Poetry is the most useless thing every culture has. No society has ever discovered it and then given it up. This piece sorts the evidence by age — rhyme at three, metre as a memory architecture at nine, the naming of feeling at fifteen, chills at forty, and the verse that outlasts almost everything else.
Poetry is the most useless thing every culture has.
It builds nothing, cures nothing, feeds nobody. And yet no society has ever discovered it and then decided to stop. Long before writing, people were arranging words to a beat — and they were doing it for a reason.
This piece is not about which poems are good. It sorts what the research shows poetry actually does inside a human head, and it does different things at different ages.
The first thing a baby hears from its mother is not words. It is the shape of speech.
Fernald (1985) showed that four-month-olds prefer the way adults speak to infants over the way they speak to each other — pitch range stretched, tempo slowed, phrases repeated and drawn out. Which is to say: the most instinctive way a mother speaks to her child is already a kind of verse before it is language.
Lullabies take this one step further. A lullaby conveys nothing — its information content is close to zero. What it conveys is steady rhythm, stable pitch, and safety. The child does not yet know what the words mean, but is learning something more foundational: this language has a beat, and the beat is a safe place.
Here the evidence is at its sharpest, and it surprises most parents.
MacLean, Bryant and Bradley (1987) followed children from the age of three years three months and measured something that looks trivial: how many nursery rhymes the child knew. Then they followed those children into their first school years.
The result: the number of nursery rhymes a three-year-old knew predicted later phonological awareness and reading ability — and the relationship survived controls for IQ and social background. Bryant and colleagues (1990) traced the path in more detail and found the causal chain: rhyme and alliteration knowledge feeds phoneme detection, and phoneme detection feeds reading.
The reason is obvious once seen. To notice that cat and hat rhyme, a child must split the syllable into two parts. That is exactly the operation learning to read requires. Nursery rhyme is phonological-awareness training disguised as a game — and it exists in every culture because every culture needs it.
Goswami (2011) pushes this a layer deeper. Her work on dyslexia indicates that many affected children struggle with tracking the rhythm of speech — specifically the amplitude envelope, where sound swells and fades. In other words, what poetry exaggerates is precisely what some children find hard. This is the basis for rhythm-based interventions.
Before writing, a people’s entire knowledge had to live in heads. The question was: how do you hold thousands of lines without corruption?
The answer is verse.
Albert Lord (1960), continuing Milman Parry’s work, went to the Balkans in the 1930s to hear illiterate singers improvise epics thousands of lines long. He found they were not reciting word for word. They held a template — a system of formulae shaped to fit the metre — and rebuilt the song each time. This is also how the Iliad and Odyssey survived before anyone wrote them down.
David Rubin (1995) analysed the mechanism systematically in his book on memory in oral traditions. His conclusion can be put compactly: every formal constraint in verse is an error-correction device. Metre eliminates words of the wrong length. Rhyme eliminates words with the wrong ending. Concrete imagery holds the place of the idea. When you forget a word, three layers of constraint converge on nearly a single candidate.
This is why children memorise verse more easily than prose of the same length, and why every ancient education system began with learning poetry by heart. Not because the ancients revered literature. Because it was the best storage format available.
We should pause here, because there is a finding any article praising poetry is obliged to report.
McGlone and Tofighbakhsh (2000) gave participants unfamiliar aphorisms, some rhyming and some reworded into non-rhyming paraphrases with identical meaning. For instance, woes unite foes against woes unite enemies.
The result: people rated the rhyming version as a more accurate description of human behaviour. Same content. Only the rhyme differed.
The accepted explanation is processing fluency: statements that are easier to process feel truer, and we mistake that ease for accuracy. Obermeier and colleagues (2013) found the same mechanism in aesthetic response to poetry — metre and rhyme raise both liking and emotional response, with fluency mediating.
