The Ancient Church Orders (Maclean)

The Ancient Church Orders (Maclean)

continued

Chapter IX: Relative Position and Dates of the Church Orders

1. It was remarked in the first chapter of this book that several of the Church Orders agree so closely together that it is necessary to postulate a common original for them. This applies to those in the first division particularly, but also to those in two of the other divisions. The problem is not unlike that of the Synoptic Gospels. Let us then begin by taking the first division, and endeavour to place the manuals contained in it in their right order, and to fix their approximate dates. Here we shall find very considerable differences of opinion, and any result which we obtain must be considered tentative only. It is well, however, at the outset to make quite clear what the question is. Most of these Church Orders are parts of longer compilations; and it is often necessary to separate the question of the date of the Church Order from that of the date of the compilation. In the case of AC, indeed, it is common ground that the compiler of the eight books was also the compiler of the Church Order found in book viii. In the case of SEC (the Egyptian Heptateuch) and EthS it is very unlikely that the compilers of the whole were also the compilers of EgCO and EthCO respectively; for the same matter is gone over twice, in different ways—that is, first in EgCO (EthCO) and then in the Church Order which is nearly identical with AC viii.

2 He would be a strange compiler who published in the same work a Church Order of his own and a closely parallel Church Order adapted from AC. We must, then, be on our guard against assuming that the date of a Church Order is necessarily the date of the compilation in which it is found. On the other hand we must remember that all the writers of the first division, with, at most, a single exception, used one or more sources. And it is always possible that the Church Orders themselves have been more or less interpolated. We have then, further, to guard against the presupposition that the date of a Church Order, as a separate work, is necessarily the date of either the earliest or of the latest custom or phrase found in it. For example, the reference to persecution does not necessarily mean that AC is ante-Nicene, nor the presence of the Filioque in CH that that manual belongs to the sixth century. The question to be solved is, At what dates did these writers produce out of the materials at their hand the Church Orders, as separate books of instruction and guidance for their fellow Christians?

3 2. Hypothesis of a Lost Original. One of the Church Orders in our first division must be the source on which the rest depend, or all must depend on some lost source. The view that one of them is the source on which the rest depend is taken by Achelis (Die Can. Hipp.) and Funk, the former taking CH, the latter AC viii. as the original. Funk's hypothesis will be considered later; but it seems desirable to say here at once why in the present writer's opinion Achelis' view (at any rate in the form in which he has put it) cannot be accepted. Putting aside the questions of the passages marked by Achelis as interpolations (below p. 157 f.) and of the Hippolytean authorship, it is hardly credible that there should be such a large amount of material in CH which does not appear in any of the other Church Orders, if CH is the parent of them all. The preface (i. 1-6) would surely have left some trace behind it; the ordination prayer for bishops could hardly be the original of that in the other manuals, containing as it does so many phrases which have absolutely disappeared (above p. 75 f.); that for deacons (p. 71) has left no trace in the other Church Orders; the provisions about married and travelling presbyters (viii., ix. 55-58), the exhortation in xxxviii. 255-261, and the rules about vestments at the Eucharist (xxxvii. 201-204) are wholly wanting elsewhere; as also are several individual passages, for which see Cooper-Maclean p. 9, note.

4 We may conclude, then, that CH is not the original; indeed Achelis only maintains that it is so after omitting several passages which he marks as interpolations (not, however, most of those just noted). We shall see equal reasons later for rejecting Funk's hypothesis. If so, it follows that there must have been an original which is now lost; for no one maintains that any other extant manual is the source of all the others. One indication that these Church Orders depend on a lost Original is that they contain obscure and confused passages which it is impossible to understand completely, and which could not with any probability be held to be original in any of the manuals. Take, for instance, a phrase in EthCO, H₃, and Test about the first communion of the neophytes. Each writer has treated the phrase in a different way:

ETHCO 35 (Horner)

5 He (the bishop) shall give thanks over the bread and the cup; the bread that it may become the body of our Lord Christ, and the cup, the wine mixed, that it may become the blood of our Lord Christ, this which was shed for us and for all of us indeed who believe in him...[milk and honey]...All this the bishop shall go through to those who are baptized. And the water also of the Oblation he shall show in the bread like the inward part of man, who is soul as well as body.

H₃ p. 112

6 Gratias agat panem quidem in exemplum, quod dicit Graecus antitypum, corporis Christi; calicem uino mixtum propter antitypum, quod dicit Graecus similitudinem, sanguis quod effusum est pro omnibus qui crediderunt in eum...[milk and honey]...aquam uero in oblationem in indicium lauacri, ut et interior homo, quod est animale, similia consequantur (sic) sicut et corpus. Let the shepherd give thanks. But the bread is offered for a type of my body. Let the cup be mixed with wine—mixed with wine and water, for it is a sign of blood and of the laver; so that also the inner man, that is to say, that which is of the soul, may be counted worthy of those things which are like (them), that is to say, those things of the body also. There is nothing in any of the other Church Orders corresponding to these difficult passages about the water in the chalice. We may probably conclude that there was such a passage in the original Church Order, that it early became corrupt, and that some compilers omitted it as unintelligible, while others made the best sense they could out of it. We cannot conceive that any of the above passages is the original of the others.

