Topic Tags vs. Workflow Tags in a Research Library

A research library usually starts to strain at the same moment: you have enough papers that finding them matters, and enough projects that a single folder structure stops working. Two tagging habits solve different halves of that problem. Topic tags describe what a paper is about. Workflow tags describe where a paper sits in your process — candidate, screened, included, cited, needs a coauthor decision.

The practical answer is that you rarely need to choose. Use topic tags to make the intellectual landscape of a project searchable, and workflow tags to make the state of each record auditable. The failure mode is using one vocabulary to do both jobs: topic-only libraries lose track of screening decisions, and workflow-only libraries become a pipeline with no subject matter.

This guide compares both approaches on the same task, shows where each breaks down, and gives a decision table plus a combined workflow you can apply to a manuscript, a coauthor handoff, a peer-review response, or a reproducible search.

Before adding a single tag, name the job the library must finish. The evidence from established practice points to a few recurring jobs:

• Project grouping without duplication. Zotero collections can place one item in multiple project groupings without duplicating it, and tags can describe topic, method, status, or workflow. That flexibility is exactly why a tagging scheme needs a rule.
• Reproducible searching. Cochrane guidance recommends planning searches from eligibility criteria, using multiple sources, combining free-text terms with subject headings, documenting the process, deduplicating records, screening in stages, linking multiple reports of one study, and rerunning searches before publication.
• Transparent reporting. PRISMA 2020 supplies a reporting checklist, expanded explanations, an abstract checklist, and flow-diagram templates for original and updated systematic reviews, covering search, selection, and synthesis methods.

If your library must support any of those jobs, tags are not decoration. They are part of the method trail. A useful diagnostic question: if a coauthor opened your library tomorrow, could they tell both what a paper is about and what you decided about it? If only one of those is legible, you have a tagging gap.

Topic tags answer the question "which papers belong to this idea?" They are the vocabulary of your subject matter: a mechanism, a population, a method family, a theoretical lens. Their strength is that they survive project changes. A paper tagged for a topic remains findable when the manuscript is revised, the grant ends, or a new review begins.

Topic tags also make saved searches useful. Zotero saved searches update automatically, so a well-built topic vocabulary can keep surfacing new material without manual re-filing. That is the closest thing to a living literature map inside a personal library.

Where topic tags fail is state. A topic tag cannot tell you whether a paper has been screened, whether its full text was read, whether it was excluded and why, or whether a coauthor has signed off. If you try to encode state in topic tags, your vocabulary balloons and your subject terms become contaminated with process words.

Workflow tags answer "where is this record in my process?" They are deliberately boring and small: a status, a decision, an owner, a next action. Their strength is auditability. When a coauthor or reviewer asks how a record was handled, a workflow tag is the shortest honest answer.

Workflow tags also pair naturally with staged screening. Cochrane guidance describes screening in stages and linking multiple reports of one study; a workflow vocabulary can mark each stage and each linked report so the trail stays visible. PRISMA 2020's flow-diagram templates similarly depend on counts moving through defined stages — workflow tags give you the record-level state that those counts summarize.

Where workflow tags fail is meaning. A library full of "to-read" and "included" tells you nothing about the intellectual content. Workflow tags also age badly if they are not pruned: statuses from a finished project linger and confuse the next one.

Take one concrete task: assembling the evidence base for a manuscript section while a coauthor reviews your selections.

Read the table as a division of labor, not a contest. The task needs both columns: topic tags to define the evidence base, workflow tags to prove how it was assembled.

Topic-only failure. You can retrieve every paper on a mechanism but cannot reconstruct why three were excluded. When a coauthor asks about a gap, you re-read papers you already screened. The method trail exists only in your memory.

Workflow-only failure. You can prove every record passed through screening but cannot group them by idea. Writing the manuscript means manually re-sorting a pipeline into themes.

Mixed-vocabulary failure. You use one tag list for both jobs, so "excluded" sits beside "neuroinflammation" and neither search behaves predictably. This is the most common and most fixable problem: separate the vocabularies, then let each do its job.

Start from eligibility criteria, as Cochrane guidance recommends, and derive your topic tags from them. If a population, intervention, or method family appears in your criteria, it deserves a topic tag. This keeps the vocabulary tied to the question rather than to whatever you happened to read first.

Cochrane guidance recommends multiple sources and combining free-text terms with subject headings. Use a scholarly search interface to run those queries and capture the records you retrieve — for example, WisPaper search for finding academic papers, or the WisPaper Scholar Search entry point if you want a public explanation of the search capability first.

Bring the records into one place. In WisPaper, My Library is where you save or upload papers and organize them. This is the moment to apply both vocabularies: topic tags for subject, workflow tags for state.

Cochrane guidance calls for deduplicating records and linking multiple reports of one study. Zotero's Duplicate Items collection supports review and merging, which is a practical way to handle duplicates before tagging. Link multi-report studies with a workflow tag so the linkage survives screening.

Apply workflow tags at each stage — candidate, screened, full text assessed, included, excluded with reason. Because Cochrane guidance describes staged screening, the tags should mirror the stages you actually run, not an idealized pipeline.

Once papers are in the library, you can interrogate them. Library QA supports asking questions based on papers uploaded to My Library or added from search results. Use it to check what the assembled evidence says before you draft, and record the questions you asked as part of the trail.

Before the reference list is finalized, screen the BibTeX. TrueCite checks BibTeX against real academic databases to flag hallucinated references. A workflow tag marking "citation verified" closes the loop between the library and the manuscript.

Cochrane guidance recommends rerunning searches before publication. Because saved searches update automatically in Zotero, a topic-tag-driven search can surface new records; tag them with a fresh workflow status so the update is visible in the trail.

A handoff is the clearest test of a tagging scheme. Before you pass the library to a coauthor, check three things:

1. Scope is legible. Topic tags let the coauthor see the intellectual boundaries of the evidence base.
2. State is legible. Workflow tags let them see which records are settled and which need a decision.
3. The trail is reproducible. Search documentation, deduplication, and staged screening are recorded, consistent with Cochrane guidance and reportable against PRISMA 2020's checklist and flow-diagram templates.

If a coauthor can answer "what is in scope?" and "what still needs doing?" without opening each record, the handoff is ready.

A tagging scheme is finished when it passes four checks:

• Two vocabularies, no overlap. No tag does both a topic and a workflow job.
• Every included record has a state. You can trace each record from candidate to included or excluded.
• Saved searches behave predictably. Topic terms drive discovery; status terms do not pollute it.
• The trail matches a reporting standard. Your search, selection, and synthesis steps can be described against PRISMA 2020's checklist, and your search process follows Cochrane's documentation guidance.

If any check fails, fix the vocabulary before adding more papers. Tag debt compounds faster than citation debt.