Hunter-gatherers, farmers and... something in between

When it comes to thinking about the origins and expansion of agriculture, one is influenced by the powerful appeal of the word revolution. We were all taught at school that there was an Agricultural, also called Neolithic, Revolution that started somewhere in the Middle East and then propagated to the adjacent areas. This hearth of farming and pastoralism would eventually become the so-called 'cradle of civilization'. There were other places in the world where agriculture and animal domestication developed independently (South Asia, China, the Americas) but not at such an early date and with such far-reaching consequences. From the perspective of today, it is logical to see the events associated with agriculture as a revolution that involved profound changes in every possible sphere of human economy and society. But in what ways was it a revolution for the human populations who were involved in the process? I have recently read an enlightening book that analyses this issue in detail: Barker, Graeme (2006). The Agricultural Revolution in Prehistory. Why did Foragers become Farmers? Oxford: OUP. It's the kind of book that anyone interested in prehistoric languages should read.

It would be difficult to summarize the book in a few lines. It is a comprehensive study of the whole process of foraging-farming transition in all the world's regions. One of the main conclusions of this study is that there is not a general pattern for this transition but a variety of possible scenarios. According to the traditional view, the one established since Gordon Childe's coining of the term Neolithic Revolution, agriculture and farming were transmitted from its core area by a process of demic diffusion of wave of advance, whereby groups of farmers, pushed by demographic or climatic pressure, spread into new territories either displacing the local population or causing a process of quick acculturation. This model was proposed for all domestication areas in the world, especially for the one that involved the Fertile Crescent and its contiguous areas of propagation, especially Europe and Africa. It seems, however, that the archaeological evidence does not support this kind of model on a general basis.

First, it seems that most of the research in this area in the last century has been biased by a series of preconceived ideas about the topic. Most archaeological research has focused on South-East Asia (the Levant and Fertile Crescent region), whereas other areas have not been researched so thoroughly. On the other hand, the archeological evidence was traditionally interpreted on the lines of a simple contrast between foraging and farming populations. However, the use of more modern techniques, for example the study of the DNA of animals and plants or the analysis of sediments or pollen deposits, and the better understanding of the hunter-gatherer communities of modern times provided by anthropological studies, offer a much more complex picture of the process. In any case, it was not a single event, or the result of a single expansion process: (p. 378) "The traditional model of Neolithic agricultural colonists from South-West Asia spreading inexorably across Europe is extremely difficult to reconcile with the complexity of the evidence now available for the beginnings of agro-pastoral farming here". Something similar can be said about East and South-East Asia: (p. 229): "The central and eastern Pacific was a 'melting pot' of local domestications and cultigen acquisition from both west and east, not a one-way movement of agricultural colonists", and about South Asia. What we have in general is a variety of subsistence strategies that were adopted at different times and places. In some cases there was a quick transition to a farming-pastoralist economy, in others there was a long coexistence of foraging economy with some forms of basic farming. At the beginning of the Holocene, most human populations were acquainted with more or less elaborate techniques of plant collection and processing that paved the way, especially at the psychological level, for the later introduction of farming. The patterns of nomadism or sedentism also varied considerably. Some of the early farming populations show significant patterns of mobility, whereas some foraging populations were more sedentary than previously acknowledged. A few decades ago, the animal and plant remains found in archaeological sites were confidently understood as domesticated or wild species, which allowed a very simple explanation of the facts. A closer analysis, with more accurate technologies and a more multidisciplinary approach, tells us a different story. Our understanding of prehistoric societies is changing. The contrast between farmers and hunter-gatherers is not as clear-cut as once thought, and it seems that there were (and in some places of the world there still are) many possibilities in between: (p. 413) "a major problem with the demic diffusion model (...) of agriculture has been its focus on the transition to farming as some kind of unique sequence of movements in an otherwise static world".

