Poems and songs in fiction

Call this a crossover post. It’s about writing, but it concerns verse, whether poetry or song, so I think it fits on Fridays.

Occasionally, especially in fantasy literature, there may be need for a song or poem. Some authors will insert the lyrics of songs (real or fictional) into the narrative as a way to set the scene or build the world, while others might instead use poetic verse at the beginning of a chapter or elsewhere. Either way, that’s where we’re going today.

Danger zone

First off, let me just say this: if you don’t know what you’re doing when it comes to poetry and songwriting, you probably shouldn’t bother. A lot of readers simply don’t care, and they’ll skip over the verse portions to get back to the text. I mean, how many people read all the songs in Lord of the Rings? (And this whole post is basically Tolkien’s fault, if you think about it.)

But let’s say you want to give it a shot. Okay. Good for you. Why? That’s not me being facetious; it’s a serious question. You need to ask yourself why you want to do this in the first place. Is it to make the world seem more “real”? To inject a bit of humor? To show off your skills? Your reasoning will play a large role in determining the most appropriate time and place for inserting verse into narrative.

Timing is everything. If you are putting excerpts of in-story verse into your creation, then you want to do it at the right time. Probably not near the climax, for example. Instead, early on might be better, when there’s not as much action going on. Call it the story’s downtime. (Another option is introducing the song or poem early, then using it as a plot point later on. A Song of Ice and Fire, for example, does this to great effect with “The Rains of Castamere”.)

Rock and roll fantasy

Genre also plays a big role, and this extends beyond written stories, into video and audio productions. Fantasy, of course, will tend to have “period” music, which almost always means folksy rhyming couplets sung by bards with lutes. For futuristic science fiction, the trope seems to be a kind of electronic fusion style, with non-Western instruments played over computer-generated beats, rarely with lyrics.

Subverting these expectations, of course, can be genuinely useful, even if only for humor. Imagine, if you will, a space station where the music of choice is country, or a medieval-style culture whose favorite style is battle rap. (That could even be the basis of a magic system, come to think of it.) Plenty of opportunity both for inserting lyrics into the narrative and creating something unique.

On a more serious note, the style does play a role. Tolkien’s endless songs set the tone of Lord of the Rings and The Hobbit. Drinking songs, for instance, illustrate the camaraderie and shared culture of the people singing them—maybe they’ll one day turn into anthems. Epic poetry might not work as well as simple prose for an infodump, but it does twice the work: not only do you learn about the world, but you get a little more immersed in it.

Not another love song

Song and verse can also be a way of learning about the characters. This might work better in modern-day settings, where you can simply use real-world song titles, or future eras of even heavier musical herding. But even the Middle Ages had its songwriters, right?

Romance is the obvious reasoning for this. Honestly, I have to admit that I can’t come up with too many more. Still, it’s an interesting avenue, and it gives you a reason to showcase less-than-stellar songwriting. After all, if the character isn’t that good, then what better way to illustrate that fact than by showing how bad he really is?

Dreams and songs

All this mostly applies to songs, but it can work for poems, too. In a purely visual medium, they’re practically the same thing. And it’s a lot easier to say that pitiful attempt at rhyme you wrote to impress a girl was just a poem, not something meant to be set to music.

Poems, however, can also work somewhere else in a story, somewhere songs can’t: in literature. Here, I’m talking about “internal” literature, that of the setting’s culture. In fantasy settings, where literacy is presumed to be rarer than today, poetry might be more common than prose as a window into the past. (Think Beowulf, for instance, then convert that to the setting’s assumptions.) A well-timed verse could even hold a clue to a mystery, or a secret.

Whatever you choose, whatever purpose you give your story verses, the best advice I can give is simple: don’t overdo it. And if you’re not sure what you’re doing, that might be further simplified to: don’t do it. It takes a skilled hand to toss something like that into a story and not make it look out of place. Give it a reason to exist. Make it worth reading. Don’t throw 15 stanzas into a scene because Tolkien did it. (Even I skipped that one!)

Used judiciously, verse can work. It can give a story a depth that is hard to achieve any other way. The hardest part is knowing how to use it.

Meter, language, and culture

In verse, one of the most important parts of the work is its rhythm, the pattern of sounds and syllables and the way they “flow” together. Aesthetically, we perceive a poetic or musical utterance that is out of rhythm as dissonant, like a dancer tripping over his own feet. If you have any ear for music at all, you’ve probably noticed this. Think of when somebody you know tries to fit a bunch of words into the pattern of a familiar song, but he gets them all wrong. The line doesn’t scan, and it sounds awful.

