Links

Posting introduction and Table of Contents are on the Title Page.
Reference material on the Vennemann Th. Handbook: Indo-European Northwest Contact and Prehistory 1M

The posting's notes and explanations, added to the text of the author and not noted specially, are highlighted in blue font, shown in (blue italics) in parentheses and in blue boxes. Page numbers are shown at the end of the page in blue.

CHAPTER TWO — PERIPHERAL MOVEMENTS

Neolithic Origins

The exact course of events leading to the development of the Neolithic (i.e. farming) farming economy is not a critical issue here. It is enough just to set the probable times and places of its beginning in order to chart its subsequent spread with some degree of accuracy, or predictability.
Still, a brief summary helps to explain these times and places. This summation is somewhat original, but it is built only from well-known facts.

Farming is simply a logical extension of a way of life that developed in a series of quite natural steps. Within the scope of a hunting-and-gathering economy an extreme specialization on a limited range of food sources is reasonable and practical — if such foods occur in great abundance. One example of this is the Natufian cereal-grain harvesters known from about 11,000 years ago in the Near East (which has popularly come to be known as the "Middle East" in recent years). In dense natural stands of grain a year's food supply can be gathered in just three weeks. The required technology for harvesting, threshing, storing, and grinding would easily be developed in a short time when their need becomes evident.
27

More important is the fact that a tribal society of 500 people, more or less, can now find sufficient food in as little as 500 sq km instead of the more usual ten times that area. On the one hand this promotes the ease of sedentary life, and on the other hand it severely restricts wide-ranging hunting practices. Cereals do not provide a satisfactory range of amino acids, so a dependable source of animal protein must be found to make up for diminishing hunting possibilities. Capture and control of some native animals is the obvious solution. Sheep and goats are available in the Near East, and are ideal herd animals for this purpose. Their domestication appears to coincide with intensive cereal harvesting, and logically precedes the earliest actual farming.

This combination provides a way of life that, at first glance, would seem to be ideal and potentially lasting. It has the advantages of little need for movement, allowing permanent living structures with accumulation of property, and a dense population to resist inroads by other people. Actually it is unstable, and a new factor soon appears to upset the equilibrium. The labor requirements for tending animals and for the rapid harvest now make children for the first time an economic asset rather than a liability. Abnormal population pressure is an automatic result.

When the natural stands of grains are fully exploited, any expansion of the population, and extension of this way of life, is possible only by deliberately cultivating these grains in new places. The idea that seeds grew into plants would have been common knowledge to all hunting and gathering peoples. What was needed was a reason to encourage such growth artificially. The development of farming was not something new in itself, but rather it was a method of continuing a familiar way of life (harvesting) in the face of changed circumstances (overpopulation).
28

It is also true that minor climatic changes may have shifted natural grains out of the territory of a given tribe, thus forcing them to cultivate in order to continue their way of life. Moving with the grains was not an option because other tribes would have occupied the desired territory. Both factors, overpopulation and shifting vegetation zones, would lead people into cultivation simply in order to continue the way they had been living. This is a cultural application of Romer's Rule from paleontology, where an innovation is initially conservative and only later is developed into something truly revolutionary.

The natural distribution of wild emmer wheat can be used for the most reasonable indication of where the commitment to cereal-grain farming began, as shown in Fig. 6 (from Zohary 1969, cited by Dennel 1983). Farming would begin within an area of natural grain where the stands are sparse, or in patches where it is missing. Around the periphery of such stands is the major place where deliberate farming would serve to artificially extend the otherwise natural harvest. Emmer wheat is probably the best food source, and is one of those most consistently found in early Neolithic (i.e. farming) sites. One might presume it was possession of this grain that tipped the balance between continued foraging and a commitment to farming. This presumption is well supported by the natural distribution of this grain. It occurs in two major locations in the Near East that correspond exactly with my postulated sources for the two major language families that originated in this area. These are thus the original homelands of the Afro-Asian and Indo-European peoples (locations P and A in Fig. 6).

Fig. 6. Neolithic (i.e. farming) origin locations. Recent areas of wild emmer wheat are shown, with the presumption that they were similarly distributed 10,000 years ago. These are labeled by the initial letters of Palestine, Anatolia, Iran, and Caucasus (uncertain).

30

Archeological evidence today shows the earliest Neolithic (i.e. farming) to be within and around these two locations. The earliest of all appears to be in the southern, or Palestinian, location (Wenke 1980). Exact timing of each small advance in farming technology for the area is not now possible. Still, it would be quite acceptable to postulate the spread of farming from the periphery of the Palestine location at just 150 years before it spread out from the Anatolian location. (This last-named area is actually half inside modern Iraq, but its significance to the present study is concentrated more in its Turkish part.)

This timing, if correct, suggests a south-to-north spread of the basic idea by cultural or stimulus diffusion at the rate of 400 km in those 150 years. This is twice as fast as farming itself should have spread. It is also possible that the process developed independently in each place at these slightly different times. It is their close proximity in time and space that suggests a connection.

It is not immediately evident exactly what essential "idea" spread from Palestine to Anatolia. Certainly there was no discovery that seeds grow into plants — if hunters know that copulation leads to babies, then understanding plant reproduction would be child's play. It is not likely that any equipment was critical, as most of this should have been available, or at least easy to develop. The required herd animals were abundant in both areas some time earlier. Possibly the essential discovery was the calendar — knowledge of how to determine the exact time of year that the seeds should be planted in order to achieve the maximum harvest. (I wish to thank Diane Horton for this thought.)

The scenario postulated here is that the correct sowing times, procedures, and equipment were developed first in the Palestinian emmer wheat location. The timing would be that farming began its spread from here at 8150 BC. This is not to be taken as a definite date — it could have varied by a few centuries either way. It is chosen in part for its consistency with all other dates used in this reconstruction. If this date of 8150 BC is changed, then all other dates for the initial Neolithic (i.e. farming) spread must be correspondingly shifted.

By the date of 8000 BC the farmers out of the Palestine location should have spread their settlements 200 km in all directions beyond the core area of emmer wheat. Also at 8000 BC the knowledge of when to plant (or some other critical item) is presumed to have reached the western edge of the Anatolian emmer wheat location. At this time the people on this edge would likewise begin their own expansion at the same basic rate of 200 km every 150 years.
32

The calendrical knowledge probably spread out from the Palestine location faster than the expansion of the people themselves — perhaps at twice their rate. Alternatively, this knowledge may have reached the edge of the Anatolian location directly ahead of the spreading farmers — always 200 km ahead of the actual settlement frontier. This knowledge of the planting calendar would be of use only to those harvesters inhabiting the edge of a natural emmer wheat stand who are in a position to expand outward under population pressure. In other circumstances foragers normally show little interest in copying the way of life of neighboring farmers.

Selecting 150 years as the lead time for Palestinian farming was not just based on archeological suggestions. This difference is just the amount of time that automatically results in the formation of a subsequent boundary between Palestinian and Anatolian farmers that corresponds to the later language line between Afro-Asians and Indo-Europeans. It also turns out that this same lead time ultimately results in a much later meeting between North African and European farmers at the other end of the Mediterranean, in the Pyrenees Mountains, which is also linguistically critical. The archeological data are not precise enough to say that this gap was just 150 years, but it could not have been much different. This may eventually be determined exactly.

Palestinian Farmers

The spread of farming out of the Palestine location, beginning at 8150 BC, is followed here only in those details that are of significance to the development of European languages (Fig. 7). It is assumed from the outset that this expansion was by the original Afro-Asian speakers. This assignment was not arbitrarily made, but it became evident from observing the automatic distribution of farmers from this source, and how this distribution compares with that of the recent Afro-Asian language family.
33

Fig. 7. Neolithic expansion from Palestine. Heavy lines trace the two routes of Afro-Asian spread that are stopped where they meet expanding Indo-Europeans — in Syria and the Pyrenees. Dates are given at each 200-km unit of progress. Afro-Asian territory is left blank, while other speech areas are stippled.

The northern edge of the Palestine location is just 400 km from the western tip of the Anatolian emmer wheat location. At 8000 BC Afro-Asian farmers would have covered half this distance, just at the time the presumed Indo-Europeans would have begun their spread from here. In 75 more years they would meet — Afro-Asians at 300 km from their source and Indo-Europeans at 100 km from their source. This meeting line would then trend to the north-west and reach the Mediterranean in 7800 BC at its northeastern corner. This effectively ends the spread of Afro- Asian farmers at this point, and gives farmers from the Anatolian location free access through the rest of Anatolia. This leads ultimately to their unimpeded entry into Europe, hence the first specific reason to think these are the original Indo-Europeans.

33

Moving to the east, it may be expected that farming would have extended progressively outward from along the southern edge of the Anatolian location. The calendrical knowledge would not have spread so easily over the Syrian Desert where native cereals were not common. In that case the Afro-Asian frontier would soon move close to the southern edge of the Anatolian emmer wheat location and continue to the head of the Persian Gulf. This automatically results in a language line that runs very nearly along the present Afro-Asian boundary. There is no obvious reason why this language line should have moved greatly in the last 10,000 years. (Compare the line as shown in Figs. 1 and 7.)

Afro-Asian Spread

Further expansion of the Afro-Asian farmers would be into all accessible places to the south and west of their initial area of occupation. Their move into Arabia and the main body of Africa has no relevance to the present study. To the west, in the most direct line across northern Africa, these farmers would have advanced to Gibraltar and beyond (Fig. 7). At a steady rate of 200 km in every 150 years they would cover the 4000 km to Gibraltar in 3000 years, arriving there at 5150 BC.

The practice of pastoralism with sheep and goats, but without farming, is viable by itself. It should also have spread out from the Palestinian location into the drier areas that could not be tilled. It is quite likely that a different language group was involved in this economy, but it doesn’t matter here. These shepherds should have spread just twice as fast as the farmers, possibly reaching Gibraltar 1500 years earlier, at 6650 BC. There is no reason to suppose they ever developed seaworthy craft, and so did not transport themselves or their stock into Europe. These people would ethnically disappear, at least in northern Africa, when the tide of farmers eventually rolled over and incorporated them.
34

A similar group of shepherds would also be at Gibraltar when the farmers arrived from northern Africa. These are the Sea Shepherds or Boat People who developed in Europe and who spread along the northern Mediterranean and Atlantic coasts. (These are not to be confused with the "Sea People" of 1200 BC in the eastern Mediterranean, who do not appear in this reconstruction.) Unlike their African counterparts, these European pastoralists had seaworthy craft. While the African shepherds could not enter Europe for lack of boats, the European shepherds could not enter Africa because all the suitable environments would already be occupied. The Sea Shepherds are described in a later chapter; their importance at this point is that they were in a position, since 5915 BC, to transport the African farmers and their equipment into Iberia.

