Use the following persistent identifier: http://nrs.harvard.edu/urn-3:hul.ebook:CHS_Benveniste.Indo-European_Language_and_Society.1973.
Chapter 8. Royalty and Nobility
The king in Germanic (Engl. king, Germ. König, etc.) is the one who is born, that is “well born,” “noble” (from the root *gen– ‘be born’). But the noble has another name, which is extremely instructive, e.g. Germ. edel, originally *atalo-, derived from *atta ‘foster-father’: this designation for the nobleman implies that the great Indo-European families practiced fosterage. In fact, the use in Homeric poetry of the terms átta, atalós, atitállō seems to confirm this hypothesis.
To pursue this description in the western part of the Indo-European world, we now consider the name of the “king” and the “noble” in the Germanic world.
The designation of the king, exemplified in English king, German König, etc. goes back to *kun–ing–az. This is a derivative in –ing from the root kun-, cf. Got. kuni ‘race, family’, a nominal form derived from the root *gen– ‘be born’, which belongs to the same group as Lat. gens and Greek génos. The “king” is so named in virtue of his birth as “he of the lineage,” he who represents it and is its head. In fact every time that his birth is specified it turns out to be noble. Reges ex nobilitate … sumunt, as Tacitus remarks of the Germans (Germ. VII, 1). In this conception the “king” is considered as the representative of the members of his tribe.
Quite different is the Germanic conception of the “noble,” which is expressed by the German edel, and it poses a much more difficult problem. The word appears in Old English, in Middle English, and in Old High German in forms which do not show great differences from those in use today. They all go back to an ancient *atalo-, cf. Old Norse edal, which alternates with uodal, corresponding to German Adel ‘the nobility’. This reconstructed Germanic form *atalo– has no etymological connections and appears to be quite isolated. However, there is a form which corresponds to it but has an entirely different sense: this is the Greek atalós (ἀταλός) ‘childish, infantile, puerile’. This adjective may be linked with the verb atállō (ἀτάλλω) the translation of which would be “play like a child, jump, amuse oneself.” Finally we have a reduplicated present atitállō (ἀτιτάλλω) ‘feed a child, rear it’. All this is not very precise in Greek itself; but the main point is that it is difficult to see any point of contact with the notion designated by the Germanic group. Because of this disparity of meaning, the etymological dictionaries dismiss this connection.
All the same it is worth while giving close scrutiny to the sense of the Greek words. Our research will lead to another realm of the vocabulary, but we shall still be dealing with institutions.
While the verb atállō is hardly attested at all, we have numerous examples of atitállō, and it has a much more precise sense than “rear, feed.” Certainly it is used together with tréphō ‘feed, bring up’: e.g. Il. 24, 60 “I fed him and reared him”; but we may also quote Odyssey 18, 323: “she had brought him up like a child.” These two passages contain the essential significance: “rear like a child,” that is, as if he were a member of the family, which was not actually the case. In all the examples the verb is exclusively applied to a child who is not one’s own child, like Hera for Achilles’ mother (Il. 24, 60). It was never used in speaking of one’s own child. Hesiod also takes it in this sense (Theog. 480).
We now see what this verb refers to. It denotes an institution which is known under the scientific term of “fosterage,” the use of a foster-parent. This is a very important custom, particularly in Celtic and Scandinavian society, and it was the rule in the case of royal children. Noble families had the custom of entrusting their children to another family to be reared until a certain age. This was a real relationship, often stronger than the blood tie, which was established between the two families. In the ancient Scandinavian law codes there are laws, called gragas, which define the status of the child so entrusted and the conduct of the parents who are to rear it. Among the Celts the fact is well known from historic traditions and the legends. Normally the royal children are entrusted to another family, generally that of the mother, that is, to the maternal grandfather of the child. There is a special term to designate the foster-father: this is aite, which corresponds to the Latin atta, the Greek átta, and the verb which designates this custom is in Scandinavian fostra. Hubert, in his book on the Celts, cites many witnesses to this institution. Fosterage is also well attested among the Caucasian nobility, especially in Georgia.
We may now posit the existence of this institution in Greece itself, where it is to be recognized in the verb atitállō. There must have been other terms relating to this notion, but they have been preserved only by chance. Thus we have an inscription from Gortyn in Crete which presents the word atitáltas (ἀτιτάλτας), which certainly designates the tropheús ‘foster-father’.
Now that we have determined the institutional sense of this verb, we find traditions which may be connected with it. We recall how Achilles was brought up by Phoenix (Il. 9,485-495) or, according to a different tradition, by Chiron. If we explored mythical and legendary traditions, we would be sure to discover other confirmations: the essential point is to be able to identify and designate this custom. We may be sure that atitállō was applied solely to children reared outside their own family, whatever the reason may have been, whether to escape from some danger or to be brought up in a certain tradition.
We may now proceed to an examination of this root *atalo– of the Greek adjective. It has a striking resemblance to the Tocharian ātäl, but this word simply means “man” and it is not possible to tell whether this is not a simple coincidence. The formation itself of atalós suggests that it is a derivative in –lo– from the word which is represented by átta, a word denoting “father,” which is known all over the Indo-European world: e.g. Got. atta, Lat. atta ‘father’ ; Gr. átta, Skt. attī, feminine, a familiar term for the elder sister, Irl. aite, Hittite attaš ‘father’ (the word pater does not appear in Hittite).
The form atta is always regarded, because of its geminated consonant, as a word of the child’s language (cf. pappa, mamma). 
However the Irish form aite takes on a special significance because the institution of fosterage still existed in Ireland in historical times: aite is the term for the foster-father and not for the natural father. It is perhaps not an accident that Telemachus addresses Eumaeus by the term átta, if átta was the specific name for foster-father in Greek.
At the conclusion of this study we return to the Germanic Edel. If it was the tradition of great families, particularly royal families, to entrust their children to foster-fathers, it might follow that the very fact of being so brought up would imply a degree of nobility. Edel in that case would simply have meant the “nursling,” with the implication that children brought up by foster-parents could only be of noble birth. This would give precision to the relationship indicated by OHG adal ‘race’ and OE adelu ‘noble origin’, etc. In this way some scattered fragments of a prehistoric tradition would, on this hypothesis, find their original unity and the correspondence of form would agree with the sense posited.
[ back ] 1. On atta see Book Two, Chapter One.