READ IN French
Anthropology / Experiment Review

Vol. 3, NO. 2 / August 2017

Skulls That Speak

Antoine Balzeau

Letters to the Editors

In response to “Skulls That Speak

Hugo Reyes-Centeno, Katerina Harvati, et Gerhard Jäger, “Tracking Modern Human Population History from Linguistic and Cranial Phenotype,” Nature Scientific Reports 6, no. 36,645 (2016), doi:10.1038/srep36645.

In their paper, Hugo Reyes-Centeno, Katerina Harvati, and Gerhard Jäger ask whether anatomy, linguistics, and geography tell the same story about the shape of the human skull. They studied two hundred and sixty-five skulls from eleven populations in Africa, Asia, and Oceania. By establishing reference points on the outer surface of the skulls, they were able to characterize their morphology in quantitative terms. It was then possible to make comparisons between individuals and among populations. Variations in the face, temporal bone, and the skull as a whole, were closely examined. The vocabulary of more than eight hundred languages and dialects from the geographical areas where the skulls originated were also analyzed. Distinctions between the languages were used to derive a list of forty words, which in turn correspond to words currently in use. These words have the same sense in all languages, are resistant to change, and are independent of cultural variations.

Any assessment of this paper involves a consideration of the links between linguistics and genetics, between the shape of the skull and environmental data. The relationship between phenetics and genetics is also of interest. A brief tour of the osteological collections is in order, simply to evaluate how well suited this material is for bioenvironmental studies.

Language and Genes

It is generally accepted that languages change quickly—too quickly, perhaps, for the detection of structures that might date back several thousand years. By using anatomical data and various statistical techniques, it becomes possible, according to Reyes-Centeno, Harvati and Jäger, to determine a fixed linguistic point far in the past. Their study justifies this approach. This is a good thing, because it opens the door to new analytical perspectives. Still, some caution should be exercised.

Nonetheless, it remains extremely difficult, if not impossible, to document human languages over long periods of time. Such an enterprise is inevitably compromised by indirect data. The relationship between linguistics and anatomy is difficult to detect in our ancestors. It is likely inaccessible for prehistoric times, but some ancient link doubtless exists.

Charles Darwin suggested that human populations and languages evolved together. Doubts remain about this co-evolution. Since its appearance about three hundred thousand years ago, our species has undergone a series of events that have shaped the genetic diversity we see today. There have been periods of expansion and crises that are reflected in the genetic make-up of today’s populations. The first Homo sapiens to leave the African continent did so more than one hundred thousand years ago. These early emigrants were probably not inculpated in the rise of current Eurasians. There was a major genetic bottleneck some sixty thousand years ago, which was followed by an expansion of the gene pools we now find today. A second upheaval occurred during the Neolithic period, when many new genes appeared, governing lactose tolerance or lightness of skin. We now know that Neanderthals, Denisovans, and perhaps even another human species, left traces in our genome. Was our lexicon also enriched on the occasion of their exchanges? We simply cannot know. Genetic diversity within today’s human populations diminishes with their distance from East Africa. Conversely, the variation among populations is greater when the geographical distance increases.

All of this speaks to a link between genetics and geography.

When the languages of human populations are compared, phonemic and geographical distance also seem to be correlated. The results obtained by studies of variation within distinct populations are generally inconsistent, in part because there are large linguistic dichotomies between continents. Is there a link between language and genetics? Such a link may be dependent on geographical distance, expressing a complex relationship among multiple dependent variables.1

Ever-expanding databases are now available for hundreds of languages. Analysis suggests that the relationship documented by Reyes-Centeno, Harvati, and Jäger may well be valid. Further research is needed to unravel the links between genes, languages, and geography.

The origins of these correlations remain an utter mystery.

Skull Shape and the Environment

Cranial anatomy is the basis for the definition of the many species of hominids. Homo sapiens does not possess its own holotype. Linnaeus did not consider a holotype necessary and no one has dared to propose one since. Homo sapiens has never had a completely satisfactory morphological definition.2 We possess a number of anatomical features that clearly distinguish us from other hominids: a bony chin on the mandible, the shape of our cranial vault, and well-developed parietal lobes. These characteristics have been present for around three hundred thousand years.3

The relationship between skull anatomy and the environment was a subject of interest for Reyes-Centeno, Harvati, and Jäger. There are many difficulties involved. Consider the development of the sinuses: an adaptation, it has been suggested, to a cold climate. This could explain both the broad face and large sinuses of the Neanderthals. In fact, Neanderthals did not possess large sinuses in relation to their faces. Their skull was not an adaptation to an austere climate.4 Furthermore, the size of the sinuses of modern man does not seem to be correlated with ambient temperature.