These Afro-Asians would be the first serious agriculturalists in Iberia — the Sea Shepherds appear not to have engaged in any significant farming. The farmers' expansion over Iberia would be an extension of their spread across northern Africa, and which continued at the same rate. Here they no doubt overran and absorbed the Sea Shepherds in their coastal locations.

Across Iberia to the north-east the farmers would progress four units of 200 km each in the next 600 years. In 4550 BC they would reach the lower Ebro River and meet an incoming wave of farmers from the opposite direction. These new farmers ultimately stemmed from the Anatolian location, and are here identified as Indo-Europeans. The two groups would have had essentially the same economic base of crops, domestic animals, and technology. Their methods of land utilization and consequent population density would have been the same. Neither group would have any significant advantage over the other, and each was living on familiar home territory. Their meeting line becomes the second place where people of Palestinian and Anatolian origins encountered one another. This time it was 4800 km from the Afro-Asian starting point. A stable language boundary would exist here until some other factor forces a change.
35

Over the next 50 years this meeting line between two farming peoples would arc around to the north-west and end against the Pyrenees. This leaves a triangular area in northeastern Iberia in Indo-European hands. This site ultimately correlates with the origin place of Catalan, the recent language of Catalonia.

Along a different path through Iberia, somewhat to the west of the one just described, the Afro-Asian farmers would advance to the western end of the Pyrenees. The Indo-Europeans reach this area measurably later, so the farmers from Iberia cross the mountains. Their predictable meeting line is about 50 km up the modern French coast, and arcs inland about 100 km to where it reaches the northern flank of the Pyrenees. This meeting line of 4450 BC so accurately follows the recent Basque language boundary that its identification appears unavoidable. It has been suggested that Basques once occupied much more territory in the past — the Gaulic province of Aquitaine. It is also possible that this larger area was only controlled by Basque speakers long enough to establish some place names.

This identification of Basque with Afro-Asian is based on geography rather than on any good linguistic evidence. Tovar (1970) and others would distantly associate Basque with Hamitic, and most authorities tentatively include Hamitic in Afro-Asian. These tenuous connections conform well with this reconstruction. Basques are known to have been in their present location since at least 3000 BC according to Wilbur (1980); there is no evidence against extending this back another 1500 years.

A short segment of the Afro-Asian frontier would also follow the mountain divide in its west-central part. By 4500 BC all of Iberia (except for its northeastern corner) would be occupied by these farmers of presumed Afro-Asian affiliation. They would naturally subdivide into a dialect mesh over the next few thousand years. They would also have included large areas that had been thinly populated by the much earlier Sea Shepherds.
More recently their incorporation into the Empire of Carthagewould likely introduce Phoenician in many places.
36

Finally, an influx of Celtic settlers in the last few centuries BC would be added to help create a very complex linguistic picture for the peninsula. Almost all of this becomes irrelevant somewhat later when Latin became the nearly universal language of the area.

These and still later events are discussed in other parts of this study. The major point established here is the exact course of the first meeting line between Afro-Asian and Indo-European farmers in Europe. Any other accounting for this language line seems to involve far more speculation than is offered here.

Caucasian Farmers

It is calculated here that farming began to spread out from the western edge of the Anatolian location at 8000 BC, just 150 years after it began from the Palestinian location. From the Anatolian source the Neolithic spread (i.e. farming, not necessarily with Neolithic toolbox) would have been mainly to the west and east, and only to a lesser degree to the north and south. Before detailing the major Indo-European expansion it is necessary to deal with the northern flank. Here there appears to have been a third source for the Neolithic (i.e. farming) that must be pinned down.

To the north-east of the Anatolian location there is another known stand of wild emmer wheat. It is in a large valley that surrounds the northwestern corner of modern Iran (Fig. 6, location I). An obvious idea would be to associate this location with the origin of the Caucasian language family. While this might be true, there are some good reasons to doubt this, and instead to place the Caucasian origin well to the north, at location C in Fig. 6. (The reader is reminded not to confuse Caucasian languages with Caucasoid race.)

Fig. 8. Neolithic (i.e. farming) expansion from the Caucasus. Heavy lines trace the routes of Caucasian and Altaic spread from their proposed sources. The transfer of pastoralism to Altaic speakers is indicated by the letter T at the end of the dashed line where deep soils begin. Dates are given at each 200-km unit of progress. Caucasian and Altaic territory is left blank, while other speech areas are stippled.

38

The Iranian location, marked I in Fig. 6, is entirely occupied by an Indo-European speech (Armenian) today, rather than Caucasian — a difficult though not impossible problem to get around. A language that spread from this location would also tend to put its speakers much more to the east than is the present Caucasian distribution. More likely, the expanding Neolithic (i.e. farming) wave from the Anatolian location simply overran most of this area before its occupants had time to build up a significant farming population. Also, it is a closed valley that would be entered by the presumed Indo-Europeans from its open end, thus making egress difficult.

The present Caucasian languages are found on the north and south sides of the Caucasus Mountains throughout all but the eastern end of the range. The simplest accounting for this would be to postulate yet another stand of wild emmer wheat, or its functional equivalent, to be found on the south side of the mountains, toward their western end (location C in Fig. 6).

Except for heavier rainfall, ecological conditions here are similar to those at the other wheat locations. According to Chard (1975:224) there is wild emmer wheat within this general area of the Soviet Union, but he gave no map. If deliberate cultivation appeared and spread out from here just 200 years later than from the presumed Anatolian source, then all of the major language distributions work out correctly. This delay of 200 years is otherwise unsupported by any direct data that I am aware of. It also appears to imply an independent development of the planting calendar because the distance involved is too great for the transmission speed suggested above.

Given this presumed starting time of 7800 BC, Caucasian­speaking farmers would have spread out in all directions until they encountered barriers of geography or other farmers. They would expand their frontier to the south and then west along the southern shore of the Black Sea for a distance of 400 km in 300 years. Farmers from the Anatolian location presumably started 200 years earlier and expanded their holdings directly north (among other directions). With the standard delays imposed for cold winters here, they would reach the same point on the Black Sea coast in 500 years. They meet at 7500 BC, with the presumed Indo-Europeans (or rather with the actual Caucasian­speaking forefathers of Hatti and Urartans) being in the process of occupying most of Anatolia, and covering all access to the west. The Caucasians would occupy just the northeastern corner of modern Turkey (i.e. most of the N. Anatolia) along the Black Sea, much  (i.e. west of) as they do today (Fig. 8).
39

The timing difference of just 200 years is critical here, and might someday be directly determined. If the Indo-European lead time were any less, the Caucasian farmers would have had access to western Anatolia and eventually into Europe proper. Clearly this did not happen. If the Indo-European lead time were any greater, then the next-described result of the Caucasian spread would not have been completed. In the following section it is shown that this result did occur.

Altaic Shepherds

The only other direction of Caucasian expansion that is of any concern here is that to the north-west along the eastern shore of the Black Sea (Black Sea before 5600 BC was a freshwater lake about 2/3 of its present size). After advancing some 280 km along this route, and using 350 years, they would enter the deep steppe soils that could not be farmed with the simple plow in use at that time. (This area would remain largely unfarmed for another 8000 years.) At 7450 BC these farmers must halt their advance, but their practice of animal husbandry would automatically be picked up by the local Mesolithic peoples. These would be the natives of the Kuban region, just to the east of the Sea of Azov (Fig. 8).

This is the first part of the Eurasian Steppe land that would be reached by people with domestic sheep and goats. From here, people with sheep herding and other Neolithic traits, except for serious farming, were able to spread over the entire grassland and semidesert area from the Ukraine to Manchuria. Almost the whole area is today inhabited by speakers of the Altaic language family, so-named from the Altai Mountains located near its center.
40

It is usually assumed that their languages originated here, on the border between the Turkic (western) and Mongolic (eastern) branches of the family. There is no obvious mechanism whereby the original language type of this entire area could have been replaced by another single type at any time after the introduction of this pastoral economy. The obvious conclusion is that the Altaic family actually originated with the Mesolithic inhabitants of the Kuban.

Altaic expansion to the east is of no concern here. To the north and west the sheep-herding economy would be very successful throughout the relatively open country. Their progress would halt at the fully forested lands where sheep are not suited. Sheep occur in most of Europe today only in association with farmers; previously there were no wild forms in these forests.

Altaic expansion within Europe would be at the rate of 75 years, plus temperature and daylight penalties, for every 200 km. At this speed they reach the fully forested zone in modern Russia long before any Neolithic (i.e. farming; Mesolithic period in Europe is dated from 9660 to 5000 BC, overlapping farming period) farmers enter these forests. Traveling due north out of the Caucasian source, these shepherds would reach the border of the forest zone at 6565 BC; this is some 2500 years before the European farmers reach that point, coming from the west with their own sheep.

If these Altaics had long ago abandoned all serious farming, which is most likely, they would be unable to penetrate these forests. They did not even have the option of reverting to a hunting and gathering economy because this niche was already filled here by the native inhabitants. If there had been any significant spread of Altaic speech into these forests there should be clear indication of this in recent linguistic distributions, but there is none.

The most important route of Altaic expansion is that which follows the most direct line around the Black Sea, through the Crimea, and into the eastern Balkans, on to where the grasslands and deep soils end. At pastoralist advance rates, these shepherds would reach the border of this zone, just inside the northeastern corner of modern Bulgaria, at 6825 BC.
41

(Putative) Indo-European farmers, following standard calculations, should arrive at this same line from the opposite direction (from the south) at 6830 BC. The two meetings with (putative) Indo-Europeans, first by the Caucasians in northeastern Turkey, then by the Altaics in Bulgaria, are exactly timed and serve to determine the gap of 200 years between their beginnings. (See Fig. 8.)

