Extending the body

Back, heart, foot, face, hand, neck, eye. Body-part words like these are some of the most basic in the lexicon. They are old words, often short, and usually learned early. They also range widely, showing up where you might not expect them. Humans have backs, yes—but so do books, televisions, and lecture halls. People have hearts—much like cities, problems, and matters. The foot is an appendage, but also a basic unit of length, not to mention part of a mountain, staircase, or bed. Many body-part terms are nouns that moonlight as verbs—consider face, hand, neck, and eye, not to mention shoulder, stomach, and head. These lexical contortions are legion, and they are by no means a peculiarity of English. Across the world’s languages, body-part terms may be the most commonly, widely, and enthusiastically extended meaning cluster of all. 

The core meaning of a body-part term is, by definition, anatomical. This is the meaning that children learn first, the one listed first in the dictionary, the one that most people would offer up first if asked. But, eventually, that core meaning gets extended, whether through metaphor, metonymy, or some other process; these extensions themselves often get further extended; over time, senses splinter off, fan out, and stack up. Consider the case of back. The first entry in the Oxford English Dictionary identifies the word as a noun with an anatomical meaning; it then lists an additional 61 senses of the word in the noun form alone, all with discernible connections to the first. These include, for example, the back of a knife (i.e., the part opposite its cutting edge) and the figurative use of back to mean protection or defense. Add in the meanings of back as an adjective, adverb, and verb and the count reaches 130.

Other body-part words are similarly branchy, but they don’t branch out willy-nilly. Across languages, what you find is that such words tend to venture in certain directions, popping up in some places much more than others. Here are six domains in which anatomical terms regularly rear their heads:

Parts of objects. Body part terms are commonly extended to characterize the “anatomy” of inanimate objects. As mentioned, English speakers widely ascribe backs and fronts, not to mention feet and heads, to things like chairs and beds. That’s just the beginning. Rivers have mouths; peaches have shoulders; books have spines; needles have eyes. (Houses have etymological eyes, too; window is rooted in a compound of wind and eye.) English might seem especially enthusiastic in drawing on body-part terms to describe things that, strictly speaking, do not have bodies. But other languages go further. Speakers of Isthmus Zapotec, a language of Southern Mexico, could readily point to a car’s head, stomach, nose, eyeballs, and more. Relative to English and European languages, Isthmus Zapotec—like languages across Mesoamerica—applies a richer set of body-part terms to a wider range of inanimate objects.

Spatial relations. Body-part terms are regularly used, not only for referring to specific anatomical regions, but also for referring to the space near those regions. In this way, speakers use body-part terms to locate one thing with respect to another. Take front. At its core, it refers to the forward plane of person or thing. But when we say “Let’s meet in front of the fountain,” we do not mean that we’ll be directly in contact with the forward plane of the fountain. We mean we’ll be near that forward part. Similar relational uses of body-part words are found the world over. Surveying languages of Africa and Oceania, Bernd Heine reports on the widespread extensions of words for ‘face’ and ‘eye’ to mean the spatial region in front of a person (or analogous object); of ‘back’ or ‘buttocks’ to mean the spatial region behind; and of ‘head’ to mean the spatial region above. 

Extending the body.jpeg

More surprising, perhaps, is the use of body parts to express cardinal directions (north, south, east, west). In Ambrym, a language of Vanuatu, the term for east is ‘in front of us’ and for west is ‘behind us.’ The likely reason is that in Ambrym—and other cultures where such associations have been observed—there is a “canonical orientation” facing the sunrise, such that a person’s front side becomes associated with the east. Some scholars trace the English word north to a Proto-Indo-European word meaning ‘left,’ suggesting a similar orientational ideal. A 2012 paper by Rafael Núñez and Carlos Cornejo presents an especially detailed treatment of the the east-facing “canonical orientation” in the Aymara, an indigenous group of the Andes. Aymara speakers might describe an object to the east of another as a “little in front” or to the west as "a bit behind.” But these are not isolated figures of speech—Aymara people also arrange their towns and houses in a way that reflects the centrality of the sunrise. 

Temporal relations. Much as body-part terms are used to capture relations between objects in space, they are also used metaphorically to capture relations between events in time. English speakers think back on lost youth, or anticipate the months ahead. Shakespeare, meditating on the nature of memory, referred to the “dark backward” of time. Again, this is not just an English thing. Body-part terms—and especially ‘fronts’ and ‘backs’—are used the world over for expressing temporal relations. Martin Haspelmath’s (1997) cross-linguistic study of temporal adverbials notes several languages—including Tamil, Maori, Greenlandic, and Albanian—in which the spatial words for ‘front’ and ‘back’ are identical to the temporal words for ‘before’ and ‘after.’ Strictly speaking, this is not the case in English, as our words before and after are by now mostly specialized for temporal relations. They do still they bear the mark of their bodily origins, however. An increasingly archaic meaning of before is ‘in front of,’ as in “How dare you appear before the throne in such rags!” After is cognate of aft, meaning ‘behind,’ which is still used in a spatial sense on ships and planes.

