Two Thousand Years
of Doing It Wrong?

What this page is asking

The claim under examination is not a caricature and it is not rare. It is said warmly, from platforms, by people who are trying to help, and it usually sounds something like this. No church, ministry or person is named anywhere on this page, living or dead — and the passage below is composed by this page, in the register in which the claim is actually made. It is not a quotation from anybody.

Concede at once what is true in it, because a great deal is. Worship can be performed rather than offered. Words can be said by a whole congregation without a single mind behind them. A form can become a shell, and when it does, Scripture is savage about it — this people draw near me with their mouth, and with their lips do honour me, but have removed their heart far from me (Isaiah 29:13), a sentence Christ quotes against the most liturgically careful people of his day (Matthew 15:8). The historic churches did not need to be told this by anybody; it is in the lessons they read out loud every year. This page is not a defence of ceremony for its own sake, it does not argue that older is holier, and it does not hold that a form works by being performed. Any reading of it as a brief for empty ritual is a misreading, and the tradition it describes supplies the refutation.

The question is narrower, and it has two halves that are usually run together. The first is historical: is it true that the earliest Christians worshipped without an inherited order, and that the ordered liturgies came later, as a decline? The second is theological and practical: if a congregation discards the inherited order, what does it lose — not in atmosphere, but in what the service teaches, protects and hands on? The first question has an answer that the evidence settles fairly quickly. The second is the one worth arguing about.

Two positions, labelled, and used consistently throughout

Most of the heat in this argument comes from each side attacking a position the other does not hold. So the page labels two and keeps them apart from here to the end.

The page holds L and tests F. Two things follow from having only two labels, and both need saying on the first screen. The first is that holding L is not holding that a rite works of itself, that antiquity settles anything by itself, or that a congregation reciting words it does not mean has worshipped. Nobody in the tradition described here has ever held that, and the page argues the point rather than assuming it — see Part XIII, which turns the page’s own best argument round and tests it to destruction. The second is that testing F is not attacking the people who hold it. F is, among other things, a reaction to something real, and a good deal of what it reacted against deserved the reaction.

What the page concludes, stated up front

The prompt behind this page

Everything below was produced by putting the following brief to a frontier language model with web search, which assembled and verified the sources. It is published so you can run it yourself, change it, sharpen it, or reverse it and see what comes back. Reversing it is the interesting experiment: ask for the strongest possible case for Position F, from the same primary sources, and see which parts of this page survive. If you get a materially different answer, that is information, and worth publishing too.

Write a detailed, source-first investigation of the following question, in
British English, for a general reader who is willing to work. Quote
scripture from the Authorised Version throughout, so that every citation
can be checked free of charge. Date and attribute every other quotation:
give the work, the edition or translation, the book, chapter, section,
article or paragraph number, and where possible the page. Give the
Greek, Hebrew, Latin or Aramaic wherever an English translation hides
what a word is doing. Where the date, authorship or text of a source you
rely on is disputed by scholars, say so at the point of use rather than
burying it in a note.

  THE QUESTION. What are the downsides of informal worship that discards
  liturgy? The Church has worshipped with an inherited order for two
  thousand years. Why?

  THE THESIS UNDER TEST. That the earliest Christian worship for which
  we have direct evidence is already ordered, textual and eucharistic;
  that the liturgical shape grew continuously out of synagogue and
  Temple worship; that the New Testament itself quotes liturgical
  material fixed before the documents quoting it were written; that the
  Reformation reformed inherited liturgies rather than replacing them
  with an informal service; and that a congregation which discards the
  inherited order does not thereby have no order, but an unwritten one,
  with specific and predictable consequences.

Distinguish two positions throughout and use the labels consistently:

  L. THE LITURGICAL. Christian worship has a given shape - Scripture
     read at length and in course, prayer in common in words the
     congregation can join, confession and absolution, the creed, the
     Lord's Prayer, and the Supper as the ordinary Sunday climax. The
     forms are inherited rather than composed weekly; they are
     reformable and have been reformed; the shape is apostolic rather
     than the minister's to invent.
  F. THE FREE. Fixed forms are dispensable at best and an impediment to
     sincerity at worst. Each congregation should order its worship as
     seems best now. Spontaneity is nearer the Spirit than recitation.
     The sermon is the centre; the Supper is occasional. What the first
     Christians had was freedom, and the liturgies are a later
     accretion.

Hold L and test F. Do not let holding L slide into defending ceremony
for its own sake, antiquity as self-justifying, or the idea that a rite
works by being performed: state on the first screen that the tradition
itself supplies the refutation of all three, and cite Isaiah 29:13 and
Matthew 15:8 doing it.

SCOPE LIMITS, BINDING. Name no living person, church, ministry or
organisation anywhere. Every named figure must be deceased and quoted
from a dated published source. Where an example of the worship under
examination is needed, compose it and label it as composed at the point
of use. Repeat no third-party allegation of misconduct against anybody.
Answer objections inside the argument, at the point where each arises,
in the words their proponents would use - no separate "case against this
page" section. Exclude music entirely: instruments, hymnody, choruses,
style and volume are a different enquiry. Exclude the Puritan regulative
principle as a topic, stating that it is a serious and distinct case
deserving its own treatment. Do not adjudicate between the liturgical
traditions - Eastern, Roman, Lutheran, Anglican, Reformed - on the
points where they disagree with each other, and say so. Give no advice
about staying in or leaving a church, and make no clinical claims.

Work through, at minimum:

 1. IS THERE SUCH A THING AS WORSHIP WITHOUT A LITURGY? Take the
    typical informal order - welcome, songs arranged to build and
    soften, announcements, offering, long sermon, response, dismissal -
    and build an interactive that makes the reader say what each
    element teaches by its position before revealing it. Argue
    structurally, never by attributing motives. Establish that the real
    question is not whether a church has a liturgy but who wrote it,
    whether it is written down, and whether anyone may object to it.

 2. WAS EARLY CHRISTIAN WORSHIP LITURGICAL? The Didache (7-10, 14, 8),
    Justin Martyr First Apology 65-67, Pliny to Trajan 10.96, Cyprian
    On the Lord's Prayer 31, Egeria, the Apostolic Tradition. On the
    Apostolic Tradition give the dating dispute honestly: the
    attribution to Hippolytus c.215 is now widely contested, with
    Bradshaw, Johnson and Phillips arguing for composite material from
    the mid-second to the fourth century compiled c.375-400, and the
    page must not lean on it where an undisputed witness will do.
    Build an interactive that gives the reader liturgical texts,
    some ancient and some from modern service books, and asks him to
    date them.

 3. DID CHRISTIAN LITURGY GROW OUT OF BIBLICAL WORSHIP? Synagogue and
    Temple: Luke 4:16-20, Acts 13:15, Acts 15:21, Acts 2:46, Acts 3:1,
    Matthew 26:30, the Shema, the Amidah, the berakah form behind
    Didache 9-10, the Psalter as the church's first hymnbook, the
    Sanctus from Isaiah 6:3 by way of Revelation 4:8. Make the reader
    match each Christian element to its predecessor.

 4. DOES THE NEW TESTAMENT CONTAIN LITURGICAL MATERIAL? Philippians
    2:6-11, Colossians 1:15-20, 1 Timothy 3:16, Ephesians 5:14,
    2 Timothy 2:11-13, 1 Corinthians 15:3-5 and 11:23-26 with the
    paredoka/parelabon vocabulary of received tradition, Maranatha at
    1 Corinthians 16:22 and its Greek equivalent at Revelation 22:20,
    the congregational Amen at 1 Corinthians 14:16, the triune
    baptismal formula at Matthew 28:19, the benediction at
    2 Corinthians 13:14, the Lucan canticles, the doxologies. Build an
    interactive that asks the reader to distinguish an author's own
    composition from quoted material, and give the internal evidence
    that scholars use - rhythm, rare vocabulary, relative-clause
    openings, retained Aramaic - rather than asserting the conclusion.

 5. WHY DID THE SHAPE BECOME NEARLY UNIVERSAL? Set out the common
    structure of the historic rites East and West and treat the
    schisms of 431 and 451 as a natural experiment: churches out of
    communion cannot have borrowed from each other afterwards. Use the
    Anaphora of Addai and Mari, which has the opening dialogue and the
    Sanctus and, in its oldest manuscripts, no institution narrative.
    Be careful about what convergence proves and what it does not.

