WHY is one French noun masculine and another feminine? This question has puzzled generations of students who have studied French. The sun and moon rise and set in similar ways, yet one is masculine and the other feminine. The words “hand” and “fist” essentially refer to the same body part, yet one is masculine and the other feminine. But which is which? And why? Lacking any obvious explanation of this seemingly arbitrary phenomenon, we sometimes resort to poking fun at it: “Is ‘computer’ masculine or feminine?” The answer, of course, is either – with perfectly plausible reasons for both.
For readers without any experience of gender in French or any other European language (and gender is also present in various African and Australian Aboriginal languages), the phenomenon can best be described as a systematic way of organising nouns into a limited number of subsets. It might surprise you to know that Old English was also a gender language, and has left us with such remnants as singular/plural contrasts (house is/houses are, this/these/those) and a complex pronoun system (I/my/mine, his/her/its/their) that involves cumbersome “agreements” requiring certain changes particularly to verbs. As native speakers, though, we don’t find any of this difficult.
People who aren’t fluent in French are largely unaware that in some cases – après-midi (afternoon) and couple (couple), for example – speakers are free to choose which gender they use. Nor do they realise that French dictionaries don’t always agree on gender: pamplemousse (pomelo, grapefruit), for instance, is feminine in some sources (Académie française dictionaries, writers), and masculine in others (Larousse dictionaries, botanists), while Dictionnaires Le Robert offer alternative genders in their Petit Robert but masculine only in Le Robert pour Tous. Some nouns – orgue (organ), for example – change gender from singular to plural. And not even native speakers agree on the gender of some nouns, as an informative post from the HeiDeas blog reveals:
Fifty-six native French speakers, asked to assign the gender of 93 masculine words, uniformly agreed on only 17 of them. Asked to assign the gender of 50 feminine words, they uniformly agreed [on] only 1 of them. Some of the words had been anecdotally identified as tricky cases, but others were plain old common nouns.
It’s not that native speakers don’t know the gender of nouns, says HeiDeas. It’s just that they don’t agree. And then there is the challenge of synonyms with different genders – river can be rivière (feminine) or fleuve (masculine) – for which French is infamous. As for words borrowed from other languages, let us not venture far into that quagmire. Most are masculine – but not all.
Unfortunately, agreement principles in French are even more complicated – and none more so than those relating to gender. English speakers often recite the French nonsense expression la plume de ma tante (the pen/quill of my aunt) to indicate some familiarity with French gender and its principles of agreement. We readily recognise the relationship between ma, a feminine agreement form, and tante, a female person (just as we would recognise mon oncle as “my uncle,” a male person). But the precise relationship between la (a feminine agreement form similar to ma) and plume is not clear. What it does indicate is that, while “gender” and “sex” may be used interchangeably in English, in French “feminine” is not equivalent to “female,” and “masculine” is not the same as “male.”
My book Gender Assignment and Word-Final Pronunciation in French: Two Classification Systems sets out a radical explanation of how gender works in French. The answer, I argue, lies in contrasts in meaning, or binary opposites, where one part is associated with masculine and the other with feminine. We are used to these either/or opposites because we use them all the time in everyday life: dead or alive, dark or light, open or closed, full or empty, odd or even. My explanation fits an extensive vocabulary of French nouns that deal with the natural world and beyond. It produces only one single exception – to which we’ll return – for which gender is determined not by meaning, but by the presence of adjectives.
Dead:alive opposites form important semantic distinctions associated with contrasting genders. Things that are “dead” – cadavers, corpses, wood – are masculine, as are things that never existed or no longer exist, including mirages, shipwrecks, dinosaurs and recently extinct species such as aurochs (ancestor of our domestic cattle, extinct in 1627) and tarpan (a wild Eurasian horse, extinct for around a century) plus several extinct birds, including the great auk, the moa and the dodo. This correlation between “extinct” and masculine gender is supported by the fate of the North American passenger pigeon, whose name changed from the feminine tourtre to the masculine pigeon migrateur following its extinction in 1914.