This is one expression of the idea of meter. Now, if you remember English classes in high school, you probably recall those endless lessons about classical meter. You know, iambic pentameter, and…nobody older than sixteen can ever name any of the others. I know I can’t without looking them up.

But meter isn’t just important for reading Shakespeare and pleasing English teachers. (Personally, I didn’t care for either of those activities.) It’s actually an integral part of language as a whole, and verse in particular.

Unit of measurement

Now, I’m not going into a long digression about the kinds of meter and all that. I just don’t want to, and it doesn’t really help the discussion. Instead, this post will focus on the generalities of meter and how they relate to a language and a culture. That, I believe, is more enlightening than trying to memorize the difference between an iamb and a trochee or whatever.

Meter is the rhythm of language, as I said, and each language has its own, even before you start talking about poetry. Some languages, for instance, have a fixed stress position; according to WALS Chapter 14, about 56% are like this. Of these, most have a fixed stress on the penultimate (second to last) or initial syllable of a word, while the final syllable is a fairly distant 3rd place.

Now, if a language requires stress to be in a certain place, then that’s something verse has to work around. It forces a kind of “template” on meter. Initial and penultimate stress, for instance, both imply a stressed-unstressed (trochaic) pattern, while final stress gives the opposite (iambic) pattern. That doesn’t always have to be true, of course, but there’s an obvious pressure to do things that way.

A lot of other languages, however, are like English. Their stress can move around. In English, for example, affixes can and do affect stress: compare nation (stress on first syllable) to nationality (stress on the third from the last).

In such languages, there’s no obvious “default” meter. Now, there can be other factors determining which rhythms are acceptable, but any rules are more along the lines of convention than law. Thus, English has a rich history of different meter types, and these have come to be associated with different authors, styles, and even moods.

If you’re making your own language (and culture), then this knowledge gives you an edge on pursuing more creative aspects of your work. Say you’ve already decided that your conlang has a fixed penultimate stress. Well, the speakers will naturally gravitate towards poetic meters which go well with that.

Length of a meter

Now, that’s only half the problem. For a lot of verse, not only are you worrying about the rhythmic pattern of “heavy” and “light” syllables, but also how many there are. That’s where the “pentameter” part of “iambic pentameter” comes in: one line of that meter is five feet long, each of which is an iamb. (One meter, of course, is about 3.28 feet, but that’s neither here nor there. And so ends your daily dose of puns.)

English, thanks to its startling propensity for muttering, shortening, and just plain ignoring unstressed syllables, gets a lot of leeway in this department, but there’s only so far you can really go. Really, the length of a line is going to be measured in stresses, however a language counts them, and a verse in a consistent meter will have lines containing generally the same number of those stresses. (That’s where the “poor scanning” thing comes in: the words you’re trying to fit into the rhythm don’t line up with the expected stress pattern, so it sounds forced and unnatural.)

In languages without overt stress, the principle is roughly the same. Pitch-accent languages may be able to substitute high pitch or pitch drops for stress. Tonal languages are a bit harder, but there will be patterns there, too. And we can even talk about line length separate from meter, as in Japanese haiku, with their fixed structure of three lines, containing, in order, five, seven, and five morae. (A mora is like a syllable, except that, e.g., long vowels count as two morae instead of one. Some languages use these as the basic constituent of meter, rather than syllables.)

A musical note

Meter carries over from poetry into song, too. After all, they both come from the same foundations. In Western music, for example, 4/4 time is common, and the usual pattern is a heavier “stress” on the first and third beats of a measure. If you consider each beat to be a syllable (or mora, if your language is doing that), then that maps quite naturally onto a trochaic rhythm. Put the emphasis on beats two and four instead, and you’ve got something iambic.

There are a few easy ways you can use that to your advantage. One, you can assume that the off beats are unstressed syllables, making each measure two metrical feet, if we’re using 4/4 time. (Something like, say, 6/4 would be different, but the math should be easy.) Another option is to take one beat as the primary stress, turning each measure into a single foot.

Once you get outside the Western musical tradition, this pattern tends to break down, but not the basic idea of it. Beats match up with stresses, pitch changes, or whatever a language uses. That’s not even a cultural thing, really. It’s pretty much hardwired into our brains. The cultural aspect is all in deciding what patterns we want to hear.


Writing verse is hard. I can’t do it—that’s why I’m writing about it instead. Setting it to music is even more difficult. But the basics aren’t impossible to understand. Verse, whether in poetry or song, comes down to following (and knowing when not to follow) simple patterns in language. Meter is one of those patterns, an imprint upon our poetry, a kind of paint-by-numbers outline. If you’ve ever tapped your foot to a song or been entranced by a chant or speech, you know how powerful those patterns can be.

