Reading triage: how to decide what to read fully, skim, or skip

The volume of text a typical knowledge worker encounters daily — emails, reports, research papers, news, documentation, Slack threads — far exceeds what can be read carefully. The question is not whether to read selectively, but whether to select deliberately or by default.

Reading triage is a deliberate allocation system: for each document that reaches your attention, a brief initial assessment determines the reading depth it merits — full reading, structured skimming, or deliberate skipping. The decision is made once, quickly, based on explicit criteria. The alternative — reading everything slowly until time runs out, then abandoning the rest — produces poor coverage of high-priority material and wasted time on low-priority material.

Carver's five gears: a vocabulary for reading depth

Ronald Carver (1990) identified five distinct reading processes based on decade of eye-tracking and reading rate research. These are not arbitrary speed categories but genuinely different cognitive modes, each with characteristic eye movement patterns, comprehension profiles, and appropriate use cases.

Gear 1 — Memorising (~138 WPM) Extremely slow, deliberate processing with many backward regressions and long fixations. Appropriate for material that must be retained precisely: legal definitions, mathematical proofs, safety procedures, precise instructions. Not appropriate for general non-fiction — applying Gear 1 to all reading is the most common cause of reading backlog.

Gear 2 — Learning (~200 WPM) Careful reading for unfamiliar or difficult content. Appropriate for technical material outside your domain expertise, foundational learning in a new field, or any text where you lack the prior knowledge to fill gaps from context. The slower pace is not inefficiency — it is calibrated to the comprehension load.

Gear 3 — Rauding (~300 WPM) Normal comfortable reading — the automatic gear for familiar content at comfortable difficulty. Most adults spend most of their reading time in this gear regardless of content complexity, including for material that would benefit from Gear 4 or 5. Raising awareness of the other gears is itself a comprehension and efficiency gain.

Gear 4 — Skimming (~450 WPM) Selective fixation targeting content words and sentence beginnings, skipping function words and body sentences. Appropriate for: documents where the main claim is more important than the evidence, familiar-topic updates, triage assessment passes, and pre-reads before a full Gear-3 reading. Comprehension at Gear 4 is typically 50–70% of main ideas.

Gear 5 — Scanning (~600–650 WPM) Very sparse fixation for specific targets — a date, a name, a figure, a keyword. Not reading in the comprehension sense but search behaviour. Appropriate for locating specific information in a known document, skimming indexes and tables of contents, and rapid document surveys.

Effective reading triage means assigning each document the correct gear before beginning — not defaulting to Gear 3 for everything.

The triage decision matrix

A practical triage framework operates on two dimensions: relevance (how closely does this document address your current goals?) and replaceability (how easily could you get this information elsewhere, or from a summary?).

Replaceability is often underestimated. Most business updates, meeting summaries, and informational emails are either superseded by subsequent events or covered more efficiently by asking a colleague. Reading them is often redundant.

A third dimension — time sensitivity — applies to time-critical decisions: read in full even if replaceability is high, if the decision window closes before a summary will be available.

Document-type heuristics

The matrix above requires two judgements per document. For common document types, standard heuristics reduce the judgement load:

These are defaults, not rules. Apply the matrix when the heuristic is uncertain.

The 3-minute triage protocol

For any document where the heuristic is ambiguous, apply a structured 3-minute triage:

Minute 1 — Survey structure: Read title, all headings, first paragraph, final paragraph, and any callout boxes or bolded terms. This is the SQ3R Survey step — it builds a minimal schema and reveals the document's argument skeleton. At the end of Minute 1, you know the document's main claim and approximate structure.

Minute 2 — Assess goal relevance: Answer: does this document's main claim advance any of my current active goals? If yes, which ones specifically? If not, is there a secondary section that does? If neither: the document is low-relevance.

Minute 3 — Decide and act: Based on relevance and replaceability, assign a gear and begin or schedule accordingly. If skipping: record that the document existed, in case you need to revisit later.

The SQ3R method's Survey step is the core of triage Minute 1. The metacognitive goal-setting protocol makes Minute 2 faster by keeping your current active goals explicitly loaded before your reading session begins.

Triage failure modes

Over-triaging low-stakes documents: Spending 3 minutes triaging a 2-minute email costs more than reading the email. Reserve the protocol for documents longer than 5 minutes of full reading time.

Using triage to avoid difficult reading: Triage finds documents that do not merit full reading. It is not a mechanism for classifying difficult-but-necessary material as skippable. Gear 2 discomfort is a signal that the content requires full engagement — not a triage flag.

No follow-through on flagged documents: Triage only saves time if skipped documents are genuinely not needed and full-read documents are actually read. A backlog of "skimmed" documents that still require full reading has simply moved the problem, not solved it.

Integrating with diagonal reading

Diagonal reading technique is the execution mechanism for Gear 4 triage scans. The diagonal path provides a systematic route through the document; the triage question provides the relevance filter. Run a diagonal scan as your triage pass, not a random skim — the systematic path ensures coverage of sentence beginnings across the full document, which is sufficient to establish main claims and structure.

The Diagonal Reader tool makes Gear 5 survey scans visual and calibrated. Set a high step angle and low word density for a fast triage pass through any pasted text. The content-word highlighting makes structural terms, numbers, and proper names immediately visible. Free, no account required.

For a complete evidence-based system integrating triage, diagonal reading, and metacognitive goal-setting, the Diagonal Reading course covers all of this in Lesson 5: multi-mode reading strategy. Six lessons, free, no account required.

Further reading in this series

• What is diagonal reading? The evidence-based guide to structured skimming
• F-pattern reading: what eye-tracking research reveals
• How to skim read effectively without losing comprehension
• The SQ3R method: why surveying before you read improves comprehension
• Metacognitive reading: why knowing your goal makes you faster

References

• Carver, R. P. (1990). Reading Rate: A Review of Research and Theory. Academic Press.
• Kintsch, W. (1988). The role of knowledge in discourse comprehension. Psychological Review, 95(2), 163–182.
• Klimovich, M., Tiffin-Richards, S. P., & Richter, T. (2023). Does speed-reading training work, and if so, why? Journal of Research in Reading, 46(2), 123–142. https://doi.org/10.1111/1467-9817.12417
• Rayner, K. (1998). Eye movements in reading and information processing: 20 years of research. Psychological Bulletin, 124(3), 372–422. https://doi.org/10.1037/0033-2909.124.3.372
• Rayner, K., Schotter, E. R., Masson, M. E. J., Potter, M. C., & Treiman, R. (2016). So much to read, so little time. Psychological Science in the Public Interest, 17(1), 4–34.
• Robinson, F. P. (1946). Effective Study. Harper & Brothers.