Language is a funny thing. We’re all expert users of it, but quite what language is and how it works remains a mystery to most of us.
That’s why an article like Phil Campbell’s ‘Deadly, dull, and boring’ is such a godsend.1 God’s word is a preacher’s core business, and therefore language is the preacher’s most basic tool. Anything that can help us understand a little more about how language functions and how we can use it better—particularly something that highlights the differences between spoken and written language so that we can preach better—is a great thing!
Of course, since this is a response article, you know there’s a ‘but’ coming! So before I burst dent the bubble, I want to stress that I did find the article very stimulating, and several of the points were a good rebuke to my own lazy preaching habits. Tip #10 (“work towards your key text”) is one in particular that I’m now more conscious of in my sermon preparation.2
But (and here it is!), I also found a number of aspects of the article, from the point of view of understanding language, a little disappointing.
Linguistics is the study of language in all its forms—what language is, how it works, how we use it to achieve our social goals, etc. This means that the difference between spoken and written language is one topic about which linguists have a lot to say. It also means that practitioners like Rudolph Flesch may have helpful insights, but may also get some things very wrong.
What do linguists know, then, about the difference between spoken and written language? Here are a few observations, specifically relating to some of the claims in ‘Deadly, dull, and boring’.
Tip #4 in the article is to “use shorter sentences”, implying that in spoken communication listeners can’t follow the thread if it’s presented in sentences that are too long. On the basis of the Reader’s Digest style guide, it is suggested that sermon sentences be kept to 17 to 20 words.3
There are a number of problems here. The first and most obvious being that the Reader’s Digest style guide relates to written language, not spoken4 This makes it singularly unhelpful for training us in ‘natural scripting’. However, more significant is what we know about the way ideas are packaged in spoken language, and how this differs from written language.
The article claims “we package ideas in sentences”.5 This is not quite correct. We actually package ideas in clauses, and we connect ideas to each other in clause complexes.6 This is not being picky; it’s very important for understanding how writing and speaking are different. The fact is, a sentence is a rhetorical unit (a matter of communicative style), not a grammatical unit (a matter of meaning). Moreover, a sentence only exists in written language; there is no such thing as a sentence in speech.
How you decide to punctuate your text when you write is, to a certain extent, a matter of personal preference. I love using semi-colons; my wife does not. She prefers to separate her lists of ideas with full stops. However, if we were to read our differently-punctuated texts aloud, you would have no way of telling which was which; you don’t hear the punctuation. What you hear and unpack are the words and the grammatical relationships between them—clauses and clause complexes.
But here’s the interesting bit: spoken language is actually characterized by much longer clause complexes than written language. To see why, we need to understand a bit more about clauses and how they are used.
Linguists divide clauses into three types: independent clauses, dependent clauses, and embedded clauses.
Independent clauses are (as the name suggests) clauses that can stand-alone:
1. The author devised a compelling example.7
Dependent clauses are clauses that have been connected to an independent clause in such a way as to explain or amplify it:
2. The author devised a compelling example, using all his wit and imagination.8
Embedded clauses are used as a part of another clause, to explain or describe one element within it:9
3. The author devised a compelling example consisting of three separate parts.
Understanding the difference between dependent and embedded clauses can be tricky, but it is important for understanding how we communicate, and particularly how we speak. The difference is that a dependent clause modifies the whole of another clause, while an embedded clause only modifies one part of it.
One simple test to work out which is which is to see if you can move the dependent/embedded clause to another position in the sentence. If it still makes sense, it’s dependent; if it doesn’t make sense, or if the meaning noticeably changes, it’s embedded:
4. The author, using all his wit and imagination, devised a compelling example (dependent).
5. The author, consisting of three separate parts, devised a compelling example (embedded).
If your eyes have begun to glaze over, rest assured that I’ve taken you on that brief linguistic journey for a reason. If you compare examples 2 and 4 above, you’ll notice that one is more ‘written’ in style (4) and the other is more ‘spoken’ (2). What this shows is that one of the differences between spoken and written language is the way our clauses are combined.
