Émile Benveniste, Indo-European Language and Society
Book 1: Economy
Section 1: Livestock and Wealth
1: Male and Sire 2: A Lexical Opposition in Need of Revision: sus and porcus 3: Próbaton and the Homeric Economy 4: pecu and pecunia Section 2: Giving and Taking
5: Gift and Exchange 6: Giving, Taking, and Receiving 7: Hospitality 8: Personal Loyalty Section 3: Purchase
9: Two Ways of Buying 10: Purchase and Redemption 11: An Occupation without a Name - Commerce Section 4: Economic Obligations
12: Accountancy and Valuation 13: Hiring and Leasing 14: Price and Wages 15: Credence and Belief 16: Lending, Borrowing, and Debt 17: Gratuitousness and Gratefulness Book 2: The Vocabulary of Kinship
Introduction 1: The Importance of the Concept of Paternity 2: Status of the Mother and Matrilineal Descent 3: The Principle of Exogamy and its Applications 4: The Indo-European Expression for "Marriage" 5: Kinship Resulting from Marriage 6: Formation and Suffixation of the Terms for Kinship 7: Words Derived from the Terms for Kinship Book 3: Social Status
1: Tripartition of Functions 2: The Four Divisions of Society 3: The Free Man 4: phílos 5: The Slave and the Stranger 6: Cities and Communities Book 4: Royalty and its Privileges
1: rex 2: xsay- and Iranian Kingship 3: Hellenic Kingship 4: The Authority of the King 5: Honour and Honours 6: Magic Power 7: Krátos 8: Royalty and Nobility 9: The King and his People Book 5: Law
1: themis 2: dike 3: ius and the Oath in Rome 4: *med- and the Concept of Measure 5: fas 6: The censor and auctoritas 7: The quaestor and the *prex 8: The Oath in Greece Book 6: Religion
1: The 'Sacred' 2: The Libation 3: The Sacrifice 4: The Vow 5: Prayer and Supplication 6: The Latin Vocabulary of Signs and Omens 7: Religion and Superstition
Chapter 4: *med- and the Concept of Measure
In historical times the root *med- designated a great variety of different things: “govern,” “think,” “care for,” “measure.” The primary meaning cannot be determined simply by reducing all these to a vague common denominator nor by a confused agglomeration of the historically attested senses. It can be defined as “measure”—not “measurement,” but “moderation” (Lat. modus, modestus)—designed to restore order in a sick body (Lat. medeor ‘care for’, medicus), in the universe (Hom. Zeùs (Idēthen) medéōn ‘Zeus the moderator’), in human affairs, including the most serious like war, or everyday things like a meal. Finally, the man who knows the mḗdea (Hom. mḗdea eidṓs) is not a thinker, a philosopher, he is one of those “chiefs and moderators” (Hom. hēgḗtores ēdè médontes) who in every circumstance know how to take the tried and tested measures which are necessary. *Med-, therefore, belongs to the same register of terms as ius and díkē: it is the established rule, not of justice but of order, which it is the function of the magistrate to formulate: Osc. med-díss (cf. iu-dex).
As has been observed several times in the course of our previous discussions, neighboring dialects may have different expressions for essential ideas. This is the case, once again, for the term iudex, which was coined by Latin.
We do not find elsewhere a comparable term: not only is ius as a term of law unknown among the Indo-European languages apart from Latin, but even within the Italic group the idea is designated by a different root. As a correspondent of the Latin iudex we have already cited the Latinized Oscan term med-dix. The sense is the same: he is the supreme magistrate who, besides the function of judge, holds authority over the community. Oscan resorts to a different stem, med-, to form a compound analogous to Latin iu-dex. The original form meddíss is incidentally not isolated in Oscan. In spite of the scantiness of our information about this language we possess a series of derivatives. We have meddikíai ‘in iudicio’, medicatinom (accusative singular) ‘iudicationem’, medicim ‘iudicium’, and finally, built directly upon meddix, meddixud (ablative) ‘iudicio’.
Meddix is also used in certain other dialects of the Italic group of which only rare and short inscriptions survive (Paelignian, Volscian). The substantive med-, which is the first element of the Oscan compound, appears in Umbrian as mers, which is translated as “ius” or “fas”, while the derived adjective mersto- is equated with “iustus.”
The root *med-, which here takes the place of the Latin ius, is not unknown in Latin, where it is represented by the family of medeor (medeo), which also comprises the frequentative present meditor. It provided in Italic a new expression for the notion of law which we propose to examine and try to determine in its exact sense.
At first sight it is difficult to see, if we take Latin medeor ‘heal’ as our starting point, how we can arrive at a term which designates the exercise of a magistracy. But the variety of the senses of *med- is still wider and must be considered as a whole. We must begin by listing the various forms together with the senses attached to each in order to see how all these senses diverged and the origin which can be recognized in them all.
