A Historical Commentary on St. Paul's Epistle to the Galatians

By W. M. Ramsay

Part 2

Historical Commentary

Chapter 34

The Use of Diatheke in the Pauline Epistles

The Biblical idea, which is usually rendered in the English Version by the word “Covenant,” is an exceptionally important one. It does not belong to our purpose to discuss it from a theological point of view, or to describe its origin and development in the religious life of the Hebrews. Roughly speaking, Paul took the word Diatheke to indicate a certain gracious act of God, in the exercise of His own absolute power, towards His chosen people, conferring certain privileges upon them on certain conditions which they are expected to fulfil in their life and conduct: His chosen people being first the Jews, and in due course all nations, whom the Jews ought to train and instruct. That original act of God may be called a Promise, or a Covenant; but no single word expresses fully its nature and character; every name leaves much to the imagination and thought, the knowledge and experience, of individual men, so that each man must make his own conception of the thing which is meant. Every word has a misleading connotation, due to its ordinary sense “after the manner of men”; and that ought to be stripped off when one applies the word to the Divine act.

The Greek word Diatheke, which was most widely in use to designate that Divine act, was frequently used in ordinary society to indicate a certain common act of legal character, viz., a Will or Testament. This connotation was distinctly detrimental, when the Greeks attempted to understand the Biblical idea, and to conceive in its purity the character of the Divine act. We have to study the action and language of Paul in the face of this difficulty. He had to convey to his Greek-speaking converts from Paganism as clear a conception as possible of the Divine act; and he was not entirely free to use whatever words he chose, for there was already in existence a certain customary series of terms, employed for centuries by Greek-speaking Hebrews. The word Diatheke, which we have to study, occurs nearly 300 times in the Septuagint Version of the Old Testament, and thirty-three times in the New (chiefly in Paul and in Hebrews).

Now the history and sense of the Greek Diatheke is exceedingly obscure. The Diatheke was a different thing at different periods and in different parts of the Greek world. Yet Paul in some cases is clearly trying to use the recognised ordinary sense of Diatheke “after the manner of men,” in order to aid his readers to picture to themselves the Divine Diatheke, as we have seen that he is doing in Gal 3:15-17. In trying to grasp his meaning we find so little trustworthy information about the Greek usage, that we must attempt to treat the subject a little more accurately and less vaguely than the commentators. Most writers on “Covenant” discuss the theological and philosophical side very elaborately, and confine themselves to a few vague and not very accurate words about the Greek use of the word Diatheke.

I touch upon the subject with reluctance and diffidence. It lies beyond the special sphere of my knowledge, among the obscurest mysteries of Greek law and of theological theory; and I shall be grateful for any corrections of, or useful additions to, the statements made in the following paragraphs.

The Septuagint translators found themselves confronted with a difficult problem, when they had to select a Greek word to translate the Hebrew berīth. The Hebrew word, denoting primarily an agreement, private or public, among men, guaranteed and confirmed by weighty and solemn oaths on both sides, had become almost a technical term to denote the promises made, and confirmed by repetition, by God to the ancestors of the Hebrew people, especially Abraham, and, in a much less degree, Isaac and Jacob. As Professor A. B. Davidson says,1 it “had become a religious term in the sense of a one-sided engagement on the part of God”. This sense was peculiar and unique. Nothing like it was known to the Greeks, and therefore there was no Greek word to correspond to it. Accordingly, the translators were compelled to take some Greek word, which hitherto had denoted something else, and apply it to their purpose. The word selected must necessarily be encumbered by associations connected with its recognised meaning, and, therefore, must be to a certain degree unsuitable. The problem was to find the least unsuitable word.

A word which in some respects corresponded well to the sense required was Syntheke, συνθήκη, which brought out the binding force and legal solemnity of the idea. But it was unsuitable, because it implied pointedly that the two persons concerned in the Syntheke stand more or less on a footing of equality (though not necessarily on perfect equality), each joining in the act with a certain degree of power and voluntary action. But in the Biblical idea the power and the action lie entirely on one side. God gives the assurance, binds Himself by the promise, and initiates alone the whole agreement. The other side merely accepts the agreement, and has simply to fulfil the conditions, which are often unexpressed, for God foresees the course of events, and knows how far the future action of the chosen recipients will fulfil the conditions. The Biblical idea was one-sided, but Syntheke was two-sided essentially.