In natural speech, we less frequently interrupt an independent clause by inserting a dependent clause into the middle of it, as in example 4. Constant interruption like that makes it more difficult to follow the flow of ideas. That is one reason why clause complexes in written language tend to be shorter: when we write, we like to arrange our clauses artistically, using maximum variety from one clause complex to the next. This playing around with the clause order makes the ideas more difficult to follow, and so forces us to limit the length of our clause complexes.
In speech, however, there is almost no limit to the number of independent and dependent clauses we can link together, as long as we put them into linear, non-overlapping chains.
We need to be more careful with our embedded clauses when writing in a spoken style. These are the things that can add great descriptive depth to a clause, but can also muddy the communicative waters:
The author sitting at his desk and gazing distractedly out the window while systematically chewing the ends off all his pencils devised a compelling example bearing all the hallmarks of clear and efficient expression.
Notice, by the way, that I’ve managed all of that descriptive embellishment without using a single ‘who’, ‘that’ or ‘which’. The first descriptor could have begun “who was sitting at his desk”, however in speech we often discard ‘who’, ‘that’ and ‘which’ while keeping the lengthy descriptors. The trick is to identify and deal appropriately with them. We could just delete them. However, if we’re preaching on narrative (and therefore retelling a story), it would suck the life and colour out of our sermons if we lost them all. Here are two other possible solutions:
- Repeat the thing described after its descriptor: The author sitting at his desk and gazing distractedly out the window while systematically chewing the ends off all his pencils—that author devised a compelling example.
- Turn it into two or more connected clauses: The author was sitting at his desk and gazing distractedly out the window while systematically chewing the ends off all his pencils, and while he was in this state of distraction he devised a compelling example—an example that bore all the hall-marks of clear and efficient expression.
These examples are, of course, very silly! The serious point is that it is a mistake to think that ‘short’ equals spoken and ‘long’ equals written: it is the other way around. However, when it comes to spoken language, it is the kind of long that is the key: long, linear strings of independent and dependent clauses, with any embedded clauses carefully managed.
This leads to a couple of other tips I have for writing in a spoken language style:
- Work out a punctuation system that gets you away from full stops and that preserves the lengthy clause chains that are characteristic of speech. My sermon text is liberally sprinkled with dashes (—), ellipses (…), and semi-colons.
- Use simple-sounding conjunctions, begin your clauses with them, and use them repetitively (“and… and then… and then…”).
All that being said, though, sentence length in spoken language is one of the more minor issues we need to understand if we’re to prepare engaging sermons.
If you have a child attending an Australian school (and possibly elsewhere), you will have noticed some changes to the curriculum since you were in school. One area that is new is how children are taught to understand and produce a whole variety of ‘texts’.10
The curriculum now reflects some of the great gains linguistics has made in understanding the many different ‘languages’ we all use every day. Note the use of the plural—languages. Every different context we find ourselves in requires a different kind of linguistic expression, and this goes far beyond the spoken/written distinction. In fact, when we speak about the difference between spoken and written language, as if those are the only two options, we over-simplify the matter so much as to be misleading.
The way you speak to your neighbour on the bus is very different from the way you speak when giving a business presentation, which is very different to the way you speak as coach of the football team at half-time. To be competent language users (and the vast majority of us are), we need to be skilled in all of these different language types, and many more besides. Linguists (and the school curriculum) call these different languages we need to master ‘text types’.
What unifies texts into a single text type is not their genre (sci-fi or non-fiction) or their mode (written or spoken), but their shared social purpose. Language is, after all, a tool of social interaction; we use language to achieve certain social ends. Sometimes we want to describe for our listeners something that happened; other times we want to persuade them of a point of view; at still others we want to teach them how to do something. What linguists have discovered is that each of these different social goals is written into the language we use. You know what my social goal is because you can recognize it from the language choices I make—both the words and the structure of the text.