7 Or, again, take as an example the treatment of the honorary presbyterate of confessors (above, p. 87 ff.). In all these Orders, with the exception of AC (which in this instance could not possibly be the original of the rest, and which in any case has not become corrupted in the MSS) the passage is hopelessly confused, though the general sense is clear. Here again we can hardly believe that any one of them is the parent of all the others. The same thing is seen from some passages which are not confused. Take the passage about soldiers (CH xiii. f. 71—75, EgCO 41, EthCO 28 f., Test ii. 2, AC viii. 32, ConstH 22; see above, p. 97). CH says that no one who has received the power of M. 8 killing or is a soldier is to be received. Those who, being soldiers, are commanded to fight, but have abstained from evil talk and have not placed crowns on their heads, etc. (may be received); a magistrate not clothed with justice is to be excommunicated; no Christian is to become a soldier except by force, and if so he is not to shed blood. EgCO says that a soldier who is in authority is not to kill men; if commanded to do so he is not to hasten to the work, nor to swear; one who has power over the sword or a magistrate of a city who is clothed with purple, unless he leaves (his occupation), is to be rejected; no catechumen or believer is to become a soldier. EthCO says that a soldier of the prince is not to be received, and if ordered to kill he is not to do so; a soldier, whether baptized or a catechumen,...and a magistrate with the sword, or chief of prefects, and he who is clad in red is to be rejected; no catechumen or believer is to become a soldier. [H₂ has a lacuna here.] Test says that a soldier or one in authority is to be taught not to oppress or kill or rob... but if they wish to be baptized they must leave (their occupation). No catechumen or believer is to become a soldier. For AC and ConstH, which remove the prohibition, see above, p. 97. In these passages the meaning is quite clear. Yet the wording is such that we can hardly conceive any one of them to be the source of the rest.

9 It does not, of course, necessarily follow that each of the extant Church Orders is independent of all the others, except in being derived from a common original. There is much evidence, for example, that the Test writer had a manual very like H₃ before him when he was writing*. But it seems clear that those Orders which are now extant are not all in one line of tradition. For example, it is commonly said that Test is derived from the Church Order of the Egyptian Church, and this is very possibly true, if we take that phrase in a wide sense, as meaning some Church Order written in Egypt; but Test is certainly independent of our EgCO in some things; e.g. in the formula of submission to God at baptism and in the baptismal creed; in much of the baptismal section (ii. 8) Test is unlike EgCO and nearer CH and H; so the passage (ii. 11) about alms for a widow is in Test and CH, not in EgCO; that quoted above about the water in the chalice is in Test, H₃, and EthCO, not in EgCO.

10 It appears, then, to be a fundamental mistake to assume that, if one of these Church Orders has earlier features than another, it must be its direct ancestor. Indeed it would not be surprising if it proved to be the case that most of the extant manuals had only an indirect relationship with one another, as cousins rather than as parents or children. At any rate it is probable that many forms of the Original Church Order were extant in widely different circles—Montanist, Catholic, anti-Arian, semi-Arian, * Similarly it is manifest that ApCO was known to, and had a great influence upon, him. Another very confused passage (see CH xxxii. 160—163, and Arendzen's translation of the Arabic in JThSt iv. 282), which illustrates the necessity of our present hypothesis of a lost Original. 10—2 perhaps Arian—and that each author dealt very freely with his materials, adapting them as he thought best for his own immediate surroundings. It is not improbable that many of the writers were bishops, drawing up rules for their own dioceses.

11 It may be permissible to hazard a conjecture with regard to the 'Lost Church Order.' It is quite possible that it was the work of the great Hippolytus, Bishop of Portus, and that this was the reason why two of the extant manuals (CH, ConstH) received their titles. We know that Hippolytus wrote a work called The Apostolic Tradition (ἀποστολικὴ παράδοσις); this is mentioned on the Chair of Hippolytus at Rome* immediately after the treatise Of Gifts (περὶ χαρισμάτων). Perhaps these two works came together in the earliest MSS, and were read as one book, which would account for the arrangement of ConstH and AC viii (where the sections Of Gifts are followed immediately by the Church Order), and also for the prologue of H₃ (below, p. 162 f.). Lightfoot, indeed, regards these two phrases as one single title (περὶ χ. ἀπ. παράδοσις), but he thinks that perhaps Hippolytus drew up some canons, and that early canonists prefixed to their canons some sentences of Hippolytus from the περὶ χαρισμάτων (Clement, ii. 400; see also his remarks on p. 402 and his account of the chair at pp. 440 ff. 324 ff.). Either of these hypotheses would account

12 * A marble chair, with part of a seated figure, was discovered near Rome (close to San Lorenzo) in 1551, and was proved to be that of Hippolytus by the names of the works inscribed on the back. It is in the Lateran Museum. The date, according to Lightfoot, is c. A.D. 236. for the phenomena.—It is probable that the Lost Church Order contained at least the outline of a Liturgy, and also ordination prayers and rules for the organisation of the Christian community. 13 3. Date of the Apostolic Constitutions. It will be convenient to begin with this work, as there is general agreement that it was compiled, and that the Church Order in bk. viii. was composed, by a single writer in the latter half of the fourth century. The indications of date come from the book itself, as we must dismiss the idea that Epiphanius quotes AC; he is really quoting a form of Didasc older than AC. That the book is a uniform whole, by a writer who first interpolated Didasc (i.—vi.), then D (vii. 1—32), and added some liturgical matter (vii. 33—end), a treatise On Gifts (viii. 1, 2), and a form of our Church Order (viii. 3—end), is seen by an examination of the phraseology, style, and other characteristics; see the careful summary in Brightman LEW p. xxiv ff. The date of AC is determined very closely by the festal cycle (see above, p. 129 f.). This would have merely given us a terminus a quo, but for the obvious fact that the AC writer is endeavouring to lead the way in separating Epiphany and Christmas and in developing the cycle; and therefore we are brought to a date as early in the last quarter of the century as possible, say c. A.D. 375. That the work is post-Constantinian is seen e.g. by vi. 24, which speaks of the conversion of the Roman Empire, and in which the reference to the Romans is in striking contrast to that in || Didasc (DidCA i. 364).

