Uncovering Discourses of Representation in Young Adult Fiction (DoRA)

The aims of the project are to understand the ‘mirrors and windows’ in literature that young adults read by analysing the linguistic patterns in the books, and the responses of young adults to that same reading material.

What we’re aiming to do

We’re interested in the ways in which language is used in young adult literature to construct characters, storylines and fictional worlds that young people access through their reading repertoires. 

A key question for us is: what worldviews are presented to young people through their reading of current best-selling YA fiction, and what are their responses to these worldviews?

The project aims to reveal the ‘windows and mirrors’ in the top 50 best-selling young adult novels through Corpus Assisted Discourse Analysis and to explore secondary school students’ perspectives on key findings through focus groups and surveys with students and their school librarians.

The literal function of a mirror is to see your own reflection. A window is designed to be looked through, and a sliding glass door can be opened or closed to allow us to walk from one space (inside) to another (outside). In a foundational text on children’s literature by Rudine Sims Bishop (1990), she argues that children’s literature should offer ‘mirrors, windows and sliding glass doors’ onto the world.  That is to say, the books children and young adults read should provide opportunities to see other world views and experiences (windows) and equally they should be able to see themselves reflected in the stories that they read (mirrors).  We take this often-cited metaphor as a way to frame the project aims, that is, to identify and investigate the worldviews (windows and mirrors) in best-selling young adult fiction. 

How we’re doing this

We’ve built a corpus (or electronically-stored, organised set of texts) of the 50 most commercially-successful books for young adults (aged 11-18 years) sold in the UK over a 5-year period (2017-2022). We’re exploring this corpus using corpus software tools to reveal the ‘mirrors and windows’ woven through the language choices of characters and storylines in these books. 
The second phase of the research is to seek the thoughts, views and opinions of secondary school students and their school librarian on some of these findings and their experiences of reading. 
The third phase involves a questionnaire of a wider group of students and will build on phases 1 and 2.

How is this research different to other projects?

Most research in this area focuses on a single book and takes literary approaches to the analysis conducted by adults. Young people’s responses and perceptions of young adult fiction are largely unknown. This linguistically-focused project is quantitatively investigating a corpus of 50 books. In addition, key findings will be shared with groups of young people, and thus the corpus linguistics work will be complemented by a qualitative exploration of the young people’s responses to the ideological messages woven through their reading repertoires. 

We decided to take the most commercially successful Young Adult (YA) fiction in a recent five year period as the basis of what young people read. We recognise that some young people will have read many of the best-sellers and others fewer. To access the best-sellers, we purchased a Nielsen-compiled list of the 100 best-selling books in the UK between 2017-2022, with data drawn from multiple retailers. For this pilot study, we focus on the top 50 books. Some changes were made to the original Nielsen list, for example, we excluded duplicate entries from books with more than one edition, three books which were out of print and two World Book Day editions (as these are very cheap and might skew the results). This meant our selection extended beyond the 50-mark in the original list and the rankings were adjusted accordingly.

The top 10

Here are the resulting top ten best-selling YA fiction books from 2017-2022:-

1. One of Us is Lying by Karen McManus
2. They Both Die at the End by Adam Silvera
3. The Book of Dust, Volume Two: The Secret Commonwealth by Philip Pullman
4. Midnight Sun by Stephanie Meyer
5. The Book of Dust, Volume One: La Belle Sauvage by Philip Pullman
6. A Good Girl’s Guide to Murder by Holly Jackson
7. The Hate U Give by Angie Thomas
8. Thirteen Reasons Why by Jay Asher
9. Two Can Keep a Secret by Karen McManus
10. One of Us is Next by Karen McManus

Building the corpus

The start of our investigation involved building a corpus of the 50 YA books. A corpus is an electronically-stored, organised collection of texts. Having a corpus of YA texts enables us to use corpus software to explore language patterns across all 50 books. It is also possible to build sub-corpora, for example all books written by women, or all fantasy books. The corpus analysis is a fundamental part of the research, as we are interested in the patterns of meanings present across young people’s reading repertoire, rather than a more traditional literary approach to understanding a single book or small number of texts. Before compiling the corpus, we asked for publisher permission to convert the 50 e-books we had bought into a machine-readable format suitable for corpus software. The YA Fiction corpus currently contains 5 million words.

Corpus analysis

We are currently exploring the corpus using corpus software tools including Wmatrix, AntConc and WordSmith to reveal the ‘mirrors and windows’ woven through the language choices of characters and storylines in these books. We are pursuing four initial areas of interest: 1) language patterns across male or female authorship, 2) representations of health and ill health, 3) representations of worries, and 4) constructions of genders. Brief snapshots of each of these areas are given below.

Initial findings 1: language patterns across male or female authorship

Initial findings from semantic analysis show that the books written by men (n=16) and the books written by women (n=34) create quite different worlds. For example, the books in the top 50 YA fiction authored by women have more words around emotion and feelings (e.g. mood, tone, happy, selfish, frustrating) and paying attention (focused, concentrate, ignored, distract); this contributes to creating worlds which are more relational and express more affective states. In contrast, the books authored by men include more words featuring animals (birds, sharks) and places (river, desert, world), and actions such as warfare (gun, officer); this helps to create a more physical world with physical actions.