Latin medeo (medeor) ‘heal’ has as a derived noun the word medicus ‘doctor’, and this was the basis for a numerous group of derivatives such as medicare, (medicari), medicatio, medicina, medicamentum, and remedium. Here the sense of med- is narrowly specialized. This medical sense, oddly enough, coincides with what we observe in Iranian: Avestan vi-mad ‘doctor’ (with the preverb vi- underlining the idea of separation). In Irish, on the contrary, midiur (with a middle inflection like the Latin medeor) means “I judge” and, with the preverb con-, con-midathar ‘he exercises authority, he possesses power, he dominates’; this Celtic *med- also gives rise to a derived abstract mess (*med-tu) ‘iudicium’. This approximates to the sense found in Oscan.
On the other hand we are far from this sense with the Greek forms, which are numerous and constitute a unitary group: médomai (μέδομαι) ‘take care of’, which in the form of the present active is hardly attested except in the participial form médōn, Homeric medéōn, ‘the chief’. We must also include in the group the name of a measure, médimnos.
Another series, closely connected, hardly differs from the preceding except by the length of the radical vowel: mḗdomai (μήδομαι), ‘meditate, reflect, invent’ and the neuter noun *mē̂dos, which is attested only in the plural, Hom. mḗdea ‘designs, thoughts’; to mḗdomai corresponds an old agent noun mḗstōr ‘counselor’. The feminine counterpart of this agent noun is –mḗstra which appears in a famous name, Klutai-mḗstra ‘she who takes decisions in a celebrated way’, which was remodeled to Klutaimnēstra.
The root is also represented in Germanic by well-known verbs which have persisted down to the present day: Got. mitan ‘to measure’, OHG mezzan, Germ. messen, with the same sense; and, with a derived form of the present stem, *medā-, Got. miton, OHG. mezzōn ‘reflect, make plans’, cf. Germ. ermessen. A substantive is evidence for an ancient ablaut form: OHG māz, Germ. Mass, ‘measure’. We find a correspondent in Armenian in mit, genitive mti (a stem in -i) ‘thought’, a substantive with the root vowel ē, corresponding in form to the Greek *mē̂dos (mḗdea).
We must list in a category by itself the Latin present tense meditor, which has diverged so far from the sense of medeor that it has become a distinct verb, the primary sense of which is “meditate, reflect” but which soon took on the sense “practice, exercise oneself in, study.” Scholars agree in attributing this development to the influence of the Greek word meletân ‘to exercise oneself’: the Romans in certain words of their vocabulary were used to an alternation of d/l, which had originated within Latin phonology or was of dialect origin, examples being oleo/odor, dingua/lingua. Because of this, meditor was formally equated with the Greek meletân and rapidly acquired the senses of the Greek verb.
Latin presents a final series of forms characterized by a stem *med- but with the ο-grade of the ablaut alternations with *med-. First modus, a derivative of the same type as Greek lóg-os as contrasted with lég-ō. From modus we get the adjective modestus and the verb moderor, moderari. Modestus actually presupposes a neuter noun *modus, gen. *moderis, in the same relation as scelestus is to scelus. This substantive subsequently passed into the thematic declension in -o with the animate gender.
We now have surveyed the whole group of forms. The types of formation are all clear: they do not call for any particular comment, and they correspond satisfactorily. Only the sense is something of a problem. The very fact that the root has produced in neighboring languages terms of different meanings makes us hesitate to decide which of these meanings should predominate in our reconstruction. Shall we choose “to heal,” as might be suggested by the agreement of Latin and Iranian; or is it “to measure,” as in Germanic, or “to attend to, to reflect,” as in Greek?
In general *med- is translated as “think, reflect.” And from this a number of technical meanings are derived: “weigh, measure, judge” or “care for a sick person” or again “to govern.”
Once again with the problem which interests us now, we are faced with the questions encountered every time it is necessary to define the sense of an Indo-European root.
(1) Generally the meaning given to the root is the vaguest sense, the one which is most general, in order that this may be capable of divergence into a variety of special meanings.
But the fact is that “to care for” is one notion, and “to govern” is another. In the Indo-European vocabulary “to reflect,” or “to measure,” or “to govern,” or “to care for” are so many distinct concepts which can neither coexist in the same forms nor be derived from one another. Besides, for a notion of such general scope as “to think” there are traditional terms: in particular we have the root *men-. Now it is obvious that the sense of the terms which have been cited does not permit us to merge *med- with *men-, for *med- does not indicate simply a mental activity, a process of reflection, as *men- does.
