Reading time: 8 minutes
Authors can show aspects of characterisation in many ways. For example, authors may describe how characters look and act and the ways in which others think and talk about them. Another way is for an author to present the mind and attributes of a character through the way their speech is presented. This digest summarises a study by Maeve Eberhardt from the University of Vermont, who looks specifically at the distribution of direct speech reporting verbs (verbs used to frame a character’s exact words, such as ‘said’, ‘replied’, ‘argued’) in J.K. Rowling’s Harry Potter novels. Eberhardt draws on a corpus of 1.1 million words across all seven books to examine the direct speech reporting verbs from two of the series’ main characters, Ron and Hermione. This study explores not what is said but the ways in which characters’ words are framed and how these portrayals may evoke and reproduce cultural gendered stereotypes.
Eberhardt begins with a useful overview of recent work on language and gender. On the one hand, she reminds us about some myths and stereotypes about male and female talk such as ‘women speak more than men’ and ‘men are more confrontational’. These fall under the heading of the ‘Men are from Mars, women are from Venus’ myth after the book of the same name by John Gray (1992). As Eberhardt outlines – there are actually far more similarities than differences in male and female talk. Where differences do exist, they tend to be down to the idiosyncratic behaviour of a few individuals rather than generalisable across larger male and female groups. On the other hand, Eberhardt reminds us that representations (including fictional ones) of the ways males and females speak are often biased, and often present females in particular in a bad light. For example, she quotes a study that examined the verbs used to describe the speech of male and female politicians in Canada which found those used to describe the speech of females were far more negative.
Eberhardt also suggests that children’s literature is a particularly interesting genre for the study of gender representation since the ways it portrays male and female characters may influence how young readers see the world. This may, in turn, reproduce crude gender stereotypes such as the ideas that female characters are weak and emotional, and that male characters are strong yet lack emotion.
Eberhardt’s study builds on previous work by Sally Hunt (2015) on gender representation in the Harry Potter series. Hunt found there were some differences in how actions of male and female characters were presented, and discovered that, generally, female characters were presented as more passive than their male counterparts.
Eberhardt analyses all instances of direct speech attributed to Ron and Hermione (including instances where the pronouns ‘he’ and ‘she’ are used to introduce the reporting verb). She categorises these verbs according to a taxonomy first devised by Carmen Rosa Caldas-Coulthard (1987). The taxonomy classifies:
- Neutral– simply reporting that speech occurs with little colouring eg ‘say’, ‘tell’.
- Structuring– building and developing a dialogue, eg ‘ask’, ‘inquire’.
- Metapropositional– showing degrees of force which include assertive. eg ‘suggest’, ‘correct’; directive e.g. ‘urge’ ‘instruct’; expressive ‘accuse’ ‘complain’.
- Metalinguistic– relating to the act of telling, eg ‘narrate’ ‘quote’.
- Prosodic– describing pitch, loudness or duration, eg ‘scream’.
- Paralinguistic– showing the manner of speech, eg ‘whisper’, ‘sob’.
- Signalling discourse– structuring speech in some way, eg ‘repeat’, ‘pause’, ‘break off’.
1. Groups of verbs
- Ron has more reporting verbs than Hermione (2,154 and 1,937 respectively).
- The characters have almost the same number of reporting verbs of different types based on Caldas-Coulthard’s taxonomy (80 for Ron, 82 for Hermione).
- Ron has more neutral verbs than Hermione (77% compared to 73%).
Eberhardt argues that the greater number of non-neutral reporting verbs for Hermione demonstrates that Rowling colours and mediates readers’ possible interpretations of Hermione’s speech more than she does Ron’s. Eberhardt uses the term ‘manufactured’ (2017: 241) to describe Hermione’s speech, arguing that this greater focus on how she speaks follows a trend in popular media and fiction to specifically focus on how females look, speak and act.
Eberhardt also uses statistical tests to show that Ron has more expressive reporting verbs than Hermione but that overall there is little difference in the way characters are represented with their use of any particular type of verb. Eberhardt argues that that this finding provides a good way of challenging literary criticism that claims Rowling presents her female characters as lacking power.
2. Unique verbs
Eberhardt finds a different pattern in her exploration of the unique reporting verbs used by Rowling for Ron and Hermione. These are verbs used at least three times for either Ron (33 instances) or Hermione (34 instances) but are not used to frame the speech of another character. Interestingly, Hermione’s unique reporting verbs are related to a high-pitched voice (such as ‘scream’, ‘squeal’, ‘shriek’ and ‘squeak’) or show degrees of sadness and/or helplessness (such as ‘wail’ and ‘whimper’). Ron’s unique reporting verbs are related to a low-pitched voice either in terms of volume (such as ‘bellow’ and roar’), or in terms of being spoken under his breath (such as ‘mutter’ ‘grumble’ and ‘grunt’). Eberhardt suggests these show emotional disengagement. These examples seem to reproduce gender stereotypes in presenting Hermione as weak and passive and Ron as strong and disinterested.