The precise timing of the (putative) Indo-European and Caucasian meeting in Anatolia has already been described, and this allows for no significant leeway in either direction. The same is true for the Altaic encounter in the Balkans. If the pastoralists had arrived earlier in northeastern Bulgaria, the less-than-dense forest there would have allowed some further advance of an archaic Turkic language. Just such a language, Gagauz, is found today in two Balkan enclaves — in eastern Romania and just inside northeastern Bulgaria, but not beyond. They probably date from this time. (Gagauz is usually assumed to be a recent import from Turkey, but its deviant nature and location fit much better with the interpretation given here.)

If the Altaic shepherds had arrived in (Danube) Bulgaria much later, then the Indo-European farmers would have transferred their pastoralism to the local Mesolithics. These would necessarily expand northward to meet the incoming Altaics at some point.

This would then create an unknown ethnic group of which no trace remains. It would also make the Gagauz difficult to explain. (In my earlier study in 1981 I did not consider the possibility of a Caucasian origin of the Neolithic (i.e. farming) at all, and assumed that the Altaics arose from this Balkan, source. This was clearly incorrect even though the results are largely the same.)

The importance of this part of the geographic reconstruction lies in determining the ethnic identity of the early inhabitants of the Pontic Steppes, From the earliest Neolithic (i.e. farming), until the area was largely overrun by Slavic farmers in the Christian Era, these nomads were necessarily Altaics. From Kurgans through Scythians to Sarmatians the great bulk of the population would have spoken Altaic (Turkic) languages. The fact that Iranian personal names have been identified among Scythian leaders (Crossland 1967) is no proof of what the popular speech may have been. (If personal names were used as our only guide, we would conclude that much of Europe today is Hebrew speaking.)

43

This reasoning effectively removes the archeologically well- known Kurgans of southern Russia from being a possible source of Indo-European speech. Leaving aside the problems of how the Kurgans could have spread their language, and the fact that the initial spread of farming is the only reasonable mechanism for a change of this magnitude, we would still have the problem of explaining what happened to the Kurgans themselves. If they had been the source of Indo-European speech they would doubtless have spread it over the entire Eurasian grasslands as well. (The existence of Tocharian might be used to argue in favor of this.) Then some natural mechanism must be found to replace this supposed Kurgan speech with Altaic throughout its range. Nothing of the kind seems to be available, especially after the development of horse-based military procedures in this area.

Indo-Europeans in Asia

The next major non-Indo-European language group to deal with is the one I call Aegean. This is the speech that was apparently spread by the Sea Shepherds that have already been mentioned. (Again I apologize to the reader for using a name similar to the later "Sea People," with whom there is no connection.) Before these Sea Shepherds can be described it is necessary to review their background and origin. This stems from the Neolithic farmers who first advanced their frontier toward Europe from a Near Eastern source. These would be the early Indo-Europeans who transferred their sheep-raising technology to some local natives in the Aegean area, much like the Caucasian farmers did to their Altaic neighbors.
43

Three sources for the origin of Neolithic (i.e. farming) farming have been tentatively located in the Near East. The starting times for their expansions and their initial contacts with one another have also been laid out in the geographically most probable pattern. The spreads from two of these sources, Palestine and the Caucasus, have already been followed through as far as they should have affected the languages of Europe.

From the perspective of the overall Neolithic (i.e. farming) invasion of Europe, the Palestinian source of presumed Afro-Asians produced something like a left flank of this invasion. The Sea Shepherds may almost be viewed as an adjunct to this left flank. The Caucasus source, with its Altaic off-shoot, forms something of a right flank of this invasion. Still another group, the yet-to- be-described Uralics, are an adjunct to this right flank. These two flanks almost surround the core of Europe. The main weight of the central invasion, by direct reasoning and by elimination, must have been the Indo-Europeans themselves.

The spread of farmers from the Anatolian location has been reconstructed in terms of its meeting lines with those farmers stemming from Palestine to the south and from the Caucasus to the north. These lines are in reasonable agreement with the recent distributions of the language families that are involved. The central group of presumed Indo-Europeans would have access to all areas both east and west of its source, and between the areas occupied by people from the other two sources on the north and south. This (presumed) Indo-European spread may now be followed through its initial stages to just short of the actual invasion of Europe.

(Presumed) Indo-Europeans would have unimpeded access to the east over all of modern Iran and into the Indian subcontinent. Their (Indo-Europeans) identification with the first farmers in that area would appear to be unavoidable. The writings of the Harappan civilization accordingly would have to be an early form of Indo-European? their decipherment could be an important test of my ideas. The supposed introduction of this speech into India from invasions around 1800 BC is unsupported by any direct evidence (now it is, though). Most indirect evidence, including the present study, argues against the likelihood of a total language replacement being effected by a relatively small number of primitive invaders at this time. (The brief appearance of Indic writings in Mesopotamia in 1800 BC may represent a late military incursion, rather than a record of these people on their way to India.)
44

The (presumed) Indo-European spread that is of more concern here is the west through Anatolia, where they also had unimpeded access. The most direct route to Europe would first be straight west in order to skirt the central highlands where winter temperatures would cause delays. Then in western Anatolia a northwesterly course would bring these farmers into Europe just short of 7100 BC. Within that time frame most of the rest of Anatolia would also be settled.

One important result of this proposed initial occupation by (presumed) Indo-European farmers is that we need not look elsewhere for the origin of the old Anatolian languages. Hittite, Palaic, and Luwian were all based here since the arrival of the first Neolithic farmers (That is funny, how do we know if the first writing in the whole world appeared 3500 years later, at 3600 BC? Hittite writing, no earlier than ca. 18th century BC; Palaic ca. 1600 BC, Luwian 14th-13th cc. BC; all extinct before 600 BC). Later Anatolian languages like Lycian, Phrygian, and Armenian are probably also equally native to the region (Phrygian ~ Armenian came from Balkans according to contemporary sources). The ultimate fate of the Anatolian languages with the introduction of Turkish is covered in a later chapter.

A large area of grassland in central Anatolia was probably not easily plowed, and local Mesolithic people here should have adopted sheep herding. (Presumed) Indo-European farmers then would have spread around both sides of this area, meeting each other again to the north-west of it. The ethnic identity in this grassland area is almost certainly Hattic. A short digression into some later historic events is necessary at this point.

Singer (1981) has established the Hattic homeland (ca. 18th c. BC – ca. 12th c. BC) to be exactly here in the Halys River basin (now Kizil Irmak), which in turn happens to be the core area of the ancient Hittite Empire. This odd contrast of an (presumed) Indo-European empire based on a non-Indo-European homeland has been neatly solved. Steiner (1981) showed that the Hittite Empire was in fact Hattic-speaking in terms of its rulers, religion, and core population. The Indo-European Hittite (more properly "Nesite") was merely the diplomatic and commercial language of central Anatolia from earlier times that continued to be used by the Hattic rulers. Once horses came into use in Asia Minor, this pastureland of central Anatolia would automatically become the basis of a major military power. It would be of non-Indo-European speech only if the (presumed) Indo-Europeans were the farmers who could not have settled this particular area.
45

Sea Shepherds

The next event of importance is the presumed transfer of some Neolithic (i.e. farming) traits from the (presumed) Indo-Europeans to some of the local Mesolithic inhabitants off the southwestern coast of Anatolia. This is the most reasonable starting place for the Sea Shepherds or Boat People, as I variously call them. This is an archeologically well-known group who spread early Neolithic (farming or polished stone tools?) traits around much of the European coastal areas.

The development of seaworthy craft may be expected among any Mesolithic peoples living where there are numerous islands within sight of the coast and of each other. These invite exploration and exploitation, and the boats themselves can also support fishing. Boats could have been devised anywhere that bodies of water occur, but any degree of sophistication depends on their usefulness. The Aegean is one of the most ideal locations for such a development, and there is archeological information that supports this as far back as 6500 BC, and possibly farther (Dennell 1983:116). This is the first such maritime zone that would be encountered by early Neolithic farmers.

The advancing front of (presumed) Indo-Europeans along the Mediterranean coast of Anatolia would reach the first significant Aegean island, Rhodes, in 7300 BC. Most of these islands are not well suited to cereal-grain farming on a large scale, but would be excellent for sheep herding. It would appear natural that the farmers would be delayed here, and that the Mesolithic natives would adopt their domestic sheep keeping. This would involve a transfer of mainly the pastoral aspect of the Neolithic (i.e. farming) economy, just as in the Caucasian-Altaic transfer and the others that are indicated here.
46

These new Sea Shepherds would have the opportunity to spread around the coasts and all accessible islands, which means the entire Aegean Sea, ahead of the advance of Neolithic farming. They would occupy any and all places along the coasts that offered good sheep pasturage. They would not tend to move far inland because of the forests, and also so as not to lose the other major aspect of their economy — fishing.

The ethnic identity of the Sea Shepherds is not that of any recognized linguistic group of today. It would most likely relate to the Pelasgian language of Greece which is known from geographic place names that are not Greek. It may well be recorded in the Minoan hieroglyphs, and just possibly in Linear A. This Aegean language would be carried by the Sea Shepherds in the course of their expansion, changing over time, and be distributed sporadically around the seacoasts of much of Europe. The fishing aspect of their economy would be lost whenever they spread inland. Their sheep herding, without farming, was only sporadically viable in mostly forested Europe. Eventually they would be met by the (presumed) Indo-European and Afro-Asian farmers, and were overrun and easily incorporated. They would probably have no more than minor linguistic impact except for one location near the end of their range. (See Fig. 9 for their distribution.)

Fig. 9. Spread of the Sea Shepherds. The dashed line traces the progress of Aegean Boat People, from the Island of Rhodes around the Mediterranean and Atlantic coasts of Europe, and ending in the Baltic. Suggested dates are given at each 200-km unit of advance. Cross-hatching indicates areas that were occupied well after the initial wave had passed through.

The starting time for the Sea Shepherd economy would be at 7300 BC when (presumed) Indo-European farmers first reached this area of developed boats. I would allow 200 years for the integration of this way of life and for its spread throughout the Aegean. This interval is arbitrary, and its actual length would have no critical bearing on subsequent events.

(Presumed) Indo-European farmers should be at the Dardanelles about 7050 BC, presumably just 50 years before the Sea Shepherds reached that point. They (presumed Indo-European farmers) can cross into Europe with the aid of other Mesolithic boat people who would have lived there. With this timing, a new group of Sea Shepherds might have spread briefly over part of the northern Aegean.
47, 48

With slightly different timing, the developed Sea Shepherds could have been at the Dardanelles first and helped transport the (presumed) Indo-European farmers across that water barrier. In this case they would have had access into the Black Sea, finding sufficient grazing lands up to this point. Within this sea they could make no further progress as sheep herders. All of its coasts already would be occupied either by farmers or by land-based (sheep and goat) shepherds. Their ethnic identity would not persist long with their economy reduced to just fishing and commerce. Similarly, any spread of Boat People clockwise around the Mediterranean would be confined to the maritime aspect because the sheep-herding niche was also filled long before.