“Alive” is associated with feminine gender, although in quite surprising ways. (It would not be a particularly useful distinction if all living things were feminine.) It becomes significant in contrast to “dead” for words such as “recruit” (a living body that makes up numbers lost through death or injury), and “victim” (originally a living body offered for sacrifice). “Alive” is also crucial where appearances suggest otherwise. For example, fish typically turn on their sides when they are dead; but rays, skates and “flatfish,” feminine in French, swim in this side-on position and are very much alive.
In one area only, gender distributions have always been plausible, regular and uncomplicated – sex. “Male” and “female,” that is, because nouns that indicate the “male” of a species are masculine (for example, “buck,” “bull,” “drake”) and nouns that indicate “female” are feminine (for example, “doe,” “cow” and “duck”). Learners can use these principles to generate the correct gender – for animal and bird species, at least. (For humans they are less consistent.) The basis on which we come to apply these “male” and “female” terms for each species is another matter, particularly for birds; as we shall see for pairs of bird species, we rely on different aspects of appearance and certain qualities associated with one sex over the other.
But what happens when we want neither “male” nor “female” but simply the name of the species? Only one of each pair can serve (cow rather than bull, for instance, in English), but which will it be? In this move from sex to species, we discover something remarkable about French gender, and begin to unlock the secrets hidden in this code.
Ducks and drakes
Both English and French have different male and female terms for certain species of birds, though not always for the same species. English makes a distinction between “swan” and “cob” that French does not, while French makes distinctions that English does not, as we can see in this complete list of French terms:
French terms for male and female birds
faisan “male pheasant” | faisane “female pheasant”
paon “peacock” | paonne “peahen”
coq “rooster” | poule “hen”
dindon “male turkey” | dinde “female turkey”
canard “drake” | cane “duck”
merle “male blackbird” | merlette “female blackbird”
jars “gander” | oie “goose”
pigeon “male pigeon” | pigeonne “female pigeon”
sacret “male saker” | sacre “female saker”
Most of these birds we know well. The first four pairs are terrestrial ground-feeders we call “fowls.” We can apply these pairs of terms accurately because of their association with certain differences – size, hairiness/feathering, crests/coxcombs, plumage colouration, even parenting duties. We are less aware of the consequences of these differences. Against the brilliant, colourful, even lustrous plumage of male fowl, the brownish or speckled colouration of females may seem drab, but in their terrestrial habitats it camouflages and protects females and leaves males exposed to greater danger. High visibility is dangerous for daytime ground-dwellers. These differences are not difficult to find, given a polygamous lifestyle in which a single male lives alongside his female harem. The dominant male maintains his position by fighting off other males of his kind; the females lead a more collaborative existence.
Even when birds are out of sight, differences in voice (pitch, amount, purpose) convey male:female distinctions for each species. When they call, males have a loud, raucous voice, for early-morning crowing or gobbling that establishes their territory, for warding off other males or alerting the female harem to potential danger; but they are generally silent. In contrast, the quieter but more constant calls of females maintain a social pecking order, and keep their broods in touch as they engage in the constant search for food. Together, these elements provide a composite that offers constant distinction between male and female. (We use a similar composite to distinguish between “male” and “female” humans – body shape, height, hairiness/beards, hairstyle, dress, voice, gait, and so on.)
Male:female distinctions are more difficult to establish for ducks. Mature males have plumage nuptial (breeding plumage), but only at breeding time, and only for certain species; after moulting, the new growth is much the same colour as for females and immature males. Like terrestrial fowls, female ducks brood their eggs and nurture their young, while male ducks disappear to moult with other males and typically play no part in parenting. But again, such distinctions are only available at a certain time of the year. Duck species are largely silent. Some dabbling duck species quack, but not particularly loudly, and then it is only the females (dommage for Walt Disney’s Donald Duck and his three “nephews”). Other duck species have a wide variety of calls, whistles, coos, grunts and so on, but these are also infrequent.
The more aerial blackbirds are also common ground-foragers. In this species, male and female are dimorphic; that is, they have entirely different plumage colours and were once considered to be different species. Today we recognise the black as male and brown as female (suggesting that the “four-and-twenty blackbirds” in the nursery rhyme pie were an all-male affair). Male blackbirds sing loudly and joyfully, but not constantly, and only for certain months of the year.