14 It is generally agreed that AC was written in Syria; the Syro-Macedonian calendar is used in v. 14, 17, 20, where the months Xanthicus, Dystros, Gorpiaeus are named (not in || Didasc). The fact that Holy Week is not included in the Forty Days (p. 137 above) excludes Palestine, Egypt and the West. For other indications see LEW p. xxviii. But the absence of metropolitans suggests that neither Antioch itself nor its immediate neighbourhood was the home of the writer (above, p. 72 f.). 15 The forms of Church Order in SEC 63—78 and EthS 49—72, which are no doubt derived from AC (or from an earlier draft of AC), may only date from the time of those compilations. We cannot get very near to the date, but they may have been composed at any time in the fifth century, perhaps early in the century.

4. Relation of the Constitutions through Hippolytus to the Apostolic Constitutions.

16 These two documents are very closely related, whether they are the work of the same author or not. The question arises as to which document has priority. Is ConstH the original, or an epitome, of AC? We must first notice that in both documents the sections are divided among the Apostles, and that both enumerate fourteen as being present, the Twelve, St Paul and St James (AC viii. 4, ConstH 3). But in ConstH, of the fourteen only ten have sections assigned to them, while in AC all fourteen have sections, and the four omitted in ConstH have in AC those sections assigned to them which ConstH has not got. Moreover in AC viii. 29 Matthias has assigned to him the section about oil and water (which is not in ConstH) as well as viii. 30: 'I the same...' But ConstH 20 (which corresponds to AC viii. 30) is assigned to Simon the Cananaean (to whom both works assign the sections preceding that about oil and water). It begins 'I, the same Simon the Cananaean command...' Thus it appears that ConstH is an excerpt of a work in which all fourteen Apostles had sections assigned to them. Of what work is it an excerpt?

17 The simplest solution is undoubtedly that of Funk, who holds that ConstH is an epitome of AC viii. (DidCA ii. p. xiv ff. TUH 179 ff.). Funk considers that a writer other than the AC compiler shortened that work by omitting the liturgical matter, abbreviating the bishop's and presbyter's ordination prayer, and so forth, and corrected what he did not approve of, such as the ordination of a reader with laying on of hands. He holds also that one or two phrases in ConstH show posteriority. 18 The problem, however, is far from easy. The careful attention of the reader is asked for what is said above (pp. 74-80) about the development of the ordination prayer for a bishop, and the ordination of minor orders by imposition of hands. Funk's idea of successive writers becoming more antiquarian and producing simpler forms as time went on is hardly credible. If the ConstH writer was a corrector of AC, and did not approve of what seems to have been an invention of the latter, i.e. the ordination to some of the minor orders by laying on of hands, why did he not cut it out altogether as the || SEC 66 and EthS 54 (see above p. 23) have done? Why did he only forbid it in the case of readers, and leave it for subdeacons and deaconesses? Is it not much more probable that the AC compiler was gradually feeling his way? At first he prescribes the laying on of hands in the ordination of subdeacons and deaconesses, as these offices might be supposed to be derived from the diaconate, but he does not yet (in ConstH or its original) venture so far in the case of the old charismatic office of the reader. Then, later, he goes to the logical conclusion of ordaining readers in the same way. Funk seems to have overlooked the fact that the ordination to minor orders with imposition of hands is not found before AC, and to all appearance was the invention of the AC compiler. The invention did not find favour at first, at any rate in Egypt, as we see in SEC and EthS. The fact, then, of the ordination conditions being simpler in ConstH than in AC (and simpler still in EgCO, EthCO, H₂, and, as far as minor orders go, in Test) makes Funk's solution of the relation between ConstH and AC an improbable one.

19 Let us, however, consider the points of phraseology which are thought to indicate posteriority in ConstH. (a) In § 1 we read of believing 'through Christ our God on him and on his undefiled Father and the all holy and life-giving Spirit,' while in || AC viii. 1 we have simply through Christ on God.' The AC phrase is said to indicate a 'more ancient doctrine of the Trinity' and the same is said of the subsection which follows. It is not very easy to see the force of this. When the documents confessedly do not date from before the last quarter of the fourth century, we are not helped to decide on a question of priority by the presence in one or the other of an older phraseology on this subject. In this case the probability is rather the reverse of what Funk supposes. The AC writer, a pronounced subordinationist, probably altered the phrase which he had at first written (as we read it in ConstH) so as not to name the Son first (cf. above, p. 118); and if he is the same as Pseudo-Ignatius we must remember that his semi-Arian tendencies increased as time went on (p. 124), for the interpolated Ignatian Epistles are certainly later than AC (Lightfoot, Ignatius i. 263 f.). (b) The phrase in the bishop's ordination prayer about the Spirit given to or through Jesus and to the Apostles (above, p. 75) is thought to show the priority of AC over ConstH. This does not follow.

