Principle 01
Conceptual Metaphor: You Think You're Describing — You're Framing
We Think in Metaphor, Not Just Speak in It
Lakoff · Frame
The Principle in One Line
Metaphor is not only a feature of language — it's a feature of cognition. We borrow a familiar domain (the source) to grasp an abstract one (the target). "Argument is war" makes us defend a position, attack a flaw, win the debate. Swap the metaphor and thought switches tracks. So: choosing a metaphor is choosing the lens your reader will see the world through.
"The essence of metaphor is understanding and experiencing one kind of thing in terms of another."
— Lakoff & Johnson, Metaphors We Live By, Ch. 1
Why It Works
Lakoff's subversive finding: metaphor lives not in the dictionary but in thought. "Argument is war" isn't one figure of speech — it's an entire system we don't notice, the reason we say a claim is "indefensible" or that we "surrendered." If a culture saw argument as a dance rather than a war, people would have an entirely different experience of it. For the writer this means: every metaphor you use quietly installs a frame. Treat "tech debt" as a moral stain and the team feels guilt and avoids it; treat it as financial leverage and the team weighs and repays.
Source · what you borrowTarget · what you discuss
Battlefield · enemy
→
Debate · opponent
Attack / defend / surrender
→
Rebut / justify / concede
Win the battle
→
Win the argument
"Argument is war": the source's structure is imported wholesale — along with all its implications
Before → After
We've piled up far too much tech debt; it's a disgrace to the team and must be wiped out at once.
Tech debt is like a loan: we borrowed speed up front and now pay interest monthly. This quarter we won't repay principal — just the three highest-interest items, to put the system back in "positive cash flow."
Swapping the "moral stain" frame for a "finance" frame — the team moves from guilt to trade-off.
We're drowning in bugs and need to fight our way through this backlog.
Think of the backlog as a garden, not a battlefield: we'll never "win" it, but we can keep weeding the beds that matter most.
The "war" frame is exhausting and endless; the "gardening" frame brings priorities and sustainability.
When to Use + Common Mistakes
- ✓ Strategy narratives, change communication, team culture, naming abstractions (project codenames, metric names) — anywhere you want to quietly shift how a reader thinks
- Trap: unconsciously inheriting someone else's frame ("war," "race"), importing all its implications with it
- Trap: a pretty metaphor that pulls against the action you want — a "death march" codename deflates the team before work even starts
Key References
George Lakoff & Mark Johnson, Metaphors We Live By (1980, founding text of conceptual metaphor theory) · George Lakoff, Don't Think of an Elephant! (metaphor framing applied to political communication)
This Week's Exercise
Pick a metaphor your team uses daily ("race," "war," "moat," "debt"). Write down its three hidden implications: what to do, who the enemy is, what counts as winning. Then force-swap in a different metaphor and rewrite the three. See which lens is closer to the action you actually want.
Question: in your last big decision, which unspoken metaphor was at the wheel?
Principle 02
What Makes a Metaphor Work: The Vehicle Must Sit Closer Than the Tenor
Explain the Unknown with the Known
Aristotle · Craft
The Principle in One Line
A metaphor is a bridge: from the reader's known shore (the vehicle) to an unfamiliar one (the tenor). For the bridge to hold, the vehicle must be more concrete, more familiar, more sensory than the tenor — explain what people haven't seen with what they see every day. Reverse it and you explain the strange with the strange; the bridge collapses.
"The greatest thing by far is to be a master of metaphor… it is the one thing that cannot be learnt from others; and it is also a sign of genius, since a good metaphor implies an intuitive perception of the similarity in dissimilars."
— Aristotle, Poetics, 1459a
Why It Works
The rhetorician I.A. Richards split a metaphor into three: the tenor (what you're explaining), the vehicle (what you borrow), and the ground (the structure they share). The whole craft lies in making the ground align on what matters, while keeping the vehicle inside the reader's comfort zone. Bad metaphors die two ways: the vehicle is even stranger than the tenor (explaining love through quantum entanglement), or the ground misaligns where it counts — "a neural net is like a human brain" has fooled millions; it resembles a brain in "learning" but not at all in mechanism, and readers slide down the wrong ground into misunderstanding.
Tenorthe subjectthe unfamiliar
thing you explain
Vehiclethe imagethe familiar
thing you borrow
Groundshared structurewhere success or
failure is decided
The three parts: whether the ground aligns on what matters decides if the bridge holds
Before → After
A vector database performs semantic retrieval by embedding text into a high-dimensional space and computing cosine similarity.
A vector database is like a library that shelves books by meaning: ask for "books about courage" and it hands you the closest ones — no title match needed. (Under the hood, text becomes high-dimensional coordinates and it measures the distances.)
Give the familiar library first, the technical line second — vehicle ahead, jargon behind.
A load balancer distributes incoming requests across multiple backend servers per a scheduling algorithm.