This is not a piece of trivia. It is why advertising slogans rhyme, why propaganda rhymes, and why false proverbs are so hard to kill. Poetry is persuasive beyond its truth value — and knowing that is part of reading it honestly.
At fifteen something new happens: the child has feelings for which no word has yet been supplied.
Lieberman and colleagues (2007) ran what is now a classic experiment. Participants viewed strongly emotional faces in a scanner. When they simply looked, the amygdala — the brain’s alarm — fired hard. When they were asked to choose a word for the emotion on the face, amygdala activity dropped, while right ventrolateral prefrontal cortex rose.
The phenomenon is called affect labelling. Torre and Lieberman (2018) argue it functions as a form of implicit emotion regulation: you do not suppress the feeling, you name it, and it subsides.
Now look at poetry in that light. Poetry is the precise naming of feeling, performed by the people best at it. A teenager who reads a line that says exactly what they have been carrying is not merely appreciating literature. They are receiving a label for something that previously had no shape — and along with the label, the news that they are not the first.
On writing poetry, the claim should be sized correctly. Pennebaker (1997) showed that writing about emotional experience benefits mental and physical health. But Frattaroli’s meta-analysis (2006), pooling over two hundred studies, found the effect to be small — real and consistent, but far more modest than the popular retelling. Writing helps. It is not a cure.
There is a phenomenon everyone knows and few assume is measurable: the shiver at the right line arriving.
Wassiliwizky and colleagues (2017) measured it. They played poetry to participants while recording skin conductance, filming the skin for visible goosebumps, and scanning. They found three notable things.
One: poetry produced chills in most participants, with measurable physiological correlates — not imagination.
Two: those responses coincided with activity in the brain’s reward circuitry, including the nucleus accumbens — the same system that responds to music, to good food, to things we want.
Three, and this is the loveliest detail: the emotional peaks fell at the ends of stanzas, exactly where the formal structure closes. Form is not the container the feeling arrives in. Form is where the feeling happens.
Miall and Kuiken (1994) describe the complementary mechanism: foregrounding. When a line breaks ordinary usage — an odd word order, an unlikely comparison — the reader slows, and the slowing drags emotional response behind it. Poetry works by making language just difficult enough that you have to look at it.
An old paradox: if humans avoid suffering, why do we seek out poems about loss, and why do we call them beautiful?
Menninghaus and colleagues (2017) proposed the Distancing-Embracing model. Art does two opposite things at once. It distances: we know this is a poem and not our own funeral, so the sadness does not threaten. And it embraces: precisely because it is safe, we allow ourselves to feel more deeply than we do in life.
The result is a rare state — grief felt fully without being damaged by it. For adults, that may be poetry’s largest gift: a place to rehearse what life will eventually require you to feel for real.
Anyone who has sat with someone in dementia has seen it: a person who no longer knows their child’s name will sing an old song through, or recite a passage learned seventy years earlier.
This is not a coincidence. Cuddy and Duffin (2005) documented that recognition of familiar music is often preserved even where verbal memory has badly failed. Jacobsen and colleagues (2015), in Brain, found part of the anatomical reason: the region associated with long-term musical memory sits in an area comparatively spared in Alzheimer’s disease.
Verse learned in childhood is a near relative of music in exactly this sense — it too has rhythm, it too has pitch contour, it too is stored as a motor sequence rather than a fact. That is why the old lines stay after so much else has gone.
The uncertainty should be stated plainly: the Cochrane review by Woods and colleagues (2018) on reminiscence therapy in dementia concluded that benefits are small and of low certainty. Poetry does not treat dementia, and should not be sold as doing so.
But that is not the claim. The claim is that when language withdraws, the last thing to withdraw is the thing with a beat — which means the poems learned at six will be the last ones standing. Worth choosing carefully.
Every language makes verse the way its structure permits, and Vietnamese does something distinctive.
