Monday, November 29, 2010

Understanding Keita's "Coastal Northern African"

Keita's term of "coastal northern African" in his earlier publications has been a repeat subject of inquiry over the years. Indeed Keita's terms can be confusing, thanks in no small measure to his blanket terms of "coastal northern African pattern" vs. "southern"/"Africoid" pattern, in lieu of old racial constructs, which he rejects as being restrictive such that they overlook the true variability of populations under assessment, since they rely on highly idealized archetypes. However Keita's terms, perhaps while not as restrictive, can also be misleading if details of his analysis are not examined carefully.

The term "coastal northern African pattern" is to be understood as an average (of a combined sample—as denoted by the 'centroid' value, as opposed to individual crania) pattern that assumes an intermediary position between the "southern"/"Africoid" patterns and the "west European" patterns. Perhaps the following piece from Keita can serve as a cue...

This **northern modal pattern**, which can be **called coastal northern African**, is noted in general terms to be intermediate, by the centroid scores of Function I, to equatorial African and northern European phenotypes - Keita, Further studies of crania from ancient northern Africa: an analysis of crania from first dynasty Egyptian tombs, using discriminant functions, 1992.

Keita claims that the Maghrebi series contains crania with affinity to both those in the "Tropical African"/"Southern" series and the "European" series, as well as crania with "intermediary" patterns between those seen in these two entities. The latter would seem to be a code word for "hybrids", which he does in fact use in his article. He goes onto to claim that the Maghrebi series assumes a position closest to the northern Egyptian series, i.e. the Sediment and E series. Funny thing here is that Keita attributes this to possible Neolithic migration from "Europe" as well as that from the African interior spurred by desiccation of the Sahara, but then implies that part of the diversity [the "European"-like part] in the Maghrebi series is most likely attributable to Phoenician migration. It is funny in the sense that his Maghrebi samples were supposed to be pre-Punic specimens, and pre-Roman.

Keita acknowledges the potential cosmopolitan backdrop of the "E" series, but is less certain about that of the Sedment series. The Maghrebi series is closest to the "E" series, of all the Egyptian series, and in fact, Keita himself makes a note of this. It is closest to the "E" series than it is to the Sedment series, and the "E" series in turn, by centroid value, is closer to the European series than the Sedment series. This is why Keita's blanket term of "coastal northern African pattern" on these series can be misleading, if not examined at a greater depth. Looking back, the present author of this blog noted Keita's own acknowledgement of the need for making distinctions between these three "northern" cranial series:

By centroid scores,...

The Lachish series is found to plot nearest the Maghreb and “E” series, both of whose centroids plot nearer the Romano-British groups than any of the other series; the D2 value between these series is significant as previously noted.

Examination of the classification results (when Lachish is run as an unknown) shows that the “E” series receives the plurality, with the Maghreb series receiving a very small percentage. The results seem to indicate that the morphometric patterns of crania in the Lachish series show a great range of variation with many crania classifying into Egyptian and Nubian series, even when Lachisch is available as a choice. - Keita, An Analysis of Crania From Tell-Duweir Using Multiple Discriminant Functions, 1988.

And importantly...

The notable classification of Lachish crania into the northern Egyptian, but not Maghreb, series suggests that it is not helpful to stereotypically generalize about morphometrics of people in “North Africa.” This Maghreb series is actually quite morphmetrically heterogeneous (Keita, 1983). - Keita, An Analysis of Crania From Tell-Duweir Using Multiple Discriminant Functions, 1988.

Two things to note:

1)By comparison, the Lachish crania misclassified more so into the northern Egyptian samples [particularly the "E" series] than Maghrebi sample, amongst the northern [coastal] African crania.

2)Maghreb crania is notably heterogeneous, as is the "E" series, yet more Lachish crania fell into the "E" series than the Maghreb samples.

If both the "E" series and the Maghrebi series are heterogeneous, and at the same time both show discernable trends from those of more southward African crania, then what does this say about coastal north African crania? Well, the implications are spelled out in Keita's own words; see again:

suggests that it is not helpful to stereotypically generalize about morphometrics of people in “North Africa.”

Note that the study data also shows both Maghreb and the "E" series to be yet dicernable from the "Sedment" series of northern Egypt.


In summation, the Discriminant plot (using centroid values; see immediately above—click on all images for greater resolution) shows that the Lachish, a Levantine-based series, assumes its closest position to the Maghrebi and the "E" series. In fact, while the Sedment series is close to the Maghrebi and the Lachish propents.

It isn't clear what the specific modal pattern for the Sedment series entails from Keita's perspective, which is where the territorial maps may come in handy, but he claims that the appearance of the "northern pattern" in the "south" [which would be Upper Egypt] increases over time and characterizes ensuing diversity as being one with a bimodal contrast within it. He attributes this to what he calls "the presence of real northerners [A. Egyptians]"...

The notable increase in northern pattern crania in the south, from Badari times, might have a selection explanation, but the essentially bimodal nature of the presence of the contrasting trends suggests the presence of “real” northerners. - Keita, Further studies of crania from ancient northern Africa: an analysis of crania from first dynasty Egyptian tombs, using discriminant functions, 1992.