For the last three pairs – geese, pigeons, sakers – male and female of each species are so similar that, for the most part, we cannot tell them apart. Mature male geese are larger than females, while female sakers are larger than males, but distance makes this difficult to judge. There is one characteristic that distinguishes females from males of each species: laying eggs, a process in which males have some involvement, but only indirectly. These male and female distinctions for geese, pigeons and sakers remain significant only for breeders. Others, including native French speakers, would simply use the general or “unmarked” term for the species.
From sex to species
The question is, which term for each pair comes to serve as the general term for the species? In English, we typically use the female – with one exception, “peacock” rather than “peahen.” In French it is the male – with two exceptions, oie, the female goose, and sacre, the female saker. Sacre is highly unusual. As its meaning changes from “female saker” to the species “saker,” its gender changes to masculine. This change does not occur for oie.
How can we account for these irregular outcomes? What motivates the extraordinary change in gender for sacre? Since the “generic” sense for each species must now encompass both male and female, we would expect the gender to reflect some characteristic shared by all members of the species, something that humans can easily identify. One characteristic stands out – response to threat or intrusion, a characteristic that humans would immediately notice and exploit for food or sport, or both.
Despite their heavy-bodied frames, fowl are excellent flyers, yet they prefer to run across the ground in response to threat rather than take to the air – putting them at risk from faster (or better armed) predators. Ducks are not wary of humans, and nor are pigeons or blackbirds, whose initial response is to hop or run a short distance, only to take to the air at the last moment, and even then not flying far. Single male blackbirds rely on intimidation to send off other males.
A striking contrast
In contrast, geese collaborate to attack intruders. To guard over the flock, they set sentries whose cry of alarm at the slightest hint of danger alerts the flock. The other geese respond in unison, and the resulting cacophony is sufficient to repel most predators, particularly those depending on surprise. Their loud, repetitive honking helps the flock keep in touch when members spread out to feed, or during long migratory flights.
These adaptations offer a “protective” environment for individuals and the flock. This relationship between “protective” and feminine gender is consistent with female fowls being protected by their colourations. On the other hand, responses to threat for the previous terrestrial and aquatic fowls increase the likely harm for individuals and possibly others in the flock. Again, the relationship between “harmful” and masculine is consistent with the greater danger of bright colourations for male fowls.
For the sacre, the change in gender is tied to a change in meaning. When the noun is feminine, it means “female saker”; but when it becomes masculine, its meaning becomes generic: “saker,” a kind of falcon, one among many diurnal birds of prey that feed in daylight hours rather than nocturnally. Diurnal birds of prey are typically masculine: vulture, condor, various falcon species (merlin, lanner, hobby and peregrine), harrier, hawk, goshawk, sparrow hawk, kite, lammergeier, secretary bird, bald eagle – although four, including osprey, are feminine. The nocturnal set, which we call owls, are mostly feminine; the exceptions are the “horned owls,” whose ear tufts look much like horns. This attribute is strongly associated with male/masculine in the animal world (but not for ruminants that are typically “horned”).
But why would diurnal hunters be masculine and nocturnal hunters feminine? Owls have the same excellent eyesight as diurnal birds of prey, but they have other adaptations: excellent hearing (through feathered facial discs/ear tufts that amplify and locate sound waves), and flight feathers rigged for “silent running” that muffle the clap of beating wings, giving prey no advance warning to take cover. Where diurnal birds of prey are restricted to daylight hunting, owls can extend their hunting time when needs must. The form and habit of these “feminine” birds are flexible and adaptable, qualities that are not present for the inflexible, less adaptable, “masculine” set.
In French culture (and our own), these different male:female terms are restricted to bird species we have kept captive as farmyard birds in keeps or cages, for food, sport, communication and/or profit. For non-domesticated birds, we only need a generic term. (This pattern is also found for other animal species more generally.)
