Body parts
Because of our common physiology, the body is an obvious place to look for universally shared aspects of meaning. But cross-linguistic investigation reveals that there is actually huge variety in the extensions of human body-part terms. Languages certainly do not all divide the body up in the same way for the purposes of reference. This is a significant finding. If languages don’t treat something as basic and universal as the body identically, what is the likelihood that there will be semantic universals in less basic domains?
A first surprising fact is that not all languages even have a term for ‘body’. Tidore (Papuan, Indonesia) has a word for ‘fl esh’, but this does not cover the body as a whole, and no other word in the language does either. In order to refer to the body as a whole, Tidore speakers must either use the Indonesian word badan ‘body’, or speak non-literally using a Tidore word, mansia which literally means ‘person, human’ (van Staden 2006: 330). Absence of a word referring to the body is not confined to Tidore. In Thaayoore (Pama-Nyungan, Australia), the word with the closest extensional range to English body is pam-minj. This doesn’t just cover the physical body, however. It also denotes many other related ideas, such as people’s tracks, their voice, and their shadow, none of which can be referred to with body in English (Gaby 2006). Pam-minj in Thaayoore thus spills over the familiar boundaries of the physical body to encompass entirely non corporeal referents. The same phenomenon can be repeated at the level of body parts. Thus, the word for ‘bottom’ in Thaayoore is also the word for ‘excreta’, and in Jahai (Mon-Khmer, Malaysia/Southern Thailand; Burenhult 2006: 168), ‘belly’ and ‘excreta’ are expressed by the same word. To Western ears, this is a highly unfamiliar polysemy, that strikingly suggests that the divisions of the world presupposed in familiar languages are by no means universal.
The world’s languages display, then, remarkable variety in their treatment of the different parts of the body. Sometimes, for example, there simply are no words for (to our minds) important body parts. Jahai has no word for ‘mouth’, ‘face’ or ‘leg’ (Burenhult 2006). Instead, there are many morphologically simple terms for more detailed body parts: wε̃s ‘frontal tuber’, nus ‘upper lip’ and mŋka? ‘molar tooth’ (Burenhult 2006: 167). Jahai also lacks terms for ‘arm’ and ‘leg’, and the word for ‘head’, kuy, only refers to the part of the head covered by head-hair (Burenhult 2006: 169). These facts disprove the assumption that languages always favour lexicalization of body-part categories at the level of the limb or other major ‘whole’ body part (head, trunk).
Let’s now consider how different languages segment the parts of the body. The body is obviously a continuous whole, but is made of perceptually discontinuous parts: the trunk, limbs and head, at the highest level of structure, with each of these parts presenting easily distinguish able subparts (chest, back; fingers, elbows; mouth, ears). Since these divisions are so perceptually salient, a natural hypothesis is that all languages exploit them for the purpose of body-part labelling. This hypothesis has been tested through the use of a body-colouring task, in which consultants are given an image of a body and asked to shade in the region that corresponds to a given body-part term. Experiments using this kind of method have partly confirmed the hypothesis that visual discontinuities like those of the limb-joints guide the referential range of body-part terms. For example, no languages are known in which perceptual discontinuities play no role in the segmentation of body-part terminology. Most languages respect the discontinuity of the joints of the arm and leg in the extensions of at least some of their terms for these regions of the body (Enfi eld et al. 2006: 141). In other words, there are always some words in a language which divide the body at the shoulder/elbow/wrist or hip/knee/ankle, even if not all the language’s body-part terms do. For example, many languages, like Punjabi (Indo-European, India) and American Sign Language (signed; USA), distinguish upper and lower leg (Enfield et al. 2006: 141), with the knee constituting the point of division. Yélî Dnye (isolate; Papua New Guinea) is an example of a language which does not observe the joint-division below the knee: a single term covers the lower leg and the foot, ignoring the discontinuity of the ankle. A single term for leg/foot and another for arm/hand is a common situation in Austronesian and Papuan languages (van Staden 2006: 327).
QUESTION Use a line drawing outline of a human body to colour in the boundaries between the different body-part terms in your own language. How easy is it to designate clear boundaries? Are there any cases where the boundaries are unclear?
Two interesting exceptions to the relevance of visual discontinuity in body-part naming have been noted in the recent literature. The first comes from Tidore. Here, the word yohu, translated as ‘leg’, does not actually correspond fully to that English body part, since it does not include the upper thigh. According to van Staden (2006: 327), this is explained by the fact that exposure of the upper part of the thigh is considered indecent, and it must be covered in public. It is therefore not a perceptual, but a cultural discontinuity that determines the limits of this body-part term. The second exception is the Jahai word cŋĩ ŋ, which refers to a spectacle shaped area around both eyes (Burenhult 2006: 167) – again, not an extension with any obvious visual determinant, although there may well be socio-cultural factors behind the salience of this region of the face which explain its lexicalization.
Finally, many investigators of cross-linguistic body-part terminology report widespread inconsistency among speakers of the same language about the extension of body-part terms. For example, speakers of Lavukaleve (East Papuan; Solomon Islands) were divided over whether the word vatu ‘head’ also includes the meaning ‘face’ (Terrill 2006: 3070). This uncertainty may partly be a product of the body-colouring task, the unfamiliarity of which may induce a higher level of self-consciousness and hence hesitation than spontaneous, unmonitored language use. But it might also indicate that it is wrong in many cases to imagine that body part terms have a single fixed, circumscribed meaning in a language community: both within and between speakers, there may be considerable variation in the extensions of terms in this semantic domain. It would be a mistake here, as in most other areas in semantics, to imagine that everything is cut and dried.
