Over the years I have regularly been driven to distraction by the habit, especially among adults talking to young children, of referring to virtually all non-human animals as “he”. It’s the “default male” principle in action: if the sex of a living being is unknown or unspecified we automatically treat it as male. That impulse is so strong, I’ve often observed it being acted on in cases where the animal was unmistakably female—a lioness, for instance, whose lack of a mane distinguishes her clearly from the male members of her species, or a garden bird whose plumage does the same. Among the birds that frequent my own garden is one we jocularly call “the blackbird’s wife”: that appellation also treats the male blackbird as the default, but at least it acknowledges that female blackbirds (which are, of course, brown) exist.
So, I was excited when I saw that the data-crunchers at The Pudding had been looking into some of the details of this pattern. Specifically, they’d been trying to ascertain whether the default maleing of animals is more likely to happen with some species than others, if so which species those are, and what that might tell us about the logic behind it.
My own thoughts about this are based on randomly observing, over a period of many years, ordinary people’s interactions in public places like zoos and parks—an approach that produces only anecdata, since the observation is unsystematic and the sample cannot be controlled. The people at The Pudding are made of sterner methodological stuff: they decided to analyse data from children’s books that feature animals. Their sample comprised around 300 titles that were highly rated on Goodreads. There were around 30 items for every decade since 1950, plus some still-popular classics published before that date. To qualify for inclusion a title had to mention a species of animal that was also mentioned in at least nine other items, since a single book about, say, Judy the Unicorn cannot tell you if there’s a general tendency to imagine unicorns as female.
The researchers were aware that the patterns found in this sample would reflect authors’ conscious decisions, and that these might differ from the more spontaneous decisions made by ordinary speakers. So they also ran an experiment in which 1300 people were asked to continue a story that began
and then the [animal] said “I must go to the river”. Upon arriving…
What appeared in the [animal] slot was randomly selected from a list of seven animals that commonly appear in children’s fiction: bear, bird, cat, pig, duck, mouse and dog. The question was whether the frequency with which respondents chose he, she or it as the next word would vary depending on which animal their prompt referred to.
So, what did the researchers find? In their sample of published children’s books it turned out that most species skewed male: the only ones that didn’t were birds, ducks and cats—and I confess to being slightly surprised by the cats. Though it’s a cultural cliché that dogs are boys and cats are girls, the iconic fictional felines of my own youth were all male: they included the cartoon characters Top Cat, Felix, Sylvester and Tom (in Tom and Jerry), Orlando the Marmalade Cat, and the pantomime stalwarts Puss in Boots and Dick Whittington’s cat (which is female in some older versions of the story, but is always male in the panto). Maybe that reflects another pattern that’s very clear in the children’s fiction sample, where male animals overall were twice as numerous as female ones. Popular stories about both humans and animals are more likely to feature male central characters–in part because of the publishing industry’s belief that boys won’t read books featuring female protagonists. But in the case of animals the bias seems to be even stronger. And it was stronger still in the responses to the story completion task, where none of the seven species presented skewed female. He was selected three times more often than she, despite the fact that women respondents outnumbered men. There were also some uses of ungendered forms such as “it”, occasionally “they”, or a noun phrase like “the bear”, but these were all less common than he.
What about the gendering of different species? In the fiction corpus the “malest” animals were frogs, wolves, foxes, elephants, dogs, monkeys and bears, in that order, while the least male were the aforementioned birds, ducks and cats; pigs and mice could swing either way. In the experimental story-completion task, where only seven species were in the mix, the malest animals were bears, followed by mice and dogs, while the least male were cats and then birds—though the birds still skewed slightly male and the cats were right on the dividing line (i.e., as likely to be male as female).
This species-based hierarchy of maleness does not appear to follow any single logic. You look at the list and think “OK, so big and/or fierce animals like elephants, wolves and bears are male—that kind of makes sense”; but then you think, “in that case why are frogs male, in fact the malest of the male?” Can that maybe be put down to the influence of a small number of canonical male frogs that have come to serve as our mental prototypes, like the fairytale Frog Prince, the Froggy Who Would A-Wooing Go and Kermit from the Muppets? It sounds vaguely plausible, but you have to wonder why it doesn’t work for cats, who remain at least somewhat female in the cultural imagination even though so many canonical cats are male.