Yet the history of the Greek rendering of the Old Testament shows that Syntheke must have been felt to have some claim, for the later translators, Aquila, Theodotion and Symmachus, use that word in a number of cases, where the Septuagint version has Diatheke.2 The reason for this change, as we shall see, lies in the gradual development of meaning and character in the ordinary use of Diatheke. The word, as used in the early part of the third century B.C., was a closer and better representative of the Biblical term than it was in its later development. The development was partly in the line of natural growth in Greek Will-making (and that growth seems to have been more rapid in Egypt than in Asia Minor and Syria), partly in the way of assimilation of Roman ideas on Wills.

The word ἐπαγγελία, Promise, might also have been selected. It had the advantage of expressing strongly that the action and the initiative proceeded solely from one side in free grace. But it lacked entirely the idea of bond, of solemn guarantee, of the binding force of oaths and religious sanctity, which was absolutely indispensable. It was used, for example, to indicate the public promises, made by a candidate for public office, as to what he would do when elected; there was no binding force in those promises beyond dread of the unpopularity likely to accrue, if they were not carried out at least to some extent, and they were recognised generally as the stock-in-trade of a candidate, made to be broken as far as was safe. Hence the word is very rarely used in the Old Testament, and never to represent berīth.

In the New Testament, on the other hand, it is rather common. Paul seems to have liked it, as expressing the perfect voluntariness of the act of God.3 It made the “Covenant” an act of God’s grace, wholly undeserved by any previous conduct on the part of the recipients. Hence he even speaks of “the covenants of the promise” (Eph 2:12), i.e., the solemn, binding, holy engagement of God’s voluntary grace and kindness, where he requires the two Greek words, when he desires to bring out very clearly and thoroughly the two sides in the Biblical ideas, the binding force and the free grace.

It is characteristic of the change of spirit that the Old Testament uses only the word indicating binding, inexorable legal force, the New Testament prefers the word indicating free, undeserved kindness and grace.

The word Diatheke was fixed upon by the Septuagint translators to represent berīth. This resolve must have been formed at the beginning of their work. They took the word in spite of its associations with human business on the ground of its character as a whole. Now the word Diatheke went through a rapid course of development during the period B.C. 300 to A.D. 100 or 200; but the Septuagint translators, taking the word about B.C. 285, found it without any of the connotation derived from the changes that affected it after B.C. 300. It had such marked advantages over any other word in Greek for their purposes that their choice could hardly have been doubtful.

In the first place, the ancient Diatheke was a solemn and binding covenant, guaranteed by the authority of the whole people and their gods. It was originally executed verbally before the assembled people as a solemn religious act, the people being parties to it; and even in Greek-Egyptian Wills of the late third or second centuries B.C., when the Diatheke had become a private document, the reigning sovereigns were made parties to it, and named executors of it:4 this was, of course, a mere form, a sort of legal fiction, substituted for the old fact that the public authority was actually a party to the Diatheke. The word was therefore well suited to express the binding irrevocable solemnity of the word uttered by God.

In the second place, the Diatheke was primarily an arrangement for the devolution of religious duties and rights, and not merely a bequeathing of money and property.5  The heir by Diatheke was bound to carry on the religion of the family, just as if he had been a son by nature, and was placed there for that purpose. The term was therefore well suited to describe God’s promise of a religious inheritance to His chosen people.

In the third place, the maker of the ordinary Diatheke had full power in his hands; and the party benefited by the Diatheke exercised no authority in the making of it. The latter had only to fulfil the conditions, and he succeeded to the advantages of the Diatheke. The act of God was of the same one-sided type.

In the fourth place, while the noun διαθήκη is confined almost exclusively to the sense of the disposition of one’s property and duties by Will,6 the verb διατίθεσθαι, has a wider sense, and is used in the sense of “to dispose of one’s property by sale,” and in various other senses of the term “dispose” or “arrange”; but in every case the one single party disposes with authority.