Initial findings 2: representations of health and ill health

To investigate health, we created a subcorpus of books featuring character’s health (n=10) and contrasted this with the remaining books (n=40). We first explored the frequently-used pronoun ‘I’ to see what verbs follow this in each corpus. In the health corpus, verbs following ‘I’ included realised, wondered, hesitated, suggesting that characters in this corpus were described as more reflective than characters in the non-health corpus but more work is needed to follow up on this.

Initial findings 3: representations of worries

In terms of what characters worry about, all instances of the words worried, anxious and concerned from across the 50 books were identified, alongside five words either side. For each instance, the nature of the worry/ anxiety or concern was coded using an iterative and inductive approach driven from the data. The initial findings suggest that characters are concerned about other people’s worries, about their safety and their health, but less concerned about others’ thoughts, words or reactions, their romantic relationships, finances and appearance. The data is now being explored in more detail to include objects of worry, anxiety and concern that were, in the first phase, not available through the short textual extracts.

Initial findings 4: constructions of genders

An investigation to understand how gender is represented through the descriptions of characters and their actions. As an example, to show the kinds of areas to be explored, we looked the verbs that occur with she and he. The most common verbs reflected quite stereotypical depictions of characters, with female characters who flounce, vomit, choke, miss, realise, listen and yawn, whereas male characters chuckle, deserve, admit, expect, own, and refuse. Further explorations in terms of the representations of female and male characters will expand this line of research.

Taking the findings to participants

The above four sets of findings were shared with young people via focus groups and the librarians in interviews. The brief snapshots presented here offer a range of areas we felt would be of interest and which would generate discussion about representations in the fiction young adults read. The analytic work of the project is ongoing to investigate these areas of interest in more depth.

Focus groups

In order to understand young people’s responses, we shared some of our initial findings (see above) from the corpus analysis with young people. We first gave a presentation of our findings to self-selected students from two year groups and then split into smaller groups for discussion on the findings. Each year group included a mix of girls and boys.

As an example, here is a response from one girl in Year 8 (12-13 years old) talking about what characters worry about:

’I feel like it doesn't really reflect what people actually feel. because usually teenagers, especially. They're gonna worry about what they look like and all clothes that they wear. And I think thought that people read these books because they like not having to worry about the clothes they wear or how they look or what someone else gonna think of them. And they like worrying about other people and other characters instead of just themselves.’

In some contrast, a year 12 girl focused her response on the impact of what characters worry about:

‘‘…society kind of influences what you should be worrying about more. But I feel like books can also do that in the sense that, you know, they can also tell you what you should be worrying about.’

Librarian interviews

We interviewed two secondary school librarians in the same school and shared the same initial findings.

As an example of the comments made, below is what one librarian said when asked about the stereotypical female and male characters:

‘storylines can be totally diverse. I mean, a lot of those are science fiction and fantasy novels, and But they're still reinforcing the stereotypes of society. I mean, there's really no need for that.’

And a comment about the impact of stereotypes on young people, and her thoughts on her future reading practices:

‘They become hidden philosophy, […], they're feeding into the psyche, without, you know, they're enjoying the story, but then all of this [stereotyping] is going in behind the scenes, without them even being aware that that's what they're I taking on board. […]But I think now, when I read a book, I will be more aware of what it's saying to me about what categories and putting people in, what language it's using around the characters, definitely.’

Reflecting on the safe space of books, one librarian commented:

‘…they have a lot of pressure on them at school, when they read stories like that, either it's never happened to them, or it's really close to home. And they love that, why wouldn't they? And they feel like they're part of something else don't they? And I think that's why the younger ones really like that or it's something they would never talk about at home. We have lots of families that don't have a lot of even English books at home. Or, you know, language, it's a language barrier. And so they would, they would never talk about that at home. So they found reading a story, it's a safe place.’

So far we have presented at the Applied Linguistics and Language Research Group at The Open University (September 2022 and November 2023) and at the international Corpus Linguistics conference in July 2023.

Conference presentation: Leedham, Maria; Hunt, Sally and Mukherjee, Sarah Jane (2023). Mirrors and windows in Young Adult fiction. In: The twelfth international Corpus Linguistics conference (CL2023), 3-6 Jul 2023, Lancaster, UK.

Education Blogpost: Teaching and research synergy: Windows, mirrors and doors in young adult fiction | Research Conversations: by and for Education researchers (open.ac.uk)

Drawing on the success of the pilot study described above, we are planning an expansion of the project, subject to funding. We will continue to use corpus linguistics to investigate patterns of meaning and will expand the corpus to include the most commercially successful 250 books within a 5 year period resulting in a corpus of an estimated 25 million words. The new expanded project will include an online questionnaire to all four UK nations, and focus groups with students from year 8 and year 12 and interviews with their school librarians in 10 schools. Three of those schools will be invited to work with us to co-develop materials for use in schools across the UK.