Measurement. Body-part terms also commonly pop up in the sphere of measurement. This is readily seen in the case of foot, a workhorse unit in the imperial system. It is also evident in hand (formerly a general measure, now used primarily to measure the height of horses) and finger (used colloquially to measure spirits, as in, “I’ll take three fingers of whiskey”). Similar examples are found far beyond the West and perhaps universally. In a recent paper, Dedre Gentner and I analyzed an ethnographic database spanning 114 cultures and found that small-scale length units are overwhelmingly drawn from the body. The most commonly used body spans in the cultures we analyzed were the ‘cubit’ (referring to the forearm) and ‘fathom’ (referring to the outstretched arms). In at least one case, a body part was used to express longer distances. The Ojibwe of North America considered the sun’s daily course, from sunrise to sunset, to consist of eight hand-stretches. (If you extend your arm and superimpose your stretched hand over the sun’s arc, you’ll find this to be roughly accurate.) Speakers could thus express the length of a journey in terms of how many ‘hand-stretches’ it was—a half-day’s journey would be four, for instance.

Number. Another neck of the lexicon in which body parts emerge is number terminology. Across cultures and eras, number words have frequently derived from terms for fingers, hands, toes, and feet. This is particularly common for the numerals ‘five’ and above. In Hup, a language of the Vaupes region of the Amazon, ‘five’ is expressed as ‘one hand’ and ‘six’ as ‘one finger stands up’; ‘fifteen’ is rendered as ‘one foot finish’ and ‘twenty’ as ‘both feet finish.’ Similar patterns are found in “young” numeral systems worldwide. But, generally, it seems that number systems lose their connection to body-part words through time and use. As a result, speakers of global languages like English might find such appendage-rich number systems a bit fanciful. But the connections are not entirely lost. Our word digit refers to both appendages and numerals, and some scholars posit a common ancestry to our words fist and five.

Emotion and cognition. A last semantic domain in which we regularly encounter body-part words is that of feelings and thoughts. English speakers talk of using our head or our heart; we speak of having lacking guts or backbone. It seems that most cultures have such bodily associations, but the mappings vary in their specifics. One example of such variation concerns the part of the body most strongly associated with memory and understanding. While in European languages it is most often the eye, in Australian Aboriginal languages it is the ear. In Jiwarli, for instance, the word kurlga doubles as noun for ‘ear’ and a particle for ‘remember.’ In Kuuk Thaayorre, a forgetful person may be said to have a ‘bad ear.’ Australian languages also furnish another example of such variation. For English speakers the prime seat of emotion is the heart, or sometimes the guts. In Kuuk Thaayorre it is the belly, or, secondarily, the liver, throat, or heart.

These are six semantic domains in which the use of body-part words is widely attested across the globe. But this is little more than a start—the body gets extended in innumerable other ways, too. In Mixtecan languages, words for ‘foot,’ can also mean ‘for the benefit of,’ ‘in exchange for,’ ‘because’; words for ‘face’ can also mean ‘in the presence of,’ ‘than,’ and ‘if.’ In Australian Aboriginal languages, words for kin are often associated with body parts. Thus one might mention the ‘beard’ when referring to one’s paternal grandfather, the ‘knee’ when referring to a father-in-law, or thebackbone’ when referring to one’s maternal grandmother. In many languages, body part terms are used in endearments. Speakers fondly address loved ones as ‘my heart’ (in Spanish) or ‘my eyes’ (in Swahili); elsewhere, they might use ‘my liver’ (in Arabic) or ‘my intestines’ (in Amharic).

The patterns above suggest that body-part words are among the most extendable words there are. Why would this be? Proponents of embodied cognition have long argued that bodily experience is the basis of our higher-order thinking—that it furnishes our foundational schemas and our guiding metaphors, that it lays the groundwork for even our most abstract conceptual systems. The experience of walking may account for why many cultures understand the future as something in front; visceral sensations could account for why we associate cognition with our head and intuition with our gut.

But lived experience does not seem to be the whole story. The body likely gets extended for number terminology because, conveniently, we have a bunch of similar-enough appendages. The body probably enters into measurement because it is portable; because it involves salient spans that can be aligned with to-be-measured objects; and because it is human scale. The body may be widely used for describing the “anatomy” of objects because it is a highly familiar visual template. And we don’t have to live a visual template for it to be compelling: we refer to the tail of a line or the wings of a building without really knowing what it’s like to have tails or wings. In all these cases, what makes the body so extendable is its utter familiarity. When we want to capture a new idea—no matter the sphere or scale—we often reach first for what we know best. And often what we know best is right under our noses.