 6. WAS THE EUCHARIST THE CENTRE OF SUNDAY WORSHIP? Acts 2:42 and
    2:46, Acts 20:7 with attention to the purpose clause,
    1 Corinthians 10:16-17 and 11:17-34, Didache 14, Justin 67, Pliny
    10.96. Then trace when and why the sermon displaced the Table in
    parts of Protestantism.

 7. DID THE REFORMATION REJECT LITURGY? Luther's Formula Missae (1523)
    and Deutsche Messe (1526); Augsburg Confession XXIV and VIII and
    the Apology; Cranmer 1549, 1552, 1662; Calvin's Form of Prayers
    (1542) and Institutes IV.17.43-46 on weekly communion and the
    Geneva council's refusal; and the Westminster Directory of 1645 -
    which is itself a prescribed order with a lectio continua rule.
    Let the Directory's preface state the case against set forms in
    its own words.

 8-12. THE FIVE COSTS. Take each in turn, argue it structurally, and
    give the historic practice that was shaped to prevent it:
    personality displacing the office (the Donatist settlement,
    Augsburg VIII); doctrinal drift (lex orandi lex credendi from
    Prosper, Basil's On the Holy Spirit and the doxology dispute of
    375, the creed's entry into the Sunday service); emotion as the
    measure of successful worship (Finney's Lectures on Revivals of
    Religion, 1835, quoted exactly, as the documented origin of the
    service-as-means); Scripture becoming selective (the lectionary
    against the preacher's private canon - build an interactive that
    makes the reader mark which books of the Bible he has heard
    preached in the last year); and the congregation becoming an
    audience (count what the people actually say aloud in each kind of
    service).

13. WHAT DOES TWO THOUSAND YEARS ACTUALLY PROVE? Turn the page's own
    strongest argument round and test it. Longevity is not truth.
    State plainly what continuity does and does not establish, concede
    what the tradition has transmitted badly, and show the tradition
    reforming itself. The page must be willing to lose this section.

14. A DIAGNOSTIC. Close with an instrument the reader can run on a
    service, scoring it on several axes. It must assess a service and
    never a person, must be capable of returning a clean result for a
    simple service in a house or a field, and must recommend no course
    of action.

Write it as a single self-contained interactive HTML page with at least
ten interactive sections that make the reader commit to an answer
before the sources are revealed. At least two must be capable of
showing the reader that his instinct is wrong, and at least one must
use the reader's own answers against him. Publish this prompt on the
page.

Four instructions did most of the shaping. Excluding music kept the page on structure and theology instead of on taste, which is where this argument normally ends up and stays. Requiring the dating dispute over the Apostolic Tradition to be stated at the point of use cost the page its most convenient early witness and made the remaining case stronger, because Cyprian at Carthage in 252 is not disputed by anybody. Requiring Part XIII to be willing to lose is the instruction that stops a page like this becoming an advertisement. And forbidding a separate counter-case section forced every objection to be met where it actually arises — which is harder to write and much harder to dodge.

What do you currently think?

Ten questions, four minutes. Answer from what you believe now, not from what you suspect this page wants. Your answers come back at the end alongside what the sources say. Most readers miss the same three, and in all three the surprise runs the same way: the thing assumed to be a late medieval accretion turns out to be attested earlier than the New Testament canon was settled.

What follows

The liturgy you already have

Begin with the claim at its weakest point, which is also the point nobody argues about once it is put plainly. A congregation that says it has no liturgy has an order of service. It is not written in a book, it was not adopted by any synod, and most of the people following it could not say where it came from — but it exists, it is remarkably stable from week to week, and a visitor could predict it after two Sundays. Here it is. This order is composed by this page, as a representative example; it is not taken from any actual church.

The interesting question is not which of these is a liturgy. Both are. The interesting question is what each element teaches, because a service teaches by its structure whether or not anybody intends it to. Position of an element in a service says what that element is for. Frequency says what matters. What the congregation is given to do says who the congregation is.

So work through the informal order one element at a time. For each, say what you think its position in the sequence communicates — not what the church believes, which may be entirely orthodox, but what the shape itself says week after week to someone who has no theology yet. Commit to an answer before you read the note.

0 of 8 answered.

Two things fall out of that exercise, and neither of them is an accusation. The first is that the informal order is a coherent theology — you can state it in a sentence. God is encountered chiefly through feeling and through instruction; the assembly gathers to be moved and then taught; the decisive moment is inside the individual and happens near the end. That is a real position, held by real Christians, and it can be defended. The point is that it is being taught structurally, every week, to people who have never heard it stated and cannot therefore agree or disagree with it.

The second is the one that matters for the rest of this page. Nobody wrote it down, so nobody can object to it. A congregation which dislikes the 1662 communion service can open the book, find the sentence it objects to, take it to a synod and get it changed; this has happened repeatedly and is why there are so many revisions. A congregation which dislikes the shape of an unwritten order has nothing to point at. There is no text, no author, no minutes, no procedure. The order is simply how things are done, which is the most secure form a rule can take.

That is a fair distinction and the page accepts it: sequence and prescribed text are different things, and the objection is right that only the second is what “liturgy” usually means in the sentence being examined. But hold the two apart and the argument does not get easier for Position F; it gets harder, for a reason that is worth stating carefully. Prescribed text is the part that can be checked. A written prayer is available for inspection before it is prayed; a congregation can read it at home, disagree with it, and say why. An unwritten order is not text, so it cannot be inspected — and yet it is doing all the same work, shaping expectation, allocating time, deciding what the room is for. The objection therefore concedes the sequence and defends the disappearance of the only part that could be audited. Whether that trade is worth making is the subject of Parts VIII to XII; that it is a trade, and not an escape from liturgy into freedom, is the finding of this one.

What they actually did on a Sunday

Now the historical half, which is the half that can be settled. The claim is that the first Christians worshipped freely and that ordered liturgy came later. There are only so many documents; they are short; they are free; and most of them can be read in an evening. Take them in order of date and let them speak.

The Didache

The Didache, or Teaching of the Twelve Apostles, is a church order from Syria or Palestine. Dates proposed for it run from around the year 50 to around 150, with many scholars placing at least its core in the first century — which would make parts of it contemporary with, or earlier than, some of the New Testament. Nothing in the argument below depends on the earliest dating; even on the latest, it is a document of the same generation as the last apostolic writings. Here is what it prescribes.

Baptize into the name of the Father, and of the Son, and of the Holy Spirit, in living water.

Didache 7, Roberts–Donaldson translation. A formula, not a description — and the same triune words as Matthew 28:19. The chapter goes on to legislate for cold water, warm water, and pouring where immersion is impossible, which is the prose of a church that already has settled practice and edge cases.

Pray this three times each day.

Didache 8, of the Lord’s Prayer, which the chapter quotes in full immediately before. A fixed text, at fixed hours, prescribed for every Christian. If reciting given words at appointed times is a later corruption of free worship, this is the corruption arriving before the ink is dry.

We thank thee, our Father, for the holy vine of David thy servant, which thou madest known to us through Jesus thy servant…

We thank thee, our Father, for the life and knowledge which thou madest known to us through Jesus thy servant… As this broken bread was scattered upon the mountains, and was gathered together and became one, so let thy Church be gathered together from the ends of the earth into thy kingdom.

Didache 9, the thanksgivings over the cup and over the broken bread, with a further thanksgiving after the meal in chapter 10 which ends Maranatha. Amen. These are texts, given in words, to be said. Chapter 10 then adds a line that has caused two centuries of argument: but permit the prophets to make thanksgiving as much as they desire — a licensed exception, which is worth noticing precisely because it is framed as an exception to a rule.

Every Lord’s day gather yourselves together, and break bread, and give thanksgiving after having confessed your transgressions, that your sacrifice may be pure.

Didache 14. Weekly. Eucharistic. Preceded by confession. Three of the elements Position F treats as medieval, in the earliest church order we possess.

Pliny to Trajan, 112

The next witness is a hostile one, which makes him valuable. Pliny the Younger, governor of Bithynia, writes to the emperor Trajan for instructions on how to handle Christians, and reports what he has extracted from those he has interrogated and tortured. He describes their worship as evidence of how harmless or otherwise they are; he has no interest in flattering them and no theology to protect.

