Western Europe has been inhabited for tens of thousands of years, but we know nothing of the languages spoken there before the introduction of writing into the area in the first millennium BC. Writing originated in the Middle East; by around 500 BC it had reached both Spain (via North Africa) and Italy (via Greece). Since a number of written texts have survived from this period, it is from roughly this date that we can begin to get some idea of what languages were spoken in the area.

The vast majority of the modern languages of Europe (including English, Spanish and French, for example) all belong to a single huge family called the Indo-European family. What this means is that all these languages are descended from a common ancestor -- that is, they all started off as no more than regional dialects of a single language. The various Indo-European languages have been spreading across Europe from east to west for thousands of years. The appearance of writing around 500 BC allows us to form a picture of the linguistic position in western Europe at the time.

So far as we know, the first Indo-European people to reach western Europe were the Celts. By 500 BC Celtic languages were spoken in Austria, Switzerland, southern Germany, northern Italy, most of France, much of Spain, Britain and Ireland. These languages had completely displaced the earlier languages that had previously been spoken in the same areas, and we know nothing about these earlier languages.

A second group of Indo-European languages, which we call Italic, was spoken in much of Italy; one of these languages, Latin, was destined later to become the most important language in Europe, but in 500 BC it was only the local language of the small city of Rome. The most important language of Italy at that time was Etruscan, spoken in what is now Tuscany. Etruscan was not an Indo-European language and it is not related to any other language we know of. With the spread of Roman power in the succeeding centuries, the Etruscans began speaking Latin and abandoned their ancestral language, which died out, leaving behind only some written texts which we can read only to a certain extent.

In Spain, the linguistic position was rather complicated. Much of central and northern Spain was occupied by the Celtic people who we call the Celtiberians. These Celts had writing, and they left behind some written texts, including the famous bronze tablet of Botorrita, which we can read only partly. The Mediterranean coast of Spain (and also a small part of southern France) was occupied by a quite different people who we call the Iberians. The Iberians too had writing, and they have bequeathed us a sizable number of written texts in their Iberian language. For a long time we could make no sense of these, but, in the first half of the 20th century, the Spanish linguist Manuel Gómez Moreno succeeded in figuring out the phonetic values of the characters, and so we can now read Iberian to the extent of being able to pronounce it. However, we still can't make the slightest sense of the texts, because Iberian has turned out to be a completely unknown language: it is certainly not Indo-European, and in fact we are confident that Iberian is not discoverably related to any other known language (including Basque -- see below).

In southwestern Spain and southern Portugal we find a few texts written in yet another unknown language; this language is sometimes called "Tartessian", a label which is completely meaningless. The little information we have been able to extract about this language suggests that it was again not related to anything else we know about. About the northwest of Spain we know almost nothing, since we have no significant written texts. Finally, there were a number of Greek and Punic colonies along the Spanish coast; these of course spoke Greek (which is Indo-European) and Punic (which is not; it's a form of the Semitic language Phoenician, introduced from the Middle East).

Writing did not arrive in Gaul (France) until the Roman conquest of Gaul in the first century BC. In his memoirs of the military campaign, the Roman general Julius Caesar tells us that Gaul was divided into three parts, occupied by three different peoples. Two of these peoples were Celtic, and they spoke Celtic languages which, as I remarked above, had already displaced the earlier languages of Gaul. But the third part was different.

The southwestern part of Gaul, from the Pyrenees to the Garonne, was inhabited by a people the Romans called Aquitani, or Aquitanians, and these Aquitanians, Caesar tells us, were entirely distinct from their Celtic neighbors.

As we shall see, there is good evidence that the Aquitanian language was also spoken in the Pyrenees themselves, at least as far east as the valley of Arán, in territory which is today Catalan-speaking, including Andorra. There is also evidence that Aquitanian was spoken south of the Pyrenees, at least in eastern Navarre. We suspect that Aquitanian was also spoken in at least part of Gipuzkoa, but we have no direct evidence for this, since no Aquitanian texts have ever been found there (in fact, there are hardly any texts at all from Gipuzkoa at this period).

The Aquitanians did not have writing at the time of the Roman conquest, but, after that conquest, they learned to write in Latin. We have a sizeable number of Latin texts written by the speakers of Aquitanian during the Roman period, and, crucially, these texts contain a large number of Aquitanian names: about 400 personal names and about 70 names of divinities, most of them found in votive and funerary inscriptions; these inscriptions typically identify the sex and the parents of the people referred to, a fact which is highly convenient.

Now here's the crunch: many of those Aquitanian names are unmistakably Basque. Consequently, we are now satisfied that Aquitanian was an ancestral form of Basque: modern Basque is the direct descendant of that Aquitanian language spoken in southwestern Gaul and in most of the Pyrenees, with (so far as we know) only a rather modest extension into Spain, in eastern Navarre and probably Gipuzkoa. Hence, in origin, Basque was primarily a language of Gaul which later spread west and south into Spain, into the remainder of the modern Basque Country. In the early Roman period, in Bizkaia, in Araba, and in western Navarre, we find evidence only for Indo-European speech: not a single Aquitanian name is recorded in this area. We therefore believe that Basque must have spread into these territories (and beyond) only later, probably after the collapse of Roman power in the area (see below).

Here is a sample of some Aquitanian personal names and elements of names; where relevant, all are found exclusively in the names of individuals of the appropriate sex.