Claude: source papers and notes illustrating long-reading critique & essay feedback
An editorial illustration of the workflow.

Download the templates

Free files, ready to edit. No account required.

  • Blank reverse-outline worksheetCSV · 187 bytes
    Download
  • Filled reverse-outline exampleCSV · 2 KB
    Download
  • Blank text form for editing or printingTXT · 2 KB
    Download

Open a blank file for your own work. Keep examples separate from your evidence and follow the instructions below.

On this page

Make the outline describe the draft you have

A reverse outline starts after a draft exists. Number its paragraphs, describe each main point and name its job in the argument. Topic and function are different: “reservations” names a subject; “defines the proposed reservation rule” tells you what the paragraph contributes.

Write the current overall claim separately. Compare it with the paragraph map: which paragraphs support it, which repeat a job and which change the subject? Cornell, Brandeis and George Mason describe this as a way to inspect the actual organization and its relationship to the argument. A plan you intended to follow cannot tell you what the draft actually contains.

Try the six paragraphs below before reading the suggested table. Use the blank CSV or draw five columns on paper: paragraph, point, function, link to claim and revision decision. Write a short label, not another summary paragraph. The filled CSV records one reasoned diagnosis, not an official marking scheme.

Supporting reading: [1] [2] [3]

The constructed draft: a community workshop booking policy

Entirely fictional practice draft written for this exercise. The workshop, proposal and situation are invented; no real policy, study, participant data, student submission or AI test output is being reported. P1–P6 identify the original paragraphs and remain stable when we discuss a different order. The writer’s task is to make a qualified recommendation and address an objection, not to demonstrate that a policy has already succeeded.

P1 — The community workshop should try reserving some places in advance while keeping places for walk-in visitors. A trial would let people plan a visit without making advance booking the only way to take part. The question is how to offer some certainty without closing the door to someone whose plans change on the day.

P2 — A reserved place would give a visitor a clearer reason to prepare for the session rather than arrive unsure whether there is space. That is a reason to consider reservations, although it does not show that attendance would increase. The workshop should also redesign its posters with brighter colors and a larger title. A poster that stands out would make the whole booking policy easier to accept.

P3 — The proposal is a temporary arrangement, not an immediate switch to fully booked sessions. Some places would remain available on arrival, and a person could request a reservation without using an online form. Before starting, the organizers would need a rule for releasing a reserved place when its holder does not arrive.

P4 — An objection is that any reservation scheme could still leave a spontaneous visitor outside while a place is held for somebody else. Keeping walk-in places does not answer every version of that problem. The organizers would need to decide when an absent visitor loses a reservation and explain that rule before people make the trip.

P5 — In this imagined workshop, visitors share a room and a limited set of tools during a scheduled session. The current arrangement admits people on arrival until the session is full. This example supplies no attendance records, survey or cost estimate. Its purpose is to compare two ways of allocating places, not to report that either one has already worked.

P6 — Advance booking is therefore the fairest arrangement, and every place should be reserved before the session begins. The workshop should introduce the new rule permanently. Visitors would then know exactly what to expect, and the organizers would no longer need to reconsider how places are allocated.

A reverse outline of P1–P6

Scroll the table sideways to view all columns.

Original paragraph IDs, topics, functions and revision problems
ParagraphTopicFunction in this draftProblem to inspect
P1Partial-reservation trialIntroduces a qualified recommendation and evaluation questionUseful thesis; needs brief context before or within it.
P2Planning certainty plus poster redesignOffers a rationale, then shifts to promotionTwo jobs; final claim about acceptance lacks support.
P3Pilot conditions and access routesDefines the proposed policy and an unresolved ruleReaders need these boundaries before assessing benefits.
P4Absent holders and spontaneous visitorsRaises an objection and a condition the policy must addressUseful challenge; overlaps P3 on releasing places.
P5Current allocation and evidence boundarySupplies context and identifies the example’s limitsArrives after readers have evaluated the proposal.
P6Permanent reservations for every placeAttempts a conclusion but substitutes a stronger policyConflicts with P1 and exceeds the reasoning supplied.

Compare your labels with this map. Your wording may differ while describing the same work. If you think P3 or P4 serves more than one function, identify the sentences that justify the distinction. The table is an interpretation of the visible draft, not a claim that every paragraph must fit exactly one label.