On alliteration and assonance

When most people think about verse, they tend to think of rhyme first and foremost. Understandable, since that’s the defining quality of so much poetry. But there’s a whole other side of the word to explore, a front-end counterpart to the back-end rhyme.


Alliteration is the repetition of a sound at the beginning of a word, a mirror image to rhyming. It’s not quite as obvious these days, as rhyme and rhythm have won our hearts and minds, but it has an illustrious history. Some of the earliest Anglo-Saxon verse was composed using alliteration, as were epics from around the Western world. Classics such as “The Raven” and “Rime of the Ancient Mariner” have sections of alliterative verse, as do children’s nursery rhymes. Peter Piper probably needed something to catch the spit from all those P sounds. And who can forget all those old cartoons with hilariously alliterative newspaper headlines? Those were a thing, and they still are in places.

Echoes of alliteration are all around us. Like rhyme, the reason borders on the psychological. In oration, the beginning of the word tends to be more forceful than the end, more evocative. So punctuating your point with purpose (see what I did there?) helps to get your message stuck in the minds of your listeners. They can “latch on” to the repetition. Wikipedia’s article on alliteration uses King’s “I Have a Dream” speech as an example: “not by the color of their skin but by the content of their character.” Notice how the hard K sounds beginning each of the “core” words grab your attention.

To be alliterative, you don’t have to use the same sound at the beginning of every word. The rules of English simply can’t accommodate that. (Newspapers cheated by removing extraneous words such as “a” and “the”.) It’s the content words that are most important, especially the adjectives and nouns. However, alliteration tends to be stricter than rhyme in what’s considered the “same” sound. Voicing differences change the quality of the sound, so they’re out. Clusters are in the same boat. On the other hand, sometimes an unstressed syllable (like un- or a-) can be ignored for the purposes of alliteration.


Alliteration is concerned with consonant sounds. (I did it again!) Assonance is different; it’s all about the vowels. What’s more, it’s not limited to the beginnings of words. Rather, it’s a vowel sound repeated throughout a phrase or line of verse. Vowel rhyming can be considered a form of assonance, but it’s so much more than that.

Assonance pops up everywhere there are vowels, which means everywhere. It’s very well suited to small utterances, such as a single line of a song or a proverb. As with alliteration, it’s not an absolute requirement for all the vowels to be the same, but those that are need to be essentially identical. And it’s the content words that are most important. Schwas, ineffectual as they are, don’t even appear on the radar; a and the aren’t going to mess up assonance. But any other vowel is fair game, in English or whatever language you’re using.

In conlangs

Alliteration and assonance are perfectly usable in any context, and they can be made to fit any language. They might not be quite as permissive as rhyme, but they can have a greater lyrical effect when used properly. (And sparingly. Don’t overdo it.)

These literary devices work best in languages with patterns of stress. That stress can be fixed, but that narrows your options slightly. Inflectional languages with fixed final stress are probably the worst for alliteration, while initial stress gives the most “punch”. For assonance, it’s not so vital, but you want to make sure your vowels aren’t being forced to a fit a pattern.

Both alliteration and assonance are easiest to accomplish in languages with smaller phonemic inventories. That shouldn’t be surprising. It’s far less work to find two words that both begin with a P if your only other options are B, D, K, and S. With these smaller sound sets (are you kidding me?), you can even create more complex styles of alliterative verse. Imagine a CV-type language with interwoven alliteration patterns, where the first and third words of a line start with one sound, while the second and fourth begin with a different one.

The other end of the spectrum holds English and most European languages, and it’s less amenable. You need lots of words, or you’ll have to get some help from stress and syllabics. That’s how we can have alliterative English: by ignoring those tiny, unstressed prefixes that pop up everywhere. It’s possible to make it work, but you have to try harder. But trying is what this is all about.

Rhyme in conlangs

I’ve been doing posts on here for a year now, and there’s been one glaring omission. The name of the place is “Prose Poetry Code”, but we have yet to see any actual posts about poetry. So let’s fix that by looking at how we can add a poetic touch to a constructed language by using that most famous of poetic devices: rhyme.

You most likely already know what rhyme is, so we can skip the generalities. It’s all around us, in our songs, in our video game mysteries, in our nursery rhymes and limericks and everywhere else you look. Do we need a definition?

The sound of a rhyme

From a linguistic perspective, we probably do. Yes, it’s easy to point at two words (“sing” and “thing”, for instance) and say that they rhyme. But where’s the boundary? Rhyme is a similarity in the final sounds of words or syllables, but we have to define how close these sounds must be before they’re considered to rhyme. Do “sing” and “seen” rhyme? The English phonemes /n/ and /ŋ/ aren’t too far apart, as any dialectal speech illustrates.