Vietnamese poetry is counted in tones, not stresses. English verse runs on strong and weak beats; Vietnamese verse runs on the bằng–trắc rule, level tones against oblique ones. The music of a Vietnamese line is therefore bound to the very tone system a child absorbed before saying a full sentence. The metre is the language’s own rhythm, not a coat put over it.
Lục bát is a form Vietnamese speakers know without being taught. Six syllables, then eight, with a rhyme rule that chains the couplets into an unbroken run. Almost any Vietnamese can hear a mis-formed lục bát line without knowing the rule — because they heard it from the cradle.
And there is a book. Truyện Kiều is 3,254 lines of lục bát, and it holds a place few literatures can match: a work that illiterate people knew hundreds of lines of, that gets quoted in ordinary conversation, and that has been used for divination. It is living proof of everything Rubin wrote about oral memory — a text that long could only live in a population’s heads because it is verse.
Under six: nursery rhymes, lullabies, and repetition past the point of adult boredom. Do not explain the meaning. Meaning comes later; rhythm has to come first.
Primary school: memorise, but memorise what is worth carrying — and do not make it a test. What is being built is a store, not a mark.
Adolescence: offer poems that say what they are actually feeling, even when those are not the poems in the syllabus. And do not demand analysis on first meeting — analysis before feeling is the surest way to kill a poem.
Adults: read slowly, read aloud, read again. The chills research locates the effect in sound structure, and sound structure does not surface when you read with the eye.
The old: ask what they know by heart. Then sit and listen. It is often the last door still open.
Back to the opening question. Why does every culture have poetry, and why has none abandoned it?
Because it is not useless. It is useful in ways that never show up on a payslip.
At three it installs the ear for reading. At ten it is the storage format. At fifteen it supplies words for what is happening inside. At forty it is where grief can be felt safely. At eighty it is what remains when much else has gone.
Five different jobs, one form. Which is why humans were arranging words to a beat before they could write, and are still doing it long after every other means of storage arrived.
The same article, told in plain words — for younger readers, or for anyone who wants the point quickly.
Poetry is the most useless thing every culture has.
It builds nothing, cures nothing, feeds nobody. And yet no society has ever discovered it and then decided to stop. Long before writing, people were arranging words to a beat — and they were doing it for a reason.
This piece is not about which poems are good. It sorts what the research shows poetry actually does inside a human head, and it does different things at different ages.
The first thing a baby hears from its mother is not words. It is the shape of speech.
Fernald (1985) showed that four-month-olds prefer the way adults speak to infants over the way they speak to each other — pitch range stretched, tempo slowed, phrases repeated and drawn out. Which is to say: the most instinctive way a mother speaks to her child is already a kind of verse before it is language.
Lullabies take this one step further. A lullaby conveys nothing — its information content is close to zero. What it conveys is steady rhythm, stable pitch, and safety. The child does not yet know what the words mean, but is learning something more foundational: this language has a beat, and the beat is a safe place.
Here the evidence is at its sharpest, and it surprises most parents.
MacLean, Bryant and Bradley (1987) followed children from the age of three years three months and measured something that looks trivial: how many nursery rhymes the child knew. Then they followed those children into their first school years.
The result: the number of nursery rhymes a three-year-old knew predicted later phonological awareness and reading ability — and the relationship survived controls for IQ and social background. Bryant and colleagues (1990) traced the causal chain: rhyme and alliteration knowledge feeds phoneme detection, and phoneme detection feeds reading.
The reason is obvious once seen. To notice that cat and hat rhyme, a child must split the syllable into two parts. That is exactly the operation learning to read requires. Nursery rhyme is phonological-awareness training disguised as a game — and it exists in every culture because every culture needs it.
Goswami (2011) pushes this a layer deeper. Her work on dyslexia indicates that many affected children struggle with tracking the rhythm of speech — specifically the amplitude envelope, where sound swells and fades. In other words, what poetry exaggerates is precisely what some children find hard. This is the basis for rhythm-based interventions.