(2) Often the attempt is made to reach back to the original sense of a root simply by a summation of the different senses which it comes to designate in historical times. But is it permissible to operate with such a conglomeration of ideas, each of which is distinct and presents itself in the history of each language fixed in a particular sense?
Comparatists thus practice two operations—(1) and (2)—the first of which is an abstraction which consists of emptying the meanings historically attested of all real content, the vague residue being elevated as the “primary meaning,” while the second is a juxtaposition which simply adds together all the later senses: this is no more than a figment of the mind, which has no basis in real usage. In fact a meaning such as the one we are looking for cannot be reached except by an analysis in depth of each of the historically attested meanings. Simple and distinct notions like “to judge,” “to cure,” and “to govern” simply transfer into our language a semantic system which was differently structured. They are all components of a global sense which it is our business to reconstruct in order to restore the fundamental unity of meaning.
Should we take as our starting point “to care for the sick,” a sense attested in two separate languages, Latin and Iranian? We cannot trace the sense “to measure” back to so precise a meaning. And yet it does seem that a priori (and in a confused way) it is the notion of “to measure” which predominates. This is limited in Greek to médimnos, but is more amply represented by Latin modus and in Germanic by Got. mitan, Germ. messen, etc. At the same time the notion of “reflection” crystallizes out, as we see in Gr. mḗdomai, mḗdea.
Let us begin with Latin modus. This means “measure,” but not a measure in the sense of material dimension. For the notion of “to measure” Latin uses a distinct verb, metior. Modus signifies a measure imposed on things, a measure of which one is master and which implies reflection and choice, and also presupposes a decision. In short, it is not something to do with measurement but with moderation, that is to say, a measure applied to something to which measure is unknown, a measure of limitation and constraint. This is why modus has a moral rather than a material sense. The word modestus means “he who is provided with measure, who observes measure”; moderari means “to submit to measure (what escapes it).”
Latin makes it clear to us that if *med- meant “measure,” it was quite different from *mē-, the root from which IE *mens ‘moon’, Latin mensis ‘month’ are derived, which is a measure of dimensions, a fixed and as it were passive quality—the symbol of which is the moon which measures the months. Modus appears to us in quite a different guise: a measure of constraint, presupposing reflection, premeditation, which is applied to a disorderly situation. Here we have our starting point.
Now, with the help of Greek, but giving precision to the evidence it provides, we may carry our analysis a stage further. The usual translation of the Greek *médō, considered in the light of its present participle médōn, is “protect, govern,” while the substantive use of the participle is rendered as “lord, master.” The present middle médomai is translated as “to watch over, devote oneself to something.” It is however the same verb, and it ought to admit of the same translation.
We must study on the one hand the Homeric uses of medéōn in fixed phrases, with Zeus and a place name: Idēthen medéōn, literally “who rules from Ida” (Il. 3, 276; 7, 202), cf. Dōdṓnēs medéōn (18, 234); and on the other hand the frequent expression hēgḗtores ēdè médontes (Il. 2, 79). Is it sufficient to translate the verb or its participle in these examples as “protect” or “govern”? It is clear that scholars, seeing that medéōn was applied to a personage such as Zeus, have contented themselves with a vague translation implying authority: ‘governing, ruling over’. But in the nominal group hēgḗtores ēdè médontes we must distinguish two separate notions. In the verb hēgéomai we have the notion of the conduct of operations, implying calculation and planning; in médōn we feel primarily the notion of authority and secondly—in the same way as in Latin—the notion of a directing “measure.”
Let us give further precision to this result by study of the middle médomai. This verb takes a number of objects in greater variety than in the case of médōn. Some of the terms relate to battle: polémoio medésthō (Il. 2, 384) ‘let them bethink themselves of war’; or again medṓmetha alkē̂s ‘let us think of stout resistance’ (5, 718; cf. 4, 418). But we also find médomai applied to “food”: sítou, dórpoio (24, 2), or to “return,” nóstou (Od. 11, 110; 12, 137), or more vaguely to objects of thought: e.g. in Il. 4, 21 two goddesses, Athena and Hera, “pondered evil things (kakà … medésthēn) for the Trojans.”
In this last use, médomai coincides with mḗdomai, which means fairly frequently “prepare, premeditate” (an evil fate), with reference to a god: “The whole night wise Zeus pondered evil things” (kakà … mḗdeto, 7, 478) or again “Zeus pondered their destruction” (mḗdet’ ólethron, Od. 14, 300).
Let us now consider the substantive mḗdea. It is constantly used with boulaí ‘counsel, designs’ (e.g. Il. 2, 340), or else it refers to one who knows, who is wise and inspired: pepnuména mḗdea eidṓs (Il. 7, 278; Od. 2, 38).
These are the principal uses from which scholars have extracted the sense of the verb as “premeditate, advise, dominate, busy oneself with…” and “to govern.” All these activities comprise a notion of authority, and, in the case of the substantive, the idea of sovereign decision.
