By Susan Carter

Writing about literature can cause confusion and frustration for new PhD candidates. How can they start writing a ‘literature review’ in the first few months of candidature when they are still not sure of what they are really looking for, and when they may not have finalised the scope and aims of the thesis?

Despite this, reviewing the literature is commonly the entry point to doctoral writing for many. Frequently it is begun early in the first year in order to stake out the research landscape. Early stage candidates generally work from a preliminary proposal that includes some literature. The main purpose of reading on the topic early is to ensure that their project hasn’t already been done, and to better understand methods commonly used. At the same time, the novice researcher will absorb the jargon and conventions of the discipline, and should be encouraged to do this consciously.

They must at least make notes of what they read so as to avoid forgetting who said what: writing must accompany reading. The more this early writing already includes evidence of their own critical analysis of literature, the more effort will be saved later down the track. It has to be done, but means building a literature review without a sure sense of its final definitive purpose. Some candidates are unnerved by contingency, and purpose for writing can help with that.

Here’s an approach for producing writing that is likely to be usable during a literature review early on when things still seem undecided.

I’ve written on the literature review previously, suggesting that, early on, reading is like a smash and grab or breaking and entering: you need to get in and then out again as quickly as possible, carrying away only what is valuable. A commenter moderated this to the less violent analogy of being a tourist who only takes souvenirs that are 1) valuable and 2) fit in the suitcase. Cally Guerin also has a post (citing Kamler and Thompson 2006), about having attitude when it comes to literature. This time I want to start from a different direction: not the reader’s attitude to the task of reading, but her method in recording findings to make that writing as useful as possible down the track.

I’m basing this on the convention that theses make generic moves, each of which needs support from the literature the candidate reads. The entry-level doctoral student could begin with a plan of the finished thesis or dissertation that factors in how each section of writing will advance the overall argument, and what work literature will need to do to support the final written thesis. The idea is to plot out the story that the literature must detail in, much like developing the storyline of a novel.

The introduction usually establishes a problem, limitation, or lack of understanding. Literature provides evidence that a gap exists in understanding about something that matters, to show its seriousness, and to identify exactly where the gap in knowledge lies that this thesis will fill. That gap is often not clearly defined at the start of the project, especially in non-STEM disciplines. However, any time in literature that the problem in the topic is mentioned, it can be written into a document that will eventually become the first few pages of the introduction. Literature that distinguishes where the problem is worse, or how it is mitigated slightly, or what details affect it (e.g., building material, gender, location, etc.) is also likely to fit under what I would simply call ‘the problem.’ Here, the ‘best’ literature will show that the problem really matters—that will strengthen the thesis in the eyes of readers.

Then the methods used will be defended with literature. Detail about previous approaches will build the argument for methods used—reporting these from the literature requires looking for other studies’ limitations as well as strengths and writing the story of what seems useful and what less so to the current project. Having a set of evaluation criteria in hand enables even early reading to be fitted into a coherent plan, described in a way that is likely to be useable. Chris Hart (1988) spells out the kinds of questions that literature ought to answer. Candidates can work out what questions they need to ask of literature for their own project, and might include criteria like:

  • Are any definitions useful?
  • Is their problem the same as mine?
  • Are their methods good?
  • What supports my ideas?
  • What raises new ideas or disagrees with mine?
  • What are the limitations?

Additionally, they can look at each publication’s prose for models that can be followed. Does it clearly articulate the problem, and the argument? Is it succinct and convincing? How does it defend methods and articulate a methodology? Does it use good vocabulary? How does it handle theory? Strong prose structures at the level of paragraph and sentence can be emulated by novice researchers who are unsure.

If you are in a discipline that uses direct quotation, accumulating useful quotations is essential—and while you are picking those that express ideas eloquently, you can look at syntax for how they achieve this quality. Care with page numbers is essential.

It still won’t be easy, but the candidate could begin by constructing a mock contents page for their future thesis and itemising what literature will be needed in each section gives one way to mitigate the panic that some doctoral students feel when reading and writing while still unsure of their final topic.


Hart, Chris. (1998). Doing a Literature Review: Releasing the Social Science Research Imagination. London: Sage Publications, 1998.

Kamer, B. & Thompson, P. (2006). Helping doctoral students write: Pedagogies for supervision. London: Routledge.