The Shape of a Sermon

Be honest about the anchor first. 1 Corinthians 14:40 is not a verse about sermons. It closes a long chapter about the gathered worship of a church in which several people spoke, some in tongues, some in prophecy, with too many speaking at once and too few understanding. To use it as the text of a lesson on sermon structure is to extend it, and the extension needs to be justified rather than assumed.

The justification is in the chapter's reasons. Paul does not demand order because God likes neatness. He demands it because disorder stops people understanding, and what is not understood cannot build anyone up. 'Unless you speak intelligible words with your tongue, how will anyone know what you are saying? You will just be speaking into the air' (14:9). 'But in the church, I would rather speak five coherent words to instruct others than ten thousand words in a tongue' (14:19). The principle that governs the whole chapter is that speech in the assembly exists for the hearer. A sermon that cannot be followed fails that test as surely as a tongue without an interpreter, even if every sentence in it is true.

Here is why shape matters more for a sermon than for almost any other kind of writing. A reader can go back. He can see that a chapter has four parts, skip ahead, reread a paragraph that lost him. A listener cannot. The sermon passes him once, at the speed the preacher chooses, and when he loses the thread he cannot find it again by looking. In much of the world the listener also has no printed Bible open in front of him, or cannot read the one he has, or is hearing the sermon through an interpreter a sentence behind. For every one of those hearers the shape of the sermon is the only map they will be given. A preacher who has not decided on the shape is asking them to walk a road he has not walked himself.

So the question of this lesson is not which shape is the biblical one. The Notice question in the inquiry should have settled that: the New Testament prescribes no internal form for a sermon, and the sermons it records do not share one. The question is which shape will carry this text to these hearers so that they can follow it, remember it and be changed by it. The church has developed many answers. Four of them will be set out here, because between them they cover most of what you will hear and most of what you will need, and each one will be given its strength and its characteristic failure.

The deductive sermon states its main idea near the beginning and then spends the rest of its time explaining, proving and applying it. It is the shape most people picture when they hear the word sermon: an introduction, a proposition, two or three or four main points that each develop the proposition, and a conclusion. John Broadus set it out in his Treatise on the Preparation and Delivery of Sermons (1870), which trained generations of Baptist and other preachers in America and was carried by missionaries far beyond it. The textbook that most often stands behind it today is Haddon Robinson's Biblical Preaching (1980), which you met in module 2 for its big idea. Robinson discusses several ways the idea can be developed: explained, proved, or applied, and he sets deductive arrangement beside inductive and mixed forms rather than making it the only option.

Its strengths are real, and some of them matter more outside the West than inside it. It is clear. The hearer knows where he is going, and he knows when he has arrived at each stage. It is memorable, because a stated idea with a few numbered points can be carried home and repeated to someone who was not there. It suits teaching, when the congregation needs to learn what a passage says before it can feel anything about it. And it travels well through an interpreter, because each point can be announced, repeated and returned to; an interpreter who knows the outline in advance can stay with a deductive sermon even when the preacher's sentences are long.

Its characteristic failure is that it can hand down a verdict before the hearer has seen any evidence. The proposition comes first, so the hearer who doubts it spends the rest of the sermon either resisting or switching off. It can sound like a lecture, which some texts are and most are not. Above all it can impose a form on a passage that has a different form of its own. Preach the parable of the prodigal son as three points, the father's love, the son's repentance and the brother's resentment, and every point may be true while the story has been dismantled: the hearer never feels the suspense of the son on the road, and the parable's open ending, with the elder brother still outside, is closed by the preacher's third heading.

A second failure is cosmetic but common. The deductive sermon invites alliteration, and alliteration invites distortion. Three points beginning with the same letter are easier to remember, and there is nothing wrong with that until the preacher bends the text to fit the letter. When a point is worded as it is because it rhymes and not because the passage says it, the form has started to govern the content. A useful rule: the points should be the passage's points, in the passage's order where the passage has an order, and only then made memorable. If they cannot be made to alliterate honestly, do not alliterate them. A preacher in a second language should be especially wary, since alliteration rarely survives translation and the interpreter will be left translating a pun.

The inductive sermon reverses the order. It begins with particulars, a question, a story, a problem, a detail in the text, and moves with the hearer towards the main idea, which may not be stated until near the end. The modern case for it was made by Fred Craddock in As One Without Authority (1971). His argument was that the old deductive sermon assumed an authority the pulpit no longer had in the eyes of its hearers. People no longer accepted a conclusion because a preacher announced it. So the sermon should not deliver conclusions; it should retrace the journey by which the preacher came to them, and let the hearers make that journey themselves, so that when they arrive the conclusion is theirs. Craddock's book was a starting point for what came to be called the New Homiletic, a broad movement in North American preaching from the 1970s onwards that gave much more attention to the experience of the listener.