It is not by chance that male:female terms deal largely with species that we have domesticated and have a vested interest in breeding. For all other animals, in French as in English, we simply add mâle/male or femelle/female to the generic term to indicate a specific sex. In French, the gender of these nouns is determined by the species, not by the sex, as we can see for the following two species, one masculine (M), the other feminine (F):
éléphant (M) elephant
a male elephant: un éléphant mâle (M)
a female elephant: un éléphant femelle (M)
girafe (F) giraffe
a male giraffe: une girafe mâle (F)
a female giraffe: une girafe femelle (F)
For any domesticated “farmyard” bird, French has a single generic term, the masculine volatile. Also masculine are domesticated homing and carrier pigeons – but not white doves, bred in captivity to be set free as symbols of peace (or freedom from war). Other domesticated animals are also masculine, including cat, dog, horse, sheep, cattle, donkey, pig, guinea pig and ferret (although wild creatures are not all feminine).
If we take just one of the more abstract pairs of semantic opposites, such as “protective” and “harmful,” we find an extensive set of nouns among living things whose genders correlate in the same way as in previous examples. “Protective” covers any response that protects life against the environment and competition for food, light and warmth. “Harmful” applies to the opposite, particularly responses that harm or endanger life. The various protective:harmful attributes are as varied as they are astonishing.
Some entities produce a thick, hard, protective outer shell (crab, clam, oyster, mussel, scallop, turtle/tortoise, hazelnut, walnut), while other shells are brittle (egg, snail, peanut) or protect only part (acorn, cashew); those in the first set are feminine, while those in the latter set are masculine. Some creatures are wary and immediately flee when they sense danger (viper, trout, bleak, cockroach), while others are unwary (kookaburra, sea lion, grebe, booby); the former are feminine and the latter masculine. Some can change direction unexpectedly, swerving (mouse, ant, gazelle), leaping (flea, frog, trout) or taking to the air (insects such as ladybird beetle, fly, wasp, honey bee, cicada, locust), and are feminine. Others hang or move around upside down (sloth, certain wrasse), dive headfirst into a different medium (diving ducks, various diving birds, sand eels), or rely on speed (hare, fox), each habit bringing its own risk of injury and death, and are masculine.
Seeking safety or shelter underground (rabbits, wombats) is inherently dangerous: whole colonies risk being buried alive when terrains suddenly flood, or weaken and collapse (miners and cavers well understand these risks); their entrances can admit other less welcome visitors, particularly predators from whom there is little chance of escape, particularly for their young. “Underground safety” is almost an oxymoron. The diurnal but feminine osprey has waterproofed feathers and can haul itself out of the water or swim to safety after a dunking where the masculine fish eagles do not; they easily become waterlogged and drown.
These protective/feminine and harmful/masculine correlations are consistent, regular, even predictable. Other variations include:
Protective: Includes a collaborative response to threat (crane, guinea fowl, moorhen); thorns protecting new growth (hawthorn, honey locust tree, bramble, barberry, bougainvillea); an additional prehensile grip via tail (op/possum), mouth part (lamprey, tick, some caterpillars) or fins (topknots); or the ability to spin a safety net (spider, certain caterpillars), to spread despite changes in environment (carp) or inhospitable landscapes (goat, Canadian white spruce, lavender), to use a repellent odour to ward off predators (weasel, skunk, marmot, mole, cantharis (Spanish fly), grass snake), or to use synchronous surfacing (scoter, a type of seaduck). Fieldfare, large European migratory thrushes, collaborate to ram predatory birds or “escort” them away from the colony. All these creatures have feminine names.
Harmful: Includes an inflexible set of growing conditions (most plants); the need for ready access to water (hippopotamus, elephant) in environments that make this uncertain; a fixed diet (grazing herbivores such as deer, bison, cattle) which cannot alter even in long droughts and brings the possibility of death; a fixed terrain/specific patch (chamois, other mountain goats); tails not prehensile (squirrel, monkey); arboreal creatures whose lives are always at risk as they move through trees at considerable speed; night singing without protection of group (rail, crake, nightingale, cricket); and restricted spread where disease may wipe out an entire species (feral pigeon, rock dove).
The quintessential thorned rosebush (rosier) is masculine; its thorns cannot protect new growth from sap-sucking insects, or old growth from fungal infections that kill. Many of the cuckoo family, including “brood-parasitic” Old World cuckoos, indigo birds, African whydahs and honeyguides, are “harmful” and therefore masculine because they lay their eggs in others’ nests and attempt to foist their care on non-parents at the expense of their own clutch. Some cuckoos endanger their own lives by preferring to leap, walk or run rather than fly from harm (roadrunners, whydahs, anis). A cuckoo may find itself in another’s nest but it remains true to its heritage.
