Women/ Rabbit rabbit rabbit women/ Tattle and titter/ Women prattle/ Women waffle and witter/ Men talk. Men talk.
These are the opening lines of ‘Men Talk’, a rap poem by the incomparable Liz Lochhead (you can watch her performing the whole thing here). It’s built around the familiar lexicon of disparaging terms for women’s speech: words like ‘rabbit’, ‘prattle’ and ‘witter’, which represent women’s talk as excessive, trivial and inane; and words like ‘gossip’ and ‘nag’, which represent it as malign and spiteful.
But those words are only the tip of the iceberg. If you look at the way the act of speaking is described in everything from news reports to Great Literature, you’ll soon discover that it’s persistently represented in stereotypically gendered and sexist ways.
The most neutral way to describe the act of speaking is by using the generic verb ‘say’. ‘X said’ is the reported speech equivalent of Lochhead’s ‘men talk’: it conveys no more than ‘this person uttered these words’. But writers often add colour by choosing something a bit less basic. Here’s an example, from a political sketch that appeared in the Telegraph after the second TV debate of the 2015 General Election campaign.
“Ed Miliband is scared to be bold,” scowled Ms Sturgeon. “We don’t want a pretend alternative to austerity.” “Exactly right!” squeaked Ms Bennett.
“Labour are letting the Tories off the hook!” snapped Ms Wood. The audience applauded.
Desperately Mr Miliband tried to steer the debate back to his absent foe. “Let’s not pretend there’s no difference between me and David Cameron,” he said, rather pleadingly. “There’s not a big enough difference!” barked Ms Sturgeon.
Notice that it’s the only male participant in this exchange, Ed Miliband, whose contribution is reported using the basic ‘said’ (though the writer does add some extra information with the adverbial ‘rather pleadingly’). The three women, by contrast, don’t just ‘say’ things, they ‘scowl’, ‘squeak’, ‘snap’ and ‘bark’.
These verbs aren’t literally describing how the women sounded. They’ve been chosen to help the writer tell a story, in which a hapless male is ganged up on and berated by three angry and aggressive females. If we only had the speakers’ own words to go on, we might not make that interpretation: we’re directed to it mainly by the writer’s choice of verbs (‘scowl’, ‘snap’, ‘bark’) and adverbs (‘desperately’, ‘pleadingly’). The verbs also say something about the power dynamic among the women. Whereas ‘squeaked’ casts Natalie Bennett as a small animal, ‘snapped’ and ‘barked’ suggest bigger beasts.
This example isn’t unique. When Elisabeth Gidengil and Joanna Everitt examined TV coverage of the 1993 Canadian election campaign, in which two of the five parties were led by women, they also found a tendency for men’s words to be reported with the plain and non-committal ‘he said’, while women’s were described in more elaborate, less neutral terms. Among the verbs which were only used about the women party leaders, and never about their male opponents, were ‘argue’, ‘blast’, ‘fire at’, ‘hammer away’ and ‘launch [an attack]’. There were also verbs, like ‘accuse’, which were sometimes applied to men, but occurred more frequently in relation to women. The women, in short, were systematically represented as more verbally aggressive than the men.
The researchers did consider the possibility that the women really were more aggressive, but when they analysed the five leaders’ actual speech they found no evidence to support that. The real difference, they argue, is in the way male and female speakers are judged: if women deviate from stereotypical expectations by presenting themselves as tough rather than gentle, combative rather than conciliatory, they are perceived as more aggressive than men who behave in exactly the same way. That perception, Gidengil and Everitt suggest, explains why female leaders’ speech is reported using more aggressive verbs of speaking. The contrast between ‘he said’ and ‘she blasted’ is an explicit encoding of the underlying double standard.
Do creative writers rise above these journalistic clichés? Not according to Ben Blatt, who analysed a large sample of literary and popular fiction for his book Nabokov’s Favorite Word Is Mauve. His number-crunching revealed, among other things, that male fictional characters habitually ‘mutter’, ‘shout’ and ‘chuckle’, while female characters ‘murmur’, ‘scream’ and ‘weep’. Other patterns were influenced by the sex of the author as well as the character. Male authors were more reluctant than female ones to allow their male characters to ‘sob’; and in their books it was usually female characters, not male ones, who ‘interrupted’.