They were in the habit of meeting on a certain fixed day before it was light, when they sang in alternate verses a hymn to Christ, as to a god, and bound themselves by a solemn oath, not to any wicked deeds… after which it was their custom to separate, and then reassemble to partake of food — but food of an ordinary and innocent kind.

Pliny the Younger, Letters X.96, to the Emperor Trajan, c. 112, Melmoth–Bosanquet translation. A fixed day, a fixed hour, antiphonal singing (carmen…secum invicem — verse answering verse), a binding formula, and a common meal. This is a Roman official describing a settled liturgical pattern eighty years after the crucifixion, to a man he needs to persuade of his own competence.

Justin Martyr, c. 155

And then the document that decides the question. Justin, a philosopher writing in Rome, addresses a defence of Christianity to the Emperor Antoninus Pius. Near the end he describes, for a pagan reader who has never seen one, exactly what happens at a Christian assembly. It is the earliest full description of a Sunday service outside the New Testament, and it is not a theory or an ideal; it is a public statement of fact made to a hostile authority who could check it.

And on the day called Sunday, all who live in cities or in the country gather together to one place, and the memoirs of the apostles or the writings of the prophets are read, as long as time permits; then, when the reader has ceased, the president verbally instructs, and exhorts to the imitation of these good things. Then we all rise together and pray, and, as we before said, when our prayer is ended, bread and wine and water are brought, and the president in like manner offers prayers and thanksgivings, according to his ability, and the people assent, saying Amen; and there is a distribution to each, and a participation of that over which thanks have been given, and to those who are absent a portion is sent by the deacons. And they who are well to do, and willing, give what each thinks fit; and what is collected is deposited with the president, who succours the orphans and widows and those who, through sickness or any other cause, are in want, and those who are in bonds and the strangers sojourning among us, and in a word takes care of all who are in need.

Justin Martyr, First Apology 67, c. 155, Roberts–Donaldson translation (Ante-Nicene Fathers I). Chapter 65 adds, for the service at which a newly baptised person is received: common prayers, a kiss of peace, bread and a cup of wine mixed with water brought to the president, a prayer of thanksgiving “at considerable length”, the people’s Amen, and distribution by the deacons.

That is the shape of a Lutheran, Anglican, Catholic or Orthodox service this Sunday. Not approximately: element for element, in order. Whatever else is true, the proposition that ordered, eucharistic, lectionary-based worship is a late accretion on an earlier freedom is not true. It is the earliest thing we can see.

“Lift up your hearts”, Carthage, c. 252

One more, and it is the single most useful sentence in this Part, because of how casually it is dropped. Cyprian, bishop of Carthage, is writing a devotional treatise on the Lord’s Prayer. Explaining why Christians should pray attentively, he reaches for an illustration his readers will all recognise, the way one might today mention standing for a national anthem:

For this reason also the priest, by way of preface before his prayer, prepares the minds of the brethren by saying, “Lift up your hearts.”

Cyprian, On the Lord’s Prayer 31, c. 252, Ante-Nicene Fathers V. He is not introducing the practice, defending it, or arguing for it. He is using it as a familiar fact to illustrate something else — which is how we know it was old already in North Africa in the middle of the third century. That single dialogue is still said, in the same place in the service, in every historic rite on earth. Part V is about what that means.

Before going on, test the instinct this Part is meant to disturb. Below are eight liturgical texts. Some are ancient, some medieval, some from the Reformation service books, and at least one is from living memory. Date each one.

It began in a synagogue

The reason the earliest Christian worship is already ordered is that it did not start from nothing. The first Christians were Jews who had worshipped in an ordered way all their lives, in two settings at once — the Temple, with its fixed hours, sacrifices and psalmody, and the synagogue, with its readings in course, its prayers, its blessings and its seated exposition. They did not abandon either. They kept going to both, and then did something additional in houses.

And they, continuing daily with one accord in the temple, and breaking bread from house to house, did eat their meat with gladness and singleness of heart.

Acts 2:46, Authorised Version. Both, on the same day: the Temple hours and the household breaking of bread. The Christian Sunday service is the second of these growing into the shape of the first.

The synagogue pattern is described in the New Testament so often that it can be reconstructed from it alone. Christ himself goes to Nazareth as his custom was, stands up to read, is handed the scroll, finds the place, reads, closes the book, gives it back to the attendant, sits down, and expounds it (Luke 4:16–21). At Antioch in Pisidia, Paul speaks after the reading of the law and the prophets, when the rulers of the synagogue send to ask whether he has a word of exhortation (Acts 13:15). James, at the council in Jerusalem, takes it as common ground that Moses of old time hath in every city them that preach him, being read in the synagogues every sabbath day (Acts 15:21). Peter and John go up to the Temple at the hour of prayer, being the ninth hour (Acts 3:1) — a fixed hour, kept after Pentecost.

Three structural features carry straight across. Scripture read in course rather than chosen: the synagogue worked through the law on a fixed cycle, which is why James can assume that Moses is heard in every city every week. Prayer in given words: the Shema twice daily, and a fixed sequence of benedictions. The blessing form — Blessed art thou, O Lord our God, King of the universe, who… — which is the exact literary shape of the eucharistic thanksgivings in Didache 9 and 10, and behind the Greek word εὐχαριστία, eucharistia, thanksgiving, which is what the Christians called their meal.

The Psalter came across whole. It was Israel’s hymnbook and it became the Church’s; Paul tells two congregations to sing psalms and hymns and spiritual songs (Ephesians 5:19; Colossians 3:16). And the Last Supper itself ends inside a liturgy: and when they had sung an hymn, they went out into the mount of Olives (Matthew 26:30). The hymn is the Hallel, the fixed group of psalms sung at the end of the Passover meal. The institution of the Lord’s Supper takes place within a prescribed annual liturgy, and closes with prescribed sung texts.

Work the correspondences yourself before reading on. For each element of later Christian worship, say where it came from.

0 of 8 answered.

The objection is right that the sacrificial system is finished and that Hebrews says so at length. But notice what it is the Epistle to the Hebrews says has ended: the sacrifices, the Levitical priesthood, the annual entry into the holy place. It does not say that the reading of Scripture has ended, or corporate prayer, or psalmody, or the blessing of God over bread and wine, or the assembly itself — and in the same letter it tells its readers not to forsake the assembling of ourselves together (Hebrews 10:25). The apostles evidently did not read Hebrews as abolishing ordered worship, because they went on keeping the hours of prayer in the Temple courts after Pentecost. What was fulfilled was the sacrifice. What continued was the shape of the assembly — word, prayer, praise and a common table — now with a different centre. That is exactly what the second-century evidence in Part II shows us finding.

The liturgy inside the New Testament

So far the argument has been about what Christians did. This Part is about something more direct: the New Testament documents themselves quote liturgical material that was already fixed before those documents were written. If that is right — and the evidence is internal, which means you can check it — then the liturgy is not merely early. Parts of it are earlier than the books of the New Testament that contain them.

Start with the one that is explicit, because Paul says what he is doing in so many words.

For I have received of the Lord that which also I delivered unto you, That the Lord Jesus the same night in which he was betrayed took bread: And when he had given thanks, he brake it, and said, Take, eat: this is my body, which is broken for you: this do in remembrance of me.

1 Corinthians 11:23–24, Authorised Version. “Received” is παρέλαβον and “delivered” is παρέδωκα — the technical pair for handing on a fixed tradition, the Greek equivalents of the rabbinic qibbel and masar. Paul uses the identical pair four chapters later for the resurrection formula: I delivered unto you first of all that which I also received (1 Corinthians 15:3). He is not composing. He is reciting, and saying that he is reciting.

Written in the mid-fifties, about twenty years after the events, quoting words Paul says he was given when he was converted, which is earlier still. The words of institution are therefore among the oldest fixed Christian texts we possess — older than any gospel — and Paul cites them to settle an argument about how the Corinthians conduct their assembly. That is what a liturgical text is for.

Then there is the untranslated word.

If any man love not the Lord Jesus Christ, let him be Anathema Maranatha.

1 Corinthians 16:22, Authorised Version. Maranatha is Aramaic — marana tha, “our Lord, come”. Paul is writing in Greek, to Greeks, in a Greek city, and leaves it untranslated, as if his readers will know it. Words survive untranslated in a foreign-language letter for one reason: they are formulae, said aloud, learned by sound. It appears again at the end of the Didache’s eucharistic prayers (10:6), and in Greek at the very end of the Bible: Even so, come, Lord Jesus (Revelation 22:20).