Snakes typically lie in wait for prey to pass by, and are masculine, but our Australian death adder is feminine. Herons also typically stand and wait for prey to pass by and are masculine (héron and other masculine terms); but some herons are more flexible and adaptable, using “toes” to dig up food or lure prey towards them; they are called aigrette, a feminine term. (English terms “heron” and “egret” do not fit French usage, although they may once have done so.) When food becomes scarce, those able to obtain a more constant food supply have a level of protection that can help them survive until more promising times return; others are left to starve and possibly die. Where attributes enhance safety and promote life, we find feminine gender. Where attributes are harmful, endangering the lives of individuals themselves, or their family or their species, we find masculine gender.
It may be that certain entities have one characteristic that is “harmful” and another that is “protective.” Although the “tricoloured heron” can change its foraging and feeding strategies, it builds its nests on mud flats, leaving its young with little protection from cold and predation or from drowning when water levels change abruptly. Speakers may use the feminine form in the context of feeding but use the masculine form during the breeding season and changes in weather.
We may not know these things about each entity, or be able to predict their gender. But knowing the gender can now tell us something about their characteristics; and “exceptions” will prompt us to look for some other crucial attribute associated with the contrasting gender.
What other opposites are there? How do we know which ones to look for, and what their associated genders will be? The answer lies partly in an entity itself, and partly in the meaning contributed by the lexical component(s); together they give rise to, yet limit, the range of potential opposites. For example, fruits have only a limited number of potential semantic opposites – odourless:fragrant, hard:soft, sweet:not sweet, filled:hollow, thick:thin-skinned, and so on. Using our intuitions and stereotypical associations, it is not difficult to distribute the opposites for each pair between the two genders. For hard:soft, we are more likely to associate “hard” with masculine and “soft” with feminine than vice versa – and so we find masculine for fruits that remain hard even when ripe, such as passionfruit, watermelon and quinces (where softness indicates rotting), and feminine for raspberries, red/white currants, figs and dates, where soft indicates ripe. These hard:soft semantic contrasts can also apply elsewhere – for example, hard bone (masculine), soft flesh (feminine). Hard:soft can also vary slightly, providing other sets of semantic opposites such as rigid:flexible or fixed:changeable.
Restricted:free can vary into closed:open contrasts to create distinctions between fist (masculine) and hand (feminine) although they indicate the same body part; a masculine bud opens out into feminine leaf or flower. Closed geometric figures (circle, square, rectangle and triangle) are masculine; open forms (line, parabola, curve and spiral) are feminine. The principles associated with these items become easy to understand; we can know and recall their gender assignments with ease, without even knowing the French terms. For the very first time, these semantic principles offer a systematic means of understanding French gender assignments, even synonyms with different genders, while exceptions prompt us to consider another potentially crucial attribute. An item can display multiple attributes – the salient attribute depends on other entities in its environment.
The universality of this radical account has yet to be tested, particularly for other previously unexplained gender and noun class languages. We cannot know if these patterns are those that native speakers are attuned to at a subconscious level. If various semantic features identified for French seem astonishing for a European gender language, the notion “harmful” was first identified nearly forty years ago in an Australian Aboriginal language (although not in precisely the same way) and has since been identified in several others, as have other features identified in French, including “animate,” “edible,” and semantic opposites shiny:dull.
These findings are exciting. For the very first time, this single step from sex to species offers principles we can understand, recall with ease and apply without effort once we know. It’s also exciting because it offers learners an explanation that they can use. There’s no need for memorisation; the gender becomes a given, something learners know, just like native speakers. For native speakers it offers an explanation that they find fascinating. All speakers enjoy learning something astonishing about their own language.
And the single exception? The gender of the word gens, or “peoples,” is determined by adjectives. It is a result of the artificial re-engineering of this noun from feminine to masculine by authorities in the middle ages, in the period before the Académie française was established, and resistance by native speakers to the change. It has its own remarkable story. •
Margaret à Beckett is an Adjunct Research Fellow at Monash University. Gender Assignment and Word-final Pronunciation in French: Two Classification Systems is published by Lincom GmbH.