If the "northern" Egyptian crania were a composite of contrasting trends itself, and Keita figures that contribution from the north to south lends a hand to a bimodal contrasting pattern in the latter territory, then how does one characterize the "northern" Egyptian pattern, in terms of its modality [not in terms of centroid scores]? Trimodal, mono-modal, or bimodal? Yes, Keita describes the "northern" pattern as "coastal northern African", but that says little of the nature of variation contained in the northern Egyptian pattern, since as demonstrated briefly above, the "coastal northern African" pattern is not some uniform pattern. All Keita tells us about the "Lower Egyptian" crania, which he places as a subset of the just now mentioned "northern coastal African', is simply that its pattern is "intermediate to that of various northern Europeans and West African and Khoisan series" (Keita, 1993). This, unless specified otherwise, is to be understood as a description of the said series as a composite or the average [by way of centroid scores] pattern of the individual crania taken together; as such, the descriptive tells us little else about the specific modal cranio-morpho-metric phenotypes contained in the series. For what it's worth, the few available predynastic "Lower Egyptian" skeletons that are out there, are reported to have body proportions that place them nearer to tropical African groups, as is the case with other ancient Egyptian specimens, from "Upper Egypt" (Robin 1986; Zakrezewski 2003); to recap:

Moving to the opposite geographical extremity, the very small sample populations available from northern Egypt from before the 1st Dynasty (Merimda, Maadi and Wadi Digla)turn out to be significantly different from sample populations from early Palestine and Byblos, suggesting a lack of common ancestors over a long time. If there was a south-north cline of variation along the Nile valley it did not, from this limited evidence, continue smoothly into southern Palestine. Thelimb-length proportions of males from the Egyptian sites group them with Africans rather than with Europeans. - Barry J. Kemp, Ancient Egypt: Anatomy of a Civilization, 2006.

Through all this, it should be kept in mind that Keita has not examined predynastic northern Egyptian specimens, which is a major hindrance to his analysis, leaving him to keep begging the question of the starting/original phenotypic orientation of northern Egyptians. However, again, the examinations of skeletal remains in that region of that era, we are told, group them with Africans rather than Europeans, in terms of body proportions. We are also told that these specimens were not like "Palestinian"-based examples that they were compared with [see above]

ImageShack, share photos, pictures, free image hosting, free video hosting, image hosting, video hosting, photo image hosting site, video hosting site

Upon examining one of the territorial maps (ten variables; see above), which Keita himself acknowledges give a better indication of intra-series variation, it is noticeable that the Maghrebi series is spread along a wide area, more so than any other series. Hence, as Keita also notes, they are clearly the most heterogeneous standalone series of the bunch.

The "E" series closest neighbors generally tend to be either members of the Nagada series, the Sedment series, Maghrebi, or the European series. Every series has a Maghrebi neighbor.

The Sediment series' closest neighbors are generally elements from the Nagada series, the Maghrebi series, the Gabon series and the 'E' series.

The Gabon series' closest neighbors generally tend to be elements from the Maghrebi series, followed by the Nagada, the Sedment and the Badari series.

The Badari series' closest neighbors generally tend to be elements from the Nagada series and the Maghrebi series, and on another level, the Teita series.

The Kerma series' closest neighbors generally tend to be elements from the Nagada and European series, and to some extent, the Maghrebi series.

Nagada series have neighbors with elements from mostly Kerma, followed by elements from the Gabon, the Sedment, the Maghrebi, the "E", the Badari and the Teita series.

The European series' is mostly neighbors with elements from the Maghrebi series, to be followed  by 'E' series, and to a lesser extent, the Kerma series.

On the map, there is a noticeably tendency of the Badari, the Teita and the Gabon to be concentrated   on the extreme negative half of the territorial map (again, the ten variable version) along the coordinate parallel to the 'x' Cartesian coordinate, and the Kerma, Sedment, 'E', and European series concentrated on the positive half, but the European series has the distinction of assuming a sort of boundary of the extreme positive, barring the individual Maghrebi neighbors. Elements of the Gabon and the Nile Valley series assume the role of the intermediary position along that same coordinate. The Maghrebi series encircles the entire territorial map, almost in a circumference manner, and therefore acting like the actual boundary for the entire cluster.

*This post may be updated or modified anytime.
_____________________________________________________________________
*References:

—Keita, An Analysis of Crania From Tell-Duweir Using Multiple Discriminant Functions, 1988.

—Keita, Studies of Ancient Crania From Northern Africa, 1990.

—Keita, Further studies of crania from ancient northern Africa: an analysis of crania from first dynasty Egyptian tombs, using discriminant functions, 1992.

—Keita, Studies and Comments on Ancient Egyptian Biological Relationships, 1993.

—Sonia R. Zakrzewski, Variation in Ancient Egyptian Stature and Body Proportions, 2003.

—Robin, Predynastic Egyptian stature and physical proportions, 1986.

—Barry J. Kemp, Ancient Egypt: Anatomy of a Civilization, 2006.

*Personal notes from 2006, 2010.

4 comments:

Anonymous said...

Can you create a link to a larger Image?

Mystery Solver said...

Anonymous poster,

Have you tried clicking on the image for a higher resolution? The images are posted in their original size, which is then shrunken to cut on page download time.

mghani said...

Do you believe that there is any chance notwithstanding their tropical body adaptations, that ancient Maghrebi populations and even Northern Egyptian populations were possibly Levantine or an interbreeding of Levantine populations with African populations?

Mystery Solver said...

Nakhti writes:

Do you believe that there is any chance notwithstanding their tropical body adaptations, that ancient Maghrebi populations and even Northern Egyptian populations were possibly Levantine or an interbreeding of Levantine populations with African populations?

Nope. They would have had to have shortly arrived from a tropical region, as opposed to a "sub-tropical" environment, to retain such patterns as the ancient Egyptians do. The predynastic Lower Egyptian specimens were found to have tropical body proportions, similar to the pattern observed in tropical African samples.