And then the passages that read like quoted hymns and confessions. Here the evidence is literary rather than explicit, and it should be handled carefully, because this is a judgement rather than a certainty. The criteria scholars use are these, and they are the kind you can apply yourself with a Bible open: (a) the passage is introduced by a relative pronoun or a formula such as it is a faithful saying or wherefore he saith; (b) it falls into balanced lines or couplets when the surrounding prose does not; (c) it uses vocabulary that occurs nowhere else in that writer; (d) it can be lifted out without damaging the argument around it; and (e) its theology is stated in terms the writer does not otherwise use. No one of these settles anything. Three or four together are what makes the case.

Apply them. For each passage below, decide whether the writer is composing on the spot or quoting something his readers already knew. One of the eight is included as a control — it is a famous, rhythmic, exalted passage which is nevertheless the writer’s own composition — so the exercise cannot be passed by answering the same way every time.

0 of 8 answered.

Set the results out plainly. Embedded in the New Testament, on the ordinary scholarly reading of the internal evidence, are: hymns to Christ (Philippians 2:6–11; Colossians 1:15–20; 1 Timothy 3:16), a baptismal hymn fragment (Ephesians 5:14), a creedal formula about the resurrection with a list of witnesses (1 Corinthians 15:3–5), the words of institution (1 Corinthians 11:23–26), a faithful saying in four balanced lines (2 Timothy 2:11–13), a triune baptismal formula (Matthew 28:19), a trinitarian benediction still said at the end of services (2 Corinthians 13:14), an Aramaic prayer-acclamation (1 Corinthians 16:22), the congregation’s Amen at the thanksgiving (1 Corinthians 14:16), four canticles which the Church has sung daily ever since (Luke 1:46–55, 1:68–79, 2:14, 2:29–32), doxologies (Romans 16:27; 1 Timothy 1:17; Jude 24–25), and the Sanctus, taken from Isaiah 6:3 and sung by the living creatures in Revelation 4:8. Add the Lord’s Prayer, given as a text to be said (Matthew 6:9; Luke 11:2), and the instruction that letters be read aloud in the assembly and exchanged between churches (1 Thessalonians 5:27; Colossians 4:16).

That is not a church which had no liturgy. That is a church whose liturgy is quoted in its own scriptures.

Both halves of that deserve a straight answer. On 1 Corinthians 14: the chapter is indeed the fullest picture we have of a first-century assembly, and it is genuinely more participatory than most services of any kind today. But it is worth noticing what Paul does with the freedom he describes. He regulates it, at length, and the chapter is the most procedural passage in the New Testament. Two or three at the most, and in turn; one interpreter, or silence; the prophets to speak two or three and the others to judge; let the first hold his peace if something is revealed to another; for God is not the author of confusion, but of peace; and the conclusion in verse 40 — Let all things be done decently and in order, κατὰ τάξιν, according to an arrangement. The one chapter in the New Testament devoted to Spirit-given speech in the assembly is the one chapter that lays down rules for it. Paul’s remedy for a disordered meeting is not more spontaneity.

On “vain repetitions” (Matthew 6:7): the Greek is βατταλογήσητε, to babble — heaping up words in the belief that volume of speech compels a god, which is what Elijah mocks on Carmel. The proof that it is not a prohibition of set forms is what Christ says two verses later: After this manner therefore pray ye: Our Father which art in heaven… He answers the warning against babbling by giving them a fixed text. Luke’s version is even more explicit: When ye pray, say… (Luke 11:2). And the first Christians took it exactly that way — which is why the Didache, within a generation or two, tells them to say it three times a day.

Why everyone ended up with the same shape

Here is a fact that ought to be more famous than it is. Take the eucharistic liturgies of the Byzantine Greeks, the Latin West, the Syriac-speaking Church of the East, the Copts of Egypt, the Armenians and the Ethiopians. They are in six languages. They developed in regions separated by deserts, mountains and empires. Their peoples fought wars against each other. And they have the same structure, in the same order, with several of the same sentences.

Now the part that turns this from an observation into an argument. Several of these churches have not been in communion with one another since the fifth century. The Church of the East separated after the Council of Ephesus in 431. The Oriental Orthodox — Copts, Armenians, Ethiopians, Syriac Orthodox — separated after Chalcedon in 451. Greek East and Latin West drifted apart over centuries and formally divided in 1054. These are not branches which kept comparing notes. For most of their history they regarded each other, in varying degrees, as heretics.

Churches that have stopped speaking to each other cannot borrow from each other afterwards. Therefore any structure they share must predate the separation. That is not a theological argument; it is the same reasoning a philologist uses when two isolated languages share a word, or a textual critic uses when two independent manuscript families share a reading. Shared material in separated witnesses points backwards to a common source.

Before reading the discussion, commit to an answer. For each element below, say whether it is found in every historic rite East and West, or only in some.

0 of 8 answered.

The Addai and Mari case is worth dwelling on because it cuts both ways and the page should say so. On one hand it is the strongest single piece of evidence in this Part: a eucharistic prayer from a church separated in 431, possibly with roots in third-century Edessa, which opens with the same dialogue Cyprian quotes at Carthage in 252 and contains the Sanctus. On the other hand, its oldest manuscripts have no institution narrative at all — no “this is my body” — which is a striking exception to the common shape and an inconvenience to any argument that the shape was uniform from the beginning. It is evidence of early convergence and of early variation at the same time, and a page that quoted only the first half would be selling something.

The first half of that is a serious argument and it has teeth. A good deal of liturgical standardisation is post-Constantinian, and pressing every detail of the fourth-century rites back into the first century is exactly the overreach this page is trying to avoid. But the argument has a date problem. The common core — readings, homily, prayers, the peace, the dialogue, the thanksgiving, the people’s Amen, the meal — is attested by Justin around 155 and by Cyprian around 252, both of them writing under persecution, in different languages, on different continents, with no emperor to please and no synod to obey. Whatever Constantine did in 325, he did not do it in 155.

The second half — that agreement enforced is not agreement — is true and is properly the subject of Part XIII, where it is taken seriously rather than answered here with a phrase. Note only that the churches which suffered most from imperial coercion, and which separated from the imperial church precisely to escape it, are the ones that kept the shape most conservatively. The Assyrians were outside the empire altogether, under Persian rule, when they were praying it.

The Table, and when it moved

Ask a congregation in the informal tradition what the high point of the Sunday service is, and there is no real doubt about the answer. It is the sermon. Everything before it is preparation and everything after it is response; the length of the sermon is what the service is measured in; and the Supper, when it happens, happens perhaps monthly or quarterly, often at the end, sometimes in a separate shorter meeting. That arrangement feels so natural to those raised in it that its origin never comes up.

Work the sources first. For each of the earliest texts, say what it presents as the reason the assembly gathered.

0 of 8 answered.

The clearest single verse is the one people read past.

And upon the first day of the week, when the disciples came together to break bread, Paul preached unto them, ready to depart on the morrow; and continued his speech until midnight.

Acts 20:7, Authorised Version. Read the grammar. The purpose clause — why they came together — is to break bread. The preaching is what Paul did while they were assembled, and Luke mentions its extraordinary length as a circumstance, which is the point of the anecdote that follows: a young man falls asleep in a window and out of a third storey. Even in the one place where the New Testament records a sermon of exceptional length, the stated reason for the gathering is the meal.

Set beside it the summary of the first church’s life: And they continued stedfastly in the apostles’ doctrine and fellowship, and in breaking of bread, and in prayers (Acts 2:42) — four items, of which teaching is one and the breaking of bread another. And Paul’s rebuke at Corinth assumes it without needing to argue it: When ye come together therefore into one place, this is not to eat the Lord’s supper (1 Corinthians 11:20). The complaint is that what they are doing does not amount to the Supper — which only works as a complaint if eating the Supper is what the gathering was for.

Then the Didache: every Lord’s day…break bread. Then Justin: bread and wine and water every Sunday, brought after the prayers, with a portion carried by deacons to those who could not come. Then Pliny, from the outside: they reassemble to take food. Four independent witnesses across three generations, and in every one of them the Sunday assembly culminates in a meal.