So there’s your first “dimension” to rhyme. Clearly, there are limits, but they can be fluid. (Poetry is all about breaking the rules, isn’t it?) Most languages would allow inexact rhymes, as long as there’s enough of a connection between the sounds, but how much is necessary will depend on the language and its culture. You can go where you want on this, but a good starting point is the following set of guidelines:

  1. A sound always rhymes with itself. (This one’s obvious, but there’s always allophonic variation to worry about.)
  2. Two consonants rhyme if they differ only in voice. (You can add aspiration or palatalization here, if that’s appropriate for your conlang.)
  3. Two vowels rhyme if they differ only in length. (Again, if this is a valid distinction.)
  4. A diphthong can rhyme with its primary vocalic component. (In other words, /ei/ can rhyme with /e/ but not /i/.)
  5. Nasal consonants rhyme with any other nasal. (This is a generalization of the explanation above.)

This isn’t perfect, and it’s especially not intended to be a list of commandments from on high. Feel free to tweak a bit to give your conlang its own flavor. And if you’re using an odder phonology, look for the contrasts that make it distinct.

Where to begin, where to end

Another thing to think about is how much of a syllable is considered for a rhyme. In Chinese, for instance, it’s basically everything but an initial consonant. English, with its complicated phonological and syllabic systems, allows more freedom. Clusters can count as simplified consonants or stand on their own. Reduced or unstressed vowels can be omitted, as can consonants: “’twas”, “o’er”.

Once again, this is creativity at work, so I can’t tell you what to do. It’s your decision. Really, the only “hard” rule here is that the initial part of a syllable rarely, if ever, has to match for a rhyme. Everything else is up for grabs.

With longer words, it’s the same way, but this is a case where different languages can do things differently. Stress patterns can play a role, and so can the grammar itself. To take one example, Esperanto’s system for marking word class by means of a change in final vowels is interesting from a mechanical point of view, but it’s awful for rhyming poetry. One could argue that all nouns rhyme, which is…suboptimal. (A better option there might be to require rhyming of the penultimate syllable, since Esperanto always stresses it, ignoring the “marker” vowel altogether.)

Going in a different direction, it’s easy to see that a language with CV syllables—think something in the Polynesian family here—will tend to have very long words. With a small set of phonemes, there aren’t too many combinations, and that could lead to too much rhyming. Maybe a language like that requires multiple matching syllables, but it might just discard rhyme as a poetic device instead.

And then there’s tone. I don’t speak a tonal language, so I’ve got little to go on here, but I can see a couple of ways this plays out. Either tone is ignored for rhyming, in which case you have nothing to worry about, or it’s important. If that’s true, then you have to work out which tones are allowed to rhyme. For “level” tones (high, low, medium), you could say that they have to be within one “step”. “Contour” tones may have to end at roughly the same pitch. Why the end, you may ask? Because rhyming is inherently tied to the ends of syllables.

Different strokes

As rhyme is tied to the spoken form of a language, it will be affected by the different ways that language is spoken—in other words, dialects.

One good example of this in English is “marry”. Does it rhyme with “tarry”? Most people would say so. What about “gory”? Probably not. “Berry”? Ah, there you might have a problem. Some dialects merge the vowels in “marry” and “merry”, while most other (American) ones don’t.

Rhyming verse is made to be spoken, recited, chanted, or sung, not merely read, so this is not a theoretical problem. It’s important for anyone writing in a natural language with any significant dialectal variation. Nor is it limited to slight changes in vowel quality. What about English /r/? It disappears at the end of words in England, but not America…at least in speech. Except for country music, most singers tend to drop the R because it sounds better, which has the side effect of creating more opportunities to rhyme.

Of course, for a conlang, you probably don’t have to think about dialects unless you’re specifically creating them. Still, it might be useful to think about for more “hardcore” worldbuilding.

Sing a song

Rhyming isn’t everything in poetry. It’s not even the most important part, and many types of verse get by just fine without it. But I started with it for two reasons: it’s the easiest to explain, and it’s the simplest to build into your conlangs. In fact, you’ve probably already got it, if you look close enough. (If you’re using random generation to create your words, however, you may not have enough similar words to get good rhymes. That’s where author fiat has to come in. Get in there and make them.)

If you don’t care for rhymes, that’s not a problem. Others do, and if you’re making a language for other people to speak, such as an auxlang, you have to be prepared for it. Poetry is all about wordplay, and creativity is an unstoppable force. Whether song or spoken word, people will find ways to make things work.