20 And if it did follow, H₃ and Test would be, equally with AC (with which they here agree), older than ConstH; which supposition (though probably quite true) would upset Funk's theory of their relationship. (c) In the matter of the reader ConstH is said to show signs of being a correction of AC. We have already discussed whether the provision of ConstH or that of AC about the reader is likely to have been the older, and have seen reason for assigning the priority to ConstH. But Funk says that the wording itself shows the posteriority of ConstH. That manual omits, at the section in question, the name of the Apostle who gives the command; AC inserts it; this, however, we need not consider, as we have already reached the result that ConstH is an excerpt from a work which had the full complement of Apostolic ascriptions. But whereas the preceding and following sections of ConstH use χειροτονία, χειροτονεῖν, this section has οὐδὲ γὰρ χειροθετεῖται—those about confessors, virgins, widows have οὐ χειροτονεῖται—and this may be thought to show that the ConstH section about a reader is a correction. AC itself uses χειροτονία, χειροτονεῖν in all these sections except in those about deaconesses and readers, where neither χειροτονεῖν nor χειροθετεῖν, nor their substantives, are used. The conclusion that because ConstH uses χειροθετεῖν in the section about the reader, therefore it is a correction of AC, is an insecure one. The fact is that AC uses χειροθετεῖν of laying on of hands in general;

21 when the laying on of hands is emphasised, χειροθετεῖν is used; when ordination as a whole rite is spoken of, AC uses χειροτονεῖν. Both ConstH and AC therefore say that an exorcist or a confessor or a virgin or a widow 'is not ordained' (οὐ χειροτονεῖται)—they are not set apart by any office which could be called ordination, because their position was still entirely a charismatic one; but a reader, like a subdeacon, is in this sense 'ordained,' though he is not in ConstH, as he is in AC, ordained with laying on of hands. Therefore ConstH uses χειροθετεῖν here. It would not be quite true, or at least it would be ambiguous and inconsistent with AC usage, to say of a reader οὐ χειροτονεῖται; it is true of him, in ConstH, to say οὐ χειροθετεῖται (see appended note below). 22 The indications, then, point in the opinion of the present writer to the priority of ConstH over AC. But as the former is certainly an excerpt from a document which named all the fourteen Apostles, the solution which perhaps is the most probable is that once propounded by Harnack, that ConstH is a shortened form of an original draft of AC viii. The word χειροθεσία (or χειροθετεῖν) is used in AC for the various impositions of hands, or even (probably) where a hand is extended over the people as in a blessing. It seems to include both Confirmation and Ordination in iii. 10: laymen are not to do 'any of the priestly works such as the sacrifice (θυσία, the Eucharist) or baptism, or laying on of hands (χειροθεσία), or blessing (εὐλογία), great or small; for no one takes this honour to himself unless he be called of God, for by the laying on of the bishop's hands (διὰ γὰρ τῆς ἐπιθέσεως τῶν χειρῶν τοῦ ἐπισκόπου) such an office (ἀξία) is given.' It is used of Confirmation in ii. 32, iii. 16, vii. 44, and apparently in viii. 28f. (ἐπίσκοπος εὐλογεῖ...χειροθετεῖ, χειροτονεῖ...πρεσβύτερος εὐλογεῖ...χειροθετεῖ, οὐ χειροτονεῖ); so ConstH 19, except that the first χειροθετεῖ is omitted; in the last passage a presbyter is perhaps allowed to lay on hands in Confirmation, doubtless only in the absence of the bishop, see above, p. 107 f. It is used of the dismissal of catechumens with a blessing in vii. 39; of the reception of penitents (ii. 18, 41, 43); and of blessing in general (viii. 37, 39).

23 χειροτονία (or χειροτονεῖν) is used in AC for Christian ordination to all grades; thus, in AC iii. 11 it is used of deaconess to all minor orders, including singers and doorkeepers, who in AC are not ordained with laying on of hands (cf. also iii. 20). It is also used in ii. 27 of the 'ordination' of Aaron by God; in iii. 9 of heathen appointments of priestesses; in v. 20, vii. 16 of God's 'ordaining' the Roman empire and the Roman emperor. It is true that elsewhere than in AC χειροτονεῖν is used when laying on of hands is emphasised, as in SEC-sah 66, which says that 'it is not right to ordain (khirodonei) subdeacons, readers and deaconesses,' but this is not the case in AC. 4. As we have already seen (p. 78 above), κατάστασις is used of appointments to any grade of the ministry. 5. Date of the Canons of Hippolytus. It is in the case of this book that the widest difference of opinion exists. Achelis, who believes that it was written by Hippolytus himself, necessarily dates it very early, c. A.D. 220; but he considers that it has been considerably interpolated. Morin assigns it to Dionysius of Alexandria and identifies it with the 'diaconal' epistle 'sent to those in Rome through Hippolytus' (c. A.D. 250; Euseb. HE vi. 46). Funk, on the other hand, holds this Church Order to have been derived (through ConstH and EgCO) from AC viii. and takes the alleged interpolations to be part of the text. Harnack has lately given his adherence to this view, with some considerable modifications. Funk inclines to a date in the sixth century.

24 There are many indications in CH of an early date (Achelis, Die Can. Hipp. p. 217 ff.). Although persecutions were not going on at the time of writing—for there are no penalties for the lapsed, and the section about teachers (xii. 69 f.; see above, p. 116) would not be possible in times of persecution—yet it was probable that persecution would revive (vi. 43—47). The traces of close relationship between bishop and presbyter (p. 69 f. betray a date earlier than that of the Church Orders which have a separate prayer for ordaining bishops and presbyters; the simple nature of the ordination prayer (p. 75 f.), the fact that there is no fixed day but Sunday for the Eucharist (p. 55 ff.), the arrangement that Easter is to be kept at the same time as the Jews' Passover, and the simplicity of the festal cycle, all tend to show an early date. Some of Achelis' indications will hardly command assent, such as the fact that charismata are still held in the highest honour; for the other manuals have the same phraseology. Achelis is on firm ground when he points out the great similarity between CH and Tertullian in fasts, in baptismal ceremonial, in rules about virgins' veils, about soldiers, and about care for the Eucharistic elements; but this need prove no more, in view of similar regulations being found in other Church Orders, than that the customs of the lost Original closely agreed with those mentioned by Tertullian.

