Every dense paper has a trail. This is the map.

Do this

Finish the sentence “I am reading this paper in order to…” in writing. Maybe you need to decide whether an approach is worth adopting, learn the vocabulary of a field, check one claim before repeating it, or find a method you can reuse. The answer sets your pace: a paper read to pass a quiz deserves a different trek than one read before a purchasing decision.

Watch for

Reading because it was assigned, with no question in mind. When nothing is the point, every sentence looks equally important, and you will tire before the interesting part.

Try this

Write your reason at the top of a page. Add one thing you already believe about the topic. Then take question 1 from the 30 questions and answer it in two lines.

Suggested time box: about 2 minutes

I have walked this waypoint

Do this

Read the title, the abstract, every heading, and the conclusion, in that order, and nothing else. Then write down in your own words what you think the paper claims and how it plans to show it. Treat the guess as a bet you will settle later.

Watch for

Headings that promise more than the abstract does, and abstracts that stop short of saying what was actually shown. A mismatch is not a verdict, but it tells you where to look hard.

Try this

Open the paper anatomy and walk the suggested reading order on the mock paper. It is the same scouting routine, slowed down and labelled.

Suggested time box: about 5 minutes

I have walked this waypoint

Do this

Go through every figure and table with its caption. For each one, say aloud what is on each axis, what is being compared with what, and what the authors want you to notice. If you cannot say it, circle it. Those circles become your reading list inside the text.

Watch for

Axes that start somewhere other than zero, comparisons against a weak baseline, missing units, and captions that state a conclusion the picture does not quite show.

Try this

Pick the one figure that carries the main claim. Sketch it from memory, then check what you got wrong. Questions 14 to 18 on the questions page are built for this moment.

Suggested time box: about 10 minutes

I have walked this waypoint

Do this

Read the introduction and the conclusion properly. Then move through the middle at the speed of your finger, reading the first sentence of each paragraph and stopping only when a sentence snags. Do not stop to look anything up yet. Put a mark in the margin and keep walking.

Watch for

The urge to master every paragraph on the way up. The first pass is for shape, not for detail. If you can say what each section is for, the pass worked, even if you cannot yet explain how.

Try this

Close the paper and write a three-line summary: the problem, the move the authors make, and what they say it buys you. Compare it with your guess from waypoint 2.

Suggested time box: 10 to 15 minutes

I have walked this waypoint

Do this

On your second pass, collect every term the authors define or lean on. Write each with its meaning in your own words, and the page where it first appears. Note the ones that look ordinary but are used in a special way: in technical writing, the plain-looking words are the ones that hide the most.

Watch for

A term that quietly changes meaning between sections, an acronym used before it is spelled out, and a definition that only works because it already contains the thing being defined.

Try this

Paste a stretch of text into the reading lab and tag every definition you find. The summary builder will collect them into one list for you.

Suggested time box: 15 minutes

I have walked this waypoint

Do this

Read slowly now. Mark each central claim, then look for what it rests on: the assumptions stated, the assumptions left unstated, and the evidence offered. Draw the chain on paper if it helps. A claim with nothing underneath it is not necessarily wrong, but you should know that it stands alone.

Watch for

Sentences that begin “clearly” or “it follows that” where no step has been shown, and verbs that grow stronger as the paper goes on: “suggests” in the results becoming “proves” in the conclusion.

Try this

Use all five tags in the reading lab: Claim, Assumption, Definition, Evidence, Question. If you have many claims and few pieces of evidence, you have found where to dig.

Suggested time box: 30 minutes, longer for long papers

I have walked this waypoint

Do this

Put weight on each load-bearing claim and see what holds. Ask where the evidence came from and who chose it, what a fair comparison would look like, what the authors say they cannot show, and what the strongest competing explanation would be. Read the limitations section last and most carefully.

Watch for

Limitations that are listed but never allowed to change the conclusion, and evidence that is easy to produce rather than informative. Also watch your own reaction: you are most likely to skip scrutiny on claims you already like.

Try this

Choose three questions from the Evidence and Limits groups on the questions page, save them, and answer them in writing for the one claim you most want to rely on.

Suggested time box: about 20 minutes

I have walked this waypoint

Do this

Write a one-page summary without looking: the question, the main claim, the evidence you find convincing, the evidence you do not, the assumptions you are taking on trust, and what you will do next. Then compare it with your reason from waypoint 1. Did the paper answer what you came to ask?

Watch for

Summaries that only repeat the abstract. If nothing in yours is a judgment, you have copied the paper rather than read it. Also note the questions you still have; they are a result, not a failure.

Try this

Let the summary builder in the reading lab assemble your tagged passages into notes, then export them as Markdown or plain text and file them where you will find them again.

Suggested time box: about 10 minutes

I have walked this waypoint