Also, what is going on with pigs? Analysing the fiction sample decade by decade revealed that they used to be solidly male (like the Three Little Pigs, or Wilbur in the children’s classic Charlotte’s Web), but since the turn of the millennium there has been an upsurge in stories where sows take centre stage (think Olivia and Peppa). Clearly, pigs as a species do not have the metaphorically feminine qualities which are often said to explain the perceived femaleness of birds and cats, namely that they are small, delicate and fast-moving. Adult pigs are large, heavy creatures who eat swill and enjoy wallowing in mud. That could help to explain their historical maleness—but what explains the recent turn to femaleness? Is this a case of authors deliberately “flipping the script” in an effort to subvert traditional stereotypes?
Another question the researchers pose is whether language itself influences the gendering of animals. In languages that mark gender grammatically, are animals more likely to be perceived and represented as female if the word for them is grammatically feminine? In many cases this question seems wide of the mark, because animals, like people, can be referred to using either masculine or feminine forms depending on the individual’s sex: a Spanish cat, for example, is either a gato or a gata. But there are some animals that doesn’t apply to: according to The Pudding a Spanish frog is always a rana, and frogs are also more likely to be female in Spanish stories than they are in English ones.
In modern English, where we no longer have grammatical gender but we do still have paired male and female terms for some animals, we might wonder if it makes any difference whether the more generic term, the one we’d use to talk about the whole species, is also the female member of the pair (as with duck/drake and goose/gander) rather than the male one (as with fox/vixen). Are we more inclined to think of ducks as female because female ducks, unlike female foxes, are the unmarked case linguistically?
These are not uninteresting questions, but in the end I don’t think analysing data from children’s stories (a category in which I include The Pudding’s experimental responses, since the prompt they used was obviously modelled on that genre) can explain the phenomenon I began with—the default male-ing of live animals in real-world settings. Though there are some exceptions (books whose aim is to educate children about the natural world), most animals in kids’ stories are basically human characters in nonhuman form. They live in houses, wear clothes and deal with situations familiar to human children (e.g., Peppa Pig’s fear of spiders or Olivia’s irritation when her little brother copies her). They often have families which mimic human families, and this makes it likely that even if the central character is male, there will be female animals of the same species in supporting roles (like Peter Rabbit’s mum, Babar the elephant’s wife Celeste or Paddington Bear’s Aunt Lucy).
These animal characters are gendered, using the same conventional signifiers of masculinity and femininity, such as names, clothes, hairstyles and stereotyped personality traits, that would be used to gender them if they were human. Though they are also by implication sexed, direct signifiers of sex are typically absent: Paddington Bear, for instance, does not to my knowledge have a penis. He also doesn’t act much like a bear. The values he represents and the lessons his creator set out to teach (e.g., kindness to strangers) are human, not ursine.
By contrast, the animals we see in nature documentaries are represented as sexed but not gendered. When a bear emerges from hibernation with a couple of cubs, David Attenborough or whoever will refer to this animal as she, not because her appearance or behaviour is “feminine”, but because giving birth to cubs is something female bears do and male ones don’t.
Yet when ordinary people interact casually about animals they have seen or are seeing “in the flesh” (always provided that those animals are not already known to them as, say, their own or their acquaintances’ pets), their approach to the sex/gender question is neither the one they’ve seen modelled in nature programmes (in that they frequently ignore very obvious signifiers of female sex like the lioness’s lack of a mane) nor the one that’s familiar from folk-tales and children’s stories (anthropomorphizing animals, which also means gendering them in much the same ways we gender humans). They just “naturally” seem to gravitate towards masculine forms.
Does that mean, though, that they actually think of the animals as male? That’s not, IMHO, an easy question to answer. It’s at least possible that something else is going on—something hinted at by an observation made long ago by the feminist theorist Monique Wittig. “There are not two genders”, said Wittig, “there is only one: the feminine. …For the masculine is not the masculine but the general”.
Her point is illustrated in some studies of the way people gender human referents. For instance, in 2016 and 2017 a group of researchers investigated what pronoun English-speakers in the US and Britain would choose to complete a sentence about “the next US president” or “the next UK prime minister”. Though historically and culturally the terms president and prime minister are strongly associated with maleness/masculinity (there has never been a female US president, and by 2017 only two British women had ever served as prime minister), this research was conducted in the run-up to two national elections which women were expected to win. In the US it was generally (though wrongly) thought that Hillary Clinton would defeat Donald Trump; in the UK it was (correctly) predicted that Theresa May, the incumbent Conservative PM, would win another term in government. The researchers’ question was whether respondents’ expectation that a woman would be “the next president/prime minister” would influence their pronoun choice—that is, make them more likely to pick she over the more traditional and stereotypical he when referring to the next holder of the role.