A load balancer is the host at a busy restaurant: instead of letting everyone pile into one section, it seats each new guest at the emptiest table.
(Misleading metaphor) Our AI thinks just like a human brain.
Our AI is a very fast pattern-matcher trained on text — closer to autocomplete than to a mind.
The original misaligns where it matters; readers overestimate its "understanding."
When to Use + Common Mistakes
- ✓ Technical explanation, onboarding docs, pitches to non-technical audiences, teaching
- Trap: a vehicle harder than the tenor (metaphor as showing off); or a ground that resembles only on the surface and misleads on essentials
- Trap: stretching one metaphor too far — push it to a fourth layer and the structure breaks
Key References
Aristotle, Poetics (1459a, "the similarity in dissimilars" as the mark of genius) · I.A. Richards, The Philosophy of Rhetoric (1936, tenor / vehicle / ground) · Steven Pinker, The Sense of Style (the concrete beats the abstract)
This Week's Exercise
Pick a concept you often explain to laypeople (caching, rate limiting, compounding). Write three metaphors for it, drawing vehicles from the kitchen, from traffic, and from the body. Read them to a layperson and ask them to paraphrase — if they get it right, the ground aligned.
Question: "a neural net is like a human brain" — what does it illuminate, and what does it hide?
Principle 03
Argument by Analogy: Moving a Verdict with a Familiar Structure
Borrowing a Conviction the Reader Already Holds
Hofstadter · Argument
The Principle in One Line
Analogy isn't only an explaining tool — it's an arguing tool. When you say "A is to B as C is to D," you borrow the verdict everyone already accepts about C–D to move the verdict on A–B. An analogy's persuasive power comes from structural alignment; its danger comes from structural misalignment.
"Without concepts there can be no thought, and without analogies there can be no concepts."
— Hofstadter & Sander, Surfaces and Essences, Preface
Why It Works
Argument by analogy is powerful because it anchors an abstract claim to a conviction the reader already holds. "Shipping without tests is like driving without a seatbelt" — the reader's verdict on seatbelts is ready-made, and the analogy carries it over to tests, sparing you a statistics lecture. But borrowed force needs a clear title: does the core structure truly correspond? Where does it stop corresponding? Is that gap precisely the essential point? "National finance is like a household budget" rolls off the tongue but breaks where it counts: a household can't print money or set tax rates. The best analogists mark the boundary themselves — "the analogy ends here" — and gain credibility for it.
| Structural point | Driving (vehicle) | Shipping (tenor) |
|---|
| Normal case | You usually arrive safely | Most deploys go fine |
| Rare disaster | The one crash costs dearly | The one outage costs dearly |
| Cheap prevention | Buckle the seatbelt | Write tests / roll out gradually |
| Break point | The person is hurt | The system auto-alerts |
Stress-test the bridge: list the matches, then surface a break point — naming the boundary is a plus
Before → After
We should invest in a gradual rollout because it lowers the risk of production incidents.
A full rollout is like dumping fish straight from cold water into a hot tank — the shock kills them. A gradual rollout is "tempering": release one small bucket, let the temperature match, then add more. (Boundary: fish can't alarm you, but a system can.)
We shouldn't optimize this code path yet; it isn't the bottleneck.
Optimizing this now is like widening a driveway to fix a jam that's actually on the highway — weeks of work, and the cars still won't move. Let's profile first, then widen the real bottleneck.
When to Use + Common Mistakes
- ✓ Decision memos, arguing for a technical approach, persuasion, teaching, promotion cases
- Trap: the "false analogy" — surface-similar, essentially different, yet offered as evidence (the most common sophistry)
- Trap: leaving the boundary unmarked so readers extrapolate forever; or an analogy so clever readers pick at it and forget your claim
Key References
Douglas Hofstadter & Emmanuel Sander, Surfaces and Essences (2013, analogy as the core of cognition) · Douglas Hofstadter, "Analogy as the Core of Cognition" (2001 lecture) · Chip & Dan Heath, Made to Stick (concrete and credible)
This Week's Exercise
Take the next claim you need to argue and write an analogy ("this is like…"). Run the "bridge stress test": list two matching points, find one break point, and write it into the text — "this analogy fails at X." See whether credibility rose or fell.
Question: same claim, one good analogy versus one bad — how much does persuasion differ? Can readers tell?
Principle 04
Dodging Dead Metaphors: When a Figure Turns to Jargon
Avoiding Dead Metaphors
Orwell · Revision
The Principle in One Line
Metaphors have a lifespan. Alive, they make you see a picture; overused, they "die" into jargon that no longer summons any image — "synergize," "leverage," "circle back," "move the needle." The real danger of a dead metaphor isn't ugliness — it's that it lets you think you said something when you said nothing.
"Never use a metaphor, simile, or other figure of speech which you are used to seeing in print."