1. There is some cross-linguistic variation in how terms like front and back are assigned to inanimate objects. In English, the ‘back’ of the tree is the part further away from the viewer; in Mian, a language of Papua New Guinea, it is the side of the tree leaning toward the ground. In Chamus, a language of Kenya, the side of the tree leaning toward the ground would be the ‘front.’

2. For discussion of how Zapotec speakers parse cars—and how they use body-part terms more broadly—see Gabriela Peréz Baéz’s recent paper.

3. Len Talmy (2000) has drawn attention to the astounding—and occasionally confounding—ways English speakers use front to locate things in space (see the diagram on page 226 of his book on Toward a Cognitive Semantics, Vol. 1). Paraphrasing, if I say that Jim is in front of Jane, this could mean that he is next to the front of her body, that he is on the side of Jane closer to me, that is he closer to the front of a line than she is, or that he is closer to the front of the room were are all in.

4. At first blush it may seem a bit exotic to use ‘face’ or ‘eye’ to for this purpose, but consider that English front originally meant forehead. Not so fanciful after all. But why would ‘face,’ ‘eye,’ or ‘forehead’ be called upon for this purpose and not other salient features of our ventral surface? Or, as Heine (1997) formulates the question: “What induces people worldwide to decide that a body-part like face, rather than navel or kneecap, provides the favorite model for the developing the spatial concept ‘front’? And why not the body-part nose? Why, in fact, is the nose notoriously ignored as a source concept for spatial orientation?” (p. 47).

5. Why would an east-facing canonical orientation predominate across languages? Cecil Brown’s (1983) answer is that “the east if often of great salience for people since it is the place of the sun’s rising” (p. 136). But this just substitutes one mystery for another: Why is the sunrise more salient for people then the sunset? Note also that other canonical facing directions are attested, as Brown discusses. In Hawaiian, there is a conflation of ‘right’ with ‘north’ and of ‘left’ with ‘south,’ suggesting a west-facing ideal. Similar evidence points to a north-facing ideal in Mataca.

6. In fact, in English and many other languages, fronts and backs do double duty in the expression of temporal relations. First, they are involved in distinguishing past and future events (“those days are behind us/ahead of us”)—that is, to locate events relative to now. In this spatial model, the observer is on a path, with past events behind her and future events off ahead. Second, fronts and backs are involved in distinguishing positions in a sequence (“there is a coffee break before/after the discussion”)—that is, to locate events relative to each other. In this spatial model, there is a line-up of events, much like a queue of people all facing in one direction. Conceptually, an earlier event is in ‘front’ of a later one.

7. For a recent discussion of the association between body parts and kin terms in Australia—and how such associations are realized in the sign languages of the region—see the paper by Jennifer Green and colleagues, ‘Pointing to the Body: Kin signs in Australian indigenous sign languages.’

Further reading

Brown, C. H. (1983). Where Do Cardinal Direction Terms Come From? Anthropological Linguistics, 25(2), 121–161.

Cooperrider, K., & Gentner, D. (2019). The career of measurement. Cognition, 191. https://doi.org/10.1016/j.cognition.2019.04.011

Epps, P. (2006). Growing a numeral system: The historical development of numerals in an Amazonian language family. Diachronica, 23(2), 259–288. https://doi.org/10.1075/dia.23.2.03epp

Gaby, A. (2008). Gut feelings: Locating intellect, emotion and lifeforce in the Thaayorre body. In F. Sharfian, R. Dirven, & N. Yu (Eds.), Culture, body, and language: conceptualizations of internal body organs across cultures and languages (pp. 27–44). Mouton de Gruyter.

Heine, B. (1997). Cognitive foundations of grammar. Oxford University Press.

Heine, B., & Kuteva, T. (2002). World lexicon of grammaticalization. Cambridge University Press.

Hollenbach, B. (1995). Semantic and Syntactic Extensions of Body-Part Terms in Mixtecan: The Case of ’Face’ and ‘Foot.’ International Journal of American Linguistics, 61(2), 168–190.

Kraska-Szlenk, I. (2014). Semantic extensions of body part terms: common patterns and their interpretation. Language Sciences, 44, 15–39. https://doi.org/10.1016/j.langsci.2014.02.002

Núñez, R. E., & Cornejo, C. (2012). Facing the Sunrise: Cultural Worldview Underlying Intrinsic-Based Encoding of Absolute Frames of Reference in Aymara. Cognitive Science, 36(6), 965–991. https://doi.org/10.1111/j.1551-6709.2012.01237.x