Current human populations vary anatomically according to their geographical distribution. This much is undeniable. A link with climate, on the other hand, is more difficult to establish. Our anatomy is the result of complex processes, not just the direct influence of the temperature at sunrise, altitude, or wind force.

On the subject of diet, recent analytical progress has been impressive. Although it has already been explored to some extent, the direct relationship between nutrition and the shape of the skull requires more research. A new technique, isotopic analysis, can be used better to understand what the men of the past ate, or what they drank, or even, when it comes to strontium analysis, where they lived. The soil leaves a geochemical signature in growing bones. These methods have been used in an effort to determine the diet of australopiths and paranthrops, the African hominids of several million years ago. Progress to date has been limited. On a more positive note, these techniques have been successfully applied to determine the presence of plants in the Neanderthals’ diet.

The validity of this technique for studying current populations has not been fully explored. A pilot study has recently been carried out on African populations, comparing ways of life such as pastoralism, fishing, and agriculture.5 Several factors—likely including the isotopic characteristics of Kenyan lakes and the individual diversity of feeding behavior—mean that it is not possible to distinguish fishermen and farmers from isotopic data alone.

Living conditions, especially those shaping the growth of an individual, cannot be fully expressed by a single measure. An average temperature cannot reflect daily, seasonal, or annual variations. Similarly, a single figure for altitude or a geographic coordinate cannot describe all the movements of a man during his life. If correlations, such as those documented by Reyes-Centeno, Harvati, and Jäger are perceptible, it is because the variations in skull shape between populations are related to geography, which reflects not only climatic changes, but also the history of human migrations.

Different regions of the skull do not vary in the same way. Reyes-Centeno, Harvati, and Jäger note that geography “explains up to half of overall cranial shape variation and when considering its component parts, it explains respectively ~42%, ~15%, and ~20% of variation in facial, neurocranial, and temporal bone phenotype variation.”6 These values are dependent on the material being analyzed, the composition of the samples and their origin, as well as the methods used. The reference points reflect only part of the shape of the skull and its different zones.7

Skull and Genes

We are still far from identifying all the genes involved in the morphogenesis of the skull. The phenomena remain elusive. There has long been debate regarding the origin of the fossils from the earliest periods of the Upper Paleolithic in Europe: which among them are the first representatives of our species? The most contested issue is whether their anatomy indicates a relationship with the Neanderthals that populated the continent before the arrival of Homo sapiens.

In order to classify a fossil within a species, unique anatomical characteristics must be identified. These are termed apomorphies, or derived characteristics. So-called primitive or plesiomorphic characteristics, on the other hand, are shared by several species. The possession of an opposable thumb is a derived characteristic of primates in relation to other mammals, but it will be seen as primitive if one considers a current man in relation to a gibbon. The robust characteristics of Homo sapiens from the Upper Paleolithic in Europe have nothing to do with a possible relationship with the Neanderthals, since these traits are also found in ancient African fossils, and to some extent in other species. They are thus primitive shared characteristics.

Conversely, European Homo sapiens are classified as part of our species since they possess the same anatomical peculiarities. The Oase 1 fossil found at Peștera cu Oase in Romania and dated to thirty-seven thousand years ago is an example. This fossil has nothing to do with the Neanderthals in terms of its morphology. Yet a recent paleogenetic study has shown that this individual was the most Neanderthal of all modern humans.8 Today, around three percent of our genes have Neanderthal origins. But in this case of Oase 1, that figure was more than eight percent, with a direct Neanderthal ancestor about five generations back. In spite of these findings, as far as I am concerned, there is no discernible imprint on the morphology of his skull.

The same problem arises when analyzing current or recent human populations. In the past, pseudoscientists have sought to classify human populations according to anatomical characteristics. Consider skin color. No clear distinction can be made according to geography. Instead there is a gradient across humanity as a whole, and also within each local population. Different skin colors can be observed in Africa and the Americas under similar conditions. There is phenotypic variability in humans, but there is no discontinuity between populations, whichever characteristic is considered.

The same is true of the data studied by Reyes-Centeno, Harvati, and Jäger. The shape of the cranial vault, temporal bone, and face show many dissimilarities between the eleven populations examined in their study. This is due, in part, to a restricted set of samples. The study could not test the relationship between the data and genes since they did not possess such information for the material they were examining. In short, if the skull and genes are closely linked, the relationship cannot yet be directly tested or used in physical anthropology.