Finally, the central idea expressed in the word Diatheke fairly represented one important side of the Biblical conception. The Diatheke was the concrete expression of individual authority over property, and embodied the reaction against the former system of family authority. In a more primitive stage, property belonged to the family or the tribe, and the individual had no right to dispose of it: the development of Greek civilisation put ownership of property more and more into the hands of the individual. The tradition was that Solon passed the first law in Athens permitting the owner of property to bequeath it by a Diatheke, whereas previously the family to which the owner belonged inherited in default of children. Solon, however, gave the right of bequeathing only in default of male children, only under the form of adoption, and with the obligation that the adopted heir must marry the daughter, if there was one. Gradually the freedom of making Diatheke was widened, the individual became more and more master of his property, and its disposition and the claim even of his children became weaker. He was permitted to bequeath legacies to strangers without adoption; but these legacies seem to have been at first classed as presents or gifts (δωρεαί), not as inheritances, and were restricted in various ways7:  by common Greek custom and the feeling of society a son must inherit, and an heir was called a son.

In the cases which are most familiar to us in inscriptions legacies took, as a rule, the form of religious endowments intended to perpetuate the cult and the memory of the deceased; they are exactly on the same footing as gifts made by a living person to keep up the religion and the worship of his deceased child or relative;8 and they are often stated to be by consent of the heirs.

Hence the word Diatheke expressed strongly the absolute authority of the disposer, who in the Biblical conception was God Himself

Thus, even after the Greek Will had lost its original character of being open and public, immediately effective, and irrevocable, the word Diatheke still retained many characteristics which fitted it to be used as the rendering for berīth. But, certainly, the change in the character of the Greek Will tended to make the word less suitable.

To describe the steps in the development of the Greek Will would require a treatise; but some points bearing on the New Testament usage of Diatheke may be put together here. The new evidence gained from the many Wills of Greek settlers found in Egypt,9 from inscriptions, and from the Roman-Syrian Law-Book10 of the fifth century after Christ, has never been thoroughly collected and arranged.

The obscurity in which the subject is involved may be gathered from the words used by such a high authority as Dr. W. E. Ball:11 “It need hardly be said that St. Paul, in any metaphor based upon Will-making, could only refer to the Roman Will. The Romans were the inventors of the Will.” He speaks on the assumption that there was no Greek system of Will-making. But, as soon as we realise that in Tarsus, in Syria, in South Galatia, and at Ephesus, Paul was in the region where Greek Wills had been a familiar fact of ordinary life before a single Roman had set foot in the Eastern land, and where Greek Wills were still customary when Paul was writing, the case assumes a different aspect.

The case is complicated by the difference of custom and law in different Greek countries, and by the way in which Roman law affected Greek law in the Eastern Provinces. For example, a Greek Will of A.D. 189 in Egypt is expressed entirely in the Roman style and after Roman custom,12 and the Roman-Syrian Law-Book, while retaining many points of Greek law,13 uses various Roman terms, and observes the rule of the Lex Falcidia, B.C. 40, that three-fourths of the testator’s property is at his own disposal, but one-fourth must go to his children.14

In speaking of the Roman Will, we allude only to the highly-developed “Praetorian” Will, which had become practically universal in common life, and was the only form of Will likely to affect the Provinces. Now, whereas the Greek Diatheke came in the third century B.C. or earlier to the East with Greek settlers and soldiers and colonies, and therefore with some of the associations of its past history, the Roman Will came much later, as a fact in the law of the conquerors, and without any associations from its past history: it appeared in the East as a document which had no standing and no meaning until after the testator’s death, and was revocable by him at pleasure. Therein lay the most striking difference between the Roman will and the Greek. I confess that several high English authorities on Greek Wills in Egypt, when consulted privately, expressed the opinion that these Wills were revocable at the testator’s desire; but they have not satisfied me that the evidence justifies that opinion earlier than the Roman time and Roman influence.

The Greek Wills in Egypt went through a rapid development. The soldiers who settled there were separated from their family, and were sole masters of their fortune; and therefore the family influence on the Diatheke, and family rights over the property of the individual, which were so powerful from long-standing feeling in the surroundings of their old home, had little force in Egypt. Everything concurred to give the individual owner absolute right to dispose of his property as he pleased. The development would go on continuously through the centuries, for Egypt was a battlefield for Greeks and Romans.

In the Wills in Egypt there is often contained the provision that the testator is free to alter or invalidate. Such a provision need not have been made, if Wills were acknowledged to be revocable at the testator’s pleasure: he has to guard by a special provision against the customary presumption that the Diatheke is irrevocable. The step from the formal insertion of this provision to the assumption that the provision is to be presumed in every case, might probably be easily made as time passed; but whether the step was made before Roman influence came in to facilitate it, seems not to be proved. The only second Will known seems to repeat and confirm the first (see p. 352).