We are now in a position to give a more precise definition to this notion of a “measure” applied to things. What is involved is a measure of a technical character, of something tried and tested by long use. There is no suggestion of a procedure invented on the spur of the moment or of reflection on the part of one who has to devise his plans. This “measure” is supposed to be applicable always in certain given circumstances to solve a particular problem. Thus we are far from the notion of “reflection” in general, and no less far away from the notion “to protect” or “to govern.” To give a rough definition of *med-, we might say that it is “to take with authority measures appropriate to a present difficulty; to bring back to normal—by a tried and tested means—some particular trouble or disturbance”; and the substantive *medes- or *modo- will probably mean “the tried and tested measure which brings order into a confused situation.” The notion is not preserved everywhere in the identical form. It differs from language to language, but there is no difficulty about recognizing the original sense. We can now see that the Latin medeor and the Avestan vi-mad- do not properly mean “to heal” but rather “to treat a malady according to the rules.” This is not a simple tautology: the idea conveyed is not “to give health to a sick man” but “to submit a disturbed organism to given rules, to bring order into a state of confusion.”
In Greek we find much the same sense. The word always involves measures, ordered authoritatively, to face a particular problem by tried and tested means. Whatever the subject—war, an embarkation, or even a meal—all these require a given technique. When Zeus is called médōn, this traditional epithet relates to the power possessed by the lord of the gods to apply the “measure” in given circumstances, on the occasion of a solemn oath or when help is required. We wish to secure his intervention in the resolution of a specific difficulty, since he has the power implied by the verb médō.
Finally we come to the legal sense which is found in Oscan meddix. All the constituents of meaning can be found here, and they serve to bring out the equivalence observed between med- and ius: first we have the notion of authority, which is included in the use of dico. The central idea is that of a “measure” chosen from a traditional repertoire to be applied in a given case.
One striking fact should be pointed out: neither med- nor ius give rise to any real derivatives; this means that they were no longer living forms. What have we in Latin by way of derivatives from ius? The verb iuro no longer has the sense of ius and can be attached to it only by appeal to a prehistoric meaning. The synchronic relation is broken. Apart from this verb, all that ius yields is the adjective iustus, which is paralleled by the couple modus/modestus. All the derivatives are, in fact, drawn from iudex: iudicium, iudicari, iudicatio, etc. Similarly in Oscan we have medicatinom from meddix. The derivatives are thus made from the agent noun. We must, therefore, conclude that these two legal terms, ius and med-, represent dead and not living forms. We can buttress this observation with another fact. There does not exist in Latin any derivative of ius, either adjective or substantive, with the meaning: ‘he who is a jurist, who is learned in the law and practices the law’. There is no term *iuricus to match medicus. We have, to be sure, complex expressions, but these are mere juxtapositions: iuris prudens (and prudentia), iuris consultus, iuris peritus. We can take this fact, too, as another proof that ius was incapable of providing any derivative whatsoever.
The reason for this is probably that the law was considered exclusively as a body of formulas and the practice of law as a technique. It was not a science, and it did not give scope to invention. It was fixed in a code, in a collection of sayings, of prescriptions which had to be known and applied.
Thus the role of the supreme magistrate will have been to show the “measure” which is to be imposed in such and such a dispute. We have established that the law is a thing which has to be shown, said, or pronounced, which is expressed in parallel formations—Gr. dikaspólos, Latin iudex, Oscan meddíss, and Germanic eosago. This gives us a means of measuring one of the great changes which occurred in the languages and institutions of the Indo-European peoples when law, going beyond its technical apparatus, was constituted of moral ideas, when díkē gave rise to the adjective díkaios, when ius and iustus finally developed into the notion of iustitia.
It is necessary for law itself to be renewed and to become identified in the last resort with what is just. But it took a long time for this convergence of the notions of law and justice to come about. It was in virtue of their increasing approximation that the very designation of law was transformed so that ius has been replaced in the Romance languages by directum (derectum). The “law” (droit) is what is “straight” (droit) as opposed to what is “crooked” or “perverse” (pervers). It is in this way that directum, like the German Recht, has taken the place of ius as an institutional term, whereas in English the “right” is identified with the “law.” In English we do not study “right” (German Recht studieren); we study “law.”
All this hangs together: this historical process whereby ius evolved to iustitia and a differentiation was made between iustitia and directum is connected by obscure paths which are difficult to trace with the very way in which law was conceived in the minds of the Indo-European peoples. The study of the vocabulary of institutions gives us a glimpse of how these notions of a formal character evolved and achieved new precision, concurrently with the growing refinement of conscience, finally to engender moral notions with which in some cases they become identified.