From the southern tip of Greece the Sea Shepherds had an open ecological niche in just one direction — to the west along the northern Mediterranean shore, and ultimately up the Atlantic coast of Europe as well. Their starting time is here set at 7100 BC, just 200 years after their origin across the Aegean. Their arrival at the English Channel at or before the time of the inland farmers is archeologically known; they would be available to assist in this crossing too. Still farther along the coast it would be convenient if these Boat People had reached the Danish Islands in time (4820 BC) to assist farmers in crossing into Sweden. Local Mesolithics should have had boats, but they probably could not easily handle cargoes that by then included cattle as well as sheep.

It is a safe presumption that the earliest advancing wave of Boat People would have hugged the coasts and made no attempts at sea crossings beyond visual contact. Later developments improved their range, but this would have had no effect on the initial wave of shore settlers. Given this limitation, their entire route of 11,000 km from the Adriatic to Denmark can be laid out in 55 units of 200 km each. Between their starting date of 7100 BC and their arrival in Denmark by 4650 BC (the same time as for farmers), there are 2450 years available. This gives an average of just under 45 years for each 200-km unit of advance. Considering that these Sea Shepherds were very mobile and initially could settle only small coastal areas, this rate of spread appears to be at least reasonable.
49

A more exact schedule for their progress can be constructed by following the model used elsewhere in this study. A basic rate of one generation per 200 km can be assumed, and it would take the same 30 years as in other cases. The same penalty of one-third of this is added for each hour of reduced winter sunlight. (No freezing temperatures occur along this route.) Another penalty of five years is arbitrarily added to each 30- year generation for expected losses at sea, and this is doubled to ten years for all travel along Atlantic coasts. These factors give an internally consistent set of variations that should at least approximate the real situation. For a 200 km unit the fastest advance would have them make it in 35 years, and at their slowest along this route it would take them 65 years.

The first significant point reached by the Sea Shepherds is Gibraltar at 5915 BC. North African land-based shepherds may have been there as early as 6650 BC, but would not likely have crossed this water gap. The Sea Shepherds would find no available pasturage remaining on the African coasts, so there would be a standoff between them. When Afro-Asian farmers arrive at 5150 BC their crossing into Iberia, along with their new economy, would be facilitated. Without this transportation these farmers might have been substantially delayed here while developing their own craft, if they ever would. The 765-year difference in arrival times would allow for considerable improvement in the shipping that was available here. The frontier of the Sea Shepherds by this time had already progressed to the English Channel.

The next significant item is the arrival of the first Sea Shepherds at Dover in 5095 BC. (Presumed) Indo-European farmers from the continent would not reach Calais for almost another 600 years, at 4500 BC. Again, well-developed water craft thus should be available to transport the farmers and their stock into a new location with no expected delay. This amount of lead time of boats over farmers was not essential, but the presence of some kind of good water craft was. Here, though not at Gibraltar, the boats had to be large and strong enough to carry cattle. The crossing into Ireland would follow the same pattern somewhat later.
50

The third meeting of Sea Shepherds with the advancing front of farmers on the continent would be at the area of the Danish Islands. The schedule used here puts the Aegean Boat People in the proper place at 4700 BC, and the first farmers at 4625 BC. (This is actually more than a century earlier than the crossing at Calais-Dover, but these events are here given in the order of arrival of the boats, not the farmers.) In this case the calculated lead time is only three generations, but the probable availability of local water craft makes this close timing less critical. In any case, their overall timing was calculated just so that this meeting would take place.

Following out this schedule, the Sea Shepherds would reach the coast of modern East Germany at 4700 BC, and the (presumed) Indo-European farmers would arrive there at the same time. All possible areas to the east would have been occupied by farmers even earlier. Thus the center point of the East German Baltic coast should mark the Sea Shepherds' maximum advance around the continent. They would have had access to all of Scandinavia just before this, and may well have settled parts of Denmark and southern Sweden. It is not likely that they went much farther to the north — even today few sheep are raised in Scandinavia.

The Scandinavian farmers probably incorporated a disproportionate amount of ship building as a regular part of their culture. This would follow from their abundance of coasts, lakes, islands, and rivers where shipping would have continuing use. The Sea Shepherds' ethnic component in this may already have faded out before the next major water gap was reached in this area. This was the crossing of the Baltic Sea from eastern Sweden, through Aaland, to the southwestern coast of Finland. This crossing would occur at 3300 BC.

The Sea Shepherds may have first rounded the northern-most corner of Scotland at 4675 BC, and been in the middle of its western islands 70 years later. They would complete their coverage of the Irish coast at 4505 BC. Given the dense forests over most of the British Isles, shepherd penetration would have been minor. Only in the Moors of western Scotland would they find a large area to occupy that was at all comparable to the places they held in the Mediterranean.
51

The crossing from Scotland to Ireland by the earliest farmers would have occurred at 3725 BC. Again, the Sea Shepherds had long been there (since 4685 BC) and would have provided the shipping facilities at whatever cost may have been appropriate at the time and place.

Since there are no recent European languages that have been identified, even in part, with these Sea Shepherds, the reasons for this lengthy account should be reviewed. In terms of the mechanics of the rest of this reconstruction, their significance lies in their predictable availability for transporting the advancing frontier of farmers across all significant water barriers. Without the Boat People it might have been reasonable to impose a long time penalty for the development of such shipping by the farmers themselves, at least at some key locations like Gibraltar and- Dover. It is not obvious how long this would have taken, and thus it might have had a major effect on the timing of some farmers' advances. Fortunately for this reconstruction, no such considerations were necessary.

These Sea Shepherds are well known archeologically, especially along the Mediterranean coast. They are often credited with being a major factor in the initial spread of the Neolithic (i.e. polished stone tools, since conceptually they are not farmers) in southern Europe. Considering their ethnic fate this importance could well be questioned. Yet the fact of their existence must be recognized. Actual farming might have reached as far as Italy along their established coastal routes, though well behind the Sea Shepherds' initial frontier. Such a farming intrusion from the coasts must have been minimal, considering the absence of clear linguistic evidence to this effect. The known diversity and complex relationships of early Italic languages may argue for some influence from this source.

The parts of western Europe that are noted for the megalithic monuments correspond somewhat with the Sea Shepherds' distribution (But conflict with timing, the Kurgan Culture, responsible for megalithic kurgans, started its expansion much later). This is especially true for the abrupt stopping of both at the same place just inside the Baltic area. It is tempting to look for some connection. However, these monuments occur much farther inland than we have any reason to believe these people could ever have reached. Also, the dating of the megaliths is only after the farmers arrive, and not during the time of the Sea Shepherds only. Finally, many of these monuments were for astronomical calculations — an art that only the farmers really needed. Descendants of the Sea Shepherds, with their maritime orientation, may have figured prominently in spreading some cultural items like these, but that is about as far as one can go on this line of speculation.

Linguistic influences may not be entirely out of the picture, so I offer the following speculations. The pre-Hellenic, or Pelasgian, place names in Greece must be pre-Neolithic, as figured here. They would then be from the exact time and general area from which the Sea Shepherds originated, and thus could represent a related language. Similar traces may occur elsewhere. A well-known section of northwestern Germany and the Low Countries, and running as far west as the Somme River has many place names that are neither Celtic or Germanic. These may trace back either to the local Mesolithics or to the Sea Shepherds. The Etruscan language is still unsettled, and these Boat People are the only viable candidates other than the (presumed) Indo-European farmers themselves. Ligurian remains problematical, and a Sea Shepherd element may well be involved. Probably the best case can be made for the extinct Pictish language of Scotland that occurred in just the area where they most likely would have survived in some numbers. Beyond this, they were ephemeral and they do not appear on the last maps.

CHAPTER THREE — NEOLITHIC SPREAD IN EUROPE

 The major thrust of the Neolithic (i.e. farming) entry into the main body of Europe can now be described. Four simultaneous flanking movements of other language groups have already been dealt with: Afro-Asian sweeping across North Africa and into Iberia, Caucasian into the southeastern corner (southwestern corner presumably was not Caucasian), Altaic throughout the Pontic Steppes, and Aegean along the southern and western coasts of Europe. A fifth flanking movement of Uralics in the north does not get underway until after the (presumed) Indo-European entry, so its full description is postponed to the end of this chapter.

The overwhelming source of Neolithic farming in Europe is clearly the initial wave of immigrants from Anatolia who first entered the Balkan Peninsula about 7100 BC. Their identification as Indo-European speakers has already been discussed. Within a century of this entry the first-occupied lands would have been filled to carrying capacity with settlers, and further immigration was quickly phased out. This is the same pattern as with all frontier advances. It is mentioned here only to remind the reader that there was no need for the existence of a great population reservoir in Anatolia to feed into Europe. The initial wave of immigrant farmers is sufficient, and is all that could have entered anyway.
55

Balkans

Expanding overland at a regular rate the (presumed) Indo-European farmers would have fanned out uniformly over the southern Balkans. To the north they reached the deep-soil region in northern Bulgaria at 6830 BC, where they met the Altaic sheep herders who arrived there at about the same time from the opposite direction. Details of this encounter are given in the preceding chapter. The immediate significance is that the entire Pontic Steppe region was then closed to entry by any other ethnic group. The boundary of deep soils and grasslands, along with its inhabitants, forms a right flank bordering the (presumed) original Indo-European speech area on this continent.

To the west and south the farmers spread over and occupied the major parts of modern Greece, reaching its southern tip at 6500 BC. Differentiation of the Greek language must have had its ultimate beginning at this time. There is no subsequent event that affords anywhere near this degree of population replacement.

Further expansion in the Balkans would then begin to encounter factors of winter cold and shortened daylight hours, as described in the Rules of Movement. These would seriously affect the intruding farmers in the form of personnel losses from the new climatic factors, local carnivores, and the Mesolithic natives, among other things. The influence of these penalties may be noted on the map of progress along the major routes where dates are given at each 200-km interval (Fig. 10). This becomes even more evident on the map where the advance of the Neolithic (i.e. farming) frontier is shown in 500-year intervals (Fig. 11).

