Though it may be appealing to romantic natures, it is not exactly legitimate to suggest that everything contained in the liturgical books of the 8th—9th centuries is a mirror image of the usages in the “early Church.” Much earlier sources testify the presence of various important elements of the liturgy: the “Sursum corda” dialogue or the Sanctus (in the third-century layer of the ‘Apostolic Constitutions’); observance of the daily Hours of prayer (Terce, Sext, None in the writings of Tertullian); the existence of responsorial psalmody (e.g. in St. Augustine’s Psalm commentaries). These bits of important information do not mean that the liturgy in the third century was little more than the Sursum corda and Sanctus (for instance). They mean simply that the sources at our disposal are very sparse and fragmentary, hence frustratingly silent about what was actually done in worship services. Thus anyone who wishes to make generalizations about the Roman liturgy as a whole, must perforce take “Roman liturgy” to mean that which has survived in the full liturgical books of the 8th—9th centuries and on into our own times. All else is speculation and hypothesis — not ‘facts’ — when it comes to “early Christian liturgy.”
The Roman Liturgy emerges in the sources not in its entirety, as a completed whole, an opus perfectum, but only as the succession of its elements. It is not so much the development of liturgy that is reflected in these sources; what we have to deal with is the insufficiency of the source material. It is a rather audacious or romantic attitude to refer to “the liturgy of the time of the Apostles or Martyrs,” since the very small number of sources from these periods imposes a serious limit on the level of knowledge we can possibly attain about this early stage of the liturgy. Some elements already appear in the third century sources; some chance references can be found in the sermons of the Church Fathers; a richer source of information is the Rule of Saint Benedict. However, liturgical books including actual texts and ceremonies are not available from this early age. The little we know must be gathered through the analysis of sources that date from subsequent centuries, and by their careful comparison with the early quotations.
In contrast to this disappointing picture, when we take up the earliest surviving books used in actual liturgical practice, we find that all the essential elements and structures known from the Middle Ages and valid up to 1970, are present in them. (Of course, I speak now only of the Roman rite and not of other branches of the Western liturgy which have almost completely disappeared with the passage of time.) This statement should be understood differently as regards the different elements of the liturgy, the different seasons and days of the liturgical year, and the composition, material and arrangement of the celebration itself. While the priest’s prayers and the readings, for instance, are known in different arrangements from the (partly overlapping) collections of the 7th and 8th centuries, the repertory of Mass chants (edited by Hesbert) obligatory until very recent times is about 90 percent the same as in the earliest sources.
The liturgy reflected in the “essentially identical” source material became still more homogenous by the regulation of the Roman rite and its diffusion through all of Europe. The distribution of the pericopes, the sacramentary, the chant books and even more the structure of the main components of the liturgy all exhibit great similarity when charted in hundreds of mediaeval ritual books. In the process of regulation, new contextual values and potentialities evolved: cross-references, associations, confluence of elements all enriched and stabilized the celebration of the sacred rites.
This does not mean, however, a literal identity. The pre-conciliar rites of some religious orders allow us to perceive to a greater or lesser degree the inner variety of the Roman liturgy. The Benedictines, Cistercians, Dominicans and Norbertines preserved the liturgy of their own orders even until recent times, up to the Council and some even beyond. These liturgical families preserve great values for the whole Church, and a special personal attractiveness for those living in the given communities. But in the Middle Ages, variety within unity was not limited to the religious orders. The guardians of continuity were first of all the cathedrals. In a paradoxical way, they guaranteed both conformity to the unity of the Roman liturgy and the variety of the local rites. The liturgy of the cathedral was the norm for the parish churches of the diocese.
The geographical (or rather, institutional) differences caused no confusion for two reasons. On the one hand, both the categories of ubi and quomodo were adequately regulated. By that we mean that the liturgical areas where or in which unity was to be maintained (with the local traditions duly observed) were governed by a ‘hierarchy’ of elements supported by dogmatic and liturgical considerations. On the other hand, it was the Chapter (or the convent or the superior of the religious order) that safeguarded both the continuity and the legal changes or development, and this guarantee against any kind of arbitrariness was not at all less efficient than the activity of a far distant Curial congregation would have been.
If, knowing all this, we again ask the question what the “traditional Roman liturgy” is, the answer will be as follows: it is the liturgical practice of Rome continuously living and organically developing from the 4th century at the latest (if its basic features are meant) and fixed in the 8th—9th centuries; which preserved its identity during diffusion both geographical (in cathedrals) and institutional (in orders), as also amid the local and temporal variations regulated by the liturgical hierarchy. Or more briefly put: the Roman rite is that which emerges in the uniformity of the organic temporal and coherent spatial variety of its daughter-liturgies. The description of its content, the separation of common, general, and differing specific elements can only be achieved by analyzing the rite in its entirety, a task which exceeds the limits of this chapter. The definition given above could not avoid, of course, some superficiality, but for our purposes it is valid and sufficient.