Contents
I am not trying to make a page less white. I want a highlight to tell me what to do with the passage when I return. That might mean defining a term, checking a claim or noticing an important exception. Without that reason, a bright paragraph can be as difficult to use as an unmarked one.
For a worked example, imagine reading a fictional report about a neighborhood workshop. The report describes who attended, how the organizer collected feedback and what it did not measure. I will not supply invented results. The exercise is to keep the role of each passage visible while deciding what belongs in a note.
Read for a question before choosing a passage
I would put “What can this report tell me about planning another workshop?” at the top of the note. A paragraph about the organizer's history might be interesting without helping that question. A description of the feedback method may be much more relevant, even if it is less memorable.
My question can change while I read. If the report turns out to concern a very different activity, I would record that mismatch. It is better to save one useful limitation than to collect a page of material that only appears relevant because I have stopped naming the task.
Mark the role, not merely the importance
For this exercise, I would use short comments such as “definition,” “method,” “limit” and “question.” The wording is a personal working choice, not a validated learning technique. I would use text labels even if I also used colors, so the role survives a printout or a change of viewer.
A definition may be needed to understand the source's own terminology. It does not automatically become my project's definition. Likewise, a method description tells me how the author obtained information; it does not by itself confirm that the method fits another purpose. A good comment protects that distinction.
Keep the exception beside the claim
Suppose the fictional report says its feedback came only from people who stayed to the end. I would keep that limitation with any note about the feedback, rather than highlighting the response paragraph alone. The limit affects what I can infer about people who left or did not attend.
I would be particularly careful with small words such as “some,” “reported” and “during this session.” A summary that removes those words can become a different claim. When a passage is central to a decision, I would reopen the source before turning the note into polished prose.
Turn the selection into a usable note
Zotero's PDF tools can bring annotations into notes with citations and links back to the page. That provides a route back, which is useful for checking context. I would still add my own reason for selecting the passage; the software cannot know the question I brought to it.
Here is a demonstration annotation without any invented report findings. It deliberately separates a source description from a question I would ask. If the actual report did not answer that question, I would keep it unresolved rather than make the highlight do more work than the passage can support.
Role: limitation Source location: Example workshop report, “Feedback method” Source describes: Who was invited to provide feedback. Why I selected it: It sets the boundary of the feedback claim. My question: Whose views are absent? Before using: Reopen the passage and check the exact scope.
Take care when carrying exact wording forward
Purdue OWL's guidance makes a useful distinction between quotation, paraphrase and summary. I would preserve that distinction when transforming annotations. Copying a passage into a note does not turn its wording into my own. A close rephrasing still needs attribution to the source.
When drafting, I would decide whether exact wording matters. If not, I can explain the relevant point in my own words and retain its source. If it does matter, I would quote briefly and accurately. Either way, the note should not blur the boundary between evidence and my response to it.
Leave fewer selected passages with better explanations
At the end of the reading session, I would review the annotations tied to my question. Some might become a note; some might remain useful navigation marks; others might no longer matter. I would not turn every highlight into a permanent card just because it already has a color.
The measure here is a task I can inspect: can I explain why I selected this, find its source and state its limit? That does not prove better retention or faster writing. It does give the next stage of the project a clearer starting point than a collection of unexplained bright sentences.
Follow one paragraph through the whole reading process
I would take the fictional workshop report through a complete sequence. First comes the question: what does this document help me decide about another session? Next comes a selected paragraph describing how feedback was collected. I would mark it as method, then add a short comment explaining why the collection method affects interpretation. I would not begin by choosing a color palette. The annotation needs a purpose before its appearance matters, because the purpose determines what information must survive when the passage moves into a note.
The source's language should remain distinct from my response. If I copy exact wording, I would preserve quotation boundaries and a source location. My comment might ask whether people who left early were included. That question is my reasoning about the report; it is not an answer supplied by the passage. If the report does not address it, I would record the gap. The note should help me see that I need another check, rather than allow the question to become a fictional finding through repeated summarization.
When I move the annotation into a project note, I would add the decision context. “This matters because the proposal discusses participants' views” explains relevance. I would then narrow the draft sentence to what the report can support. If I have not established whose views were represented, I would avoid language about all attendees or the wider neighborhood. The highlight is useful precisely because it makes the limit visible. It need not produce a more confident claim to earn a place in the working document.
Before final use, I would return to the paragraph and inspect the surrounding section. A summary may have dropped a qualification or the source may distinguish between a formal survey and informal comments. I would correct the note if needed, then update the draft. This sequence is slower to describe than to perform, but the description shows where judgment enters: selection, explanation, transfer and reuse. An annotation tool can help preserve the route through those stages; it cannot take responsibility for the meaning I assign to the passage.
Use a text label when a color would become ambiguous
I would imagine printing the notes in black and white or sharing them with someone who does not know my color system. A short label such as definition, support, limitation or question can remain understandable in those circumstances. The exact vocabulary is mine to choose. What matters is that each term names a role I can explain. If I need a long legend to remember which shade means an unresolved issue, I would simplify before adding another distinction.
A paragraph can also have more than one role. The fictional report's method section may define a category and limit the interpretation of a result. I would avoid forcing it into one permanent color merely to keep the page visually neat. A brief comment can explain both functions: “Defines who was counted; limits the claim about participants.” That sentence is more useful to the project than a perfectly uniform scheme that requires the reader to guess which function mattered when the passage was selected.