So when did it move?

Not at the Reformation, which is the common assumption and is wrong; Part VII deals with that. The sermon-centred, infrequent-communion service is a later development within parts of Protestantism, and it has two main causes, neither of which is a doctrine anybody argued for.

The first is inherited. By the late Middle Ages the laity in the West communicated very rarely — often once a year, at Easter, as canon law required as a minimum — while attending Mass frequently and watching. When the Reformers restored preaching and the vernacular, they inherited congregations with no habit of receiving. Calvin ran into this directly and lost the argument to the civil authorities.

Plainly this custom which enjoins us to take communion once a year is a veritable invention of the devil… the Lord’s Table should have been spread at least once a week for the assembly of Christians.

John Calvin, Institutes of the Christian Religion IV.17.46 (1559), Battles translation. Calvin wanted weekly communion; the Geneva council granted four celebrations a year — Christmas, Easter, Pentecost and the first Sunday in September. Quarterly communion in the Reformed tradition is not Calvin’s theology. It is what a city council allowed him.

The second cause is the one traced in Part X: a revival movement which redesigned the service around producing a decision in the hearer, and for which a meal that assumes the hearer is already a Christian sits awkwardly at the climax. Once the object of the service is conversion, the sermon has to be last but one and the appeal last. The Table did not lose an argument. It lost a slot.

The first point is correct and this page grants it: as often as prescribes no interval, and no text in the New Testament commands a weekly Supper in so many words. The argument here is not from command but from description — what the earliest churches in fact did, unanimously, as reported by four independent witnesses including a hostile one. The second point, about familiarity, is a real pastoral observation, and it applies with equal force to a weekly sermon, a weekly offering and a weekly song, none of which anyone proposes to hold quarterly for fear of contempt. The argument from familiarity is never applied to the elements of the service that the tradition making it values most, which is a reason to suspect it is a conclusion in search of a premise. And on the third point, the historic traditions agree entirely that the Word is a means of grace — which is why the inherited order has more Scripture read aloud in it than any informal service, as Part XI is about to demonstrate with a count.

What the Reformers actually did

A great deal of the modern argument rests on a historical picture in which the Reformation is the moment the Church threw off liturgy — when the Bible was recovered, the ceremonies were swept away, and worship became free. It is a stirring picture. It is also, as a description of what the Reformers did to the order of service, almost exactly backwards, and the evidence is not obscure: it is in the confessions they wrote and the service books they printed.

Luther

Luther’s Formula Missae et Communionis of 1523 is an order of Mass for Wittenberg. It is in Latin. It keeps the Introit, the Kyrie, the Gloria, the Collect, the Epistle, the Gradual, the Gospel, the Nicene Creed, the Preface and its dialogue, the Sanctus, the words of institution, the Lord’s Prayer, the Agnus Dei and the distribution. What it removes is the canon of the Mass — the prayer of offering, on the theological ground that the Supper is God’s gift to us and not our offering to him. He excised one prayer and kept the service. The Deutsche Messe of 1526 supplies a German order alongside it; the Latin one went on being used in Lutheran churches and cathedrals for two centuries.

Seven years later the Lutheran princes had to state their position formally to the emperor at Augsburg, in a document intended to prove that they were not innovators. They put it flatly.

Falsely are our churches accused of abolishing the Mass; for the Mass is retained among us, and celebrated with the highest reverence. Nearly all the usual ceremonies are also preserved, save that the parts sung in Latin are interspersed here and there with German hymns, which have been added to teach the people.

Augsburg Confession XXIV (1530), of the Mass. The following section adds: we hold one communion every holy-day, and, if any desire the Sacrament, also on other days. Written by Melanchthon with Luther’s approval, and the confessional standard of Lutheranism ever since.

Cranmer

The English Reformation produced not less liturgy but a single authorised book of it: the Book of Common Prayer of 1549, revised 1552, restored and revised again in 1559 and 1662. Its preface argues that the medieval office had become so encumbered with commemorations that the reading of Scripture was constantly interrupted, and that the remedy is a plainer order in which the Bible is read through. The objection was that there was not enough Scripture in the service, and the remedy was a stricter form, not a looser one. Cranmer’s daily office reads the Old Testament through once a year and the New Testament three times, and he printed it in English so that anyone could check what was being said in his name.

Calvin

Calvin, driven out of Geneva and ministering to French refugees in Strasbourg, produced La forme des prières et chantz ecclésiastiques in 1542: a printed order with set prayers, a confession of sins, metrical psalms, the creed and a communion form. He argued in the Institutes that the Supper should be weekly and that the ancient assembly had a fixed order — word, prayers, communion, alms — taken from Acts 2:42. Geneva overruled him on frequency, as noted above. On the question of whether there should be a form, he never had any doubt.

And the party that abolished the Prayer Book

The sharpest test is the English Puritan movement, which really did reject the Prayer Book, really did dislike set forms, and in 1645 got its way: Parliament abolished the Book of Common Prayer and adopted the Westminster Assembly’s Directory for the Publick Worship of God in its place. If any movement in Christian history was going to produce worship without a shape, it was this one. Here is what its preface says about the book it is replacing — and it is the case for Position F, stated at its strongest by people who meant it.

…it hath proved an offence, not only to many of the godly at home, but also to the reformed Churches abroad… [and] an idle and unedifying ministry, which contented itself with set forms made to their hands by others.

The Directory for the Publick Worship of God (1645), preface. The objection is precise and worth respecting: that a minister who reads another man’s prayers may never learn to pray, and a congregation served by such a man is not fed. Anyone who has sat through a liturgy conducted by somebody plainly thinking about lunch knows the force of it.

And here is what the same document prescribes instead.

…ordinarily one chapter of each Testament be read at every meeting; and sometimes more, where the chapters be short, or the coherence of matter requireth it. It is requisite that all the canonical books be read over in order, that the people may be better acquainted with the whole body of the scriptures; and ordinarily, where the reading in either Testament endeth on one Lord’s day, it is to begin the next.

The Directory for the Publick Worship of God (1645), “Of Publick Reading of the Holy Scriptures”. The Directory then lays down the order of the service: prayer at the gathering, the readings and psalm, the prayer before sermon, the sermon, the prayer after sermon, a psalm, and the blessing.

Read those two quotations together and the finding is unmistakable. The most determined anti-liturgical movement in English history replaced one prescribed order with another prescribed order, and its lectionary rule is stricter than most liturgical churches keep today. What it objected to was reading prayers verbatim; what it never considered for a moment was a service without Scripture read in course, without a fixed sequence, and without a rule about what must happen. A modern informal service, judged by the Westminster Directory, fails on the reading of Scripture before it is asked about anything else.

Now test the sources. Below are six statements about worship. Say who is speaking.

That is a coherent position and it has a distinguished history; it is essentially the radical Reformation’s case, and later the Puritan one. Two things can be said in reply without settling it. First, it concedes the historical point this Part is making: the Reformation itself did not do what it is popularly said to have done, so the modern informal service cannot claim it as a precedent. Whatever it is, it is not a return to 1523. Second, the party that carried the principle furthest is the Westminster party, and the Westminster party produced the Directory — which is to say that the logical development, when it actually happened, produced a stricter rule about Scripture and a fixed order of service. The informal service is not the terminus of that road either. Its ancestry lies elsewhere, and Part X says where.

The man in the place of the office

The remaining Parts turn from history to consequence. Each takes one cost of discarding the inherited order, argues it structurally rather than by attributing motives to anybody, and sets beside it the historic practice that was shaped to prevent it. None of them is a claim about any individual’s sincerity, and none of them is an allegation about any church.

Start with the structural fact. When the order is inherited, the minister is doing something the congregation can check against a text. When it is composed weekly, he is the text. That is not a statement about his character; it is a statement about where the service comes from. If the prayers, the readings, the confession and the eucharistic words are given, then the man at the front is performing an office, and the office is larger and older than he is. If they are not given, then his judgement chooses the readings, his words carry the prayers, his personality sets the tone, and his gifts determine whether the hour goes well.

Three consequences follow, and they follow whether or not anyone wants them.

