available. Those are questions for further investigation. These details are invented for the exercise.

4. Design the argument

Decide what the reader needs to understand first and how each section develops the answer. Choose examples and comparisons that clarify the subject. If a comparison imports an idea from another field, explain its limits; resemblance alone does not establish that the same mechanism is operating.

Section plan: Explain the proposal, examine the transport constraint, identify the missing evidence, then recommend a proportionate next step.

5. Write a complete draft

Give each section a job: pose the problem, explain evidence, consider an objection, or draw a qualified conclusion. Mark missing sources and unresolved questions visibly in the working draft. A gap that remains visible can be investigated; confident prose can hide it from the next reader.

Draft to examine: Write the recommendation and its supporting paragraphs, leaving a visible question beside any claim about how many people would benefit. The completed model below shows where this draft can lead.

6. Revise the substance

Check the question, scope, evidence, and sequence before polishing individual sentences. Remove passages that no longer serve the developing argument.

Revision decision: Replace the claim that later hours will improve access with a proposal to investigate demand and transport before committing to a permanent change. The timetable raises a constraint; it does not establish that extending hours would be useless.

7. Seek appropriate review

Ask a reader to explain the argument back to you and identify where it becomes difficult to follow. Seek relevant expertise when the subject requires it. Record the concern, your response, and any resulting change. A disagreement can remain, but the reason for retaining a claim should be defensible.

Review note: “Your paragraph seems to assume everyone takes the bus.” Accept the concern: limit the statement to visitors who depend on that service and leave its prevalence unresolved.

8. Prepare the article for readers

Edit for clear language, useful headings, and consistent terminology. Follow the spelling conventions required by your publication or assignment; at AuthZ, we use Canadian spelling. Check links and citations. Inspect the published layout on small and large screens. A table or note that works in the manuscript may become difficult to use on a phone.

Edited wording: Replace “temporal availability is insufficient to establish equitable accessibility” with “later opening hours may not help visitors who cannot get home.”

9. Verify and decide on release

Reconcile the final wording with the evidence. Recheck quotations and locators, resolve material flags, and record remaining limitations. Preserve the version being released so later corrections can identify what changed.

Release note: The example can support a conditional recommendation, not a prediction of increased attendance. In a real essay, verify the opening hours and timetable, cite the records actually consulted, and state the remaining uncertainty.

The review note sent this writer back to the evidence and the wording. That return is part of the process: later stages can reveal work that an earlier pass missed.

From First Claim to Finished Paragraph

Here is how the fictional library essay could change. The model uses only the invented details above; it is not a report about an actual library.

First thesis: The library should stay open until 10 p.m. because longer hours will allow more people to use it.

Revised thesis: Before permanently extending its evening hours, the library should investigate unmet demand and transport needs, then use those findings to design a limited trial.

The first thesis assumes the benefit it needs to establish. The revision makes a narrower recommendation and explains what should inform the decision.

Finished paragraph:

Extending the library’s closing time from 8 p.m. to 10 p.m. would make the building available for two more hours, but that alone would not establish improved access. The last bus currently leaves the library stop at 8:15 p.m. Visitors who depend on that service and have no alternative way home could not take full advantage of the proposed extension. The timetable does not tell us how many people face this constraint or how many others would attend later. Before making a permanent change, the board should ask prospective users about demand and travel needs, assess staffing options, and use the findings to design a trial. This would give it a basis for judging whether later hours bring people into the library who could not otherwise use it.

The paragraph moves from a proposal to a constraint, acknowledges the limits of the evidence, and ends with an action suited to what is known. Its conclusion is still a recommendation, open to revision when better evidence becomes available.

A Citation Must Support Its Sentence

In a real version of the library essay, cite the timetable beside the departure time and the opening-hours record beside the current closing time. Keep the proposed trial identifiable as your recommendation. A source for the bus schedule does not also establish demand for the library.

For a book, that can mean a particular edition and page. For an unpaginated document, it may mean a section heading. A page number should never be supplied simply because the citation looks more authoritative with one.

When an article draws on external work, its citations should help readers identify that work and find the relevant passage. Keep source locations, access details, and verification notes with the research record. This essay describes our own editorial practice; it does not claim that independent research has established the effectiveness of the process.

Open each source at the recorded location and compare it with the final sentence, including any wording changed during revision.

Keep a Record That Can Be Used

Keep the brief, evidence note, and revision decision together. They explain why the library essay moved from promising a benefit to recommending an investigation and trial. An editor returning to the piece can see both the evidence used and the reasoning behind the change.

The amount of documentation should fit the work. A short explanation may need a brief source trail and revision note. An investigation built on disputed evidence needs more. Before release, decide whether the piece is ready, ready with stated limitations, needs revision, or should not be published. Record the reasons so another editor can understand the decision.

Put the Process to Work

Choose one paragraph from a draft. Write down its main claim, the evidence supporting it, and a plausible objection. Then ask what the reader would need to see to judge the connection for themselves.

If the connection is weak, return to the evidence or narrow the claim. If it is sound but difficult to follow, revise the explanation. If the objection changes the answer, let the answer change.

That is the work this process is meant to support: an author who can explain how the argument developed, what supports it, and why the final version deserves the reader’s attention.