/anthropology/glossary/cephalic-index
The cephalic index is the greatest breadth of the head expressed as a percentage of its greatest length. Taken on the skull, where it was first defined, it is the cranial index. Of all the measures in this glossary it is the oldest, the most often taken, and the one on which the classification of peoples came most to depend, so that every one of the 209 local varieties in the corpus is placed on it. For the same reason it is the measure whose limits have been most thoroughly examined, and this entry gives those limits their due.
Status: Preliminary. Boundaries are given from the standard handbooks; Retzius's original division and the exact cuts of the corpus's source remain to be confirmed against the primary texts.
dolicho- δολιχός long · meso- μέσος middle · brachy- βραχύς short · hyper- ὑπέρ beyond · -cephalic κεφαλή head · -cranic κρανίον skull
The classes are named for the length of the head, although the index is a breadth. A low index, which is a head long for its breadth, is dolichocephalic; a high index, a head short for its breadth, is brachycephalic; and the intermediate class, added later, is mesocephalic. Retzius named the two outer classes in 1842 from material that fell plainly into one or the other; the middle class and the graded bands marked hyper- and ultra- were introduced as larger series showed the continuity between them, the last of these being met in individuals but not in the means of populations.
The distinction between -cephalic and -cranic is not ornamental. The index is routinely taken on the living, where it is cephalic; on the skull it is cranial, and the handbooks write dolichocrany and brachycrany for the skull forms, because, as the next section explains, the two measurements differ and do not share a scale.
Greatest length is taken in the median plane from the glabella, the most prominent point between the brow ridges, to the opisthocranion, the point on the back of the head farthest from it. Greatest breadth is taken between the two euryons, the most lateral points of the braincase wherever they fall, above the supramastoid crests and clear of the zygomatic arches. Both are taken with the spreading caliper, on the living head over the scalp and on the skull directly on bone.
The living head carries scalp and hair over the bone, and since these add to breadth and to length unequally, the index of a living head runs about two units above that of the same skull. It follows that a value must be read on the scale of the technique by which it was taken: a cranial index of 79 and a cephalic index of 79 belong to differently shaped heads. The landmarks themselves were fixed by the Frankfurt Agreement of 1882, which standardised the horizontal plane used to orient the skull and the definitions of the common points, so that figures published before that date are not strictly comparable with those published after it.
The index is a continuous ratio, and the classes are bands cut into it. Since no natural break divides long heads from short, each author who cut the scale did so where his own material suggested, and the cuts do not agree. The table sets out those in common use.
| Class | Broca (1875), skull | Martin (1914), skull | Martin (1914), living | Human Phenotypes, living |
|---|---|---|---|---|
| ultradolicho- | — | –64.9 | — | individuals only |
| hyperdolicho- | — | 65.0–69.9 | –70.9 | lowest band |
| dolicho- | –75.00 | 70.0–74.9 | 71.0–75.9 | long |
| (sub-dolicho-) | 75.01–77.77 | — | — | — |
| meso- (mesati-) | 77.78–80.00 | 75.0–79.9 | 76.0–80.9 | medium |
| (sub-brachy-) | 80.01–83.33 | — | — | — |
| brachy- | 83.34– | 80.0–84.9 | 81.0–85.4 | from 81 |
| hyperbrachy- | — | 85.0–89.9 | 85.5–90.9 | highest band |
| ultrabrachy- | — | 90.0– | 91.0– | individuals only |
Broca's five classes use a different middle term, mesaticephalic, and two sub-classes that later authors dropped. Martin's scales are those of the Lehrbuch and its 1957 revision with Saller, and are the ones most twentieth-century typologists cite. Human Phenotypes, the source of the Pheno Guesser corpus, names five bands and places short-headedness at 81, which is Martin's living cut; it does not state the numerical bounds of its outer bands.
Two consequences follow for the reader. A population described as brachycephalic by a French author of the 1870s had an index above 83; the same word from a German author of the 1930s means above 80 on the skull or above 81 on the living. And a value of 80 is meso- on one scale, sub-brachy- on a second, and brachy- on a third. When two sources disagree about whether a population is round-headed, the disagreement is as likely to lie in their scales as in their measurements, and the scales should be consulted first.
The Swedish anatomist Anders Retzius introduced the ratio and the terms dolichocephalic and brachycephalic in 1842, in a scheme that also divided skulls by the profile of the jaw. Within two decades it had become the central instrument of the new craniology, for it offered what that science most wanted: a single number, cheaply taken, that varied visibly between peoples. Paul Broca and the Société d'Anthropologie de Paris refined the landmarks and cut the scale into five classes in the 1860s and 1870s, and the Frankfurt Agreement of 1882 fixed the conventions of measurement internationally.