— George Orwell, Politics and the English Language (1946)
Why It Works
Orwell distinguished three kinds: live ones (they carry a picture), fully dead ones (like "deadline" or "branch" — now ordinary words, harmless), and the worst, the "half-dead" — which has lost its image yet still strikes the pose of a metaphor, so people reach for it without thinking. English office-speak like "leverage," "synergy," "granularity," and "move the needle" is exactly this kind: once vivid, now a clever fog that hides "I haven't actually figured out what to do". The test: swap the metaphor for the literal meaning it claims; if the sentence suddenly empties, it was covering a void for you all along.
Alivecarries a picture
"like restarting a server at 3am"
Half-dead (danger)image gone, pose remains
"leverage / synergy / circle back"
Fully dead (harmless)now ordinary words
"deadline / main branch"
The lifespan spectrum: hunt the middle band — half-dead jargon
Before → After
We need to empower the front-line team, find the right lever, hit them with a combination punch, and close the growth loop.
We'll give the front-line team three concrete things: a self-serve reporting tool, a same-day data Q&A channel, and a weekly growth checklist.
Each dead metaphor is replaced by a concrete, checkable action.
Let's leverage our synergies and move the needle by thinking outside the box.
Let's merge the data and sales teams so we can ship the churn dashboard by Friday.
When to Use + Common Mistakes
- ✓ The final revision pass of any formal writing — hunt half-dead metaphors, especially in strategy / vision / OKR docs, where jargon runs thickest
- Trap: feeling professional for using dead metaphors while actually dodging a concrete commitment
- Trap: overcorrecting — replacing even harmless dead metaphors ("main branch," "deadline"); they're ordinary words now, and swapping them just adds clutter
Key References
George Orwell, Politics and the English Language (1946, on "dying metaphors") · William Zinsser, On Writing Well (the "Clutter" chapter) · Wang Xiaobo, "My Literary Lineage" (on how clichés and translationese corrupt Chinese)
This Week's Exercise
Dig out a recent OKR or strategy doc and circle every metaphor-jargon word (empower, lever, loop, granularity). For each, ask: what concrete action does it mean? Rewrite it as that action. Count how many words per page were covering "not yet figured out."
Question: jargon spreads in organizations — is it because it "saves thinking" or because it "avoids accountability"? How do you tell the two apart?
Deeper Questions
Pushing the edges, cross-language differences, fitting different forms
If metaphor can shape thought (Lakoff), can it also imprison it? When a metaphor traps you, how do you climb out?
The flip side of Lakoff is the "frame trap" — once you accept "argument is war," it's hard not to treat the other side as an enemy. Two moves: first, "swap the vehicle" — describe the same thing with three metaphors (argument as dance / as building a house / as chess), and the new ones expose what the old one hid; second, go literal — strip every metaphor and ask "what actually happened?" The most dangerous metaphor is the one you don't notice you're using.
How different are the Chinese and English metaphor systems? Why does a fine English metaphor often "die" when translated literally?
Metaphor is rooted in cultural experience, so it often fails across languages. "Low-hanging fruit," translated literally into Chinese, lacks the bodily resonance of picking and reads worse than "grabbing the easy gains." The deeper gap is at the conceptual level: "time is money" (spend / waste time) runs very deep in English; Chinese says "spend time" too, but the financial coloring is far fainter. Don't translate the vehicle — find a substitute with the same ground and a native vehicle.
AI writing loves metaphors ("riding the wave of the digital age," "opening a window") — why do they read so fake?
Because they're exactly Orwell's "half-dead metaphors" — statistically the most frequent and safest, so the model emits them by default. A good metaphor comes from one person's concrete observation ("like restarting a server at 3am"), carrying this person's fingerprint at this moment; AI's metaphors are the fingerprint-free average. The fix: either order it to "use no cliché metaphors," or delete all of its metaphors and replace them with concrete objects from your own life.
Argument by analogy is both the most powerful persuasion and the most common sophistry. As a reader, how do you quickly spot a "false analogy"?
Stress the bridge in three steps: one, find the claimed match — A is like B because both have X; two, ask "does the essential part really have X?" — false analogies often have X on the surface but not at the decisive level; three, ask "does a counter-analogy hold just as well?" — if both directions flow, the analogy has no argumentative force and is only decorating a stance. Someone who states the boundary of their own analogy is more credible than one who makes it seamless.
If dead metaphors should be dodged, why does great writing so often "reuse old words"? Where's the line between cliché and deliberate revival?
The line is whether the picture is re-lit. A cliché is reached for unconsciously and slid past unconsciously; "revival" is when the writer knows it's a stock phrase yet twists it so the image flares up again — turning "break the glass ceiling" into "she didn't break the ceiling; she found there was no ceiling, just a crowd looking up and pretending there was one." Same dead metaphor — the first is laziness, the second is craft.