In a previous post I commented on the importance of agricultural expansion in connection with language. The demic diffusion model, involving the expansion of farming-pastoralist populations into vast territories, has been applied in different linguistic areas of the world, e.g. the Indo-European area (Renfrew's Anatolian Hypothesis), and especially the Afro-Asiatic group, where it has found a more general acceptance (I'll write a post about this group of languages in the near future). The question is: can the expansion of agriculture be the main reason behind the expansion of these language groups? Some linguistic arguments have been used in order to support this view, especially the ones provided by linguistic paleontology: it has been suggested, for example that the IE and Afro-Asiatic 'proto-languages' have some common vocabulary for farming, which would imply that the speakers of this proto-language were already farmers. I have had the chance to look at some of these lists of 'farming vocabulary', e.g. the ones proposed for IE, and it seems to me that they are anything but conclusive. In many cases, they are words that could perfectly fit a foraging type of subsistence, especially if we understand foraging in more complex ways, as Barker has suggested in his book; in other cases, they can be explained as examples of language diffusion.The Afro-Asiatic and Nilo-Saharan groups of languages seem especially interesting as a testing ground for the various proposals (p. 322) "the theory that the earliest speakers [of Afro-Asiatic languages] were farmers depends on the assumption that terms identified in proto-Afro-Asiatic such as grains, grasses, and grinders must automatically refer to domesticates and their processing, rather than to wild plant collecting, whereas in the formative stages of the language group a correlation with wild grass collectors is just as likely (...) In both language groups, terminologies with indisputable agricultural connotations can only be identified in the more developed stages of the member families".

As I said, I will discuss some aspects of the Afro-Asian and Nilo-Saharan groups in some future posts.

Note: the second picture shows a hunting scene from the Cova dels Cavalls, in the Valltorta Valley near Castellón (Spain), a beautiful example of the Levantine prehistoric art. I really enjoyed visiting this place, and the nearby museum, a few years ago; the third picture, taken by me the other day, is from the Abrigo del Ciervo, near Dos Aguas (Valencia).

Populations and languages: the Strait of Gibraltar

Many years ago I made a trip to Gibraltar. At that time I was a post-graduate student at the University of Valencia, and one of the courses I took was about dialectology and sociolinguistics. We had to do some research as the final assignment of the course and in my group we decided to go to Gibraltar to do some field-work about the linguistic situation of this peculiar place. We spent three days there, with our questionnaires and interviews, and we also had time to do some sightseeing: we walked around the city, we saw the famous monkeys and we finally climbed the Rock, from where we had some spectacular views of both Spain and the African coast, which is a mere 14 km away. We can imagine that, throughout history and prehistory, many humans living on either side of the Strait must have felt curious to know about the land that they could see across the water, and this curiosity could have led to a significant movement of human populations in both directions.
The surprising fact, however, is that the Strait of Gibraltar has been a barrier for human migration in all ages, especially in prehistory. The main reason for this is geological: the Strait of Gibraltar has remained as it is now for the last 5 million years, even at the various glacial ages, where the sea level lowered significantly all over the world. We also have other types of evidence, e.g. the archaeological record, but the most important confirmation has come from population genetics. I recently read an interesting article about this subject: Bosch et al, 2001, High-Resolution Analysis of Human Y-Chromosome Variation Shows a Sharp Discontinuity and Limited Gene Flow between Northwestern Africa and the Iberian Peninsula, American Journal of Human Genetics, 68:1019-1029). In this article, the authors analysed the genetic components of various populations in Spain and Morocco, combined with other evidence from archaeology and history, and reached a series of interesting conclusions. It seems for example that in both cases, the populations of today are mostly the descendants of the people who lived in these areas in the Paleolithic, with a minor impact of migration from the Middle East, probably associated with Neolithic expansion. On the other hand, the genetic components of Iberian and NW African populations show that they come from different origins. Human settlement in Iberia is connected with the expansion of modern humans into Europe from Eurasia or Anatolia, whereas the population of NW Africa is mostly connected with components that originated in the African continent. The gene flow across the Strait of Gibraltar is not considered relevant; it can be estimated at about 5%, and it could, at least partially, show the traces of some recent historical phenomena, like the expansion of the Roman Empire or the Arabic conquest of Iberia. There’s no doubt that the Strait of Gibraltar, as a natural barrier, has played a decisive role in the distribution of human populations, both for modern humans and for older types of hominids. Instead of crossing the 14 km stretch of water that separates Africa from Europe, it took humans a few thousand years to go all the way to the Middle East and eastern Europe until they reached the Iberian Peninsula. This is what I would call a ‘Grand Tour’.