Again, in a Will dated in the year B.C. 123 15 the testator leaves all his property away from his two sons, except two beds: all the rest he bequeaths to his second wife. That looks like “cutting off the son with a bed,” a merely formal recognition of his right to a share: we remember that in Greek law the owner and father could disinherit his son, but at first, and probably for a long time or even permanently, the act of disinheritance must be performed by the father publicly, during his lifetime, and for good reasons.16 Even in the fifth century after Christ the principle remained in force in Syria, persisting from Seleucid custom and law, that the father could put away his son on good grounds. The heir by Will and adoption had a stronger legal position in Greek law than the son by nature, as we saw on p. 353.

On the other hand in Greek law, a daughter was not strictly an heiress. She had an indefeasible right to a dowry, and this could be greatly increased according as her father chose, but she was styled an ἐπίκληρος, not a κληρονόμος (as a son or adopted son was); and her dowry must not encroach seriously on the son’s portion.

In an unpublished Greek Will, found in Egypt,17 of the period of Trajan, a man leaves his property to his wife for her lifetime, and thereafter to the children of his concubine, who on their part are not free to alienate it, but must leave it to their own family. This implies a much extended power of the individual over the disposal of his property for generations; but it is probably due to the influence of Roman customs and law.

Obviously, a people who had been used to think of a Diatheke as a private document, which could be altered by its maker as he pleased, and which was unknown to any other until the maker died, when it was unsealed and became effective, would see hardly any points of agreement between that kind of act and the Promise of God to His people. The analogy of the ordinary use of Diatheke “after the manner of men” would tend to confuse their ideas rather than help them to understand the nature of God’s act. The only way to attain clearness would be to treat this word Diatheke as a technical term of the Greek Bible, unconnected with the common Will or Testament.

That is the case with the Epistle to the Romans. The word Diatheke occurs there twice, but only in strictly Biblical and Hebraic surroundings.

Similarly, the Epistle to the Hebrews was written to a people who knew only the Roman Will. “The Rabbinical Will was unknown before the Roman Conquest of Palestine, and was directly based upon the Roman model.”18 Under the rule of Herod in Palestine, as of Amyntas in Galatia,19 the new law introduced was almost certain to be Roman, not Greek. The pleadings in Rome about the comparative validity of Herod’s last Will show the Roman character: the last Will is tacitly acknowledged to be the only one valid, unless it could be shown to have been executed in a state of unsound mind.20

Even if the Epistle to the Hebrews had been addressed (as some think) to the Church in Rome, not to that in Jerusalem, that would only show more clearly how Roman is the atmosphere in which it moves. But the writer of the Epistle was a Jew, perhaps resident in Caesarea, on the theory that it was written by the Church of Caesarea to Jerusalem during Paul’s imprisonment.21

In accordance with this the word Diatheke in that Epistle is generally used in a purely Biblical and Hebraic way. But in Heb 9:15-17 the sense of Diatheke “after the manner of men” moves in the writer’s mind, “for where a Testament, Diatheke, is, there must of necessity be the death of him that made it. For a Diatheke is of force where there hath been death; for it doth never avail while he that made it liveth.”22 This thought leads him into a quaint and far-fetched train of reasoning, in order to show how there was a death connected with every Divine Diatheke. It is quite extraordinary to see how some theologians torture these words in order to escape their plain and inevitable meaning (even plainer in the Greek than in the English).

No thought of that kind can have troubled the minds of the Septuagint translators. And Paul in writing to the Galatians, does not feel it; and he assumes that the Galatians are familiar with the ordinary human Diatheke as irrevocable from the moment when it was properly executed and passed through the Record Office of the city.


[1] In Hastings’ Dict. of the Bible, I, p. 514.

[2] The Septuagint version uses συνθήκη in a few cases to represent other Greek words, and, in one case, 4 Kings 17:15, one of the texts uses it to represent berīth.

[3] Paul uses διαθήκη nine times, ἐπαγγελία twenty-five times; but in Hebrews (which is more Hebraic in its form) διαθήκη occurs seventeen times, ἐπαγγελία fourteen times.

[4] I am indebted to Messrs. Grenfell and Hunt for much information on the Wills executed by the Greek settlers in Egypt.

[5] See above, p. 341.