This particular finding caught many readers’ attention because it’s the opposite of ‘realistic’ in the everyday sense of the word (in real life women do not interrupt more than men). But gender stereotyping can function as a ‘realist’ device in the more technical sense. Even if a stereotype is factually inaccurate, if it fits with readers’ common-sense beliefs it can help to make a fictional world believable.
As feminists have often pointed out, though, our beliefs about men and women are not just things we bring from our real-life experience to our reading; they are also things we may get from our reading and take back into the non-fictional world. I was reminded of this recently when a colleague told a story about a conversation she’d had with her children, who were insisting that ‘mummies don’t go out to work’. ‘Where’, she asked them, ‘does daddy drop me off every day when he’s taking you to school?’ They answered without hesitation: ‘work’. But knowing their own mother went out to work hadn’t prevented them from absorbing the stereotype of mothers as stay-at-home parents.
Concern about what children might be absorbing from the books they read has prompted a number of feminist researchers to analyse the language used in children’s fiction—including the words used to describe characters’ speech. One researcher, Sally Hunt, analysed the verbs of speaking used in some of the most successful children’s book series of the last 75 years: Enid Blyton’s Famous Five adventures, C.S. Lewis’s Narnia Chronicles, Francine Pascal’s Sweet Valley High series and J.K. Rowling’s Harry Potter books. And despite the differences of period, genre and target audience, she found there were patterns which recurred across the sample.
One of these patterns related to the distribution of verbs which tell you what action an utterance is performing. Verbs suggesting authority were more typical of male characters, and verbs suggesting deference were more typical of female ones. ‘Ordered’, for instance, was 77% male, whereas ‘begged’ was 68% female. Some actions, like ‘giggling’, were off-limits to male characters, while others, like ‘boasting’, were virtually an all-male preserve.
Hunt also remarks on authors’ fondness for verbs which allude to the vocal qualities of speech. In her sample, male characters’ verbs often implied low pitch (e.g. ‘he bellowed/roared’) whereas female characters’ verbs emphasised high pitch (e.g. ‘she shrieked/screamed’). It’s interesting that this contrast features so prominently in books for and about children, because the physiological differences which produce it in adults do not develop until puberty. But like the ‘squeaking’ and ‘barking’ attributed to female politicians in the sketch I quoted earlier, words like ‘bellow’ and ‘scream’ are rarely intended just to evoke the sound of the speaker’s voice: they are also code for emotional responses like anger, surprise and fear. Associating these words with either male or female speakers is thus a covert way of telling readers which emotions are typical of each sex.
The representation of male and female speech in J.K. Rowling’s Harry Potter books has been investigated further by Maeve Eberhardt, who approaches the question through a detailed comparison of two characters, Hermione and Ron. As Eberhardt notes, Hermione is widely considered a ‘strong’ female character: Rowling herself has described her as the ‘brightest’ of the three friends, and she is also portrayed as morally courageous. But are her intelligence, strength and courage reflected in descriptions of her verbal behaviour?
Eberhardt’s answer is ‘yes and no’. In some respects, she finds that Hermione and Ron are treated similarly. Across the series as a whole they are given approximately equal amounts of reported speech, and the neutral ‘said’ is by far the commonest verb of speaking for both of them. The number of other verbs used to describe their speech is also approximately the same.
But the verbs themselves are not the same. When Eberhardt examined ‘unique’ verbs—those which were only ever used about one character—she found that Hermione’s tended to imply strong emotions, especially fear and sadness (they included ‘screamed’, ‘squealed’, ‘shrieked’, ‘squeaked’, ‘wailed’ and ‘whimpered’). Ron’s unique verbs, conversely, included a number (such as ‘mumbled’, ‘grumbled’ and ‘grunted’) which suggested emotional disengagement. The two characters were also distinguished by the frequency with which certain verbs were used about them. Both of them ‘whisper’ and ‘gasp’, but Hermione does those things about three times more often than Ron. He, on the other hand, does five times as much ‘muttering’ as she does, and over fifteen times as much ‘yelling’.