25 Achelis thinks that Egypt is excluded as a place of writing of CH by the fact that Agape and Eucharist are already separated, each with a ceremonial of its own, whereas in Clement of Alexandria they are still joined [but it is doubtful if this was the case, see ERE i. 170]. This, however, is a two-edged argument. If the rest of CH inclines us to assign it to Egypt, then the separation of Agape and Eucharist would lead us to a date later than Clement. 26 On the other hand there are indications of a later date: e.g. the passages referring to the 'equal Trinity' (i. 2, xix. 133), to subdeacons (vii. 49, 52, xxi. 217), to the forty days' fast of Lent and to the Holy Week fast (xx. 154, xxii. 195 ff.), and the Filioque* in the Baptismal Creed (xix. 131). Most * The Filioque as part of the Nicene Creed is found first at the Third Council of Toledo (A.D. 589) which introduced the Creed into the Eucharistic Liturgy of the West. But Western writers after Augustine had spoken of the procession from the Son. of these are considered by Achelis to be interpolations; Funk on the other hand considers them necessary to the text and ineradicable, and Harnack agrees (Chronologie ii. 507). Certainly interpolations are probable in an Arabic translation; a parallel case is the Arabic translation of Test which constantly alters and interpolates. But there is much force in Funk's argument on this head (TUH p. 224 ff.), and it is hardly satisfactory to date a document from certain features of it and then to mark all that does not agree with that date as interpolation.

27 It has already been argued that CH is not the direct parent of the other Church Orders (above, p. 143). May not the truth be that we should look on the early features as survivals from the lost Original, provided we do not make CH so late that the early features have become mere antiquarianisms put in for effect (see above, p. 6)? On the other hand, it is a fatal defect in Funk's argument for a late date that CH would be absolutely useless as a manual at the time which he suggests. It would be merely a pseudo-historic curiosity, a farrago of false antiquarianisms. The existence of the parallel Church Orders forbids such a supposition. We must therefore reject Funk's late date, while we may nevertheless incline to the idea that almost all the alleged interpolations are part of the text. Yet some passages may remain as interpolations. Such, at least, must be the Filioque in the Creed—the very place where we should expect assimilation to a later writer's own usage. This is the only phrase in the whole book which points to a date later than the middle of the fourth century. Putting together all these considerations, the first half of that century seems to be the most likely time of writing; the doctrinal position, the ordination conditions, the regulations for festivals and fasts, the existence of subdeacons, all point to a rather early year in the century. Nevertheless, it is probable that CH reproduces the lost Original more faithfully than any other of these manuals; and to this extent Achelis' theory may be held to be well founded.

28 In considering the place of writing, we must dismiss pre-suppositions of Hippolytean or Dionysian authorship, and take the book as it stands, while allowing for a very few possible interpolations. The internal evidence seems to point either to Rome or to Egypt. But in reality only one feature, the presence of the Roman Creed in the baptismal service, indicates Rome; and as the same creed is found in Test and H₃ (which are certainly Eastern) the only inference that can safely be drawn from its presence in CH, is that at an early date it made its way to the East. The other arguments adduced for a Western origin (as by Wordsworth, MG p. 22 ff. apply to the Lost Church Order which is so faithfully reproduced by CH, rather than to CH itself. On the other hand, CH bears a great likeness to EgCO, EthCO, H₃; to take only two examples among many, these four manuals agree in differing from the cognate Church Orders in saying nothing about clerical marriage or about deaconesses. These resemblances would point to Egypt as the place of writing; and this indication is to some extent confirmed by the existence of the forty days' fast in a writing probably compiled before A.D. 350, for it was in Egypt that the forty days' fast seems to have been first established (Duchesne, Chr. Worship. Eng. tr. p. 242).

29 A stronger confirmation may be found in the administration of milk and honey in CH at Baptism, a custom for which there is no evidence outside Egypt and Africa (above, p. 46). If CH had been written in Rome we should have expected to find acolytes mentioned, for even in the third century there were forty-two of these officials there (above, p. 80). The form of the salutation before the Sursum Corda in the Liturgy ['The Lord be with (you) all'] points either to Rome or to Egypt (Cooper-Maclean, p. 169). On the whole, while the evidence is far from being conclusive, the most probable conclusion seems to be that CH was written in Egypt, though, since no metropolitan is mentioned (above, p. 72 f.), not in Alexandria. The influence of Alexandria may perhaps be seen in the retention by the CH writer (from the Lost Church Order) of the very archaic ordination conditions.

30 6. Date of Egyptian and Ethiopic Church Orders and Verona Fragments, part 3. These manuals may for this purpose be considered together. None of them can with any probability be dated before the fourth century. The emphasis on the 'equal Trinity' and the form of the doxologies (pp. 115, 125) would make an earlier date very unlikely. But there are several indications that they are earlier in date than AC. The Liturgy in EthCO and H₃ is almost certainly more primitive than that in AC (above, p. 50 ff.); the absence of the Sanctus would be difficult to explain after the fourth century; the absence (unlike AC, Test) of any account of the church buildings would be natural early in that century. The ordination prayer for a bishop in AC, as well as that in Test, are almost certainly based on that in EthCO and H₃ or on something very like it (p. 77). On the other hand EgCO must probably be earlier than EthCO and H₃ if, as seems likely, it prescribes the same prayer to be used in ordaining a bishop and a presbyter (p. 70f.); and this consideration also points to its being earlier than Sar which (c. A.D. 350) has separate prayers. The doctrinal indications in all three manuals points to the first half of the fourth century (p. 114 f.). The Macedonian heresy seems not yet to have broken out. That the Arian heresy had arisen is shown by the phrases about the Holy Trinity; but the writers do not seem to have occupied themselves much with the controversy. We may probably assign these three manuals, or rather their Greek originals, to the first half of the fourth century, perhaps rather early in that period. Their home was, doubtless, Egypt.