1. The service becomes dependent on his form. A liturgy conducted by a tired, dull or grieving man is still the liturgy; the congregation receives what it came for. An unscripted service conducted by the same man is a poor service, because there was nothing in it but him.
2. Succession becomes a crisis. Where the order is inherited, a new minister steps into a shape the congregation already knows, and the discontinuity is small. Where the order is the man, replacing him means replacing the worship, which is why congregations of this kind so often fracture at a change of pastor.
3. Criticism has nowhere to go but the person. An objection to a written prayer is an objection to a text. An objection to an unscripted service is, unavoidably, an objection to the man who made it up — which makes ordinary disagreement feel like disloyalty, to both parties.

The historic churches met this question in its sharpest possible form, and early. In North Africa after the persecution under Diocletian, a controversy broke out over clergy who had handed over the scriptures to be burned and then returned to their ministry. The Donatist party held that sacraments administered by such men were void. The question was not academic: whole congregations had to know whether their baptisms had happened. The settlement the Church reached, worked out at length by Augustine, was that the validity of what is done does not rest on the worthiness of the man doing it, because it is not his to give. The office is not the man.

…the Sacraments and the Word are effectual by reason of the institution and commandment of Christ, notwithstanding they be administered by evil men.

Augsburg Confession VIII (1530), of what the Church is. The Lutheran confessions restate the ancient settlement without qualification, and the article exists to stop exactly the anxiety this Part is about: a congregation asking whether what it received on Sunday depended on the state of the minister’s soul.

Vesture belongs here too, and is usually misunderstood as ornament. A minister in a surplice or a chasuble is dressed in something that covers what he is wearing underneath — his taste, his status, his body, his individuality — and everyone in his tradition for fifteen centuries has worn the same thing. It is a uniform, and the function of a uniform is to make the wearer interchangeable. The modern alternative, in which what the man at the front wears is a personal statement, communicates the opposite by the same means.

Both halves of that are right, and the first is a genuine cost of Position L which this page does not intend to hide. A fixed form can carry an empty man for a very long time, and congregations have suffered decades of it. The historic answer is not that this does not happen but that the alternative fails differently and worse: the unscripted service does not expose the empty man, it exposes the untalented one, and those are not the same thing. A shallow preacher with charm passes; a devout one with a dull voice does not. As for character, every liturgical tradition also has canons about the fitness of its ministers, and the Donatist settlement was never a licence for bad clergy — it was an answer to a different question, namely whether the people are harmed when discipline fails. The two mechanisms are meant to work together: discipline for the man, a given form so that the people do not depend on the discipline working.

Drift, and the thing that stops it

The second cost is the one that is hardest to see from inside, because by its nature it leaves no trace. A congregation whose prayers are composed fresh each week has no fixed point against which to notice that its theology has moved. Nothing is being contradicted; there is simply nothing to contradict. Twenty years later the church prays differently, believes somewhat differently, and could not say when either changed.

That this mechanism is real, and not a theoretical worry, is documented in the most direct way possible: the early Church fought doctrinal battles over the wording of prayers, because it understood perfectly well that the words a congregation says every week are what it will end up believing.

Basil and the doxology, 375

Basil of Caesarea wrote On the Holy Spirit, one of the founding texts of trinitarian theology, for a reason he states in the first chapter. He had been using two forms of doxology in public worship — glory to the Father with the Son together with the Holy Ghost, and glory to the Father through the Son in the Holy Ghost — and had been attacked for it.

…at one time “with the Son together with the Holy Ghost,” and at another “through the Son in the Holy Ghost”…

Basil of Caesarea, On the Holy Spirit I.3, written 375 at the request of Amphilochius of Iconium, Nicene and Post-Nicene Fathers, second series, VIII. The whole treatise exists because prepositions in a liturgical formula were understood, by everyone concerned, to carry the doctrine of the Spirit. Nobody on either side thought this was a quarrel about style.

Basil goes further in the same work, and the passage is worth having in front of anyone who thinks liturgical practice is theologically inert.

…some we possess derived from written teaching; others we have received delivered to us “in a mystery” by the tradition of the apostles… Which of the saints has left us in writing the words of the invocation at the displaying of the bread of the Eucharist and the cup of blessing?

Basil, On the Holy Spirit XXVII.66. He lists signing with the cross, turning east to pray and the eucharistic invocation as practices carried by use rather than by text. Readers from different traditions will weigh that claim very differently — it is one of the standard proof texts in the Catholic and Orthodox case on tradition, and Protestants have answered it for five centuries. The page cites it for a narrower purpose: as evidence that a fourth-century bishop took the wording of worship to be doctrinally load-bearing.

“The law of praying establishes the law of believing”

The principle got its slogan in the fifth century, from Prosper of Aquitaine, arguing about grace: ut legem credendi lex statuat supplicandi — that the law of praying may establish the law of believing. It is usually quoted in the shortened form lex orandi, lex credendi.

…ut legem credendi lex statuat supplicandi.

Prosper of Aquitaine, in the document known as the Capitula Coelestini or Indiculus, chapter 8, c. 435–442 (Patrologia Latina 51:209–210). Prosper’s argument is that the Church’s universal practice of praying for the conversion of unbelievers proves what the Church believes about grace. He is using the liturgy as evidence about doctrine — which only works if the liturgy is stable enough to be evidence.

The creed is the clearest case of a doctrinal safeguard placed inside worship. The Nicene Creed was formulated at councils in 325 and 381, but its recitation in the Sunday eucharist came later and spread gradually — in the East from the late fifth and early sixth centuries, and at Rome not until 1014. It was put there for a reason. A congregation that says the creed aloud every week cannot drift far without noticing, because it is committing itself, in its own voice, to a set of sentences it did not write and cannot quietly amend.

Now the structural argument, stated without reference to any modern church. Suppose a congregation’s prayers, confessions and eucharistic words are composed afresh each week by whoever is leading. Each week’s wording reflects that person’s emphases, and each person’s emphases reflect the current preoccupations of the church and its surroundings. No single step is a departure from anything. But there is no recorded baseline, so there is no possible measurement of the cumulative movement. Drift is not a moral failure. It is what happens to any quantity that is measured only against its own previous value. The remedy is not sincerity, of which there may be a great deal; it is a fixed point outside the sequence.

And the fixed point has to be said, not merely believed. A church may hold an entirely orthodox statement of faith on its website and never say a word of it aloud together. What forms a congregation is not the document in the drawer; it is the sentences in its mouth fifty-two times a year.

This is the strongest objection in the Part and the page concedes the central fact: a liturgy transmits whatever it contains, including error, and it transmits error with the same efficiency. The Reformers said exactly this and acted on it. But notice what the concession does to the argument, rather than for it. The Reformers could remove the accretions because they were written down. They could point at a sentence, publish an argument about it, and get it changed — and the record of that argument survives, so we can still examine both sides. That is not available against an unwritten order. A form can be reformed; a habit can only be broken, and only by whoever happens to be in charge. As for Scripture being the fixed point: agreed, and Part XI is about which order of service actually puts more of it in front of a congregation.

When feeling becomes the measure

The third cost is the most delicate to argue, because it is easily heard as a sneer at people who are moved in church — which would be both unkind and stupid. Worship should move people. The Psalms are not emotionally neutral, Augustine wept at the singing in Milan, and a service that leaves a congregation entirely unaffected has failed at something. The question is not whether feeling belongs in worship. It is whether feeling becomes the test of whether worship has happened.

The difference is visible in the vocabulary a congregation uses on the way out. Was the Word read and preached, were sins confessed and forgiveness declared, were the prayers offered, was the Supper given? — these are questions with answers, and anyone present can verify them. Did we really worship? Was God moving? Did it go deep? — these are questions about an internal state, and they have no external check at all. Once the second set becomes the standard, several things follow mechanically.

1. The service must be engineered to produce the state, because it is now the deliverable: the sequence of songs, the dynamics, the lighting, the pacing of the appeal.
2. Those who do not feel it conclude that something is wrong with them — which lands hardest on the grieving, the exhausted, the ill and the temperamentally undemonstrative, that is, on precisely the people the assembly exists to carry.
3. The objective content becomes optional, because it is not what is being measured. Confession, absolution, creed, intercession and Supper can all be dropped without anybody noticing a loss, since the loss does not show up in the metric.

This is a genealogy rather than an accusation, and the genealogy is documented. The engineering of worship to produce a measurable effect in the hearer is not a recent invention of sound desks and lighting rigs. It was set out, in print, with its philosophy stated, in 1835, by a man who meant every word and was entirely candid about the method.