[6] That such was the sense of διαθήκη in ordinary Greek is attested by the lexicons and by many inscriptions. The only exception where διαθήκη seems to mean an agreement, is quoted from Aristoph., Av. 439, but is not clear. It contains a joke founded on some unknown popular story of the ape and the woman (or his wife): the story is explained by the scholiasts in the usual Aristophanic style, but little value attaches to their evidence, which has probably no real authority, but is merely gathered out of Aristophanes’s own words: it does not show why συνθήκη] (which would suit the metre) is not used rather than διαθήκη. Lightfoot says there are a few other examples of διαθήκη in that sense, but he quotes none, and they are unknown to Steph. Thesaurus; and we must require exact quotations to support so rare a use in prose. Hatch carries further the loose language into which Lightfoot (a rare thing with him) has fallen, and speaks of the Hellenistic usage of διαθήκη as being similar to that of the Septuagint.

[7] Mitteis, Reichsrecht, p. 336, quoting Caillemer, Annuaire de l’Assoc. des Ét. Gr., 1870, p. 34 f.

[8] A good example of this is given in Inscriptions d’Asie Mineure in Rev. Et. Gr., 1889, p. 18.

[9] See Professor Mahaffy, The Flinders-Petrie Papyri, introduction, p. 35 ff.; Grenfell, Erotic Greek Papyrus; Grenfell and Hunt, The Oxyrhynchus Papyri; Griech. Urkunden aus den kön. Museen, Berlin; Kenyon, Greek Papyri of Brit. Mus. (contains only one very late Will); I have seen some unpublished Wills copied by Mr. A. C. Hunt, but have not access to other publications.

[10] Bruns and Sachau, Ein Syrisch-römisches Rechtsbuch aus dem funftenn Jahrhundert, 1880.

[11] See p. 368.

[12] Mommsen in Berlin Sitzungsber, 1894, p. 48 ff.

[13] See above p. 338.

[14] The form was that the heir inherited the whole, but was obliged to pay out of the property such legacies as the testator ordered. The Lex Falcidia restricted these legacies to three-quarters.

[15] Gizeh Papyrus, No. 10,388, communicated by Messrs. Grenfell and Hunt.

[16] Mitteis, Reichsrecht, p. 336.

[17] Communicated by Messrs. Grenfell and Hunt: to be published in the Oxyrhynchos Papyri, II.

[18] Dr. Ball in Contemp. Review, Aug., 1891, p. 287. Compare above, p. 341 f.

[19] See Sections 11, 13.

[20] Josephus, Ant. Jud., 17:9, 5.

[21] Expositor, June, 1899, p. 401 ff.

[22] So R. V. in margin and the American Revisers in text. R. V. in text puts the words as a question.

Book Navigation Title Page Preface Table of Contents Religion in Asia Minor      ► Chapter 1      ► Chapter 2      ► Chapter 3      ► Chapter 4      ► Chapter 5      ► Chapter 6      ► Chapter 7      ► Chapter 8      ► Chapter 9      ► Chapter 10      ► Chapter 11      ► Chapter 12      ► Chapter 13      ► Chapter 14      ► Chapter 15      ► Chapter 16      ► Chapter 17      ► Chapter 18      ► Chapter 19      ► Chapter 20      ► Chapter 21      ► Chapter 22      ► Chapter 23 Historical Commentary      ► Section 1      ► Section 2      ► Section 3      ► Section 4      ► Section 5      ► Section 6      ► Section 7      ► Section 8      ► Section 9      ► Section 10      ► Section 11      ► Section 12      ► Section 13      ► Section 14      ► Section 15      ► Section 16      ► Section 17      ► Section 18      ► Section 19      ► Section 20      ► Section 21      ► Section 22      ► Section 23      ► Section 24      ► Section 25      ► Section 26      ► Section 27      ► Section 28      ► Section 29      ► Section 30      ► Section 31      ► Section 32      ► Section 33      ► Section 34      ► Section 35      ► Section 36      ► Section 37      ► Section 38      ► Section 39      ► Section 40      ► Section 41      ► Section 42      ► Section 43      ► Section 44      ► Section 45      ► Section 46      ► Section 47      ► Section 48      ► Section 49      ► Section 50      ► Section 51      ► Section 52      ► Section 53      ► Section 54      ► Section 55      ► Section 56      ► Section 57      ► Section 58      ► Section 59      ► Section 60      ► Section 61      ► Section 62      ► Section 63      ► Section 64