31 Funk (DidCA ii. p. xxi, TUH 132 ff.) gives two indications of a date later than AC. (a) In the section about the appointment and ordination of a bishop, EgCO and EthCO begin with 'as we have before said.' Funk thinks that these words are derived from AC viii. 4 and M. 32 refer to the foregoing books of AC, thus showing that EgCO and EthCO are later than AC. But this argument is not sound. The words are not in H₃; they seem therefore to have been no part of the original EgCO, but to be due to the compiler of SEC (who has been followed here, as often, by EthS), and to refer to ApCO, which immediately precedes in that compilation. If Funk's argument shows anything, it is that the compilers of SEC and EthS lived after the AC writer, which no one doubts. (b) H₃ has a prologue, i.e. a section between ApCO and the Church Order which follows it; and the prologue begins by referring to a tract Of Gifts. No such tract precedes in H, but one so named does precede AC viii. 3, which also refers to it. Funk's deduction is that H₃ is later than and derived from AC, and that its prologue is derived from AC viii. 3*. But there are other facts about the prologue which show that the problem is not nearly so simple as Funk thinks. The prologue and AC viii. 3 nearly agree in the first sentence ('we have now finished the first part of the discourse concerning gifts, whatever they be, which God has bestowed upon man according to his will,' AC). But after this sentence they are quite independent. Again, the H prologue reappears, in nearly but not quite identical terms throughout, in the interpolated EthS 40 A (p. 21) in quite another connexion. In both these two forms the paragraph is confused (especially in H₃) and neither could well have come directly from the other. In these two forms the language used has a much older flavour than that of AC viii. 3, and it is not easy to think of them as deriving their first sentence from AC and then proceeding on their own lines. Again,

33 * This section is not in ConstH, but reappears in SEC 63 and EthS 52 (much altered). H: Ea quidem quae uerba fuerunt, digne posuimus de donationibus, quanta quidem Deus a principio secundum propriam uoluntatem praestitit hominibus... the prologue is not found in EgCO or EthCO (the interpolated EthS 40 A is no part of EthCO), and if, as Funk says, it is 'not doubtful' that it formed originally a part of the Egyptian law book, it is hard to see why all trace of it (not only of the first sentence) has vanished from EgCO and EthCO, which begin, without preface, immediately after ApCO in the compilations where they are found (SEC, EthS). The solution of this small but puzzling problem which the present writer would offer has already been hinted at. The prologues in H₃ and in EthS 40 A clearly had a common ancestor; this seems to have become corrupted at an early date, and so it has reached us in two confused forms. One is tempted, then, to think that the Lost Church Order, which as we have already seen reason to suppose, was perhaps added on to the Hippolytean treatise περὶ χαρισμάτων (p. 148), had a preface, of which the prologues of H₃ and EthS 40 A are corrupt versions, and which gave the idea for AC viii. 3.

34 Here the prologue has been spoken of as an original part of H₃. We must always bear in mind another possibility, that the author of the Church Order (H₃) is not the same as the compiler of the whole of the Verona fragments, and that the prologue in question is due to the compiler. In that case the prologue would give us no indication of date for H₃. 7. Date of the Testament of our Lord. Here, again, we have great difference of opinion; Rahmani, the first editor, assigned it to the time of Irenaeus; Zahn gives c. 350, Wordsworth and Harnack c. 400 (Harnack would now place it later), Funk c. 475. That this Church Order cannot be dated before the fourth century appears from the existence of subdeacons in it (for it is an Eastern manual), from the forty days of Pascha' (p. 132), the elaborate church buildings (p. 36), the appointment of Saturday for public service (p. 57), the doctrinal tone (p. 117 ff.), the elaboration of the Liturgy, and the steps taken towards a (not yet existing) daily public service for all men (p. 59). See, further, Funk (TUH p. 62 ff.), who refutes Rahmani at length. On the other hand, marks of ante-Nicene date, such as the mention of persecution, and the relics of the ancient position of the reader still found in Test, can be well accounted for as having been taken from older materials; they fulfil the conditions suggested on p. 6 as not contradicting contemporary usage, and as not being mere antiquarianisms.

35 There are many positive marks of date. The doctrinal indications suggest a date before the secession of Apollinaris from the Church in A.D. 375, even before the outbreak of the Macedonian heresy (p. 119 f.); the prevalent form of doxology points to a date a good deal earlier than the end of the century, if we remember the standpoint of the writer (p. 127). The less developed prayers, especially in the Liturgy (p. 49 f.), suggest that Test is earlier than AC, as do the restriction of laying on of hands in ordination to the three highest orders (p. 78) and the festal cycle (p. 128). Funk remarks (TUH p. 82 f. that the simple festal cycle in Test only means that it and AC were not written in the same place, since we find in 'Silvia' that Christmas and Epiphany were still one at Jerusalem at the end of the fourth century. That is so; but even making every allowance for the AC writer having desired to extend the festal cycle, it is difficult to conceive that the very simple cycle of Test (Easter, Pentecost, Epiphany) was that used in any country where the manual is likely to have been written, even at the beginning, much less at the end, of the fifth century. The absence of metropolitans would be very unlikely in that century (p. 72 f.). The supposed mention of an archdeacon has been alleged (Wordsworth, MG 32, 162 n) as forbidding us to date this Church Order before A.D. 400.