Just as there are clear and identifiable differences between a fairy tale and a debate and a recipe, so there are clear and identifiable commonalities between texts of the same type. Any text that fits the description ‘a story to entertain’, for example, is a narrative text type regardless of its language, length, medium or literary style.
This is important for our preaching, because we need to know that what we’re doing in a sermon (our social goal) is very different from what we’re doing when talking with the person next to us on the bus. In other words, the language choices of a sermon text type are very different from those of a bus conversation text type. The school curriculum would call a sermon a ‘persuasive text type’11 and a typical bus conversation a ‘narrative recount’.12
This has a number of significant implications for our thinking about preaching, some of which challenge claims made in ‘Deadly, dull, and boring’.
The words you choose
Since the different social goals we have in using language are written into the language choices we make, the words we choose when preaching should be appropriate to persuasive texts.
For example, the general verbal ‘mood’ of a sermon will be very different from that of a bus conversation. In a sermon we will use many more commands and black-and-white declarations, while a bus conversation will contain a higher proportion of loosely-held opinions.
However, it goes beyond just our verbs. All our word choices should reflect the fact that a sermon is a persuasion. This means that tip #3 (“choose the shortest, most ordinary words you can”) is not entirely helpful.13 To the extent that it helps us realize that a sermon is not an academic dissertation, it’s an excellent reminder. But if it teaches us to use exactly the same words in our sermons as we would in a standard conversation, then it has led us astray.
Of course, there will be significant overlap between the vocabulary of a sermon and a bus conversation—we’re still speaking English after all! But our vocabulary choices do need to fit with the social purpose of a sermon. In particular, since a sermon is a persuasion, we should choose at least some words solely for their persuasive quality. That means that a sermon will contain some longer, out of the ordinary words.
Even the ancients knew that persuasion is most effective when the text not only expresses true ideas, but when it sounds true. The most successful sermons are judiciously seasoned with words chosen for their evocative power, or their rhythm, or even their alliterative qualities.14 These are words that can paint pictures and draw out emotions for your hearers that “the shortest, most ordinary words” can’t.
Sometimes it helps to use two words or expressions to make the one point: an evocative, seasoning one followed by a plain one which means the same thing.
Below is just one example from a recent sermon of mine where I worked hard to get my language choices to enhance the persuasive meaning of the point I was making. This paragraph relates to Revelation 18, where John gives us an anticipatory glimpse into how the people of the world will react when they’re overtaken by God’s judgement. I wanted to highlight the banality of their response and the tragic superficiality of consumerism (e.g. Rev 18:17), so I paraphrased the people in the plainest language I could find, and I surrounded that with much more eloquent language, choosing emotionally and visually evocative words and paying attention to the rhythmic balance of each phrase. I was trying to communicate my main point by both the propositions contained in the text and the linguistic contrast between the plainness of the response and the erudition of the surrounding text:
You can just hear their absolute desolation… Their insatiable lust for wealth… They’ve been so bewitched by its lure that at the final judgement of ‘the great city that rules over the kings of the earth’, all they’ll be able to think is, “What a waste of money!” They’ll be left standing in the wasteland of the aftermath of God’s judgement weeping over lost trinkets.
Of course, I don’t do this for my whole sermon; I pick and choose the key moments. But I think my hearers would have been less engaged if I had said instead:
These people are very disappointed, because when God’s final judgement does come, all they’ll be able to think about is the money they lost… and so they’ll have no sense at all of the huge thing that’s just happened.
The tone you use
That being said, as with the issue of sentence length, our vocabulary choices are also a relatively minor factor in the quest for engaging preaching. For starters, the way you use your voice is far more significant.
Picture the best and the not-so-best Bible readers in your church: why does one captivate you, while the other puts you to sleep? The sentence length and vocabulary choice is exactly the same in both cases! The difference, of course, is in the way they use their voices—the variations they create in pace, pitch and volume (what linguists call ‘prosody’).
The fact is, if you sound boring, you will be boring. Learn to speak in public as though you not only believe what you’re saying, but you’re captivated by it. If you can do that well (i.e. as is appropriate for public persuasion, rather than a private conversation), then no-one will notice the odd item of lexical inscrutability.
