image with the sign of Myriobiblos

Main Page | Library | Homage | Seminars | Book Reviews





Internet Dept.



Previous Page
Robert Taft S. J.

The Evolution of the Byzantine “Divine Liturgy”

Orientalia Christiana Periodica XLIII, Roma 1977, p. 8-30

a) The litany

Let us look at the liturgical units of this enarxis (1). We can dispense immediately with the opening blessing; it does not appear until the nth century (2). The initial litany is also out of place. In our primitive shape such intercessions occur only after the readings, thus safeguarding the priority of the divine action in the order of service: Only after God speaks to us His Word do we respond in psalmody and prayer. As a matter of fact our litany was once found just before the transfer of gifts. Its remains are still visible there in the vulgate recension of the Slavonic books (3).

But following a tendency observable in almost all liturgical traditions, these intercessions were either suppressed or moved up to the beginning of the liturgy of the Word. Thus, in the ίο-nth century sources of our liturgy we find this litany in its original place before the transfer of gifts, and also after the Little Entrance, just before the Trisagion, i.e. at the old beginning of the liturgy, before the enarxis was added. By the end of the nth century it is found also before the antiphons, i.e. at the new beginning. In the I2th century it disappears from its original place in the prayers of the faithful; in the I3th it disappears before the Trisagion, remaining only where we still find it today (4).

So our litany is really the original litany of the faithful of the Byzantine mass. The two abbreviated litanies that accompany the prayers of antiphons II and III are probably just a development of the original oremus of the two collects they now accompany.

b) The antiphons

What about these three antiphons and their collects? Where did they come from, and when were they added to the liturgy? The when is easy: some time between 630 and 730 A.D. There is no mention of them in the Mystagogy of Maximus Confessor written about 630 (5). As he describes it, the liturgy begins with the entrance into church of the people with the bishop, followed immediately by the readings. Now until at least the nth century the bishop was not present in church for the enarxis but entered only at the Little Entrance (6). Obviously, then, there was no enarxis in the time of Maximus. But just one century later our next Byzantine liturgical commentary, the Historia ecclesiastica of Patriarch St. Germanus I (†733), does mention the antiphons (7). So they first appear at the beginning of the 8th century.

But this does not mean that they are a permanent fixture at that time. Liturgies tend to be snobbish. They take their time about accepting newcomers as permanent members. Even as late as the zoth century the three antiphons had not yet won a permanent place as a fixed part of every mass (8).

Our main source for the history of how they did win it is the 10th century Typicon of the Great Church edited by Juan Mateos, S.J. of the Pontifical Oriental Institute in Rome (9). This crucial document has provided the key to almost the whole history of the Byzantine liturgy in the post-Justinian era. A typicon is not used in the actual celebration of the liturgy, but provides the directions for the correct use of the books that are, indicating the proper of mass and office, and giving, like the old ordines romani, detailed rubrics for special celebrations that occur in the liturgical cycle.

Now in the 10th century typicon of Hagia Sophia we see that the liturgy of New Rome, like that of Old Rome, was highly stational in character. On many days in the church calendar the liturgy was celebrated not just anywhere, but in some specially designated church. This church was the "station" of the day, and on more solemn feasts the crowd would gather with the clergy at some other sanctuary and process solemnly from there to the stational church for the liturgy. During the procession an antiphonal psalm would be chanted. Upon arrival at the station, the end of the antiphon would be signalled by intoning the Gloria patri that announces the conclusion of antiphonal psalmody in almost every tradition, followed by the final repetition of the antiphon or refrain, called the περισσή or "appendix". Sometimes a variant refrain would be substituted at the perisse.

The clergy recited the introit prayer before the doors of the nave—not before the doors of the sanctuary chancel as now— and then entered the church, followed by the people. Proceeding past the great ambo in the center of the nave, they went along the solea or walled-in processional way that extended from the sanctuary to the ambo (10), and took their places at the synthronon in the apse.

All this is almost the same as the opening of the contemporary Roman stational mass described in the Ordo romanus primus about 750 A.D. even to the entrance along a walled-in processional way, the so-called schola cantorum which Mathews has shown to be an exact parallel to the old Byzantine solea (11). To the best of my knowledge this surprising similarity between two liturgies presently so different in structure and ethos has never been noticed by the students of liturgy. It is but one more indication of the communality of much in early liturgy, showing again the validity of the comparative method of liturgiology first formulated by Anton Baumstark (†1948) half a century ago (12). I have already noted that the three base-traditions—Rome, Constantinople, Jerusalem—out of which have come the only two universal rites of Christendom, the Roman and the Byzantine, were all distinguished by their stational character. It is not an exaggeration to say that practically every addition to the Byzantine eucharist from Justinian until the post-iconoclast period had its origin in the stational rites of Constantinople. The antiphons will be our first example.

The old typicon tells us that on some feasts, on the way to the stational church, the stational procession would stop either in the forum or in some other church along the processional route for a rogation. On some days this prayer-service included an office of three antiphons. After this rogation the procession would continue on to the stational church, to the accompaniment of the usual processional antiphon.
But evidently the office of three antiphons was very popular, because it soon became customary to celebrate it in church before the liturgy on days when there was no stational procession. Here we see an example of a liturgical unit gradually detaching itself from the service in which it originated and becoming an integral part of another service.