36 But the Syriac suggests that the name ἀρχιδιάκονος was not used in the Greek original (p. 72), and in any case there is no mention of such an official as holding a distinct rank; the 'chief deacon' is merely one of the deacons told off for two special tasks, neither of them in the Church services. This feature really points to a date a good deal before A.D. 400. So also does the mention of a 'stole' worn by him (p. 65). Singers in Test are not yet a class or order, as in AC and at Laodicea (p. 82). The widows προκαθήμεναι are represented as being at the height of their popularity; at the time of the Laodicean council they were dying out (p. 84). The absence of monasticism and of penitential 'stations' requires some consideration. The latter gives us no certain indication (p. 94); the former points to a date before the end of the fourth century. In the latter half of the century many monastic communities were founded and were enthusiastically received. In so strict a writer as the compiler of Test we should certainly expect a reference to them if he wrote in the fifth century. The only ascetics mentioned are the 'virgins, male and female,' who are expressly described in i. 46 as individuals, helping on the work of the Church, but clearly not living in community.

37 On the other hand we have seen in the preceding chapters that Test is derived, at least in part, from a Greek manual very like H₃. We shall therefore probably be approximately right if we ascribe this Church Order to a date about A.D. 350. The country of the writer had a sea coast (i. 34) and was well watered, there being no reference, as in EgCO, to scarcity of water for the font (ii. 8); there are no interpreters, no milk and honey at baptism, no stated Wednesday and Friday fasts. It was probably therefore not Egypt, though the Liturgy is derived from an Egyptian source. The great dissimilarity between Test and AC makes Syria unlikely; and perhaps, if we take into account the Montanist influence and the position of women, we are within the bounds of probability if we fix on Asia Minor as its home. 38 Note. In the foregoing nothing has been said about the apocalyptic prelude of Test (i. 2—14a), which describes the signs to be expected before the end of the world and the coming of Antichrist, because it is quite uncertain if it was composed by the compiler of Test, or if it was an already existing tract merely prefixed by him to his work. Thus a mark of early date in the prelude is not necessarily a mark of early date in Test. A mark of late date in the prelude would, however, point to a late date for Test. There is some trace of its having had once an independent existence; at any rate we have an independent Syriac translation of it (JThSt ii. 401, Arendzen). A shorter form of it exists, which ends with 'Dexius shall be the name of Antichrist,' and Harnack suggests that this is a disguise for Decius, and that the shorter form dates from the Decian persecution. Some considerations, however, point to the shorter form being an excerpt from the longer. In any case it is doubtful if the prelude is by the Test compiler; it has some of the favourite expressions of Test, but on the other hand it has some striking differences of style.

39 Then there are in Test chapters introducing the prelude (i. 1) and joining it on to the Church Order (i. 14b—18), which have the appearance of having been added by a hand other than that of the writer of the prelude, and which show marked differences of style as compared with the prelude. It seems therefore probable that the Test compiler took an already existing tract on the End (from which he perhaps got the idea of ascribing his book to our Lord), and incorporated it in his manual (see, further, Cooper-Maclean p. 141 ff.). The prelude shows no sign of a post-Nicene date, certainly not of a date later than the fourth century. It has therefore no certain bearing either on the place or on the time of writing of Test.—The connecting chapter i. 16 has a mark of the influence of ApCO on Test. In the former Martha and Mary are mentioned, and some slight is thrown on the ministry of women. In Test (where the bias is all the other way) Martha and Mary and Salome are introduced, and not only is the slight removed, but the opportunity is taken to include women among 'all those who minister in' the 'Church' (i. 15).

40 8. Date of the Appendix to the Arabic Didascalia. It is clear that ArD-app is derived from Test, or else that Test is derived from it. But several minute points show the priority of Test. 41 Thus for example, (a) in the ordination prayer for a bishop ArD corrects the name καρδιογνώστης applied to the Father in Test (as in || CH, H₃, EthCO, ConstH), conscious that in Acts i. 24, which is here referred to, it is applied to our Lord, as it is applied in H₁ p. 48, AC ii 24, iii. 7. The corrected phrase in ArD 36 is 'O God who triest the hearts and reins.' (b) In ArD 37 the fiction of Test i. 22 ('I have said unto you') is removed by reading 'The Lord hath said'; here also the 'presbyters' are clearly an addition of ArD (above, p. 22). (c) ArD 38 has altered a difficult allusion in Test to the 'eighteen entrances' of our Lord to the Passion, which are connected with the bishop's three weeks' fast (six days a week), not understanding the reference, and has, not very happily, substituted three 'entrances,' death, resurrection, and ascension, as connected with the bishop's fast three days a week. Here Test apparently makes an allusion to which his readers had the key, though we have not; the ArD writer did not understand it, and made the best sense he could. His insertion in this section of the Wednesday and Friday fast is another sign of his being later than Test (above, p. 138 f.). (d) ArD 38 develops the Liturgy (which it only describes) by the addition of incense; here also the presbyter, instead of the deacon, brings in the Eucharistic elements (p. 44).