Fig. 10. Neolithic (i.e. farming) expansion from Anatolia. Heavy lines trace the main routes of (presumed) Indo-European spread through Asia Minor and over the central part of Europe. Dates are given at each 200-km unit of progress. (Presumed) Indo-European territory, along with that of the Uralic farmers, is left blank; all other speech areas are stippled. Heavy dashed lines show linguistic divisions where incoming peoples were partially separated by natural barriers.

 

Fig. 11. Rate of Neolithic (i.e. farming) expansion. The farming frontier in Europe is shown at 500-year intervals from its inception in Anatolia at 8000 BC. The contours are drawn from data given in Fig. 10 and from many more routes of spread that were also calculated. This map compares favorably with calibrated dates of the earliest farming sites, but this is because the advance rates were chosen partly with this goal in mind. The map omits the Neolithic (i.e. farming) of the Mediterranean that was spread by the Aegean Boat People only after their initial wave.

The advancing wave of farmers would first encounter the Hungarian Plain, roughly the eastern half of modern Hungary, at 6400 BC. The deep soils here cannot be plowed except for some choice locations mainly along the rivers. The Neolithic (i.e. farming) wave became divided into two fronts that passed around each side of this area. Natives of the Hungarian Plain would now have the opportunity to adopt domestic animals, among other things, from their new neighbors. At this time cattle were included among these animals; they apparently were domesticated in the southern Balkans just a century or so previously (Wenke 1980). Cattle would have gained rapid popularity because they could be used to pull plows, as well as provide meat, milk, and tough hides. (Cattle from Balkans to Scandinavia via Sea Shepherds?)
56, 57

Cattle are well adapted to the European forest zone, and wild forms lived there as part of the natural fauna. Sheep, on the other hand, are not well adapted to forest conditions and had no wild representatives in this area. Farmers can continue keeping sheep in Europe by clearing pasturage and otherwise caring for and protecting them. Pastoralists would often find cattle to be the more useful animal to keep under domestication here. The only advantage of sheep is to use their hair and skins for clothing. It would be automatic for the Hungarian natives to adopt the pastoral aspect of the Neolithic (i.e. farming) economy, with an emphasis on cattle, along with pottery and whatever else may have seemed useful. They would have taken up little or no farming, and no doubt put less emphasis on permanent dwellings and large amounts of personal property.

As cattle herders these eastern Hungarian natives would then spread out automatically into all available areas that were not already blocked by the advancing farmers. Their rate of advance would be twice that of farmers, just as with the sheep herders of the Eurasian grasslands. By 5800 BC the farming frontier would have passed completely around the Hungarian Plain with the two arms of this advance meeting on its northern side. This then cut off the inhabitants of the plain from further contact with their ethnic relatives who had spread out from there to the north and north-east. These cattle herders on the outside, expanding faster than the farmers, would continually outdistance them and come to occupy ever larger territories. The detailed history (with maps) of this advance of herders out of the Hungarian Plain is given in a later section.

The domestication of cattle had been so recent in the Balkans that it probably would not have spread far when the Hungarian natives picked up the practice. Cattle raising should have reached the deep soils of eastern Romania at a similar time. Here they would be adopted by the local pastoralists where they largely replaced sheep in this economy. This new domesticate would then spread eastward over the Eurasian grasslands. In areas of taller grass, cattle would predominate; in areas of less rainfall and shorter grass, sheep would continue to be the major domestic animal.
58

A two-part wave of cattle domestication then automatically spread eastward out of the Balkan area. South of the forest zone, cattle were largely substituted for sheep by the existing pastoralists, in place. Within the forest zone, cattle were introduced as the major domestic animal by the influx of the first pastoralists out of Hungary. It may be assumed that cattle replacement in the steppes progressed at about the same rate as the new pastoralists advanced in the forests. Evidently neither got far enough ahead of the other to cross the forest line and make a major ethnic intrusion into the other's area.

Continuing to follow the full Neolithic (i.e. farming) advance we may note that the meeting point north of the Hungarian Plain represents a 600-year separation of these two waves of farmers. This is not enough time for their speech to become separate languages, but a strong dialect contrast would probably exist by that time. As people advance at known rates from both sides, this meeting point would be extended into a meeting line that can be plotted with some accuracy. It runs to the north-west, curves gradually toward the north because those on the west move faster, and reaches the Baltic about 100 km west of the mouth of the Vistula (54.2°N 19.4°E) (Fig. 10) .

The continuity of the spreading dialect mesh of (presumed) Indo-European was broken by this line, especially to the north. It would not have been a major linguistic barrier initially, but as speech innovations emerged on each side some might cross and others might not, and it would have strengthened over time. The continuity from one side of this line to the other would still exist through contacts around the southern side of the Hungarian Plain. This would tend to limit the amount of divergence between the two sides in the north. From its location and early occurrence, this may be taken as the dividing line between the original Balto-Slavic in the east and Celtic in the west.

59

Eastern Europe

The spread of Neolithic farming in eastern Europe involves the entire area between the (positively non-) Celtic meeting line on the west and the Altaic pastoralists on the southeastern border. Within this area of expansion there are two crosscutting divisions — earlier and later phases, and separate paths of advance. The two are described here in that order.

The early phase is that of the (presumed) Indo-European farmers who can be followed across the entire eastern front, disregarding the separate paths. At their beginning these eastern farmers crossed from Bulgaria to Romania over a band of steppe land along the lower Danube, most of which could not be farmed by simple techniques. It was a narrow strip that was just then being infiltrated by the first Altaic sheep herders, and should have been passed over with little difficulty or delay.

Farther to the north the farmers' advance rate gradually slows down with increasing penalties from colder winters with shorter days. This eventually results in a halt of the ethnic advance well within Russia, and a transfer of the full range of Neolithic (i.e. farming) technology to the local residents.

Normally, if the native Mesolithic population begins to adopt farming practices, the oncoming wave of established farmers will roll over and incorporate them before they can substantially increase in numbers. At some point the rate of agricultural advance may become so slow that the native inhabitants are able to increase in numbers enough to hold their ground. When the natives are pastoralists instead of hunters, as in this case, the farmer's population advantage is only four to one, instead of ten to one. It can also be assumed that pastoralists will find less difficulty than would hunters in accepting the full Neolithic (i.e. farming) way of life. They would need just two full generations of unimpeded (doubling) population increase, with the new technology, in order to match numbers with the incoming people.

60

Roughly speaking, a slowdown to 450 years per 200 km can be used to mark this transfer of Neolithic (i.e. farming) technology to the local pastoralists. After this it would be just the technology that spread, not the people, so there would be no further ethnic moves while the Neolithic (i.e. farming) frontier itself continues to advance. It is assumed here that the advance of Neolithic (i.e. farming) technology within a resident population would continue at the same rate as it would if it had continued to overrun the natives.

The 450-year transfer rule generates an arc passing through the middle of Latvia and roughly enclosing Belorussia (White Russia). Various parts of this line are reached between 5000 and 4000 BC. This is reasonably close to the line that is known to have enclosed the early distribution of Baltic dialects, leaving the Uralic language family beyond it to the north and east. This marks the halt of the (presumed) Indo-European language distribution for over 4000 years. Actually the transfer of farming from (presumed) Indo-Europeans to Uralics must have taken place on the Lithuanian border, some 100 km to the south of where the above rule would put it. Most likely a denser population of cattle herders near the Baltic coast allowed for a faster matching of numbers with the intruding farmers, thus causing the transfer to happen somewhat earlier along this part of the arc.

Beyond the (presumed) Indo-European stopping line Neolithic farming technology should continue to spread most likely at its standard rates, becoming still slower in some places. Here it spread among Uralic-speaking herdsmen who had moved out from the Hungarian Plain somewhat earlier. With the farming advance continuing so slowly, it would be adopted by adjacent populations without any significant movements of people on this frontier. This farming frontier continued to advance until it reached the line of the 120-day growing season. Beyond that, pastoralism continued to be the basic economy — one that had switched from cattle to reindeer in this area.
61

The advance of Neolithic (i.e. farming) farming may now be followed along the two major paths in eastern Europe. As the original farmers spread north through Romania there would be a noticeable delay in the colder central part of the country. The crest of the eastern Carpathians can be used to mark this division, as it also served somewhat as a barrier of its own. The western and eastern arms of this advance meet to the north after a separation of 600 years. This separation was probably less than total, so it is not clear how strong a linguistic dividing line was generated beyond here. In any case, this meeting line progresses almost exactly along the recent Polish-Russian division, and then turns automatically to the north-east to continue nearly along the Lithuanian and Latvian borders against Russian. This coincidence is so close as to suggest that it resurfaced when more recent languages spread over the area. (See dashed line in Fig. 10.)

The earlier spread of Uralic pastoralists is described in the next section, but one aspect of it must be anticipated here. They begin in Hungary, and appear automatically to have divided themselves into three major paths as they spread north and east over much of eastern Europe. The dividing line between their left path (Finnics) and their center path (Samoyeds (i.e. Nenetses)) largely follows the meeting line just described (Fig. 12). This old Uralic division began 200 km to the west of the Polish-Russian line, but it joins that line in northern Poland and they run together to the end of the (presumed) Indo-European farmer limit. A substratum of Uralic contrasts can be added to the (presumed) Indo-European contrasts that are here indicated on each side of this line. This comes to mark the division between what might be called Coastal and Interior Baltic (terms invented here), and in recent times is the dividing line between Baltic and Slavic.

The left path of Neolithic (i.e. farming) advance in eastern Europe runs generally northward between the Baltic coast and this persistent ethnic dividing line paralleling that coast, some 200 km inland. Farming practices transfer to Uralic natives at the Latvian border and continue along this path to the north and north-east. Farming then curves eastward around the head of the Gulf of Finland and finally moves into Finland itself, crossing the modern border at 2500 BC. A short distance west of here the advance would be stopped, at about 2300 BC, when it meets a similar advance of local Neolithic (i.e. farming) coming from the opposite direction — out of Sweden. (See Fig. 10 for the dashed line paralleling the Baltic coast, and Fig. 12 for the (presumed) Indo-European stopping line.)
62

Finland should not show any ethnic break at this meeting line between eastern and western influences. On both sides it was native Finnic peoples who were adopting the full Neolithic (i.e. farming) practices on their own territory. There should, however, be some kind of a technological contrast here. The two paths that met in southern Finland carried technologies that last parted company in 6500 BC in the Balkans. Over 4000 years of separate development, even on the conservative frontier zone all the way, should have led to recognizable differences in art and style.