3. Verbal modification
Finally, Eberhardt explores how each character’s speech is modified through the use of adverbials– words such as ‘loudly’ or ‘quietly’ that tell us more about how speech is performed. Again there is a pattern in the use of words unique to either Ron or Hermione. Modifiers of Hermione’s speech tend to be related to ones such as ‘shrilly’, ‘briskly’, ‘timidly’ and ‘sadly’ which are related to feelings, whereas modifiers of Ron’s speech are largely related to what he knows rather than what he feels. These include ‘incredulously’ and ‘doubtfully’. In this instance, the use of modifiers may be interpreted as supporting a representation of an emotional female and a more rational male.
Overall, Eberhardt’s article shows some interesting similarities and differences in the way the speech of Ron and Hermione is presented, and her corpus-based approach means she is able to identify patterns across long stretches of text rather than just from a few extracts. As Eberhardt claims at the end of her article, patterns of words across a whole novel may well reinforce binary notions of gender and associate male and female characters with certain traits and behaviours. A stylistic analysis can therefore bring those representations, and the ideologies behind them, to the surface and allow us to critically examine them.
Caldas-Coulthard, R.C. (1987) ‘Reported speech in written narrative texts’, in M. Coulthard (ed.) Discussing Discourse. Birmingham: University of Birmingham Press, pp. 149-167.
Gray, J. (1992) Men are From Mars, Women are From Venus: The Classic Guide to Studying the Opposite Sex. New York, NY: Harper Collins.
Hunt, S. (2015) ‘Represenations of gender and agency in the Harry Potter series’, in P. Baker and T. McEnery (eds) Corpora and Discourse Studies: Integrating Discourse and Corpora. New York, NY: Palgrave Macmillan, pp. 288-284.
This is a digest of the following article:
Eberhardt, M. (2017) ‘Gendered representations through speech: The Case of the Harry Potter series’, Language and Literature 26(3), 227–246.
Using this in teaching and learning
You can use this research digest to support your teaching of AQA’s AS and A-level Language and Literature specification. Below are some suggestions.
Examine representation of male and female characters
Students could examine the representation of male and female characters in the novel they are studying (Imagined worlds, A-level; Views and voices, AS-level) Questions could include:
- Are there patterns in the ways the speech of characters is presented through the use of direct speech reporting verbs and modifiers?
- How does this relate to gender and its representation in the novel as a whole?
- Does this representation change across the novel at all? For example, is there a difference between the presentation of a character’s direct speech at the beginning and the end of a novel? If so, why might this be?
- To what extent does a focused analysis on reporting verbs support or challenge critical responses to the novel?
Conduct an A-level NEA investigation
For their NEA investigation (Making Connections, A-level), students could explore the use of speech verbs in the representation of gender in a literary and a non-literary text. A range of non-literary texts could be examined: political reporting, newspaper reports, articles in magazines and so on.
Examine presentation of gender for a younger readership
Students could take an example of children’s or young adult (YA) literature and examine how the text presents gender for a younger readership (see also our blog post on Peter Stockwell’s article The positioned reader). This would provide a good opportunity for students to consider the merits of choosing a children’s/YA literary text for their NEA investigation (Making Connections, A-level) together with some non-literary material that represented children and gender in particular ways.
Use CLiC to explore data
Students could use corpus software to explore patterns in larger stretches of data and across whole texts. The CLiC.bham.ac.uk tool is a good place to start. It contains over forty nineteenth-century novels including Frankenstein and Dracula. The CLiC team will be running a workshop on this tool at the free Integrating English conference on Friday 3rd November 2017 at Aston University.
1. Baker, Paul (2014) Using Corpora to Analyse Gender. London: Bloomsbury.
This offers a very useful introduction to both corpus methods and the concept of representation, specifically in the study of gendered discourse. It also includes chapters analysing the language use of male and female academics, male bias in gendered paired words (eg man/woman), the representation of gay people in articles in a middle-market tabloid and representation and gender in personal dating ads.
2. Leech, Geoffrey and Short, Mick (2007) Style in Fiction: A Linguistic Introduction to English Fictional Prose, 2nd edn. London: Longman.
This is the second edition of a classic text on the language of prose fiction. Chapter 10 offers a seminal introduction to the presentation of speech and thought.
3. Stephens, John (1992) Language and Ideology in Children’s Fiction. London: Longman.
Stephens’ book is a fascinating study of how literature written for children projects ways of seeing the world and themselves as individuals within it. It argues that writing for children necessarily involves presenting (often indirectly) clear messages about how to behave, think and act.
4. Sunderland, Jane (2010) Language, Gender and Children’s Fiction. London: Continuum.
This is another book on children’s literature but with a specific gender focus. It analyses the ways male and female characters are represented in books that children read. Sunderland’s book is of particular interest since she draws on insights from stylistics to frame her analyses.