The answer turned out to be “yes and no”. The researchers did find that in the course of each campaign, rising expectations of a female victory correlated with a reduction in the frequency with which respondents chose he; but that didn’t prompt a corresponding increase in the frequency of she. Even when the belief that Clinton would win was at its height in the US, he remained more frequent than she. In the UK, the belief that May would be the next prime minister favoured the use not of she, but of the gender-neutral singular they. When the experiment was later repeated during the 2020 mayoral election in Paris, where by the second round all the male candidates had been eliminated, the pattern was much the same: most respondents chose either masculine or gender-neutral expressions to refer to “the next mayor”, despite knowing with certainty that the person in question would be a woman.
This year a study was published which investigated the use of gendered words (both nouns and pronouns) in children’s TV programmes broadcast in the US between 1960 and 2018 (this was a large dataset, comprising more than six thousand episodes of 98 different programmes). The researchers found that in the sample overall male words were used almost twice as frequently as female ones. They also found that over time this gap had narrowed (though by the late 2010s male words were still one and a half times as common as female ones). But that wasn’t because the use of female words had increased. On the contrary, in fact, it had decreased—but the balance had shifted because the frequency of male words had fallen even more steeply. There had been a shift away from using gendered terms in general, which had a stronger effect on the frequency of male terms because they were more numerous to begin with.
Though second-wave feminist campaigners for nonsexist language often argued that avoiding the generic masculine would represent reality more “accurately”, that doesn’t seem to be the principle on which most ordinary language-users operate. They apparently take feminine forms to signal a commitment both to the referent’s femaleness and to the relevance of its femaleness, while being willing to use masculine forms without the same commitment to maleness. In effect they do treat the masculine as representing “the general”: in contexts where sex/gender isn’t at issue they will typically reach for either masculine or neutral forms, while avoiding feminine ones.
That could shed some light on why the preference for he seems to be even stronger in talk about nonhuman animals. Someone who points out a squirrel running along a garden wall, a rat scuttling down an alley or a frog jumping out of a pond is unlikely to have any firm beliefs about its sex, or to regard that as a contextually relevant question (in this type of interaction the message is basically just “Look, there’s an [animal]!”). Any pronoun they go on to use is just a placeholder. But he is evidently felt to be a more “natural” candidate for that function than she.
So, I’m suggesting that a speaker’s choice of he might not always reflect their belief that an animal is male. But that doesn’t mean the pattern is not a problem: it’s still reinforcing the more general asymmetry Wittig criticized. Even if it feels unnatural or silly, I think there’s something to be said for using, at least sometimes, unambiguously female terms to refer to the squirrel or the rat. Not because it’s more accurate (unless you can sex a squirrel from a distance you won’t know if it’s accurate, any more than people know if the male terms they typically choose are accurate), but to challenge the habit of equating the male of the species with the species as a whole. (And if you’re thinking “wouldn’t the best solution be to use neutral terms?” I’ll reiterate a point I’ve made before, that formally neutral terms are not automatically “inclusive”. Sometimes they give only the illusion of inclusion, while in practice doing little to counter male bias.)
As for the more specific phenomenon investigated by The Pudding, the gendering of animals in children’s stories, there are many reasons to see that as a problem—in fact, a version of a more general problem with the way fictional characters are gendered. The main focus of the children’s TV study I cited earlier wasn’t the frequency of male and female words, but rather their stereotypical associations: they found a tendency for male words to be associated with agentive roles (they were more often found in the grammatical subject position in sentences, and they typically went along with words denoting actions). Female words were more associated with references to appearances, relationships and emotional states. Since this kind of sex stereotyping (men act, women appear/are acted upon) is pervasive in cultural products of all kinds (on this blog I’ve written about it in relation to the verbs of speaking used in both journalism and fiction, for instance), I’d be amazed if it didn’t also turn up in stories about animals. And since animals in children’s fiction mostly are just humans in disguise, it’s as much of a problem there as in cases where the characters are people.
Language on its own is not (at least IMO) responsible for the kind of bias I’ve been discussing, but analysing language can help to make us conscious of it. And if we are conscious of it we can make an effort not to perpetuate it—especially in what we teach our children about the world.
language: a feminist guide 