Further Neolithic (i.e. farming) advance to the north would end roughly along the line of the 120-day growing season. That would include about one-third of modern Finland. The rest of the country would stay in the hands of reindeer pastoralists for a long time.

The remainder of the Neolithic (i.e. farming) advance in eastern Europe, its right path, would come out of eastern Romania and run generally to the north-east into most of Russia. This is bounded on the west by the Baltic path of Neolithic (i.e. farming) immigration, and on the east by the Eurasian grasslands where cattle were just being adopted by the resident Altaic sheep herders. This is a broad front, almost 500 km wide near its beginning and increasing to over 800 km at its widest part.
The post-Romanian part of this path consists of the western Ukraine (the forested part), and then passes over the Pripet marshes to encompass Belorussia. These marshes probably did not slow the Neolithic (i.e. farming) advance, but were enough of a linguistic barrier to later form part of the dividing line between Slavic to the south and Baltic to the north.

63

The farming front at first should have advanced directly northward through Romania and fanned out equally to the north­west and north-east within the Ukraine. After that it automatically turned more to the north-east as its western part advanced more rapidly, and it was pressed in by the even faster advance of farmers along the Baltic path. The line of technology transfer was then reached along the northeastern boundary of modern Belorussia. (See Figs. 10 and 12.)

The continuing Neolithic (i.e. farming) advance was then of technology only, and passed through the Uralic pastoralists who were native to the area. This should be archeologically testable where early farming passes through the area, and it should be marked by a style line where ethnic identity changed from intruders to residents. From here on, the Uralic substratum is presumed to be divided into two paths of about equal width. Closer to the Baltic is the central Uralic group, here equated with the ancestral Samoyeds (i.e. Nenetses). Adjacent to the Pontic Steppes is the right branch of the Uralics, presumed to be the original Vogul/Ostiaks (Ugric speakers) (The name Ostiak is a Russified form of Türkic Asty, where as = tribe, -ty = possessive affix, i.e. genetic “tribes”, not to be confused with the Türkic dynastic tribe of Ases, with the same etrymology). Both of these pastoral groups would have adopted Neolithic farming techniques and passed them on even deeper inland, with no expectation of significant internal population movement. Increasingly severe winter conditions along these paths would suggest that local population densities did not rise as much as in the milder climates to the west. There is no indication that the internal dividing line between these two Uralic language groups ever surfaced here in later populations.

The frontier of the full Neolithic (i.e. farming) economy continued to move eastward and eventually pinched out between the converging lines of the 120-day growing season on the north and the edge of deep soils on the south. These meet on the western slopes of the Ural Mountains, and this point should have been reached about 1750 BC. Beyond the 120-day season, both north and east, there would be the reindeer-based Uralic pastoralists. To the south, in the deep soils, would be the cattle-based Altaic (horse) pastoralists. Reindeer and cattle zones meet at the Urals, and the dividing line between them continues to the east into Asia.
64

Western Europe

The Neolithic (i.e. farming) frontier advancement around the western side of the Hungarian Plain constitutes the rest of the (presumed) Indo-European source for Europe. This advance subdivides into many different routes, of which two can be considered major. These were separated from each other by the Eastern Alps into northern and southern branches. The northern branch leads to the largest area and is described first.

A direct course can be plotted from western Hungary to Denmark, then across the islands into Sweden (4500 BC). From this path people filled in the western, or (positively non-) Celtic, side of the dividing line running from Hungary to the Baltic. The rate of advance generally slows down along this route because of colder winters and shorter days. There is one minor exception at the base of the Danish peninsula where winters are warmer than before or later.

It has been calculated that the Aegean Boat People reached the Danish islands just ahead of the overland farmers. If so, they would provide the means of transportation between these islands and into Sweden. If this calculation is in error we may assume that the local Mesolithic water craft were capable of moving some of these farmers and their livestock.

This northern European path can be followed directly to the coast of eastern Sweden to where it meets the 120-day growing season line at 3050 BC. A branch of this path would turn off to the north-east and cross the lower end of the Gulf of Bothnia to reach Finland. I am assuming that shipping would continue to be available throughout Sweden, even in the absence of the Sea Shepherds, because of the area's extensive network of waterways and offshore islands. There should have been no delay gaining transportation to the Aaland Islands and on to southwestern Finland. By counting only land distances in this last 200-km unit the Neolithic (i.e. farming) people and their economy should have been able to enter Finland over a long stretch of its southern and western coasts. Shipping contact between this coastal strip and Sweden would have been maintained thereafter. Swedish speech along this strip thus may date back to 3050 BC.
65

The rate of Neolithic (i.e. farming) advance slows dramatically as Finland is entered, and its transfer to native Finnic speakers would be automatic just inland from the coastal strip. In this new manner the economy alone would progress across Finland at the rate of 550 years (the maximum) for the next 200 km. With just a small additional advance, the Neolithic (i.e. farming) way of life should meet its counterpart entering Finland from the south-east at 2300 BC. Not surprisingly, the archeological contrast between these two sides of the country has long been noted, while the only linguistic contrast is the one far to the south-west where Swedes and Finns meet (Kivikoski 1967).

The Neolithic (i.e. farming) advance can be followed up the interior of Norway with interesting results. At the latitude of Oslo the 120-day line is met, at 3500 BC, and farming necessarily stops. Cattle herding and other traits would then transfer to the native Mesolithic Scandinavians of the interior. This pastoralism then should have advanced through and over other Mesolithic hunters at the rate of 75 years, plus penalties, per 200 km. After an advance of about 270 km they would meet the southward moving Uralic (Lappish) reindeer herders in 3130 BC. This encounter is calculated to have occurred on Lake Femund, which in recent years just happens to mark the southernmost extent of the Lapps. The remainder of the meeting line between Lapps and Scandinavian herdsmen, and the full boundary of the latter may be seen in Fig. 13.

Fig. 13. Language map at first stabilization. Shortly after 3000 BC Europe had seven language families with the distributions shown here. In the south-east this stability was reached as much as 4000 years earlier. Areas of developing (presumed) Indo-European dialects are indicated, along with some linguistic barriers. Aegean languages might have survived in other places, especially in Italy and the Ligurian area. Mesolithic languages should also have survived in the Low Countries, and in the Po Valley.

The mountainous interior of the southern part of the Scandinavian Peninsula thus contained a presently unknown ethnic group (Hg I 25-35%). These were local Mesolithic Caucasoids (racially), unrelated to either the (presumed) Indo-European farmers intruding from the south nor to the Finnic-speaking Lapps (Hg N 60-85%) from the north. I shall call them "Nordics".
66

Their approximate area of 145,000 sq km could have supported about 36,000 cattle herders at 0.25 per sq km. Their relatives in the rest of lower Scandinavia were early overrun and absorbed by the Neolithic farmers (Y-DNA Hg I-farmers vs. Y-DNA Hg I-pastoralists). This part of lower Scandinavia amounted to some 186,000 sq km. With a 2% incorporation of hunters into the farmer's population in this area, this amounts to 4000 people. Native Scandinavians thus totaled about 40,000 people, as compared with 186,000 (presumed) Indo-European farmers (assuming 1.0 per sq km initially). When they later all mixed, these Nordics (Hg I) would constitute 17.6% of the total Scandinavian population, not counting the Lapps (Hg N <5%). This is by far the largest increment of native incorporation into the (presumed) Indo-European speaking population for any major area (Patently false). This could go a long way toward explaining the 30% non-Indo-European vocabulary that Lockwood (1969) says is found in basic Germanic.

For all practical purposes the narrow base of the Danish peninsula may be taken as the division between the Germanic of the northlands and the Celtic speech of the continent. That Germanic initially differentiated in just such a compact area is linguistically indicated by the uniformity of its sound shifts (Lehman 1970:2).

(Presumed) Indo-European farmers moving more nearly westward may now be followed along the path leading to the British Isles. Their progress would be more rapid than it was to the north — 250 years for each 200 km for most of the way. They should reach the Straits of Dover at 4500 BC, and cross into England with the aid of the Aegean Boat People who got there 600 years earlier. At 3700 BC they again make use of these Sea Shepherds to get into Ireland. At 3200 BC their frontier reaches land's end at the southwestern corner of Ireland.

Farming progress necessarily came to a halt after only a minor penetration of Scotland. The Moor land is good pasturage, especially for sheep, but most of it cannot be farmed even today. Most likely the Sea Shepherds settled this area and have biologically survived to this day. Their original language is now lost, although Pictish is probably its last survival, which died out by 1000 AD ()

67

Some form of sea navigation was most likely a local Mesolithic development in western Scotland, a tradition which later merged with its Aegean counterpart. Such a cluster of habitable islands just offshore should have stimulated the natives to devise shipping wherever they occur. Just as with the Danish island cluster, local boats might have been able to transport the farmers even if the Aegean Boat People had not been there — but we may never know for sure.

The (presumed) Indo-European progress across Germany and France would be unremarkable, though the Low Countries may have been settled only sparsely at this time. It should have gone roughly according to the schedule given here. The spread to the south through France would halt along an east-west line about even with the southern border of Switzerland. This line is where they would meet the southern branch of the western (presumed) Indo-European farmers — those who passed to the south of the Alps. This meeting first occurs just to the south-west of Geneva at 4800 BC, and extends due west to the Atlantic where it finally closes up at 4350 BC.

The area allocated here to the (positively non-) proto-Celtic speech community is very large, especially in its breadth from central Poland all the way to Ireland. Throughout most of this area there were no dividing lines at natural obstacles that might have created separations. The only constrictions are those which separate Ireland from Great Britain, and it from the continent. The presence of considerable coastal navigation, where the Boat People presumably merged with the local farmers, would have reduced the impact of these constrictions. There were few major enclaves within the area to set one part off in contrast to another, as with Scandinavia. The only obvious enclave was that of the Sea Shepherds in Scotland. The region in and around the Low Countries, with its unusual place names, remains an enigma.