Eberhardt also looked at the use of adverbials to modify verbs of speaking (as in ‘he said, rather pleadingly’). Since Hermione is meant to be the clever one, you might expect her adverbials to include a high proportion relating to intellectual or mental states (e.g. ‘thoughtfully’, ‘logically’, ‘sceptically’). But in fact most of them are about her feelings: her unique adverbs do include ‘seriously’, but that occurs less often than either ‘timidly’ or ‘sadly’. And the most frequent of the adverbial modifiers which are only applied to Hermione’s speech is that old sexist cliché ‘shrilly’.
Since this study only compares two characters, it might be argued that the patterns it uncovers have less to do with their gender than with their distinctive qualities as individuals. What the reported speech verbs tell us is simply that Ron is the kind of person who mutters and grumbles, while Hermione is the kind who wails and shrieks. But I don’t think that argument will wash, given that other studies, like Sally Hunt’s and Ben Blatt’s, have found the same general patterns, and even some of the same specific word-choices, in a range of other texts. Generations of male fictional characters have expressed themselves by muttering and bellowing, while their female counterparts have screamed and spoken ‘shrilly’. These verbal and vocal habits could not be less individual, or more gender-stereotyped.
They are also remarkably persistent. A children’s writer starting out today wouldn’t be able to build a successful career on stories like the ones I read as a child, in which boys had adventures and girls helped mummy make the sandwiches. That kind of sexism is much less common now. Yet successful writers can still present children with a linguistic division of labour –boys giving orders and girls asking questions, boys bellowing and roaring while girls scream, squeal and giggle—that doesn’t seem to have changed since the 1950s.
I’m not accusing these authors of deliberately reproducing stereotypes. I’d be surprised if they had any conscious awareness of the patterns revealed by analyses of their work. But if we accept that those patterns both reflect and perpetuate sexism, perhaps we should be challenging writers to make a conscious effort to break away from them.
For those who want to avoid sexist clichés, whether in fiction or journalism, the research I’ve discussed suggests several top tips:
- Check you’re not consistently pairing minimal descriptions of male speech (‘he said’) with highly elaborate descriptions of female speech (‘she enunciated crisply’/ ‘she gasped in horror’).
- Go easy on the vocalisation verbs (like ‘growled’ or ‘squeaked’) which differentiate male and female speakers overtly by pitch and covertly by emotional state. And you’re going to use them, don’t make a habit of picking more ‘extreme’ ones for female speakers (if a boy ‘shouts’ or ‘yells’, why does a girl have to ‘shriek’?)
- Try not to give all the ‘thinky’ verbs to male speakers and all the ‘feely’ verbs to female ones.
- Watch out for the speech act trap–don’t let conversations be all about male speakers ‘asserting’, ‘instructing’ and ‘explaining’ while female ones ‘ask’, ‘suggest’ and ‘agree’.
The way I’ve phrased these points (‘go easy on X’, ‘don’t make a habit of Y’) is deliberate: they aren’t meant to be blanket prohibitions. As I’ve said a million times on this blog, context is all: any word–even ‘shrilly’–can be the right word if the context calls for it. What you want to avoid is not specific words, it’s the kind of regular pattern that results from the habitual, unthinking repetition of the same stereotypical formulas.
Precisely because we’ve encountered them so often, phrases like ‘he muttered’ and ‘she murmured’ or ‘he yelled’ and ‘she screamed’ may seem obvious and ‘natural’; but really there’s nothing natural about them. On the contrary, they are products of our cultural obsession with magnifying gender differences, or imposing them where they don’t exist. In reality, men and women use language to perform the same acts and express the same emotions. Girls give orders and boys make suggestions; women chuckle and men have been known to scream. If we can cope with this complexity in our face-to-face encounters, why can’t it be reflected in our descriptions of the way people talk?