At the Moth StorySLAM in New York, a woman named Dena takes the stage. English is her third language. Her vocabulary is small. Her sentences are short and grammatically simple. Her accent makes some words hard to understand on the first listen. The native English speakers who told stories earlier in the night used more sophisticated language, more varied sentence structure, more literary turns of phrase. Dena's prose is plain by comparison.
Dena wins. Repeatedly. Across multiple StorySLAMs.1
Matthew Dicks names this the Dena principle — the empirical observation that structural choices in storytelling dominate sentence-level craft by such a wide margin that a storyteller working in a third language with simple sentences will defeat native speakers with polished prose, as long as the third-language storyteller's structural choices are sound. The principle is one of the clearest hierarchy-of-craft statements in the oral storytelling tradition.
The implication for storytellers working at every skill level: time spent polishing sentences while structural choices remain weak is time spent on a lower-leverage variable. The leverage is in the structural choices — where to start, where to end, what to compress, what to land on, how to deploy the engagement techniques. Sentence-level craft is real, but it sits below structure in the dependency stack.
The principle has three components.
1. Structural choices set the storytelling ceiling. The story's shape — where it starts, where it ends, what is included, what is omitted, where the change-moment lands — determines the maximum possible audience response. Strong structure produces a high ceiling; weak structure produces a low ceiling, regardless of sentence-level quality.
2. Sentence-level craft modulates within the ceiling. Polished sentences make a story slightly better; clumsy sentences make a story slightly worse. The modulation operates within the range the structural choices have already established. A clumsy-sentence version of a structurally-strong story still lands above a polished-sentence version of a structurally-weak story.
3. The asymmetry is empirically verifiable. Dena's repeated StorySLAM wins are the proof. If sentence-level craft were the dominant variable, native English speakers with greater linguistic facility would consistently defeat third-language speakers. They do not — when the third-language speaker has sound structure. The asymmetry travels across hundreds of competitive storytelling matches and across years of consistent results.
The cognitive mechanism: audiences process stories at the event level before they process them at the sentence level. When the audience-member is listening, their attention is consumed by what is happening in the story and what is about to happen in the story. Sentence-level features (word choice, syntactic variety, rhetorical flourish) operate at a layer of attention that the audience reaches only after the event-level attention is satisfied. If the event-level attention is not satisfied — if the structural choices are leaving the audience confused, bored, or unclear about what the story is for — the audience never reaches the sentence-level layer at all. The polished prose goes unappreciated because the audience is still asking what is this story about.
The opposite situation: when the structural choices are sound, the event-level attention is fully satisfied. The audience knows what the story is about, knows where it is going, is engaged in the change-arc. Now the sentence-level features become noticeable — but the audience's emotional response is already locked in by the structural choices. The sentences modulate the response upward or downward, but the response's magnitude is dominated by the structural foundation.
A second mechanism: oral storytelling is time-pressured. The audience-member cannot pause and re-read a clumsy sentence the way a reader can pause and re-read a clumsy paragraph. Spoken sentences pass through and are gone. The audience-member's tolerance for sentence-level imperfection is therefore higher than a reader's tolerance — they are not dwelling on the prose, they are tracking the events. Dena's simple English passes through cleanly enough; the audience does not pause on her vocabulary because they are tracking her story. A native speaker whose sentences are more sophisticated does not get correspondingly more credit because the audience is also not pausing on theirs — they are tracking the story.
The third mechanism: structural failure is attentional. When a story's structure is weak, the audience's attention degrades — they look at their phone, they zone out, they start a side conversation. Once the attention is lost, sentence-level craft cannot recover it; the audience is no longer listening to the sentences. Structural strength is therefore a prerequisite for sentence-level craft to operate at all. Sentences depend on attention; attention depends on structure.
The principle is the hierarchy-of-craft framework that orders all the other Dicks teachings. The Big Three (find end / find beginning / master middle) is a structural-choice framework — Dicks codifies it because structural choices are where the leverage is. The BABC model is a structural-choice template. Mount Rushmore is a structural-choice taxonomy. The Dena principle is what justifies the dominance of structural teachings in the curriculum — sentence-level instruction would be lower-leverage and is given less curriculum time accordingly.