I see no good reason why this entire (positively non-) Celtic area should not have maintained a high degree of linguistic unity for several millennia. Given the geographical circumstances this should have been no problem. Still, it is contrary to the usual view to say that Britain and Ireland have been of Celtic speech since the earliest Neolithic (i.e. farming). Historical evidence of certain Celtic movements into Britain does not mean that the original inhabitants were not also Celtic. The archeological record was formerly interpreted as showing repeated mass immigrations into Britain, but these are now discounted by many authorities.
68

If evidence could be produced for a population replacement, or even a conquest by a higher civilization prior to the Roman Empire, then we would have to allow for at least the possibility of a (positively non-) pre-Celtic Indo-European speech in Britain. It might even turn out that the Aegean sheep herders had managed a significant and lasting area of occupation elsewhere than in Scotland. But on present evidence neither of these seems likely.

The southern branch of the western flow of (presumed) Indo-European farmers passed from Yugoslavia into Italy with little delay. They would be forced to by-pass the deep soils and marshes of the Po River valley before progressing down the length of Italy and into Sicily. The sea crossing through Elba to Corsica and Sardinia was probably accomplished at a slightly later date.

The other route of advance of this southern branch continued along the Mediterranean coast and spread over southern France with no appreciable delay. These farmers expanded northward and would meet the (positively non-) proto-Celtics at the latitude of southern Switzerland. The term Ligurian was applied in early historic times to the language community that stretched across northern Italy and southern France. Its continuity would be natural at the time of first settlement by Neolithic farmers. A pronounced constriction because of the mountains on the present French-Italian border should have led to an early differentiation into eastern and western parts. Curiously, there is little historical evidence that this happened, and the same term was used for both areas into Roman times.

The division between Ligurian and (positively non-) Celtic stemmed from their split at the eastern Alps in Austria. This separation was 1200 years old when they first encountered one another again in France.
69, 70

This is twice as long as the proposed initial separation between (positively non-) Celtic and Baltic, and it should have created a significant linguistic barrier. By the time the two sides reached the Atlantic their separation was 1650 years in the past. The division was known in early historic times and it survived the Latinization of Gaul. It was later reinforced by Germanic influences in northern France that did not extend to the southern part. Today the distinction between French and Provencal is approximately on this line. A sufficiently deep analysis of Provencal should show an ultimately Italic connection long prior to Latinization.

To the south-west the (presumed) Indo-European farmers crossed the eastern end of the Pyrenees and would have progressed about 200 km into the northeastern corner of Iberia. Here they met the Afro-Asian farmers who had spread over northern Africa and moved up through the rest of Iberia. These Iberians had essentially the same Neolithic (i.e. farming) economy, although probably without domestic cattle at that time.

Toward the western end of the Pyrenees it was the Afro- Asians who reached the mountains earliest. They crossed over to meet the (presumed) Indo-Europeans on the French side at 4550 BC. At this end the transgression was much less, only about 50 km to the north and 100 km inland.

At this point the initial distribution of (presumed) Indo-European speech throughout Europe is complete. A discussion of this long- lasting, stable condition must await a description of the geographical spread of the Uralic language family.

Uralic Spread

The Uralic language family is unrelated to Indo-European, and makes up a minor part of the European linguistic picture of today. It is of interest in this study partly because of its present distribution, but also because it included much more of Europe in the past. In addition, a detailed study of the Uralic subdivisions further illustrates some of the geographic principles that are presented here.
71

Uralic languages today are spoken across much of northern Europe and Asia, and in Hungary. Within the USSR the Uralic and Altaic national minorities are now distributed with the former mostly in the forest zone and the latter in the steppes. (See the eastern part of Fig. 1.) According to most authorities the original Uralic homeland was in the Ural Mountain area, hence the name. From this central location these people supposedly spread out in all directions to reach their present distribution, and entered Hungary in 896 A.D.

I find all of this highly improbable for various reasons. A geographically central location is no evidence that this is the original site of a language group. The reason for its spread must be demonstrated — it cannot be assumed to have expanded automatically, and equally, in all directions. The penetration of central Europe in the 9th century by a northern Asian tribe is possible. But a population replacement, or even a language change, by such a tribe within a well-populated agricultural region like Hungary at that time is clearly impossible. Any such claim should be accompanied by an explanation of the mechanism whereby this change might have been accomplished.

Given these objections the actual Uralic-speaking distributions would allow only one alternative explanation — that the family originated in Hungary and spread out in the opposite direction. This poses no serious problem if the time for this origin and dispersion is put at the earliest Neolithic (i.e. farming). If this is true it means that Hungarian (Magyar) is actually the oldest in-place language in all of Europe. To test this idea it must be shown here that the present Uralic distributions can follow automatically from a Hungarian source, and this must be done through the consistent application of logical rules of population movements. It must also be shown that the subdivisions of Uralic are distributed in reasonable conformity with these rules.
72

The following is a detailed description of events that result automatically in the correct distribution of this family. In a later chapter some additional movements are described that lead to the modern distributions of each of its subdivisions. I can find no other approach that even remotely fits both the known facts and simple consistency.

The Neolithic (i.e. farming) frontier first reached the Hungarian Plain at 6400 BC, as noted earlier, according to the schedule used here. The deep soil in this area precluded farming in all but a few choice locations. At least 60,000 sq km would remain in the hands of Mesolithic hunters. With a density of 0.1 per sq km, there should have been about 6000 people here, expectably divided into 12 tribes of 500 each. The most parsimonious arrangement of this number of tribes would be a row of four across the northern border, four more just below these, then a row of three, and finally one at the southern tip (Fig. 12A). The actual number and arrangement might well have differed from this, but the row of four dialect tribes across the northern edge is the only item of significance here. Just as with other hunters, these dialects should have been mutually intelligible, but with clear distinctions from each other.

From their new neighbors to the south these hunters would adopt cattle herding and many other Neolithic (i.e. farming) traits, but little or no farming. Within 200 years (almost seven generations) this new economy should have been integrated and spread throughout the area. This is the same amount of time it would take to advance a herding frontier through the plain if it started immediately upon contact with the farmers. This is also the same time delay that was allowed for the adoption and spread of sheep herding among the Boat People throughout the Aegean Sea.

It is assumed here that all the natives of the Hungarian Plain would have seen themselves as something of a related unit in contrast to the inhabitants of the surrounding forested hill country. The innovation of cattle-based pastoralism ought to have spread freely and easily among these tribes with little transgression onto each other's territories. The eventual spread of this economy out of the plain would begin mainly with an expansion of the edge tribes overrunning their immediate hunting-gathering neighbors.
73

When this expansion gets underway at 6200 BC the farmers will have advanced one-third of the way around each side of the plain. Those cattle herders moving out to the east and west from the central section would soon be overrun by the more numerous farmers with their own cattle. Only those herders who spread out from the row of four tribal areas across the northern edge have unimpeded and unlimited space to expand into. These people would spread out and away from the oncoming farming frontier, and at about twice the rate of speed of these farmers. Eventually the farmers closed in around the northern side of the Hungarian Plain and isolated the original body of cattle herders, but many of their descendants would already have moved out and were able to continue their expansion into new areas. We now have two Neolithic (i.e. farming) frontiers in eastern Europe — fast-moving pastoralists overrunning hunters, followed by slower moving farmers who are gradually incorporating the pastoralists from behind. (See Fig. 12 for a sequence of seven maps showing this action.)

Of the four original hunting tribal areas on the northern fringe, the ultimate disposition of three of these is the major theme of what follows. As cattle herding developed, and population increased by maybe 2 1/2 times, these four tribes should have begun to subdivide into smaller areas with comparable populations. As each subdivides into two or three social groups within its original hunting territory they would also be sending surplus population outward into new areas. The linguistic differences that separated the original four hunting tribes would be perpetuated in the rapid spread of people from each of them. Linguistically speaking, four distinct dialects from the Hungarian Plain would radiate outward with these expanding populations. One can assume that each of these four groups started out with equal numbers and reproductive potential, and came to occupy comparable tracts of land. These are referred to as groups I through IV, which may be followed along paths of the same numbers.
74

A new irregularity must be introduced at this point concerning the directions of spread of the four groups. At first glance it would seem they should spread over a full range from north-west through north-east — at least 50° both west and east of due north. In fact they would spread to the west by no more than 30° from north. This consideration was based initially on examination of subsequent linguistic distributions. If Group I on the left had spread as much as 45 west of north, then its people would have gained access to Denmark. From there they should have had little difficulty crossing into Sweden (using local Mesolithic boats) and spreading over the main part of Scandinavia. These should then be the ancestors of the Lapps who switched from cattle to reindeer herding along the way.

The first problem with this scenario is that Lapps have a Finnic-related language, clearly a part of Group II that can be traced up the east side of the Baltic. A Group I origin for the Lapps would have made their language one of the major branches of Uralic, instead of just a subdivision of Finnic. The second problem relates to the 30% of non-Indo-European vocabulary that is found in the Germanic languages. The only obvious source of this vocabulary is the native population of Scandinavia. If these had been herders out of the Hungarian Plain, then the outside component in Germanic should have been of a Uralic type. Evidently it is not, as this is the first possibility that would have been looked at.

In this reconstruction we are compelled to block the westward spread of these cattle herders at about 30° west of north, and thus have them fail to reach Denmark. The only apparent reason for this would be the density of the Mesolithic hunter population in the western part of Europe where winter survival was easier. Native hunters would take a toll of domestic animals and of the herders themselves. At some level these hunters would be so numerous as to delay, and perhaps even stop, the progress of these pastoralists. Rather than try to develop a new delaying factor for this one case I simply limited the westward spread of cattle herders to the radius at 30° west of north (contradicting the set criteria of Why? to account for reality).

75, 76, 77

Each of these four dialects then would fan out of Hungary over an arc of about 20° by Mercator projection. Group I centers on 20° west of north, Group II starts out due north, Group III centers on 20° east of north, and Group IV has a heading of 40°. Group I would fan out over most of western Poland, reaching the Baltic Sea in less than 500 years. Within the next 700 years they would be ignominiously overrun by farmers and terminated without linguistic issue.

Group II moved due north at first, then automatically curved to the east. This curve results in part from the differential advance rate between themselves and the slower Group III on their right. It may also follow from the pressure of their own members edging to the east to avoid dead-ending along the Baltic coast.

Group II continues northeastward and around the Gulf of Finland, presumably crossing at the narrow part (20 km) just west of Leningrad. They should also pass around to the east of Lake Ladoga. This arm would have been nearly pinched off at this lake by Group III. Some of them should have advanced on up the west side of Lake Onega, but Group III would take its east side.

