There's too much literature and it keeps growing
Draw the edges of your review, read with a purpose, and know when a section is done.
In short
Start here, today. Write your research question at the top of a page. Below it, list the four to six things a reader must understand before your question makes sense. These are the sections of your review. Anything that doesn't serve one of them can wait. This takes about 25 minutes.
After that, work through these steps. Choose one to jump to it.
Step by step
Tick each step as you finish it. Open a step for the how.
0 of 6 steps done
Draw the boundaries
For each section, write what's in and what's out: time period, discipline, population, method, region. Keep the note short and put it in your Decision Log, so you don't have to argue with yourself every time a tangent appears. Boundaries can move later, but on purpose.
Read in layers
Start with the abstract, the introduction and the conclusion, and glance at the figures or tables. Read the whole paper only when it's central to your question or you'll be using its methods or findings in detail. Most papers earn a skim, some earn a careful read, and a handful you'll come back to again and again.
Keep a reading log
For each source, write one line: what it claims, how it knows, and how it relates to your question. A spreadsheet works, or a note attached to the reference. After twenty entries you'll start to see clusters, and after fifty you'll see which clusters you already understand well enough.
Map before you dig
Put two or three central papers into Connected Papers to see what sits nearby and which work everyone cites. Recent review articles and dissertations in your area do a similar job. Their reference lists show you what the field treats as essential, which saves you rebuilding that list from scratch.
Set a stop date for each section
Give each section of the review a date by which you'll stop searching and start writing, and put it in your weekly plan. If new papers are mostly repeating what you know before that date, stop early. You can still add a source later, but the search itself has to end.
Let alerts wait
Search alerts are useful, but read them once a month rather than as they arrive. Most new papers can be added to a near-final draft in a sentence. The rare one that changes your argument will be obvious when you see it.
Words you can use
Is this you? Signs to look for
- Every paper you read cites four more you feel you should read.
- Your downloads folder holds hundreds of PDFs you haven't opened.
- You can't say what your review needs to cover, only that it isn't covered yet.
- New papers appear every week and you worry you'll miss the one that matters.
Why this happens
A literature review doesn't have to contain everything written on your topic. Its job is narrower: to show what a reader needs to know to understand your question, why it matters and how you're answering it. Without that boundary, every reference looks relevant and the reading never ends.
The pile also grows because of how many of us were taught to read. Start at the first word, finish at the last, take careful notes on everything. That works for the ten or fifteen papers at the centre of your project. It doesn't work for the two hundred around them.
There is a point, usually earlier than people expect, where new papers stop changing your understanding. They cite the same work, use similar methods and make familiar arguments. That is saturation, and it's your signal to stop searching for that section and start writing it.
Worth a look (5)
- Connected PapersEnter a central paper and see related work laid out as a map.
- PatterPat Thomson writes often about literature work and what a review is for.
- Google ScholarA free search engine for scholarly articles, theses, books and court opinions in every field. Set up alerts so new papers on your topic come to you.
- Semantic ScholarA free search engine for research papers from a nonprofit AI institute. One-line summaries and citation links help you see how a topic fits together.
- OpenAlexA free, open index of hundreds of millions of scholarly works, authors and institutions. You can search it directly or use its data for bibliometric research.
More on finding and reading research in the resource library
When to bring someone in
If you can't decide what belongs in the review, ask your supervisor to react to your outline of sections and boundaries rather than your reading list. A subject librarian can help you build a search strategy and set up alerts, which is especially helpful when your topic crosses disciplines and the literature is scattered across databases.
Find the right person, or talk to Ify