A Collector’s Item

A limitation faced by anthropologists working with human remains is the contextual information associated with these objects. The anthropological collections of the Musée national d’Histoire naturelle in Paris contain more than eighteen thousand human skulls gathered during scientific missions in various parts of the world. The study by Reyes-Centeno , Harvati, and Jäger relies in large part on the material from this collection. But these collections are not without their limitations. The skulls collected were obviously those that were most easily accessible. Background information was not always documented, and the point of origin, not always certified. A specimen collected two hundred years ago could be that of a migrant from a few decades earlier.

On the one hand, these osteological collections allow one to appreciate human morphological diversity. They are less useful for understanding the relationship between form and factors such as geography, environment, or diet. Some studies have relied on an average geographical coordinate from a spatial data set. Latitude and longitude are then used to find average annual temperatures, altitudes, and the like. Average diet is predicted from whatever limited ethnographic data is available. The skull is silent but the language its owner spoke is extrapolated from archival information, when available, and from contemporary comparisons.

Alas, poor Yorick!

Environmental characteristics can only be inferred from osteological collections. This, in turn, poses a number of problems. Even the most advanced analytical techniques have limitations when it comes to characterizing our own diets. Oversimplifying the data used to discuss these problems is equally problematic.

The osteological collection, owing to its vast size, expresses a limitation of its own.

On Demand

Two broad areas for further research demand consideration. It would be fascinating to document the information discussed by Reyes-Centeno, Harvati, and Jäger in living populations. Together with genetic analysis, this would allow us undertake comparisons of skull shapes and the brain, drawing upon variables such as place of birth, along with lifestyle, diet, and environmental conditions. We could then better understand the links between linguistics, biology, and geography, and the causes of morphological variations in the skulls of present and past populations.

Research is also required to better understand details not reflected by current biological and genetic diversity. The period between three hundred and one hundred thousand years ago remains poorly documented. There is evidence of our presence in the Near East, Asia, perhaps in Australia, and—much more debatably—in the Americas. Yet there is no trace of this in our genetic heritage. The number of African fossils dating from between one hundred thousand and ten thousand years ago can be counted on the fingers of one hand.

The Neolithic era was a difficult period from a demographic point of view. This much we know. But we do not yet understand all the diachronic dimensions and possible relations between the different regions of the world. These events are perhaps too far in the past to examine the role of language. If preserved specimens can be found, we would be in a better position to describe their morphology.

Translated and adapted from the French by the editors.


  1. See, for example: Keith Hunley, “Reassessment of Global Gene–Language Coevolution,” Proceedings of the National Academy of Sciences of the United States of America 112, no. 7 (2015): 1,919–20; Guiseppe Longobardi et al., “Across Language Families: Genome Diversity Mirrors Linguistic Variation Within Europe,” American Journal of Physical Anthropology 157 (2015): 630–40. 
  2. Jeffrey Schwartz and Ian Tattersall, “Fossil Evidence for the Origin of Homo sapiens,” American Journal of Physical Anthropology 143 (2010): 94–121. 
  3. The oldest currently known fossils are from Omo Kibish or Herto, both in East Africa. The newly discovered remains from Jebel Irhoud in Morocco are also among the oldest reported for Homo sapiens
  4. Marlijn Noback and Katerina Harvati, “The Contribution of Diet to Global Human Cranial Variation,” Journal of Human Evolution 80 (2015): 34–50; Marlijn Noback et al., “Paranasal Sinuses: A Problematic Proxy for Climate Adaptation in Neanderthals,” Journal of Human Evolution 97 (2016), doi:10.1016/j.jhevol.2016.06.003. 
  5. Maria Ana Correia et al., “Modern Human Hair, Nail and Breath Isotopic Signals and Their Relevance to Diet Assessment in the Past,” American Journal of Physical Anthropology 162, no. S64 (2017): 151. 
  6. Hugo Reyes-Centeno, Katerina Harvati, and Gerhard Jäger, “Tracking Modern Human Population History from Linguistic and Cranial Phenotype,” Nature Scientific Reports 6, no. 36,645 (2016), doi:10.1038/srep36645. 
  7. See, for example: Christian Klingenberg, “Studying Morphological Integration and Modularity at Multiple Levels: Concepts and Analysis,” Philosophical Transactions of the Royal Society of London B 369, no. 1,649 (2014), doi:10.1098/rstb.2013.0249. 
  8. Qiaomei Fu et al., “An Early Modern Human from Romania with a Recent Neanderthal Ancestor,” Nature 524 (2015): 216–19. 

Antoine Balzeau is a researcher at the CNRS and at the Musée national d’Histoire naturelle in Paris.

More from this Contributor

More on Anthropology


Copyright © Inference 2024

ISSN #2576–4403