The inductive sermon has deep biblical roots, and its defenders point to them. Jesus' parables are inductive. The hearer is told about a farmer, a woman sweeping her house, a man going down from Jerusalem to Jericho, and only afterwards discovers that the story was about him. Nathan's parable to David in 2 Samuel 12 is the classic case: a story about a rich man and a poor man, a king who passes judgment on the rich man, and then a sentence, 'You are that man!' (12:7), which lands because David reached the verdict before he knew it was his own. A deductive Nathan, beginning with a proposition about adultery and murder, would probably have been dead by noon.

Its strengths are participation and surprise. The hearer is not resisting a claim; he is following a question, and a question engages people a statement does not. It suits narrative texts, which are themselves inductive. It suits hearers who are suspicious of authority, sceptical outsiders, or young people who have heard every proposition before. And in many oral cultures it is simply how wisdom is passed on: the elder does not announce the moral, he tells the story and lets the listener find it.

Its characteristic failures are the mirror image of the deductive sermon's. It can wander and never arrive, so that the hearer enjoys the journey and goes home with nothing in his hands. It can rely so heavily on the preacher's own experience of the text that the text becomes a prompt for the preacher's reflections. Its critics, including many who learned from Craddock, have asked whether a sermon built on the absence of authority can still say what Paul says in 2 Corinthians 5:20, 'We implore you on behalf of Christ: Be reconciled to God.' The inductive form does not forbid that sentence, but it can lose the nerve to say it. And it is harder to follow through an interpreter, because a sermon that withholds its point is harder to translate than one that announces it: the interpreter does not know what the preacher's details are for. Haddon Robinson's semi-inductive arrangement, in which the idea is reached part of the way through and then developed, is one way of keeping the strengths of both.

Eugene Lowry took the inductive impulse and gave it a definite shape in The Homiletical Plot: The Sermon as Narrative Art Form (1980, expanded edition 2001). His insight was that a sermon is an event in time, and that the most powerful shape for an event in time is a plot: something goes wrong, the trouble deepens, a resolution appears, and life is seen differently afterwards. He set out five stages, and gave them memorable nicknames. Upsetting the equilibrium, 'oops': the sermon opens a problem or ambiguity. Analysing the discrepancy, 'ugh': the problem is explored and deepened until the hearer feels its weight. Disclosing the clue to resolution, 'aha': something, usually from the text, opens a way through, often by reversing what the hearer expected. Experiencing the gospel, 'whee': the good news is announced into the problem the sermon has opened. Anticipating the consequences, 'yeah': the sermon looks at what now becomes possible. Lowry himself treated the stages as a flexible movement rather than a fixed formula, and his followers have adapted them freely; the shape matters more than the count.

See it with a text. Take Luke 15, and begin where Luke begins, with the grumbling of the Pharisees and scribes that Jesus welcomes sinners and eats with them (15:2). Oops: the people complaining are the religious people, the ones who kept the rules; the hearers in most churches are more like them than like the tax collectors. Ugh: why does grace offend those who have been faithful? Probe it honestly; the elder brother's complaint in 15:29-30 is not stupid, it is the complaint of someone who has worked hard and seen a waster rewarded. Aha: notice that the father goes out to the elder son exactly as he ran out to the younger (15:20, 15:28). The clue is that the parable has two lost sons, and the second is lost at home. Whee: the father's words, 'all that is mine is yours' (15:31), are spoken to the resentful one; grace is offered to the people who think they do not need it. Yeah: the parable ends without telling us whether the elder son went in, and so does the sermon, with the door open and the music playing.

Lowry's plot was developed in a white North American seminary, but its central instinct is old and widespread. Henry Mitchell, in Black Preaching (1970) and Celebration and Experience in Preaching (1990), described the African American sermon as moving towards celebration: a closing movement in which the good news is not merely stated but rejoiced in, so that the congregation leaves having felt the gospel and not only understood it. The tradition expects a sermon to move, often from slow and low to high and full, with the congregation answering throughout, and there is an old folk saying among its preachers about starting low, going slow, rising higher and catching fire. Whatever you make of the saying, the point it carries is structural: the sermon is a journey with a destination, and the destination is joy in what God has done.

The strengths of plotted and celebratory sermons are obvious to anyone who has sat under one: they hold attention, they engage feeling as well as mind, and they suit the many biblical texts that are themselves stories. Their characteristic failure is manufacture. A preacher can invent a problem the text does not raise in order to have something to resolve, or produce the celebration at the end of every sermon whether the text earned it or not. Lamentations does not end in celebration; neither does Psalm 88. A plot imposed on a text is as much an imposition as three points imposed on a parable, and a celebration that has not been earned by the text becomes a performance. Mitchell himself warned against celebration that does not match the message it follows, and taught that the celebration should rejoice in the very truth the text has preached, which is the right test.

