Use the following persistent identifier: http://nrs.harvard.edu/urn-3:hul.ebook:CHS_Benveniste.Indo-European_Language_and_Society.1973.
Section 3: Purchase
Chapter 9. Two Ways of Buying
Were the roots *wes– and kwrī-, which have provided the verbs for “to buy,” synonymous in Indo-European? Greek, where these two roots coexist and function in suppletion, enables us to determine the first as the designation of transaction and the second as that of payment.
To designate the “purchase,” the agreement of several languages provides us with a well-defined etymological group, that of Skt. vasna-, Gr. ô̄nos (ῶνος), Latin vēnum. The nominal form is everywhere the primary form: Skt. vasna– ‘purchase price’ furnishes a verbal form, which incidentally is rare, the denominative vasnayati ‘to haggle’, ‘to bargain’. In Greek, ô̄nos furnishes the verb ōnéomai (ὠνέομαι), while from Armenian gin (< *wesno-) a verb is derived which is phonetically gnem ‘I buy’. In Latin the noun vēnum is linked with two verbs, vēnum dare ‘to sell’ and vēnum īre ‘to go for sale, to be sold’. It should be noted that in Latin itself, the phrase vēnum dare has produced vendere ‘sell’. This close connection established between vēnum and dare is a most remarkable fact: the notion of “selling” in Latin is defined as “giving” in a certain way, the determination being expressed by vēnum.
The Indo-European term is *wesno-, a nominal form: the historical verbal forms are all denominatives either by morphological processes or by syntactic processes (Latin vēnum dare, ire); and yet *wesno– itself cannot be anything other than a derivative. We must therefore posit a prehistoric root *wes-.
We now have this root *wes– attested in Hittite; this is a fairly recent confirmation of our reconstructions: the Hittite present waši signifies “he buys.” From this same root is derived the Hittite verb usnyazi ‘he sells’, which presents the formation in –n– of the noun *wesno-. These Hittite facts are a guarantee that we have in the root *wes one of the most ancient forms of the Indo-European vocabulary.
There is another confirmation for this, but it is indirect. It is obtained by retracing to its origin the well-known Persian word bāzār, which means “market.” We have to go very far back to reconstitute the original form: Armenian has preserved the borrowed form vačaṙ, with an ṙ (trilled r) which indicates r + consonant. In Middle Iranian we find wāčarn ‘market street’ (Sogdian and Pehlevi), where the group rn explains the ṙ in Armenian. This permits us finally to reconstruct a compound *wahā–čarana, the second term denoting the process of walking or circulating, while the first term derived from *wah– (the root *wes-). The compound word therefore denotes “the place where one circulates to make purchases,” the “bazaar.” The constancy of the form is evident.
However, this complicates the Indo-European situation. For it so happens that we have testimonies of equal antiquity for the use of a different root which likewise signifies “buy.” This is the root of Skt. krīṇāmi (which derives from the root *kwrt), of modern Persian xarīdan. In lexical usage the forms of krī– have even more substance than vasna-, which is no more than a Vedic survival.
This root is found again in the language (wrongly) called Tokharian, where “trade” is called kuryar or karyar, according to the dialect; the connection with the Sanskrit root was immediately recognized. In Greek it is recognizable in the aorist príasthai, which functions as a suppletive tense form in the conjugation of ōnéomai. In Irish we have crenim ‘buy’, in Slavic, Old Russian krǐnuti; the root exists also in Baltic. It is not found in Latin, nor in Germanic, which stands on its own in this sphere of the vocabulary.
The problem thus arises, at least for Indo-Iranian and Greek, how can we explain the coexistence of two distinct etymological families to designate one and the same notion which hardly seems to admit of differentiation? While here the same operation is designated by two different verbs, it so happens that the two notions of “buying” and “selling” may be expressed by the same verb, with a variation which may be the addition of a prefix (German kaufen and verkaufen) or a tonal variation (Chinese mǎi–mài ‘buying-selling’ with two different tones), the notion itself being somehow differentiated between the two halves of the process.
It may even happen that the determination of the sense can only be made from the context: thus misthòn pherō, where misthón signifies “wages, pay,” may have the two meanings of “to pay a wage, to take a wage to somebody” and “to carry away the wage,” when speaking of the one who receives it. Thus in different contexts it may mean “pay” or “receive.”
The problem is that here, on the contrary, we have two different verbs for the operation of “buying.” The attested sense is the same for *wes– and for *kwrī-, both equally ancient, with a distribution which coincides over part of the territory. *wes– is Hittite, Indo-Iranian, Greek, Latin, and Armenian; *kwrt– is Indo-Iranian, Greek, Celtic, Slavic, and Baltic.
Most of the Indo-European languages have opted for one or the other of the roots. In one language, in Greek, the two function together: ōnéomai and príasthai are found associated in a single conjugation of complementary forms, the second supplementing the first by providing its aorist. But the two were once used separately and thus each possessed a complete conjugation. In Indo-Iranian krī-, krīṇa– is in frequent use, practically to the exclusion of the other root, represented only by vasna– and some other forms, such as the denominative verb vasnayati, which is almost obsolete. The usual verb is krī-.
In Greek the facts are more instructive. The examples in Homer and later on those of Ionic prose allow us to determine the proper value of each of these roots. We note that ōneomai, that is “buy,” after discussion with the vendor, quite often means “to seek to buy”; but príasthai has the peculiarity that it appears with an instrumental determination like kteátessi ‘goods, merchandise, possessions’. Apparently the use of this verb denotes the mode of payment, and on occasion the amount paid. While ô̄nos, ōnḗ, ô̄néomai designate “purchase in general,” “the fact of behaving as buyer,” príasthai is “to actualize the purchase by paying.”
This interpretation is confirmed by the derivatives from the two roots which are not constructed in the same way. Thus we have the adjective ōnētós, the feminine of which, ōnētḗ, is opposed to gametḗ in Homer to designate a “bought” wife, as distinguished from one who has been formally married. But *priátē does not exist: the notion of purchase in this case is specifically expressed by ōnéomai. Conversely, we have a negative adjective: apriátē ‘not bought’, which is followed by anápoinon in a passage (Il. 1, 99) where the father of the young captive whom Agamemnon holds claims his daughter and demands she should be given back to him “without the fact of príasthai and without poinḗ.” He does not want to make a transaction: she is his daughter, she must be given back to him purely and simply, without ransom (anápoinon) and also apriátēn: she does not provide an occasion for a purchase. A father should not have to pay to obtain his daughter: apriátē is on the same level as anápoinon ‘without poinḗ’, a material notion, a manner of payment.
It can now be seen how the two verbs are distinguished: príasthai is more restricted and more material; ōnéomai is the more general expression. This also emerges from the semantic opposition established between the two aspects of the operation: if one wants to say “buy” as contrasted to “sell,” it is ōnéomai and not príasthai which is used.
Purchase and payment are two different operations, or at least two different stages of the same operation in the ancient civilizations and still in some archaic civilizations of our own days: the payment follows the conclusions of the purchase and agreement on the price.