Old European (German: Alteuropäisch) is the term used by Hans Krahe (1964) for the language of the oldest reconstructed stratum of European hydronymy (river names) in Central and Western Europe.[note 1]
Krahe writes in A1, chapter III, "Introducing preface" Number 2:32 that the old European hydronomy extended from Scandinavia to South Italy, from Western Europe including the British Isles to the Baltic countries. Of the three Mediterranean peninsulas, Italy was most completely included, whilst the Balkan Peninsula was only scarcely covered. He writes that what he presents for hydronomy also applies to mountains and ranges of mountains, and continues with "Karpaten" and "Karawanken", certainly within the Slavic settlement area, omitting the Bavarian/Austrian "Karwendel" though.:12 This area is associated with the spread of the later "Western" Indo-European dialects, the Celtic, Italic, Germanic, Baltic, and Illyrian branches. Notably exempt is Greece.
Krahe located the geographical nucleus of this area as stretching from the Baltic across Western Poland and Germany to the Swiss plateau and the upper Danube north of the Alps, while he considered the Old European river names of southern France, Italy and Spain to be later imports, replacing "Aegean-Pelasgian" and Iberian substrates,:81 corresponding to Italic, Celtic and Illyrian "invasions" from about 1300 BC.
Krahe continues in III A 5, "Geographic Area and age of the paleoeuropean hydronomy" that the overwhelming majority of river and stream names originate from words which in the historical single languages cannot be found or cannot be found any more.:77 He uses mainly Indo-European roots to allow the river names in question to speak (rule 1) of which more than 10,000 are listed.
In III A 2, "Etymology and Semasiology of the paleoeuropean river names" Krahe states that the oldest strata are composed by prerequisites of nature and that the river names especially refer to the water itself (rule 2),:60 and that words referring to humans and culture are newer. Both rules are important arguments for considering the old European hydronomy of southern France and the north of the Iberian Peninsula as a result of secondary implementation (A.1. number 3) due to a postulated immigration around 1300 BC.
In "Morphology of Paleoeuropean river names" III A1 number 3, Krahe concentrates on suffixes (simples and multiples) and distinguishes eleven different ones in a table.:62-63 He attributes geographical (Central European vs. South European or Eastern), functional (for example affluent) or temporal (before or after a change of consonants or vowels) functions to the suffixes of the river names (Rule 3). For the temporal function he claims the existence of a system of phonetic changes (Lautverschiebung), however he does not include prefixes in his considerations.
Krahe's concentration on Indo-European roots and the omission of prefixes had serious negative consequences, because ever later focus was placed on those more than 10,000 roots, perhaps on old Irish, but scarcely on Gaulish and other Celtic languages or Baltic languages and completely leaving out Basque. Delamarre later included for example under Gaulish dubron only rivers with "B" (or similar) leaving out other names, which Krahe would have called Schwundstufe = loss of a letter or with inversion of letters or both.
Krahe ignored the impact of Moorish occupation in Spain, which led to frequent combinations of Arab "prefixes" (always at the beginning) on Celtic "suffixes" as seen in Guadiana (Guadi = river and Anas = bayous, muddy, as it appears in Ptolemy). The tables <comparison old European hydronyms> show that, in contradiction to Krahe's opinion, hydronyms (and toponyms) can in some cases very well be explained even by modern Irish, Welsh, or French and certainly by Gaulish.
Krahe has influenced archaeologists, linguists and particularly experts in Celtic languages:
Marija Gimbutas (Lithuanian: Marija Gimbutiene) studied in Tübingen, and received her doctorate of archeology in 1946 in the same department where Krahe lectured. Gimbutas developed the Kurgan theory.
Antonio Tovar, with preliminary studies in Berlin, later professor of the University of Salamanca, was professor for Comparative Linguistics in Tübingen from 1967 to 1979. Together with Manuel Agud and Koldo Mitxelena he prepared an unedited etymological dictionary of the Basque language. Mitxelena (in English editions he appears as Luis Michelena) was only the creator of the modern Basque language.
Other authors with a focus on or touching on the topic old European hydronomy are:
Xavier Delamarre is a French linguist whose standard work is Dictionnaire de la langue gauloise (2nd revised and augmented edition Paris, 2003), with the subtitle "Une approche linguistique du vieux-celtique continental". This is in fact the most comprehensive publication on Gaulish words. Slightly more than 800 terms appear in alphabetical order derived from Gaulish-Greek, Gaulish-Etruscan and Gaulish-Latin or solely Gaulish inscriptions, printed classical languages, coins and some terms of Celtic substrate in Occitan. He presents all cases of appearance of toponyms and hydronyms in question, cites authors and roots, showing alternatives, and classifies, if necessary, as uncertain or questionable. He shows all river name examples with prefixes. For example, see "comparison of old hydronyms" adding "water", "clear", "hard stone", etc.
The German linguist Theo Vennemann suggested in 2003 that the language of the old European hydronyms was agglutinative and Pre-Indo-European. This theory has been criticised as being seriously flawed, and the more generally accepted view is that hydronyms are of Indo-European origin.
The Spanish philologist Francisco Villar Liébana argued in 1990 for the Old European preserved in river names and confined to the hydronymic substratum in the Iberian Peninsula as yet another Indo-European layer with no immediate relationship to the Lusitanian language. However, the idea of "Old European" was criticized by Untermann in 1999 and De Hoz in 2001. Villar Liébana is a supporter of Gimbutas against the theories of Colin Renfrew. In his work, Indoeuropeos y No Indoeuropeos en la Hispania Prerromana ("Indo-Europeans and Non-Indo-Europeans in Pre-Roman Hispania") he presents a nine root "series" and a few more collective "series", mainly of toponyms (Hispanic and non-Hispanic) but also including hydronyms.
For example, in chapter IV B VII:120 Liebana discusses hydronyms of the series "uba" starting :149 with Maenuba (Pliny 3.8) = modern Veléz and with the same name affluent of the Baetis (Pliny 3.11) = Guadiamar, Salduba (close to Málaga). He compares modern rivers like Ubia, Ove, Fonte dos Ovos with, amongst others, the Danube,:149 and with historical Corduba (actual Córdoba, Andalusia).:153 Wherever "uba" appears, like in the rivers Saruba = actual Saar (river), an affluent of Mosel, Spanish fuente Sarobals (Huesca), Sarrubian (Huesca), he acknowledges only "uba" and not the root "Dan" in Danubius (corresponding to Dnieper and Dniester) or the root "Sar" in others, which all are Indo-European roots.
The map on the right shows Old European hydronymic maps for the root *var-, *ver- (figure 3, entitled Karte 5). (maps for *al-, *alm-(figure 1, entitled Karte 2) and *Sal-, *Salm- (figure 2, entitled Karte 4) are not shown)
Another example is the old river name Isar