42 (e) The Mystagogia is on the whole shorter in ArD 39 than in Test. This seems to be chiefly because the former writer has removed the quasi-'Apollinarian' phrases of the latter. On the other hand the address of Death is slightly longer in ArD. The long concluding thanksgiving in Test (spoken by our Lord) could hardly have been deduced from the short phrase in ArD: 'I thank thee, O King, with the word (sermone) with which all creation has been created by thee. This is the word (uerbum) which is in us through the Spirit, who speaks with thee only.' See also p. 37. 43 These and other considerations lead us to consider these chapters of ArD as derived from Test. As Funk is of opinion that the unpublished part of ArD is derived from AC, the probability is that the ArD compiler (or a later writer) added on to his form of the Didascalia this fragmentary Church Order, adapted from Test, soon after A.D. 400. For the probability of some of these additional chapters being found in EthD, see above, p. 22. 9. Date of the Didache. The early date of this manual is seen from the undeveloped nature of the Liturgy and of the Church Organisation (above, pp. 62, 67), and also from the absence of references to Montanism (though Dr Bigg denied this last statement*). This will fix D before A.D. 157. Harnack thinks that D quotes 'Barnabas,' which he dates c. A.D. 130. But the dependence on 'Barnabas' and the date of that epistle are alike doubtful. Other writers are inclined to place D even earlier, at the beginning of the second, or at the end of the first, century. It is probable that it is the product of some remote district, perhaps of Syria, apart from the ordinary stream of Church life.—This Church Order has been so often edited and commented on that it is here dealt with very briefly. Those who desire more information about it may be referred to the editions of C. Taylor (1886) and Harnack (1884), to Harnack's

44 * Dr Bigg stands almost alone in assigning D to the fourth century. discussion in his Chronologie (1897), and to Dr Salmon's article on the subject in DCB. 10. Date of the Didascalia. It is a little difficult to determine this, because it is not certain that even the oldest form of it which we possess (the Syriac) is the original. H₁ is probably interpolated, if only in a slight degree. We must certainly dismiss Bickell's idea that Didasc-syr is an abridgment of, and therefore later than, AC i.—vi. Parallel passages constantly show AC as describing later, never as giving earlier, customs; some examples may be seen at p. 31 f. The points which we may consider in arriving at the date are these. The arrangement of a Christian assembly is described, but not elaborate church buildings; only Sunday assemblies are mentioned. Pascha is to be determined by following the Jewish computation; a Holy Week fast is described. Harnack at one time considered that the work was anti-Novatianist (see above, p. 114), and assigned it to the last half of the third century; Wordsworth (and now Harnack also agrees with him) places it in the first half of that century (MG p. 39), but Harnack thinks that the sections about clemency to sinners are interpolated. Funk (DidCA i. p. iv f.) and Achelis agree with Harnack's earlier view. Thus, while the exact date is quite uncertain, we can only be on safe ground if we date the work some time in the third century. The Holy Week fast may perhaps incline us to the later date. Its original country is doubtful; perhaps Syria.

45 11. Date of the Apostolic Church Order. This manual has some extremely ancient features, such as the position of the reader above the deacon in § 19, though not in § 1 (see p. 86). As § 22 is to some extent a repetition of the direction of § 20 (on deacons), Harnack suggests (SApC p. 7 ff., and in his edition of D, p. 212 ff.) that §§ 16—28 consist of two very early law books, each of the second century; these he calls 'Source A' and 'Source B.' He ascribed the Church Order itself to c. A.D. 300; but now with Funk, he is inclined to put it a little later (DidCA ii. p. xliv). In § 1 we find the word 'eparchy' which in the sense of an ecclesiastical province is not found before Eusebius; but there is nothing to show that the writer is referring to ecclesiastical rather than civil divisions. Subdeacons are not mentioned. This Church Order perhaps comes from Asia Minor (above, p. 26).

46 12. Date of the Edessene Canons. The reference in can. 25 to Christian Emperors shows that this work is post-Nicene. But it does not seem to have been written long after the conversion of the Empire, and may perhaps be ascribed to the second or third quarter of the fourth century. It came from Edessa. 13. Date of the Apostolic Canons. Several of these canons are taken from those of the Council of Antioch in Encaen., A.D. 341; this therefore is the earliest possible date. Whether compiled by the AC writer or not, they are closely connected with AC, and we may probably therefore assign them to the last quarter of the fourth century or to the beginning of the fifth, and believe that they were written in Syria.

Summary

47 It will be convenient to sum up the results which have been tentatively reached in this volume. Reasons have been given for thinking that a large number of the Church Orders are descended from a single Original, now lost, and that this may even have been the work of Hippolytus. The extant Church Orders seem not to be all in one line of tradition. Neither the Canons of Hippolytus nor the Apostolic Constitutions can with much probability be considered as the parent of the rest. With regard to the former, a date early in the fourth century is here suggested for the work as we now have it, excepting only one or two trifling interpolations; the theory of a largely interpolated work of the era of Hippolytus now appears to the present writer to be as little tenable as that of a late composition of the sixth century. Yet the Canons probably reproduce several primitive features of the lost Original with great fidelity. It is suggested that most of the extant Church Orders here placed in the first division, or rather their Greek originals, date from the fourth century; the Canons of Hippolytus, and the three forms of the Church Order of Egypt ('Egyptian,' Ethiopic' and 'Verona fragments') having been probably compiled early in the century, the Testament of our Lord about the middle, and the Apostolic Constitutions towards the end of it. The compilations in which most of these Church Orders are found may, at least in some cases, be a good deal later.

48 We have now come to the end of the task proposed in the first chapter of this book. The contents of the Church Orders have been considered from various points of view, and conclusions have been drawn as to their date and home. With regard to these, it will doubtless be disappointing to some that there is a certain amount of uncertainty about the results arrived at. But it is better to speak cautiously than to be over-positive, when, by the nature of the case, it can only be a question between what is more probable and what is less probable. The result of the investigation which does appear to be certain is that this whole class of literature is an important one, and cannot safely be neglected. It is hoped that the present volume may at least point out to the English student the value of the Church Orders, and the best method of studying them.