It pairs with strategic listening: the meta-skill is biased toward identifying structural moves in consumed artifacts precisely because structure is where the leverage is. A strategic listener who extracted only sentence-level moves would be building a low-leverage library.
It contrasts with writing vs telling as different arts: in writing, sentence-level craft carries more weight than in telling because the reader can dwell on prose. The Dena principle is specific to oral storytelling; for written storytelling the structure-vs-sentence ratio shifts somewhat (though structure still dominates, the margin is smaller).
It connects to no-one-wants-to-hear-anything mindset: if you assume the audience is uninterested by default, you cannot afford to lose them to structural confusion in service of sentence-level prettiness. The mindset's attention-budget calculation pushes the storyteller toward structural soundness as the first priority.
Dena's winning Moth story (paraphrased from Dicks's account) is structurally a textbook example of compressed change-arc with strong audience-identification setup.
Structural choice 1 — where to start: Dena starts not with backstory but in the middle of action — the moment she is at the immigration office facing the bureaucrat who will decide her fate. The audience is immediately in the scene. No throat-clearing exposition. The opening compresses her entire pre-story life into a single line of context delivered as the bureaucrat looks at her file.
Structural choice 2 — what to compress: Years of struggle leading up to this moment are compressed to fragments — a sentence about her sister, a sentence about her job in Russia, a sentence about her decision to leave. Each fragment carries weight; none is elaborated. The compression keeps the runtime tight.
Structural choice 3 — where the change-moment lands: Mid-story, the bureaucrat asks Dena a question that requires her to lie or to tell the truth. She tells the truth. The truth costs her the immigration status she had been working years to secure. The change-moment is her recognition that telling the truth mattered to her more than the status mattered to her — a five-second moment of internal change rendered through her decision.
Structural choice 4 — where to end: Dena does not end with the bureaucratic outcome (whether she got status or not). She ends with the realization itself — the heart-landing of I had not known I was that person until that moment. The bureaucratic outcome is left ambiguous. The audience leaves not knowing what happened next but knowing what changed inside her.
Sentence-level features: Dena's actual sentences are short. I came to America. I was alone. My sister was in Russia. No complex syntactic structures. Vocabulary in the 2,000-most-common-English-words range. Some Russian-accented pronunciation that requires audience-effort to parse.
Audience response: The audience does not register the sentence-level limitations because they are entirely absorbed in the structural arc. Dena wins the StorySLAM that night. She returns to win others. The structural strength carries her across the threshold; the sentence-level simplicity is invisible to the audience because the structural attention has fully consumed their processing.
The case proves the principle: structural choices set the ceiling; Dena's structural choices set her ceiling at high enough to win competitive Moth StorySLAMs; her sentence-level limitations modulate her work within that ceiling but do not lower the ceiling itself.
You have a draft you have been working on for three weeks. You have polished individual sentences extensively. You have rewritten paragraphs for cadence. You have hunted out clichés and replaced them with fresher word choices. The draft reads well. You have tested it on two trusted listeners. They said it was fine — which is the response that should worry you.
It is 9:30am on a Sunday. You print the draft. You set the sentence-polish drafts aside in a folder. You take a blank sheet of paper. You write three questions across the top: where does it start, where does it end, what changes.
You answer each question for your current draft. Where does it start: with backstory about my childhood. Where does it end: with the resolution of the external plot. What changes: I'm not sure — I think the protagonist learns something but I cannot name the moment.
The three answers are the diagnosis. The starting choice is wrong (backstory openings cost attention; should start in scene). The ending choice is wrong (external-plot resolution is rarely the strongest available landing; the internal-change moment is). The change is unnamed (if you cannot name the five-second moment of change, the structure is incomplete regardless of how polished the sentences are).
You take the polished draft and you do not edit any sentence. Instead you cut the first 400 words of backstory and replace them with the scene-opening you should have started with. You cut the final 200 words of external-plot wrap-up and replace them with the internal-change-moment landing. You locate the five-second moment in the middle of the draft and you tighten the surrounding paragraphs to point at it. You leave every other sentence alone.
The new draft is structurally restructured but sentence-identical to the previous version in the middle 60% of the piece. You read it aloud. The new draft hits much harder than the polished-but-structurally-weak version did. The leverage was in the structural moves the whole time. The three weeks of sentence-polishing produced incremental improvement; the two hours of structural restructuring produced step-change improvement.