Two problems cannot be solved by reordering alone. P2 needs a content decision about its poster tangent. P6 changes both the scope and certainty of the recommendation. Moving either paragraph without revising those choices would preserve the problem.

One reasoned sequence: P5 → P1 → P3 → P2 → P4 → P6

Begin with P5’s brief account of the current system, then P1’s qualified proposal. This gives readers the situation before asking them to assess a change. Define the mixed-access trial in P3 before explaining its possible planning advantage in P2. The benefit is easier to judge when readers know which policy is under discussion.

Keep the planning rationale from P2, but remove the poster sentences. Poster design could belong in a separate communication plan; the draft has not shown how it establishes fair allocation. A focused replacement for P2 could read: “A reserved place could help a visitor decide whether to prepare for a session. That is a planning advantage to investigate, not evidence of increased attendance. Its value must be weighed against access for visitors who cannot commit in advance.” This is constructed revision prose, not a product output.

Use P4 to test the proposal rather than repeat it. P3 can state that a release rule is needed; P4 can examine why that rule creates a tradeoff between a reservation holder and a visitor already waiting. The draft contains no basis for selecting a particular release time, so do not invent a threshold and present it as established.

Rewrite P6 rather than simply move it. One defensible ending is: “A limited trial of partial reservations is worth considering if walk-in access and a clear release rule are retained. The proposal should remain provisional until the organizers can assess who obtains a place and who is turned away.” This preserves P1’s boundaries. It does not demonstrate that the trial is feasible, affordable or fairer. Those questions would require further work.

Another defensible choice: keep P1 first and integrate P5

A reader who already knows the workshop may prefer the recommendation first. Keep P1 as the opening, add only P5’s essential context there, and retain a brief statement that this exercise supplies no outcome evidence. Then use P3 → revised P2 → P4 → revised P6. P5’s original ID remains in the change record even though its sentences no longer form a separate paragraph.

This alternative makes the proposal visible earlier and avoids a background paragraph that repeats familiar information. The first sequence is clearer for a newcomer; the integrated opening may suit an informed audience. Choose using the assignment and reader, then check transitions after moving text. Neither arrangement is a universal correct structure or a grade prediction.

Transfer the decisions to your own draft

Use the blank download for any number of paragraphs; add rows rather than forcing a longer essay into six. Name what each paragraph currently does, then add a decision: retain, move, split, combine, cut, rewrite or investigate. A label such as “needs work” is less useful than “move the definition before the comparison that depends on it.”

After editing, make a fresh outline. Check whether the revised paragraph actually performs its intended job and whether the conclusion still matches the supported claim. Reordering may expose a missing reason; it cannot supply evidence or turn an unsupported recommendation into a finding.

If AI assistance is permitted, use it as a second reading rather than a substitute for your map. A bounded optional prompt is: “Label the main point and function of each numbered paragraph. Quote the words that support your label. Mark uncertainty; do not rewrite the draft or invent evidence.” This prompt is an untested suggestion. Compare the labels with the text, reject inaccurate ones and follow course disclosure requirements.

Supporting reading: [1] [3]

Questions about this topic

How is a reverse outline different from my original outline?

Your original outline records a plan. A reverse outline describes the paragraphs that actually exist, so it can reveal an unplanned tangent, missing step or conclusion that no longer matches the argument.

Must I use the P5–P1 order shown here?

No. The example compares two defensible arrangements for different readers. Explain the reason for a move and check the revised text; an order alone is not a universal answer.

Can moving paragraphs fix every problem?

No. In this draft P2 needs a focus decision and P6 needs a narrower conclusion. A missing reason, unsupported claim or contradictory policy requires substantive revision.

Are these real student or AI-generated essay results?

No. The draft, paragraph map and proposed revisions are constructed teaching material. They are not a real student submission, a product test or research evidence about workshop policies.

What do the downloads contain?

The blank CSV contains P1–P6 rows and five editable fields. The filled CSV records one interpretation of this draft and its revision decisions. A plain-text blank form is also provided for editing or printing. These resources do not grade your work.

Sources and editorial notes

The supporting reading below informs our editorial guidance. Exercises and scenarios illustrate a method; they do not report a completed product test or measured learning outcome. Read our editorial policy or send a correction with a supporting source to info@ai-essay.com.

  1. Cornell Knight Institute: Reverse Outlining
  2. Brandeis Writing Program: Reverse Outline handout
  3. George Mason Writing Center: Reverse Outlining