A revival is not a miracle, nor dependent on a miracle, in any sense. It is a purely philosophical result of the right use of the constituted means — as much so as any other effect produced by the application of means.

Charles Grandison Finney, Lectures on Revivals of Religion, Lecture I, “What a Revival of Religion Is” (New York, 1835). Finney’s comparison in the same lecture is agricultural: a revival follows the use of appropriate means as naturally as a crop follows the farmer’s. He is not being cynical; he is arguing against a fatalism he thought was paralysing the churches, and his “new measures” — the protracted meeting, the anxious seat, the direct appeal for a decision — were designed accordingly. Finney died in 1875.

Grant the sincerity entirely, and then look at what the sentence does to the shape of a service. If the assembly is a mechanism for producing an effect, then its order is an engineering question, and every element earns its place by contribution to the effect. Confession of sin stays if it softens the hearer and goes if it dampens the room. The creed contributes nothing measurable. Scripture read at length without comment is dead air. The Supper, which declares that the thing has already been accomplished outside the hearer, sits awkwardly at the end of a sequence designed to bring him to a decision. None of these is removed by a vote. Each simply fails a test it was never designed to pass.

The inherited order is built the other way round. Its elements are there because of what they are — confession because sin must be confessed, absolution because forgiveness must be declared aloud, the readings because the Church must hear the whole of Scripture, the creed because the faith must be confessed, the Supper because Christ instituted it — and the congregation’s feelings on a given Sunday are not consulted, because they are not the point. A liturgy is what you do when you do not feel like it. That is its most unfashionable feature and its most useful one: on the Sundays when a man has nothing whatever to bring, the words are still there, and they do not require him to generate anything before he can join in.

The first half is fair and the page does not dispute it: Scripture is full of bodily, noisy, unembarrassed worship, and a tradition that has become embarrassed by emotion has a problem of its own which liturgy does not cure. Nothing in this Part argues for restraint as a virtue. The distinction being drawn is not between warm and cold worship but between feeling as fruit and feeling as test — between a service whose objective content moves people, and a service whose objective content is selected by whether it moves people. The first has no upper limit on emotion; the second has no floor under content. On numbers: which churches are growing is a question about many things, and a page arguing that truth is not settled by longevity in Part XIII cannot turn round and settle it by attendance here.

The Bible you never hear

The fourth cost is the one most easily measured, and the measurement is uncomfortable for everybody, including liturgical churches. A lectionary is a fixed cycle of readings. Its whole purpose is to take the choice away — so that a congregation hears Leviticus and Lamentations and the hard sayings of Christ and the passages about money and the passages about judgement, not because anybody selected them, but because it is the week for them.

Where there is no cycle, somebody chooses. That somebody is almost always the preacher, choosing what he will preach on next, which means the congregation hears his canon. It is nobody’s fault and it is not deliberate. Every preacher has texts he understands, texts that suit his gifts, texts that answer the questions he is interested in, and texts he has never been able to make anything of. Over a decade that produces a Bible with holes in it, and neither he nor the congregation can see the holes, because the holes consist of things that were never mentioned.

So measure it. Below are the sixty-six books. Mark every book you have heard preached from — as the text of a sermon, not quoted in passing — in the last twelve months. Be honest; nobody sees this but you.

Whatever number came out, hold it against the rules the tradition set itself. Justin’s congregation in Rome heard the prophets and the apostles read “as long as time permits”. The Westminster divines — the anti-liturgical party — required a chapter of each Testament at every meeting and the whole canon in course. Cranmer’s daily office takes the Old Testament through once a year and the New Testament three times. A modern liturgical Sunday typically has three or four readings plus a psalm, running to several hundred verses a month.

Set against that, a service with one reading of six verses, chosen by the preacher as the springboard for the sermon, is not a service in which the Bible has a large place. It is a service in which one man’s reading of the Bible has a large place, which is a different thing, and the difference is invisible from inside because the sermon is full of Scripture — verses gathered from all over, in support of the point being made. There is a separate dossier on that habit and what it does to a text: Dossier 05.

The books most often missing from the list are predictable and it is worth saying which, because the pattern is the argument. Leviticus, Numbers, Deuteronomy. Joshua through 2 Chronicles as continuous narrative rather than as anecdotes. Job. Ecclesiastes. Song of Solomon. The bulk of the writing prophets: Isaiah beyond a handful of chapters, Jeremiah, Ezekiel, Hosea, Amos, Micah, Habakkuk, Zephaniah, Haggai, Zechariah. Lamentations. Most of the minor epistles. Revelation beyond chapters 1 to 3 and 21 to 22. Which is to say: the lament, the law, the covenant history, the judgement oracles and the wisdom that does not resolve. A congregation that never hears these has been handed a smaller God and a smaller account of human life, and has no way of knowing it.

The first half is a strong reply and this page grants it: consecutive exposition of whole books is a real discipline with real results, it is far better than topical preaching against this particular failure, and a congregation which has been taken through Romans in forty sermons has heard something no lectionary will give it. But apply the test to the practice rather than to the principle. Which books get chosen for the long series? Over a decade, the answer is usually Genesis, a Gospel, Acts, Romans, the prison epistles, perhaps a psalm sequence — and the hard three-quarters of the Old Testament remains unpreached, because a series has to be chosen and somebody chooses it. Consecutive exposition removes the arbitrariness within a book; it does not touch the arbitrariness between books, which is where the gaps in your grid came from.

On the second half: yes, the preacher is meant to know his congregation, and a lectionary that could never be departed from would be an absurdity — every liturgical tradition has provision for occasions, and nobody reads the appointed lesson at a funeral. The point of the cycle is not that the preacher never chooses. It is that the default is not his choice, so that the congregation hears what nobody would have chosen for it, which is the only way anyone ever hears anything uncomfortable.

Congregation or audience

The fifth cost is the most ironic, because it is the exact reverse of what each side believes about the other. Informal worship understands itself as participatory and the historic liturgy as something done to a passive congregation by a man in robes. Test it the only way it can be tested: count what the people in the seats actually say out loud.

Take the 1662 Order for the Administration of the Lord’s Supper, and list the parts the congregation says or joins in saying. The responses at the opening dialogue. The response after each commandment. The Gloria in excelsis. The Nicene Creed. The confession of sins, said by all. The Sanctus. The Lord’s Prayer, twice. The Amen after each collect and at the end of the great prayer. The Gloria at the end. In the daily office, add the psalms — said or sung antiphonally, verse about, by the whole congregation, right through the Psalter each month.

Before the figure: guess. Roughly how many words does the congregation say aloud, together, in a full 1662 communion service?

Now do the same count for a service in the informal order described in Part I. Songs aside — and singing is genuine participation, which this page is not discounting — the congregation typically says: nothing. There may be a corporate prayer read from a screen, an amen, a greeting to a neighbour. The spoken content of the service, all of it, comes from the platform. The service that describes itself as participatory is, in the strict sense of who speaks, the more platform-centred of the two.

Paul assumes the other arrangement so completely that he uses it as the premise of an argument about tongues:

Else when thou shalt bless with the spirit, how shall he that occupieth the room of the unlearned say Amen at thy giving of thanks, seeing he understandeth not what thou sayest?

1 Corinthians 14:16, Authorised Version. The whole argument depends on the congregation having a part — the Amen — which it cannot say if it cannot follow what is being prayed. Intelligibility matters to Paul because the people are supposed to be joining in. Justin, a century later, records the same Amen in the same place in the service.

The deeper point is not arithmetical. A liturgy gives the congregation words it did not have to invent and does not have to feel, which is what makes it usable by a man in trouble. The bereaved, the frightened, the doubting and the merely tired can say the creed and mean it without generating anything. They can confess sins in the church’s words when their own have failed. They can say the Lord’s Prayer when they cannot pray. Given words carry people who have nothing to give. A service in which the congregation contributes only what it can produce on the day is, precisely, a service that asks the most from those with the least.

The first sentence is true and important: recitation without attention is not worship, and every liturgical tradition contains people doing it. But the objection proves less than it appears to, because the same applies to every other element of any service: a congregation can sing with its mind elsewhere, listen to a sermon with its mind elsewhere, and feel deeply moved about nothing in particular. Inattention is not a property of fixed forms. As for the second sentence — this page would be glad to see more of what it describes, and notes only that it mostly describes activity around the service rather than the service itself, and that the chapter it appeals to is the one that ends let all things be done decently and in order. The historic answer is not either/or. It is that the assembly should have both a given voice and a free one, and that a church which has only the second has given its silent members nothing to say.