Before writing, a people's entire knowledge had to live in heads. The question was: how do you hold thousands of lines without corruption?
The answer is verse.
Albert Lord (1960), continuing Milman Parry's work, went to the Balkans in the 1930s to hear illiterate singers improvise epics thousands of lines long. He found they were not reciting word for word. They held a template — a system of formulae shaped to fit the metre — and rebuilt the song each time. This is also how the Iliad and Odyssey survived before anyone wrote them down.
David Rubin (1995) analysed the mechanism systematically. His conclusion can be put compactly: every formal constraint in verse is an error-correction device. Metre eliminates words of the wrong length. Rhyme eliminates words with the wrong ending. Concrete imagery holds the place of the idea. When you forget a word, three layers of constraint converge on nearly a single candidate.
This is why children memorise verse more easily than prose of the same length, and why every ancient education system began with learning poetry by heart. Not because the ancients revered literature. Because it was the best storage format available.
We should pause here, because there is a finding any article praising poetry is obliged to report.
McGlone and Tofighbakhsh (2000) gave participants unfamiliar aphorisms, some rhyming and some reworded into non-rhyming paraphrases with identical meaning. For instance, woes unite foes against woes unite enemies.
The result: people rated the rhyming version as a more accurate description of human behaviour. Same content. Only the rhyme differed.
The accepted explanation is processing fluency: statements that are easier to process feel truer, and we mistake that ease for accuracy. Obermeier and colleagues (2013) found the same mechanism in aesthetic response to poetry — metre and rhyme raise both liking and emotional response, with fluency mediating.
This is not a piece of trivia. It is why advertising slogans rhyme, why propaganda rhymes, and why false proverbs are so hard to kill. Poetry is persuasive beyond its truth value — and knowing that is part of reading it honestly.
At fifteen something new happens: the child has feelings for which no word has yet been supplied.
Lieberman and colleagues (2007) ran what is now a classic experiment. Participants viewed strongly emotional faces in a scanner. When they simply looked, the amygdala — the brain's alarm — fired hard. When they were asked to choose a word for the emotion on the face, amygdala activity dropped, while right ventrolateral prefrontal cortex rose.
The phenomenon is called affect labelling. Torre and Lieberman (2018) argue it functions as a form of implicit emotion regulation: you do not suppress the feeling, you name it, and it subsides.
Now look at poetry in that light. Poetry is the precise naming of feeling, performed by the people best at it. A teenager who reads a line that says exactly what they have been carrying is not merely appreciating literature. They are receiving a label for something that previously had no shape — and along with the label, the news that they are not the first.
On writing poetry, the claim should be sized correctly. Pennebaker (1997) showed that writing about emotional experience benefits mental and physical health. But Frattaroli's meta-analysis (2006), pooling over two hundred studies, found the effect to be small — real and consistent, but far more modest than the popular retelling. Writing helps. It is not a cure.
There is a phenomenon everyone knows and few assume is measurable: the shiver at the right line arriving.
Wassiliwizky and colleagues (2017) measured it. They played poetry to participants while recording skin conductance, filming the skin for visible goosebumps, and scanning. They found three notable things.
One: poetry produced chills in most participants, with measurable physiological correlates — not imagination.
Two: those responses coincided with activity in the brain's reward circuitry, including the nucleus accumbens — the same system that responds to music, to good food, to things we want.
Three, and this is the loveliest detail: the emotional peaks fell at the ends of stanzas, exactly where the formal structure closes. Form is not the container the feeling arrives in. Form is where the feeling happens.
Miall and Kuiken (1994) describe the complementary mechanism: foregrounding. When a line breaks ordinary usage — an odd word order, an unlikely comparison — the reader slows, and the slowing drags emotional response behind it. Poetry works by making language just difficult enough that you have to look at it.
An old paradox: if humans avoid suffering, why do we seek out poems about loss, and why do we call them beautiful?