Now, what are the linguistic consequences of all this? Is there also a linguistic barrier as well? Has this language barrier existed from prehistoric times? In a previous post I wrote about the expansion of Arabic as a consequence of the Islamic Empire. The main conclusion I reached was that Arabic dialects are spoken today only in areas where other Afro-Asiatic languages (formerly known as Hamito-Semitic) were already spoken before the arrival of the Arabs, and not in areas where there were other types of languages, e.g. in Persia or Iberia. I’m not sure if anyone had realised this simple fact before, but it looks quite clear in my opinion. The important factor here is affinity. The language of the conquerors (in this case Arabic) has a varying degree of influence on the languages of the conquered depending on the affinity between them. When the Arabs arrived in northern Africa they found Berber-speaking populations, and Berber languages belong to the Afro-Asiatic group. The subsequent process of hybridization led to the linguistic situation that we find in the area today, with a series of dialects which are considered regional variations of Arabic (with the exception of the areas where Berber languages have survived until today). What about Iberia? The languages spoken in this territory were quite different from Arabic; they were connected with Latin, an Indo-European language belonging to the Italic group. The Islamic conquest brought about a process of hybridization, with a significant exchange of linguistic (mainly lexical) material in both directions, as can be seen in the vocabulary of Spanish, Portuguese and other Ibero-Romance languages, and also in many features of the Hispano-Arabic dialect. However, Arabic and Romance languages were always perceived as something different. There were not enough opportunities for hybridization to produce significant hybrids between them; people spoke one of the languages, or both, but not a mixture of them (except perhaps in some local, pidgin-like cases). Another example of the importance of affinity in situations of language contact can be seen in the Roman conquest. The influence of the Romans was linguistically relevant in the Iberian Peninsula, where there was already a background of Indo-European languages, whereas it was rather insignificant in northern Africa, with no Indo-European background (see this post for more details and some maps).

It seems therefore that the population/language distribution in NW Africa and Iberia corresponds to a pattern that dates back to Paleolithic times, when modern humans arrived in these areas via different routes. The Strait of Gibraltar, as a natural barrier, was the main factor behind the whole process, limiting the possibilities of genetic or cultural exchange. Later developments, associated with the rise and fall of empires and the expansion of religions, were not strong enough to change the overall picture.

Notes on the illustrations:
- First picture: The Strait of Gibraltar from Spain. Source: OjoDigital (here).
- Second picture: The Strait of Gibraltar and the Alboran Sea. Source: NASA (here).

The expansion of Arabic

abic So far, this blog has focused mainly on European languages, but it is clear that the Continuity/Hybridization Model (another name for the Continuity Theory) can be applied to other groups of languages. Today’s post is about a non-European, and also non-Indo-European language: Arabic, which belongs to the Semitic group (Afro-Asiatic family).

Arabic dialects are spoken today in many areas of the Near East and Northern Africa, as we can see in the following map:

(Source: Wikipedia. Click here for a larger image and further details)

This linguistic situation is a direct consequence of a historical fact: the emergence of the Islamic Empire, which originated in Arabia in the 7th c. AD and quickly expanded to many other territories, carrying with it both Islam and Arabic. The next image shows the extension of this empire at different stages:

(Source: Wikipedia. Click here for a larger image and further details).

Both maps look quite similar but we can spot some significant differences between them. We only find Arabic dialects in places where Semitic languages, or at least other Afro-Asiatic languages, were already spoken before the conquest (e.g. northern Africa, Mesopotamia, Arabia). On the contrary, there are some areas of the Empire where no Arabic dialects have survived, e.g. Persia, Kurdistan, south-eastern Anatolia and the Iberian Peninsula. The languages spoken in these areas before the arrival of the Arabs, and the languages spoken there today are not Semitic or Afro-Asiatic; they belong to other groups (mainly Indo-European). Historical events such as military conquest, and the dominance of an intrusive elite over extensive territories, do not seem to affect the basic pattern of continuity, at least at the language-group level. If we apply this type of analysis to other processes of expansion the conclusions are very similar, as we have seen, for example, in a series of posts about language distribution in the Roman Empire (you can read them here: What the Romans spoke; Language continuity in Europe (II): Switzerland; Romance Languages before the Romans). In the context of ancient times, the language of the conquerors (associated with social prestige and political power, and also with religion) is always a very influential factor, but there is not a generalized process of language substitution. What we find instead is a process of language hybridization. If the language of the conquered has a high degree of affinity with the language of the conquerors, i.e. if they belong to the same language group, it is much more likely that the final result will be a dialect of the new language. In the absence of this kinship, the normal scenario is the continuity of the pre-existing dialects. In fact, language continuity is always present: Afro-Asiatic dialects were spoken in northern Africa and the Near-East before the Roman conquest, and are still spoken today, in the form of Arab dialects created through a process of hybridization.

Last Edit: 13th October 2008.