Note however that these three antiphons celebrated in church before non stational liturgies were a combination of the three rogation antiphons with a fourth antiphon, the processional antiphon to the church. For example, in the typicon for New Year's Day—September 1, at that time—there are two liturgies prescribed, one in the Church of the Theotokos in Chalkoprateia, one in Hagia Sophia (13). The one in Chalkoprateia was stational preceded by an office of three antiphons in the forum, followed by the procession to Chalkoprateia for mass to the accompaniment of another, fourth antiphon. But the liturgy in Hagia Sophia in memory of St. Stephen the Stylite begins right there with an office of three antiphons. And at the third antiphon two refrains, that of the saint, and that of the fourth processional antiphon from the stational service of Chalkoprateia, are both sung. What they have done is simply fuse together the third antiphon of the devotional service with the introit antiphon, probably because three antiphons, not four, was the customary liturgical unit in the offices of the Great Church. So the rogational office of three antiphons and the introit antiphon are two different things, which explains why toda3r we have four orations—three antiphon prayers plus an introit prayer—with only three antiphons.

Up until the 10th century the three antiphons were not a necessary part of every liturgy. Even after this date the patriarch still does not enter the church until the third antiphon, because this is the old introit of the liturgy (14). And even today vigil masses in the Byzantine rite, in which mass is preceded by vespers, have no antiphons at all but begin with the Trisagion. It is said that in such masses, vespers replace the Liturgy of the Word. They replace nothing, but are joined to the mass at its old beginning, the Trisagion, thus illustrating Baumstark's law that older usages are preserved in more solemn seasons and rites (15).

Today the three antiphons have been reduced to a few scraps of their original form, and the troparia after the third antiphon have been so multiplied as to take on an independent existence detached from the psalmody which they were originally destined to serve as refrains. This exemplifies another common development in liturgical history: the process whereby ecclesiastical compositions multiply and eventually suffocate the scriptural element of a liturgical chant, forcing, in turn, the decomposition of the original liturgical unit, so that what we are left with is simply débris, bits and scraps of this and that, a verse here, a refrain there, that evince no recognizable form or unity until they are painstakingly reconstituted into their original structures by piecing together the remaining scraps, then filling in the blanks, sort of like doing a jig-saw puzzle with only a tenth of the pieces (16). This is why the study of liturgical units and their mutual articulation within larger ritual structures is so crucial in the reconstruction of pristine liturgical forms,

c) The Trisagion

We see another example of this in our next piece, the Trisagion. Today it is chanted as follows:
Holy God, holy, mighty, holy, immortal, have mercy on us
(three times). Glory be to the Father. . . now and always and unto ages of
ages, amen.
Holy, immortal, have mercy on us. Holy God, holy, mighty, holy, immortal, have mercy on us.

From what we know of the structure of Byzantine antiphonal psalmody, it seems that we have here the incipit and finale of an antiphonal psalm, i.e. the opening triple repetition of the ' complete refrain, then the concluding doxology, the άκροτελεύτιον, and final repetition of the refrain (perisse), with the intervening
psalm verses suppressed (17). Now we first hear of the Trisagion in the 5th century, when it was apparently used as a processional antiphon during stational services in Constantinople (18). Early in the 6th century we see it at the beginning of mass. This chant, then, is the remains of the original, invariable introït antiphon of our mass, to which at a later date first one, then three variable antiphons were appended.

So at the beginning of the 5th century our liturgy opened with the entrance of the clergy and people into church without ceremony or, apparently, accompanying chant. By the 6th century this introit had been ritualized by the addition of an element from the stational processions, an antiphonal psalm with the Trisagion as its fixed refrain. About a century later, undoubtedly as a result of further developments in the stational rites, all but the refrain of this antiphon was suppressed in favor of a more recent stational antiphon that provided more variety for this rapidly expanding rite.

Why wasn't the original fixed refrain just suppressed, or retained as an occasional variant? Probably because of its immense popularity as testified to by the legends of its origins in divine revelation, because it had become a liturgical element common throughout the whole East, and because of the role it played in the Monophysite controversy.

d) The ektene:

One further element that entered the liturgy from the stational services is the ektene or litany that immediately follows the gospel. It is sometimes referred to in modern versions as the "ecumenic" or "universal" prayer for all needs — i.e. the oratio fidelium of the Byzantine mass. It is no such thing, as should be obvious from its position before the dismissal of the catechumens. Common prayer with their participation was excluded, which is why they were first dismissed, and not because they mustn't receive communion, as is often thought. They were also dismissed at non-eucharistic services, where there was no risk of them going to communion. In our l0th century typicon this ektene or penitential litany was chanted after the gospel in stational rogation services, and rubrics prescribe the same practice after the gospel of the Liturgy of the Word on certain days of the year (19). This can be taken perhaps as a remnant of a previous stage of evolution when this litany was gradually gaining a foothold in the mass, where it appears for the first time in the 8th century codex Barberini 336.