I would treat the role as dependent on the question. The same paragraph could be background for one essay and central evidence for another. That is a reason to retain source-specific observations separately from project-specific use. The source record can describe what the passage contains. The project note can explain why it matters now. If I copy the same highlight into another project, I would reconsider the explanation rather than assume the old label gives it the same role everywhere.
I would also permit a mark to remain merely navigational. Perhaps it identifies a useful section to revisit without contributing to the current argument. That can be enough. Converting every annotation into a permanent knowledge card would add work without necessarily improving retrieval. I would inspect the small set related to the current question, decide which deserve a note and leave the rest clearly scoped. The test is usable context, not the proportion of the document covered by colors or comments.
Handle a definition and a limitation differently
For a definition, I would note whose usage it describes and where the source applies it. A term that sounds familiar can have a specific meaning in a particular document. In the fictional report, “participant” might refer to one defined group. I would not import that word into my proposal without checking whether the proposed event uses it the same way. A short definition note can protect a draft from relying on an ordinary-language assumption that the source did not make.
For a limitation, I would identify the claim it constrains. A note saying only “small sample” or “unclear method” may be too vague to use. I would write what I cannot infer and why that matters to the decision. The limitation should travel with the supporting point, not sit in an unrelated folder where it will be forgotten when the attractive result is copied into a draft. Keeping both together makes the eventual prose easier to assess.
These roles suggest different next actions. A definition may need a precise quotation or a careful paraphrase. A limitation may need another source, a narrower sentence or a decision to omit the claim. I would avoid treating all annotations as a queue of passages to summarize. The reader's next use determines the work. A well-chosen deletion can improve an argument just as much as an additional citation can, especially when the proposed claim was broader than the material available.
I would review an ambiguous annotation by writing two possible uses and checking which the passage supports. One use might describe the report's own terminology; another might make a claim about a whole community. If only the first is supported, the note should say so. This is a concrete way to inspect my own interpretation without pretending the act of highlighting has verified it. The mark remains a starting point for reasoning, rather than a sign that the relevant judgment has already been completed.
Check an exported annotation before depending on it
If I expect to use annotations outside the reading application, I would test a small representative export. The sample should include an ordinary highlight, a comment and the route back to the source. I would inspect what the destination actually displays. A file can open successfully while omitting comments or making source locations hard to follow. That is a workflow problem to discover before annotating an entire project in a way its collaborators cannot use.
Zotero documents annotation handling and export in its own PDF workflow. I would follow those current instructions for the actual application, while treating the reading method in this article as an independent editorial choice. A software feature can preserve a citation or page link without determining whether the selected passage supports my sentence. I want the exported artifact to carry enough context for that judgment to remain possible in the next environment.
For the imaginary workshop project, I would export a short note and ask whether another person can distinguish the source passage from my question about it. If the formatting collapses those boundaries, I would revise the export or add clear labels. I would not rely on color alone to preserve attribution. The result should be understandable as a document, not only to someone who watched the original annotation session and remembers which part of the interface represented which kind of text.
I would then close the loop by drafting one supported paragraph and checking it against the annotation and source. If the route is clear, the workflow has served its immediate purpose. If it takes several searches to identify the passage, I would repair the source pointer before expanding the system. A smaller set of annotations that reliably reaches usable writing is a better foundation than a large collection whose visual order hides an uncertain connection to the material it was meant to preserve.
Read the resulting note without the colored page
I would close the source view and inspect the note in plain form. Does it identify the passage, preserve attribution and explain why the selection matters? Does it keep my question separate from the source's answer? This check reveals whether the annotation's usefulness depends entirely on a visual arrangement that may disappear when the material moves into a document or another application.
For the fictional workshop report, the note should still distinguish the collection method from the proposed use of its feedback. If that distinction disappears, I would repair the wording before drafting. The useful output is not a transcript of every mark on the page. It is a small set of explanations that helps the writer make a bounded claim and return to the relevant source location when needed.
I would keep the question at the end specific enough to answer. “Who was invited to provide feedback?” gives a future reading session a direction. “Understand this better” may be a reasonable feeling, but it does not identify the next inspection. A highlight has earned its role when it can guide that inspection or support a clearly scoped paragraph after the color itself has stopped doing any work.
Frequently Asked Questions
Q: How many colors should I use?
A: Use a set you can consistently explain. One color with short text comments may be enough. This article does not claim a particular color system improves memory or reading speed.
Q: Should I highlight while reading or afterward?
A: Either can work as a personal workflow. I would first establish the question and preserve enough context. If early selections distract from understanding the whole passage, a later pass may be more suitable.
Q: Can an annotation replace a source note?
A: Sometimes a brief comment is sufficient for retrieval. For an argument, you may need a fuller note that connects passages and states their limits. Keep a reliable route back in either case.
Q: Will exported annotations appear in every PDF reader?
A: Not necessarily. Zotero documents how its annotations are stored and exported. Check the intended export method and inspect a sample in the destination reader before relying on it.
Q: What if I highlighted almost the whole page?
A: Return to your question and identify the few passages that answer it or limit the answer. You do not need to delete everything, but give the important selections a specific reason before adding more.
One thing to take forward
I would keep the next highlight only after naming its job. A definition, a piece of support, a limitation and an unanswered question are all worth noticing. They become much easier to use when the note says which one I found.
Sources & further reading
Something needs correcting? Write to the editors with the article link.