The pattern is: when a draft reads fine but does not land hard, the diagnosis is almost always structural, not sentence-level. Resist the urge to polish sentences further. Audit the structure first. Fix the structure. The sentences will then carry their own weight without further intervention.
Your draft reads fluently but listeners respond mildly. The sentence-level fluency is operating but the structural choices are leaving the audience un-grabbed. This is the classic Dena-principle failure mode in reverse — polished prose, weak structure.
You spend more drafting time on sentences than on structure. Hour-budget audit: in your last three drafts, how many hours were spent on sentence-level revision and how many on structural decisions? If the ratio favors sentences, you are probably over-investing in the lower-leverage variable.
You cannot articulate your story's five-second moment of change. If asked what changes in this story, you have to think for more than five seconds. The unclarity is structural — the story may not have a clear change-moment, or you may not have identified it, either of which is fatal to audience response regardless of prose quality.
You feel attached to specific clever sentences and resist cutting them. The attachment is sunk-cost-on-sentence-level-craft. You spent hours on the sentence; you do not want to cut it. But if the sentence is sitting in a structural position that should be cut (it is in the backstory opening that should be replaced by a scene-opening), the attachment is preventing the higher-leverage structural fix.
Other people's polished-but-structurally-weak stories feel similar to yours. When you listen to other stories that have your same problem, you can feel the fineness of them. They are not failing — they are just not landing. If you can feel it in others, you have the diagnostic capacity to feel it in your own work; use it.
Picture the leverage hierarchy: within any craft, some variables dominate others by such a wide margin that effort allocation should follow the hierarchy rather than the immediate-feedback gradient. Several other domains have explicit equivalent hierarchies.
Software engineering / architecture vs implementation: The software engineering tradition codifies an analogous hierarchy — architectural choices dominate implementation choices. A system with sound architecture and mediocre implementation outperforms a system with brilliant implementation and weak architecture, often by orders of magnitude. Put storytelling and software engineering together and you see the Dena principle is not storytelling-specific but is a general craft-leverage principle: in any craft with multiple levels of organization, the higher organizational levels dominate the lower ones, and effort should be allocated up the hierarchy. Neither domain alone gives you the full picture: storytelling craft supplies the empirical demonstration (Dena's wins are evidence); software engineering supplies the formal theoretical account (architectural complexity theory). Pairing them makes the principle's generality visible.
Athletic performance / strategy vs technique: Elite-coach traditions distinguish between strategic choices (when to attack, what to defend, how to allocate effort across a match) and technique (the polish of individual movements). Strategy dominates — a strategically sound athlete with mediocre technique defeats a technically polished athlete with weak strategy. The cross-orientation parallel demonstrates that the leverage-hierarchy is not artifact-specific (story vs match) but is a feature of any performance domain with multiple organizational levels. The implication: storytellers should audit their craft-effort allocation the way elite athletes audit their training-effort allocation — making sure the higher-leverage variables are receiving proportional investment.
Persuasion / argument architecture vs rhetoric: Six-Minute X-Ray (Hughes) and the broader persuasion literature distinguish between argument architecture (the order of claims, the framing of the conclusion, the selection of evidence) and rhetorical craft (the polish of individual sentences, the cadence of delivery). Architecture dominates. A well-architected argument with plain rhetoric persuades; a beautifully-rhetorical argument with weak architecture does not. The pairing with the Dena principle reveals that the structure-trumps-sentences hierarchy operates symmetrically across narrative and argumentative speech acts, suggesting the underlying mechanism is about how listeners allocate attention rather than about what the speech act is doing.
The Sharpest Implication The Dena principle implies that most amateur storytelling improvement plateaus come from misallocated effort — the storyteller is investing in sentence-level polish because sentence-level polish produces visible improvement on each draft, while ignoring structural choices because structural decisions are uncomfortable to revisit (they require cutting material, reordering, rethinking what the story is for). The leverage-gradient pulls toward the lower-leverage variable because the feedback is more immediate. The discipline is to override the feedback gradient and invest in structural work even though the per-hour productivity feels lower. The storytellers who break through their plateaus almost always break through by accepting this trade.
Generative Questions