Menninghaus and colleagues (2017) proposed the Distancing-Embracing model. Art does two opposite things at once. It distances: we know this is a poem and not our own funeral, so the sadness does not threaten. And it embraces: precisely because it is safe, we allow ourselves to feel more deeply than we do in life.
The result is a rare state — grief felt fully without being damaged by it. For adults, that may be poetry's largest gift: a place to rehearse what life will eventually require you to feel for real.
Anyone who has sat with someone in dementia has seen it: a person who no longer knows their child's name will sing an old song through, or recite a passage learned seventy years earlier.
This is not a coincidence. Cuddy and Duffin (2005) documented that recognition of familiar music is often preserved even where verbal memory has badly failed. Jacobsen and colleagues (2015), in Brain, found part of the anatomical reason: the region associated with long-term musical memory sits in an area comparatively spared in Alzheimer's disease.
Verse learned in childhood is a near relative of music in exactly this sense — it too has rhythm, it too has pitch contour, it too is stored as a motor sequence rather than a fact. That is why the old lines stay after so much else has gone.
The uncertainty should be stated plainly: the Cochrane review by Woods and colleagues (2018) on reminiscence therapy in dementia concluded that benefits are small and of low certainty. Poetry does not treat dementia, and should not be sold as doing so.
But that is not the claim. The claim is that when language withdraws, the last thing to withdraw is the thing with a beat — which means the poems learned at six will be the last ones standing. Worth choosing carefully.
Every language makes verse the way its structure permits, and Vietnamese does something distinctive.
Vietnamese poetry is counted in tones, not stresses. English verse runs on strong and weak beats; Vietnamese verse runs on the bằng–trắc rule, level tones against oblique ones. The music of a Vietnamese line is therefore bound to the very tone system a child absorbed before saying a full sentence. The metre is the language's own rhythm, not a coat put over it.
Lục bát is a form Vietnamese speakers know without being taught. Six syllables, then eight, with a rhyme rule that chains the couplets into an unbroken run. Almost any Vietnamese can hear a mis-formed lục bát line without knowing the rule — because they heard it from the cradle.
And there is a book. Truyện Kiều is 3,254 lines of lục bát, and it holds a place few literatures can match: a work that illiterate people knew hundreds of lines of, that gets quoted in ordinary conversation, and that has been used for divination. It is living proof of everything Rubin wrote about oral memory — a text that long could only live in a population's heads because it is verse.
Under six: nursery rhymes, lullabies, and repetition past the point of adult boredom. Do not explain the meaning. Meaning comes later; rhythm has to come first.
Primary school: memorise, but memorise what is worth carrying — and do not make it a test. What is being built is a store, not a mark.
Adolescence: offer poems that say what they are actually feeling, even when those are not the poems in the syllabus. And do not demand analysis on first meeting — analysis before feeling is the surest way to kill a poem.
Adults: read slowly, read aloud, read again. The chills research locates the effect in sound structure, and sound structure does not surface when you read with the eye.
The old: ask what they know by heart. Then sit and listen. It is often the last door still open.
Back to the opening question. Why does every culture have poetry, and why has none abandoned it?
Because it is not useless. It is useful in ways that never show up on a payslip.
At three it installs the ear for reading. At ten it is the storage format. At fifteen it supplies words for what is happening inside. At forty it is where grief can be felt safely. At eighty it is what remains when much else has gone.
Five different jobs, one form. Which is why humans were arranging words to a beat before they could write, and are still doing it long after every other means of storage arrived.
These articles summarize well-established research in learning science and linguistics. Key sources and further reading:
The question assumes focus is a muscle you train. The strongest evidence says the training barely carries over, and the...
No time, no talent, too old — those are explanations, not causes. What can be measured is duller: one sentence with a ti...
Not one event but four, running on clocks a billion-fold apart: a wave at 400 milliseconds, a word locked in overnight,...