The fourth shape is the oldest of the four in its English form, and in many churches of the global south, especially Presbyterian and Reformed ones, it is still the default. William Perkins, the Cambridge preacher, set it out in The Art of Prophesying, published in Latin in 1592 and in English translation in the first decade of the next century. His summary has four steps: read the text distinctly from the canonical scriptures; give its sense and understanding by scripture itself; collect from the natural sense a few profitable points of doctrine; and apply those doctrines rightly to the life and manners of the hearers in plain speech. The Westminster Assembly's Directory for the Public Worship of God (1645) codified the same pattern for a generation of preachers: the doctrine is raised from the text and confirmed, and then it is brought home in uses, among them instruction, the confutation of error, exhortation to duty, warning against sin, comfort, and trial, which meant self-examination.

The word uses is the key to the plain style, and it is worth stealing whatever shape you finally choose. A Puritan sermon was not finished when the doctrine was correct; it was finished when the doctrine had been turned towards each kind of hearer in the room. The Directory tells the preacher not to rest in general doctrine but to bring it home to special use. That discipline answers the commonest weakness of sermons in every tradition, which is application so general that nobody has to change. A use addressed to the weary, then one to the careless, then one to the believer who doubts, makes the hearer locate himself.

The plain style has its failures too. It can lift doctrines out of a narrative so efficiently that the story itself disappears, which is the same danger as the deductive sermon. Its uses can multiply until the sermon is two hours long, as some Puritan sermons were. And collecting doctrines from a text can become a way of finding in the passage the doctrines you already held. Perkins' own safeguard, the natural sense, is the right one: only the doctrines the text itself teaches, read in its own setting.

So how do you choose? Four questions will usually decide it. First, what shape does the text have? Module 2 taught you to let the genre of the passage shape the sermon; a narrative wants to be told, an argument wants to be followed, a psalm wants to be prayed. Second, who will hear it? A congregation that needs teaching and has few Bibles is well served by a clear deductive or plain-style shape; a congregation that has heard every proposition and believes none of them may need to be walked to the point. Third, how will it be heard? Through an interpreter, in a second language, outdoors with traffic, or to people who will repeat it in the village tomorrow, a sermon needs a shape that can be announced and remembered. Fourth, what is the main idea and the need it meets? The shape is chosen to carry that idea to that need; if the shape does not serve it, change the shape.

No shape is neutral, and none is commanded. Chrysostom preached through whole books verse by verse, as a running commentary that turned to exhortation at the end, and the shape served a congregation in Antioch and Constantinople well. The sermons in Acts are ordered and differently ordered. What 1 Corinthians 14 requires is that whatever you choose be chosen, and chosen for the hearers' sake.

By now you have a passage, a main idea stated as subject and complement, and a need the text meets, from modules 2 and 3. You also have a sense of how the passage points to Christ and what it asks of the hearers. Today you give it a skeleton. Take one clean page in your notebook, or a new note on your phone, and work through these steps in order.

Step one. At the top of the page write three lines: the passage reference, the main idea in one sentence, and the need it meets in one sentence. Everything below must serve those three lines. If at any point you find yourself writing something that does not, cross it out, however good it is.

Step two. Read the passage aloud twice, slowly. Then write down its own movement in three to six short lines, in the order the passage itself goes. A narrative will give you scenes; an argument will give you steps, often marked by words such as therefore, for and but; a psalm will give you turns, such as the turn from complaint to trust. This is the passage's shape, and it is the first candidate for your sermon's shape.

Step three. Choose one of the four shapes in this lesson, deductive, inductive, plot or plain style, and write one sentence saying why it fits this text and these hearers. If you are not sure, choose deductive for a teaching text and a plot for a story; both are honest starting points.

Step four. Write the skeleton: three to five moves, each one a full sentence, not a heading. Main point two: grace is a person, is a heading. The father runs to the son before the son has finished his speech, so grace reaches us before our repentance is complete, is a move. Full sentences force you to say what you mean, and they are what an interpreter needs. Under each move, write the verses it comes from.

Step five. Write a one-line transition between each move, the sentence that will carry the hearer from one to the next: so if that is true, what about the son who stayed home? Transitions are where listeners get lost, and a sermon heard once needs them more than an essay does.

Step six. Test the skeleton out loud. Stand up and preach the whole sermon in sixty seconds using only the skeleton. If you cannot, the shape is not yet clear to you, and it will not be clear to anyone else. Then ask one honest question: could a farmer, a market trader or a teenager in your church, hearing this once and without a Bible open, tell you afterwards what the main idea was? If not, simplify the shape before you add anything to it. Keep this page. The next four lessons add an introduction and a conclusion, illustrations, language written for the ear, and a test against the people who will hear it, and by the end of the module it will be a full outline.