What two thousand years proves

This is the Part the page would be dishonest without, because the argument it has been making has a well-known failure mode and the page has been flirting with it for twelve sections. Longevity is not truth. Practices survive for many reasons, most of which have nothing to do with being right: institutional inertia, the cost of change, the interests of those who administer them, the sheer difficulty of coordinating a new arrangement among millions of people. Slavery lasted longer than the Roman canon. Astrology has outlived several sciences. “We have always done it this way” is the standard defence of every bad practice that has ever needed defending, and a Christian tradition which recites the creed weekly should be the last to forget that its own reformers had to break some very old habits indeed.

So test the arguments rather than assuming them. Below are eight inferences, some of which this page has made and some of which it has not. For each, say whether it is valid as stated.

0 of 8 answered.

Stated precisely, then, here is what the continuity establishes and what it does not.

Which leaves the argument standing where it ought to stand: not on age, but on what the order does. The case for Position L in this page is that each element of the inherited order performs a function that can be named, defended from Scripture, and tested — and that when the element goes, the function goes with it and something else moves into the space. Scripture read in course means a congregation hears what nobody would have chosen. Given prayers mean the people have a voice and a fixed point to measure drift against. Confession and absolution mean sin is named and forgiveness is declared aloud rather than left to be felt. The creed means the faith is confessed rather than assumed. The Supper at the centre means the service ends in something accomplished outside the hearer rather than in something required of him. Take those away and you have not removed a set of forms; you have removed five functions, and nothing has been put in their place.

The argument is therefore defeasible, which is as it should be. If a congregation without the inherited order hears more Scripture, says more aloud, confesses sin more plainly, confesses the faith more explicitly, and communicates more often than a liturgical one — then on this page’s own criteria it is doing better, and the argument fails against it. That is a real test, it can be run on any two churches, and a defender of Position L who is unwilling to lose it has not understood his own case.

A service diagnostic

Seven questions you can run on a service. They are drawn from the functions identified in Part XIII rather than from any tradition’s ceremonial, which is why a plain service in a house, a hall or a field can score clean on every one of them, and an elaborate one can fail. It assesses a service and not a person. It produces no verdict about anybody’s sincerity, salvation or standing before God, it says nothing about whether a congregation is a true church, and it recommends no course of action whatever. It is a checklist for thinking.

If you ran it on your own church and did not like the result, the useful next step is not a conversation with anybody. It is to run it on the same church again in three months, and to run it on the service you grew up in, which is the comparison most people have never made deliberately.

Your answers, revisited

Here are the ten questions you answered before reading, with what the sources say. If you skipped the quiz, this section will be empty — go back to the top, answer them, and return.

Check it yourself

Fourteen items. Almost the whole of this subject is free and short: the Didache, Justin, Pliny, Cyprian, Basil, the Reformation confessions, the Westminster Directory, the 1662 Prayer Book and Finney are all out of copyright and online in full. If you read only three, read the first three — they are about an hour together, and between them they settle the historical half of the question without any help from this page.

1. 1. Justin Martyr, First Apology, chapters 61 and 65–67. Read this first. Fifteen minutes, free at New Advent, the Christian Classics Ethereal Library and Early Christian Writings. Four short chapters. Read them and then write out the order of the service he describes on a piece of paper, before comparing it with anything. This is the single most efficient check on this page: if Justin does not say what Part II says he says, the page is wrong and you will know within a quarter of an hour.
2. 2. The Didache, entire. Read this first. Twenty minutes, free in several translations. It is about nine pages. Note chapter 7 (the baptismal formula and its edge cases), chapter 8 (the Lord’s Prayer three times daily), chapters 9 and 10 (the eucharistic thanksgivings, and the exception permitted to prophets), and chapter 14 (the Lord’s day, confession, breaking of bread). Note also how much of it is about travelling preachers and how to tell a true one from a profiteer, which is a different dossier.
3. 3. Acts 2:41–47, Acts 20:7–12, 1 Corinthians 11:17–34 and 1 Corinthians 14, read one after another at a single sitting. Read this first. Twenty-five minutes, free. Read Acts 20:7 slowly enough to see which clause is the purpose clause. Then read 1 Corinthians 14 with a pencil and mark every imperative — every instruction Paul gives about how the meeting is to be run. The number surprises most people.
4. 4. Pliny the Younger, Letters X.96, with Trajan’s reply in X.97. Ten minutes, free. Two letters. A hostile Roman administrator describing Christian worship to his emperor in 112, including the detail that he tortured two deaconesses to find out what went on. Read it for the liturgical details and then for what it tells you about the conditions in which this worship was being conducted.
5. 5. Cyprian, On the Lord’s Prayer, especially sections 31–35. Half an hour, free at New Advent. The “Lift up your hearts” reference is in 31, dropped in passing. Read the surrounding pages to see how casually he treats it, which is the whole evidential value of the passage.
6. 6. The Directory for the Publick Worship of God (1645), entire, including the preface. The strongest material against the page’s emphasis and also its best single confirmation. An hour, free at several Reformed archives. Read the preface for the case against set forms, put at its strongest by people who paid for holding it. Then read “Of Publick Reading of the Holy Scriptures” and ask how many churches of any tradition keep that rule today.
7. 7. Augsburg Confession, Articles VII, VIII, XXIV and XXVIII; and the Apology’s Article XXIV. An hour for the lot, free at the Book of Concord and Project Wittenberg. Article XXIV is the one that ends the popular story about the Reformation abolishing the Mass. Article VIII is the Donatist settlement in one sentence.
8. 8. John Calvin, Institutes IV.17.43–46. Half an hour, free at the Christian Classics Ethereal Library in the Beveridge translation. Four short sections on the frequency of the Supper. Then look up what the Geneva council actually permitted, and notice that the practice most associated with Calvin’s name is the one he lost.
9. 9. The Book of Common Prayer (1662): the preface, the section “Concerning the Service of the Church”, and the Order for the Administration of the Lord’s Supper. An hour, free in many editions online. The preface is a Reformation argument about how to get more Scripture into a service. Then take a pencil to the communion order and mark every line the people say; that is the count in Part XII, and you should not take this page’s word for it.
10. 10. Basil of Caesarea, On the Holy Spirit, chapter I and chapter XXVII. Forty minutes, free at New Advent and CCEL. Chapter I for the occasion of the book — a fight about a doxology — and chapter XXVII for the tradition passage, which readers from different traditions will assess very differently and should read for themselves rather than through anybody’s summary.
11. 11. Charles Finney, Lectures on Revivals of Religion (1835), Lecture I. Half an hour, free. Read the whole lecture rather than the sentence quoted here. Finney is lucid, forceful and completely open about his method, and the experience of reading him is not what most readers expect: he is not a cynic, and that is what makes the lecture historically important.
12. 12. The Divine Liturgy of St John Chrysostom, entire, in any standard translation. An hour to read, free in several versions, and better still attended. Read it for the structure, and count the congregation’s parts. Whatever conclusion you come to about the theology, it will be hard to maintain afterwards that the people have nothing to do.
13. 13. Martin Luther, Formula Missae et Communionis (1523) and Deutsche Messe (1526). An hour and a half for both, in Luther’s Works volume 53 and in various online editions. Read the 1523 order against a medieval Mass and mark what is gone; the list is much shorter than most people expect.
14. 14. Philippians 2:5–11, Colossians 1:15–20, 1 Timothy 3:16, Ephesians 5:14 and 2 Timothy 2:11–13, read one after another aloud. Ten minutes, free. Aloud is the point: the reason scholars think these are quoted rather than composed is something you can hear. Then read the paragraphs on either side of each of them, and hear the difference.

On the Apostolic Tradition: if you want the scholarly dispute rather than this page’s one-paragraph summary of it, the standard modern treatment is the commentary by Paul F. Bradshaw, Maxwell E. Johnson and L. Edward Phillips (Hermeneia, 2002), which argues the composite case. It is not free and it is not light, and it is cited here so that readers can see that the page’s caution about that source is somebody else’s argument and not an invention.