The problem of mismatched expectations
Even more significant still, however, is the problem of mismatched expectations.
Here is where I need to tread a little carefully, because some may want to charge me with blaming the listener. In fact, if there is any blame to be laid at this point, it falls far more on the side of the preacher. But whether or not I am successful in avoiding the charge, there is a communicative reality we can’t shy away from.
If the language choices we make are shaped by the social purposes we have, then what is crucial in any effective communication is that both speaker and hearer share the same understanding as to the social purpose of the event. If you asked me to preach a sermon, but I told you a story instead, you would be rightly disappointed. In a different context, my story could have been the most brilliant story ever told in the history of storytelling, but that makes no difference: entertaining you wasn’t the purpose I was asked to fulfil.
Now this problem can occur in the other direction as well: you turn up to my church hoping to be entertained, but I give you a sermon. Again, it could have been the most brilliant sermon ever delivered, but the mismatch in expectations means that you go away frustrated, and the communicative event is a failure.
This ultimately comes down to the question: what is a sermon? I’ve already claimed that a sermon is a persuasive text type. However, that is still too broad a definition to be really useful. To be able to narrow it down more helpfully, we need to understand something of what linguists call ‘the context of situation’.15
This is a catch-all term that unites everything we can possibly say about a communicative event that gives it its unique quality in history. What gave rise to the text? Who are the participants? How many are there? What is their relationship? What is the subject matter? What shared understandings do they have? What is the tenor of discourse (i.e. giving information; requesting information; giving or receiving directions)? What is the mode of the engagement (e.g. spoken or written)?
A sermon is a persuasion, but it is a different kind of persuasion from a political campaign speech or a mother’s attempt to get her ten-year-old to cultivate good personal hygiene habits. All those different persuasive texts will share features in common, even down to structure and the types of words chosen, but there will still be significant differences that need to be taken into account. If the mother preached for 30 minutes on the virtues of regular teeth-cleaning, she would miss her communicative mark for a whole host of reasons: the number of participants (it is one-to-one, not one-to-many), their relationship (mother to child, not pastor to congregation), the subject matter (practical life skills versus the eternal word of God), etc.
Through experience, we have learned to instinctively recognize the difference between all these contexts of situation and the different texts they produce, and as skilled language users we are able to create most or all of those different texts as the situation demands. However, sometimes we need instruction in what is appropriate—both as creators and as receivers of those texts.
This is not to blame the listener; it’s to say that we preachers have a double responsibility: not just to prepare engaging sermons, but also to train our congregations in receiving sermons well. This begins with teaching them what to expect in a sermon, but goes beyond that.
For example, how do you encourage the church member who was kept up for half the night by a crying baby, and so wasn’t able to concentrate through the whole sermon the next day? Was the sermon a waste of time for them? Should we only ever preach five-minute sermons because there might be stressed and tired parents in our congregations?
This (extreme) example shows up what is a reality in all preaching: the nature of the activity (one-to-many) means that if our benchmark for effective preaching is that everyone present hears and processes every single thing the preacher says, then we’re aiming at the wrong target.
And this becomes even clearer when we think more carefully about the participants involved. How often have we taken something away from a sermon that the preacher hadn’t even thought of? The subject-matter of our preaching (God’s word) introduces a third participant into the conversation—one who speaks through the preacher’s words, yet has the power to make another (complementary) message out of them because he has the power to direct our very thought processes. He is not just external to us, as the preacher is; he is internal to us, by the working of his Spirit.
This is perhaps the biggest concept both preachers and hearers need to master if preaching is to be effective: the first and loudest voice in the conversation is God’s. Of course, this sounds so trite; surely it goes without saying? Yet when someone walks away disappointed from a sermon, it is not always down to the preacher’s failure to communicate brilliantly. They wanted something that a sermon is not intended to give them: they wanted their ears to be tickled rather than their hearts to be recreated. This doesn’t necessarily come from an ungodly mind; they just don’t know better. They need to be trained in how to listen.