1. For a complete history of the Liturgy of the Word, see J. MATEOS, La célébration de la parole dans la liturgie byzantine (Orientalia Christiana Analecta 191) Rome 1971. Mateos' study needs to be completed in certain details by the later work of Jacob (cf. previous note) in the MS tradition. Two recent studies in English, C. KUCHAREK, The Byzantine-Slav Liturgy of St. John Chrysostom, Allendale, N.J. 1974; and M. M. SOLOVEV, The Byzantine Divine Liturgy. History and Commentary, Washington, D.C. 1970; are not reliable.

2. It first appears, I believe, in the Codex S. Simeonis Siracusani (c. 1030) preserved for us in the Latin version of Ambrose Pelargus (Divina ac sacra Liturgia sancti loannis Chrysostomi. Interprète Ambrosia Pclargo Niddano, O.P. ... Worms 1541, f. Biv°). On this source see TAFT, Great Entrance, op. cit., pp. xxvii viii.

3. See A. STRITTMATTER, Notes on the Byzantine Synapte, in: Traditio X (1954) 51-108; A peculiarity of the Slavic Liturgy found in Greek Euchologies, in: Late Classical and Medieval Studies in Honor of Albert Mathias Friend, Jr., edited by K. WEITZMANN, Princeton, N.J. 1955, 197-203.

4. Loc. cit. and MATEOS, Célébration, op. cit., pp. 29-31.

5. Ch. 9, PG 91, 688 89. On this and other commentaries on the Byzantine liturgy see R. BORNERT, Les commentaires byzantins de la divine liturgie du VIIe au XVe siècle (Archives de l'Orient chrétien 9) Paris 1966.

6. Mateos' treatment of the pontifical liturgy (Célébration, op. cit., pp. 40-41) needs to be corrected somewhat on the basis of later research. See TAFT, Great Entrance, op. cit., pp. 267-9, and my review of Chr. STRUBE, Die westliche Eingangsseite der Kirchen von Konstantinopel..., in: OCP XLH (19/6) p. 300.

7. N. BORGIA (ed.) II commentario liturgico di S. Germane Patriarca Costantinopolitano e la versione latina di Anastasio Bibliotecario (Studi Liturgici I) Grottaferrata 1912, p. 21. Cf. BORNERT, Les commentaires, op. cit., p. 162.

8. The history of the antiphons is given in detail by MATEOS, Célébration, op. cit., pp. 34-71.

9. J. MATEOS (éd.), Le typicon de la Grande Eglise. Ms. Sainte-Croix No. 40. Introduction, text critique et notes, 2 vol. (Orientalia Christiana Analecta 165-6) Rome 1962-3.

10. On the solea and other aspects of the liturgical disposition of the early Byzantine church, see T. MATHEWS, The Early Churches of Constantinople: Architecture and Liturgy, University Park and London 1071 (cf. "solea" in Index); S. G. XYDIA, The Chancel Barrier, Solea and Ambo of Hagia Sophia, in: The Art Bulletin XXIX (1947) 1-24. MATEOS (Typicon II, op. cit., p. 321) wrongly identifies the early solea with the sanctuary platform, which is the solea in current Greek liturgical nomenclature. Cf. TAFT, Great Entrance, op. cit., p. 79.

11. M. ANDRIEU, Les ordines romani du haut moyen âge, II: Les textes (Ordines I-XIII) (Spicilegium Sacrum Lovaniense, études et documents, fasc. 23) Louvain 1960, pp. 74ff; T. MATHEWS, An early Roman Chancel Arrangement and its Liturgical Functions, in: Rivista di archeologia cristiana XXXVIII (1962) 73-95; G. G. WILLIS, Roman Stational Liturgy, in: Further Essays in Early Roman Liturgy (Alcuin Club Collections no. 50) London 1968, 3-87; J. A. JUNGMANN, Missarum sollemnia vol. I, part i, no. 8 (English edition, vol. I, pp. 67ff).

12. See his Vom geschichtlichen Werden der Liturgie (Ecclesia orans 10) Freiburg B. 1923; and Comparative Liturgy, op. cit.

13. MATEOS, Typicon I, op. cit., pp. 2-11; Célébration, op. cit., pp. 37ff.

14. See note 12 above.

15. Das Gesetz der Erhaltung des Alien in liturgisch hochwertiger Zeit, in: Archiv für Liturgiewissenschaft VII (1927) 1-23, Comparative Liturgy, op. cit., pp. 26 ff.

16. See TAFT, Great Entrance, op. cit., pp. 112 ff.

17. The akroteleution is the final clausula of the refrain, and was repeated after the verses of the psalm; the perisse ("appendix") is the final repetition(s) of the refrain after the doxology that signals the end of the psalm. See MATEOS, Célébration, op. cit., pp. i6ff.

18. Ibid. pp. 99-100, 112 ff.

19. MATEOS, Typicon II, op. cit., index: «ἐκτενή μεγάλη» p.. 293.

Previous Page