Somewhere along this course it may be assumed that there was a transfer from cattle to reindeer herding. I have put this at the recent boundary between mixed forest and the fully coniferous forest. (This boundary is given in many atlases with minor variations, so I used the one from Goode's World Atlas. 15th edition, simply because it had the fewest irregularities.) Along the path of Group II this event would center on the modern border of Estonia and Russia (Fig. 12D). After the change to this new major domesticate it is assumed that their population density would go down somewhat, and the area occupied by a 500-person tribe would correspondingly increase. It is also assumed that the density of local hunters would also be declining, so as to permit this. From here on, their rate of advance has been somewhat arbitrarily increased to a basic 60 years per 200 km — the time reduced from the 75 years that was used for them as cattle herders. The same rules continue to be applied for the delay factors from reduced winter temperature and daylight. Each increment of these two factors adds 20 years (again one-third of the basic rate) to the time required to advance the frontier by 200 km. The maximum application of these penalties is still held to the same four increments for each factor. The slowest rate of progress for this set of conditions now becomes 220 years per 200 km, vs the 275 years for the cattle-based economy.
78

With this new rate of advance, the herders of Group II, the presumed Finnic branch of Uralics, would have spread over all of modern Karelia and Finland. Their progress would continue around the head of the Gulf of Bothnia and then down the Scandinavian Peninsula. On the shore of this gulf, about 180 km north of Stockholm, they would encounter the (presumed) Indo-European farmers who had been advancing to the north. Both groups should reach this meeting point sometime between 3200 and 3100 BC. The exact timing is not very critical here because this is also the stopping line for farmers — the 120-day growing season. The Neolithic farmers with their boats, cattle, and greater numbers should have extended their occupation some distance beyond this meeting, but only along a thin strip of the Baltic coast.

In the interior of Scandinavia the southward progression of Group II Uralics would be stopped at Lake Femund at about 3130 BC. Here they should have met the local Nordic pastoralists who had adopted cattle herding from the incoming (presumed) Indo-European farmers. Shortly thereafter, the meeting line would automatically extend almost due west from here to the Norwegian coast. The southern farmers can also be expected to have extended their control of the Atlantic coastal strip substantially to the north.

The area inhabited by Group II Uralics was also being overrun from behind by the advancing (presumed) Indo-European farmers who had surrounded and then passed beyond the source area. The progress of these farmers would have gradually slowed and finally stopped on the Lithuania-Latvia border at 4400 BC. Farming then would have automatically transferred to these Finnic speakers who added it to their economy and passed it on through their fellows as far north as the 120-day growing-season line. This would eventually mark the division between Finnish and Lappish. The distribution of Finnic languages, from Latvia through Lappland, would remain little changed until the advent of deep-soil plowing that entered their area around 700 AD. Even then, what territory they lost in the south was more than made up by gains in the east. This is described later. The total area occupied by this linguistic unit included an impressive part of Europe, both then and now, but their numbers have never been very great.
79

The progress of Group III Uralic herders was fairly simple. They expanded out of the Hungarian Plain at 20° east of due north, and gradually curved more to the east because of the advance-rate differential with Group IV on their right. Eventually they progress almost due east, but this is largely a simple consequence of their great-circle route being plotted on a Mercator projection.

Group III Uralics should have adopted reindeer herding at the same line of full coniferous forest — about 400 km east of Leningrad. They soon reached the Arctic coast and progressed along it as far as the Taymyr Peninsula. This automatic distribution coincides almost exactly with the modern Samoyed (i.e. Nenetses) speech area after some of its European part is subtracted later by other peoples. Their identification with this branch of the Uralic family would appear to be unavoidable.

Group IV Uralics left the Hungarian Plain from its northeastern corner — a seemingly minor fact that proves to have great significance much later. Their progress centered on 40° to the east of due north, and this also follows a great-circle route that eventually heads due east into central Siberia. Their path was bounded by Group III on their left, and by the forest edge on their right. Here they should have been almost keeping pace with the spread of cattle keeping that was being adopted by the previous sheep-herding Altaics in the Steppes. Deep into Russia most of them would be expected to adopt reindeer herding upon entering the purely coniferous forest. The automatic distribution of peoples along this path, with appropriate later subtractions nearer their source, coincides almost exactly with the Vogul/Ostiak branch of the Uralic family. This linguistic identification is assumed to be the case.
80

Groups III and IV were being overrun from behind by Neolithic farmers just as Group II was. These (presumed) Indo-European farmers were able to advance over what is now Belorussia, then bogged down and transferred their farming to these Uralic cattle herders. With all three Uralic groups, Finnic, Samoyed (i.e. Nenetses), and Vogul/Ostiak, the farming technology moves through them at the rate of 450 or more years per 200 km. While this would increase their population it would likely be too little and too slow to cause any significant movements of people. The linguistic boundaries should have stayed constant after this farming transfer.

Most linguistic classifications of the Uralic family first separate Samoyed (i.e. Nenetses) from the rest, and then subdivide these into Finnic and Ugric (Magyar plus Vogul/Ostiak). This is not in accord with the three co-equal groups described above. However, other detailed analyses of Uralic do give the three groups as used here. Hajdu (1976) shows that these three are almost exactly equally distinct from one another.

The lengthy description given above is of regular and geographically predictable events. These would have-led automatically to a distribution along the three major migration paths directly into the areas that the three Uralic branches occupy in more recent times. This has involved no unexplained migrations or expansions of a "historical" nature to put these people in their correct locations. This appears to be much more than mere coincidence. Finer dating of the pertinent archeological remains should show the direction of movements (Curiously, if the distribution map is mirrored, it would place the origin of the same Uralic languages in the northern Kazakhstan).

It remains to be shown why it is Vogul/Ostiak, and not one of the other two, that should be the most closely related to the recent language of Hungary — Magyar. There is also a major shift to the east by many Finnic speakers that must be accounted for. Both of these are neatly explained in a later chapter as automatic consequences of a technological change a few thousand years after this initial dispersal.
81

Language Map

After the Neolithic (i.e. farming) economy spread to its natural limits in all possible directions, the linguistic map of Europe achieved a long-lasting stable condition. For the southeastern corner this time of stability dates back as far as 6500 BC. For most of Europe stability was reached by around 4000 BC, while in Ireland and Scandinavia a few minor ethnic movements continued through the next millennium. Fig. 13 shows the distribution of languages that automatically results from the processes that have been described up to this point.

From this time of settlement, all the way up 500 BC, there is no indication of any significant factors that would have made changes in the linguistic geography of Europe. For most of the continent this first period of stability lasted 4000 years; for some parts it was over 6000 years. This is much longer than all the time that has passed since then. The remaining linguistic movements and differentiations occurred mostly in just a 2000 year period from 500 BC to 1500 A.D. These consisted of four major events that grade into one another, and which can be separately mapped only with some arbitrariness. The only other period of relative stability is the last 500 years of European history.

The map of Europe during this first stabilization shows a number of very different languages that largely surround the continuous dialect mesh of (presumed) Indo-European intruders. In the southwestern corner Iberia is occupied by presumed speakers of the Afro-Asian family that also stretches across northern Africa back to their source area in the Near East. In a smaller northwestern corner is the Pictish speech of Scotland that may ultimately derive from the Mesolithic Boat People of the Aegean. To the north we find part of the wide-spread Uralic language family that continues through most of Finland, Estonia, Latvia, and Great Russia, as well as their surrounded source land in the Hungarian Plain. The Altaics stretch from eastern Romania, through the Ukraine, and on into central Asia. Caucasian occupies northeastern Anatolia and the Caucasus Mountain area. Brief Mesolithic survivals no doubt occurred in many places, the Scandinavian cattle herders being the most conspicuous of these, and the only one shown in Europe. The Hattics are a major survival in Asia Minor. Others would have lasted for some time in such places as the Po River valley and the Low Countries, but they were of no ultimate significance and eventually disappeared. The Aegean Sea Shepherds left settlements along much of the Mediterranean and Atlantic coast, but these too were rapidly absorbed when the full Neolithic (i.e. farming) economy arrived in each area from the landward side.
82, 83

The central core of Europe is occupied by (presumed) Indo-European speakers at this time. Their boundaries with all the other languages mentioned here are sharp and total. Only through Anatolia, and from there on into Iran and India, do they have linguistic continuity outside of Europe. The otherwise simple picture of a dialect mesh of gradual differentiation is broken in only a few distinct places, and none of these is fully bounded.

A major dividing line developed when the immigrant farmers were separated by the full length of the Alps and Eastern Alps, and this line was then extended to the Atlantic coast from their meeting point beyond that barrier. This separated (positively non-) Celtic from the Ligurian-Italic speech areas, but their linguistic continuity was maintained around the eastern end of the mountains. A similar separation occurred when the farmers passed around the Hungarian Plain, met again on its northern side, and continued this dividing line to the Baltic shore. This was the initial basis for a division between (positively non-) Celtic on the west, and Balto-Slavic on the east. A mountain and temperature barrier in the eastern Carpathians led to yet another separation whose repercussions become evident only later. The Pripet Marshes provide a partial barrier between Baltic and Slavic. Germanic tends to pinch off from (positively non-) Celtic at the constriction of southern Denmark. Reduced communication tending to separate both Ireland and Great Britain ought to have caused some divisions, but apparently these were not significant. Indo-Iranian is not distinguished here from Anatolian because this separation nowhere affects the reconstruction of European language distributions.
84

The (presumed) Indo-European dialect mesh in Europe at this time would show some divisive tendencies, but none need be called separate languages. Those at the prominent meeting lines maintained continuity back through their places of initial separation, no matter how distinct the linguistic barrier at the far end might have become. Constrictions should still have maintained continuity even if the changes were rather abrupt through the corridors. Sheer distance, combined with minor geographical barriers, would have tended to emphasize regional contrasts.

The map, of Europe at this time can be labeled with many (presumed) Indo-European "languages," but it must be noted that these are probably only dialects in a complex network, and none would necessarily be truly distinct or bounded from any other. Linguistic innovations that appeared at any point within the mesh would still tend to spread eventually over the entire system. In other words, this dialect mesh continued to evolve, or change over time, as a single unit, regardless of its slowly increasing internal diversity.

These (presumed) Indo-European dialects would not separate into independently evolving lines until something happened to create separate groups that were fully bounded from each other.