William Z. Ripley's The Races of Europe of 1899 made the index one of three criteria, beside stature and pigmentation, for a division of Europeans into Teutonic, Alpine and Mediterranean races, and mapped it across the continent; Joseph Deniker used it the following year in a finer classification. From there it ran through every typology of the following half-century, those of Eickstedt, Coon and Biasutti among them, which are the typologies the corpus's source digests. Its authority declined with that of typology itself after the 1950s, when population genetics and the study of clines displaced the type as the object of study; and by the 1960s the evidence of plasticity set out below was generally taken to disqualify it as a marker of descent.
What the index can bear is best stated under four heads: the reality of the variation it records, its relation to descent, its plasticity, and its comparability between sources.
The variation is real and geographically ordered. Population means range from about 70 to the low 90s, and the map of the index is structured: long heads in Scandinavia, the western Mediterranean, the Sahel and Australia; round heads in the Alps, the Balkans, Anatolia, Central Asia and much of Native America. At the scale of neighbouring populations the differences are consistent, and the typologists who relied on them were not measuring nothing.
The index does not, however, follow descent at the scale of the great divisions of mankind. The same class recurs in populations with no recent common history: long heads are shared by Swedes, Sudanese and Aboriginal Australians, round heads by Bavarians, Armenians and Mongols. The corpus's own source states the limit plainly, that the index "is useful to distinguish local anthropological types, but not major racial groups". Ripley's three races of Europe were built on it precisely because it cut across the other criteria, and the scheme did not outlast the discovery of how much else it cut across.
The index is, moreover, partly plastic. Franz Boas measured immigrants to New York and their American-born children in 1912 and found that the children's indices had shifted from the parents' within a single generation, and in opposite directions for different groups. Two re-analyses of his data reached different estimates of the size of the effect: Sparks and Jantz, in 2002, found the shifts small beside the differences between groups and concluded that the index remained largely heritable; Gravlee, Bernard and Leonard, in 2003, upheld Boas's finding that the shifts were real and environmental in origin. For the present purpose their agreement suffices. The index moves with conditions, and a class assigned to a population describes it at a time.
It has also drifted historically. Skull series show a broad rise of the index across Europe from the early Middle Ages, the brachycephalisation discussed by Weidenreich in 1945, and a reversal since the late nineteenth century that continues in living series; the causes of neither movement are settled. A type described as round-headed from medieval or early modern skulls need not be so today.
Custom alters it directly. Cradle-boards and firm cradling flatten the occiput and raise the index, and intentional deformation raises or lowers it by many units. Where such customs were in use, a high index records infancy as much as ancestry.
Finally, the measurement is not fully comparable across sources. Living and skull values differ by about two units; landmarks taken before 1882 differ from those taken after; and the sexes differ slightly, females running a little higher in short-headed populations and a little lower in long-headed ones. Comparisons that neglect any of these produce differences of class where there is no difference of head.
From these the judgment follows. The cephalic index is a sound descriptor of head shape and a serviceable one for distinguishing neighbouring populations at a given time. It is weak evidence of large-scale relationship and, taken alone, no evidence of it. Its class names are conventions laid over a continuum, and each carries its author's cut with it.
All 209 local varieties are placed on the index. The counts are of types whose description uses the term; a type written "(hyper-)brachycephalic" is counted under both bands, the parenthesis meaning that the extreme band is reached in part of the population.
| Class | Types | Examples |
|---|---|---|
| hyperdolichocephalic | 9 | Dinkaid, Desert Australid, Chopi-Tswana |
| dolichocephalic | 82 | Hallstatt, Arabid, Aoshima, Appalacid |
| mesocephalic | 120 | Aisto Nordid, Anglo-Saxon, Aetid |
| brachycephalic | 81 | Alföld, African Alpinoid, Armenid |
| hyperbrachycephalic | 12 | Armenid, Aralid, California Pacifid |
The source writes ranges compactly, meso- brachycephalic or dolicho-mesocephalic, for populations that straddle a cut. The form is a small acknowledgment of the point made above, that the cut is a convention and the population is not.
The height-length index, which shares this index's denominator and sets the height of the skull against it. Head size, for the absolute dimensions the index discards. The nasal index, whose boundaries differ not only by authority but by technique. The morpheme grid, for the elements of the terms.
© Encyclopaediae. Open educational reference project.