One tool I’ve found particularly helpful in this is the excellent booklet Listen Up! by Christopher Ash. I’ve even prepared a complementary set of Bible studies for our small groups to do as an excuse to get them all to read the booklet.
But beyond using a printed resource like this, there are countless informal opportunities preachers have almost every week for training our hearers. I take any complaints or criticisms I receive after the service as an opportunity for double reflection: first and foremost, for me to be able to revisit my preaching in order to improve it; but secondly, and almost as importantly, as an opportunity to help my brother or sister complainer to reflect on their own sermon listening skills.
The question of ideal sermon length arises at this point. Why does Phil Campbell feel that 23 minutes is the maximum sermon length that his congregation can deal with?16 Because he is incapable of engaging them for any longer than that? I doubt it.
In other parts of the world, sermons of double that length and more are the norm. Does this mean that Australians have lousy concentration spans? Are preachers in other places much better orators than us? Are Australians perhaps worse complainers? No!
Speakers at Australian Christian conferences always preach for longer than we are used to in church, yet no-one complains. And before anyone objects by pointing out that they’re more skilled at preaching than us mere mortals: they preach longer at conferences than they do in their own churches. Why? It’s all about expectations.
What is a sermon, and what is the purpose of a sermon in church? This is a much bigger topic for conversation than this response piece can cover, yet it has to be the first thing we discuss. If we aren’t clear on what a sermon is, and if our congregations aren’t clear on what a sermon is, then tweaking our sermon texts will help very little.
Paul saved Eutychus for another several hours of hard work engaging with God in his word. It seems that we’d rather save Eutychus from all that. By all means, let’s try to make our preaching as engaging as possible. But let’s not neglect the needs of our hearers in the attempt.
- P Campbell, ‘Deadly, dull, and boring’, The Briefing #405, May-June 2013. The article is an edited extract from Matthias Media’s new book on preaching, Saving Eutychus. I must declare up front that I haven’t yet read the book, so everything I have to say here relates to the Briefing article alone. ↩
- Ibid., p. 40. ↩
- Ibid., p. 36. ↩
- As do the books of Rudolph Flesch. ↩
- Ibid. ↩
- A clause is a unit of meaning containing one governing verbal process, and a clause complex is any number of clauses which are grammatically joined together (most commonly with conjunctions). ↩
- This is what was traditionally called a principle or main clause. ↩
- The traditional subordinate clause accounts for the majority of dependent clauses. ↩
- This is often by means of a relative clause, although not always. ↩
- A text is any discrete communication. It can be something as brief as a ‘stop’ sign, or something as long as the Summa Theologica. It can be written or spoken, or even audio-visual, like a movie or TV weather report. ↩
- A persuasive text type (persuasion for short) is an umbrella-term for a number of different methods we have for bringing others to share our point of view on an issue. School children are taught about some of these (e.g. expositions and discussions), however—strange to say!—they’re not taught the features of a sermon text type. ↩
- Describing something that happened or that will happen. ↩
- Ibid. ↩
- Yes, I disagree with the article’s claim on page 36 that “alliteration is no longer considered tasteful”—it depends on what kind of alliteration you have in mind. If you’re thinking of the sermon with three points, all starting with the letter ‘p’, then I’m inclined to agree. However, there are many other types of alliterative language. For example, consider some of the (Australian) slang idioms we use in our most unrefined, everyday speech: ‘banana benders’ (Queenslanders), ‘weekend warrior’ (army reservist), ‘lollipop lady’ (woman in charge of a school crossing), and many others. ↩
- There is a broader context linguists also speak about: ‘the context of culture’, which refers to the assumptions that are held in each cultural context about how people relate and meanings are encoded. A persuasion in Tanzania will look very different from a persuasion in a western cultural context. But since most of us are operating within our own cultural context, this aspect of language is not one we need to spend a lot of time thinking about; we know it intuitively. ↩
- Ibid. ↩