Build Understanding Between Example and Abstraction

A slogan names a pattern; a working explanation connects the moving parts. Imagine a small town where a popular café sells the same coffee for $3 one week and $3.60 a year later. That change invites questions: Did the café’s costs rise? Did customers become more willing to pay? Did something else change? One cup cannot tell us whether prices across the economy have risen or why.

The café buys beans, cups, and electricity; it pays workers and rent. Suppose a poor harvest raises the cost of coffee beans, while higher electricity bills raise the cost of running the espresso machine. The owner has choices: accept a smaller profit, find cheaper inputs, reduce costs elsewhere, or raise the menu price. Those costs create pressure, but they do not dictate a single response. The outcome depends partly on customer demand, competition, and the owner’s room to maneuver.

Change one condition. Imagine the bean harvest is poor, yet customers suddenly stop buying coffee. The café may be unable to pass the higher bean cost on to customers; it might instead absorb the loss or close. Or suppose the harvest is normal, but a shipping disruption delays beans and makes them scarce. The café may raise prices for a different immediate reason. The same visible price change can emerge from different combinations of causes.

A higher coffee price is a relative price change: coffee has become more expensive compared with some other things. Inflation, in the broad sense, is a sustained rise in prices across many goods and services. A drought affecting coffee can make coffee dearer without raising the general price level. And the café cannot establish what caused economy-wide inflation: one business cannot represent every market, and higher costs in one example do not prove that costs caused a general rise in prices. In other circumstances stronger demand may be central; demand and supply pressures may also reinforce each other.

The café’s value is local and practical: it lets us trace how changed conditions pass through costs, customer demand, and a seller’s choices to an observed price. Those relationships—not the coffee price alone—are the parts to carry forward when we compress an example into a provisional pattern.

The café’s story can now be compressed into a provisional pattern: when the cost or availability of something changes, people adjust what they offer or buy, and those choices can affect its price. Compression matters. A model small enough to carry is easier to compare with another case than a full narrative about one café, its supplier, and its customers.

But a compact model earns its usefulness only when it travels. Try it on rental housing. If many people want to live in a city while the number of available apartments changes slowly, landlords may be able to charge more. The moving parts resemble the café’s: buyers and sellers respond to conditions, and the balance between what people want and what is available puts pressure on price. The model helps explain why a surge in demand can raise rents even if landlords’ costs have not risen.

That is a successful journey, but not a complete one. Apply the same model to a city where rents are legally capped. Demand may rise, yet the posted rent cannot rise in the usual way. The pressure has not vanished; it may appear as longer queues, fewer available apartments, or competition over who gets one. Our first model—changes in supply and demand can put pressure on price—still points in a useful direction. But to explain what happens, we must add a constraint: prices are not always free to adjust, and when they cannot, other parts of the system may change instead.

This is what testing a model’s reach looks like. We do not ask only, “Does this example fit?” We ask, “What does the model explain here, what does it leave out, and what would I expect to see if the missing factor mattered?” A model that explains rising rents in one city but not the effects of rent controls needs a more specific account of how prices and other parts of the system adjust. That mismatch is not a reason to discard everything; it points to the condition the next question should examine.

A counterexample can expose an assumption that was hidden in a compact rule. Suppose someone says, “When demand rises, prices rise.” You find a case where demand rises and the price stays fixed. Perhaps a government has set a maximum price, a long-term contract fixes the price, or suppliers can quickly increase production. These possibilities do not all mean the same thing; each suggests a different condition to examine when applying the rule.

A pattern, then, is not a slogan extracted from a few examples. It is a proposed map of relationships, with a territory it covers and assumptions it relies on. Give it a new case. Notice where its guidance holds and where the case requires you to account for another condition. The goal is not to make one model explain everything, but to learn what work it can do—and let the edge of the model become the next question.

A map earns its keep by showing a few important relationships, not by resembling the territory in every detail. An analogy is a kind of map: it lets you borrow structure from something familiar to get your bearings in something unfamiliar. But every map leaves things out. The useful question is not “Are these two things the same?” It is “Which parts line up—and where does the comparison stop helping?”

Electricity offers a classic example. Imagine water flowing through a pipe. The amount of water passing a point each second resembles electric current: both describe flow. A difference in water pressure resembles voltage: a difference in electric potential that can drive current. A narrow, obstructed pipe resembles electrical resistance, which makes current harder to sustain. If you already understand pipes, this picture gives current, voltage, and resistance somewhere to attach in your mind.

The analogy does more than decorate unfamiliar words. It preserves a pattern of relationships. With the same pipe or electrical resistance, a greater pressure difference tends to produce more flow; a greater voltage tends to produce more current. A tighter bottleneck restricts water; greater resistance restricts current. The parts are not just paired one by one—the way they influence one another is paired, too. That is what makes the comparison explanatory.

Now follow the map to its edge.

Water in a pipe is made of molecules moving along a channel. In a metal wire, electric current is the organized movement of charged particles, and the electrical effects in a circuit are not simply little parcels of charge racing from a battery to a lamp. The water picture can make us imagine electrons as water droplets that leave the battery, travel through the wire, and arrive at the bulb carrying a delivery of energy. That picture is misleading. In a simple circuit, energy is transferred through the electromagnetic field around the conductors; the details are subtler than the plumbing image suggests.

The pressure comparison is useful precisely because pressure is not an amount of water, just as voltage is not an amount of charge. Voltage describes a difference in electric potential between two points—an energy difference per unit charge. The mistake is to turn either driving difference into a supply of stuff that gets used up. A circulating-water model with a pump helps keep that distinction visible: the water keeps moving around the loop while energy is transferred. It still does not explain the electromagnetic process that transfers energy in an electrical circuit.

This does not make the analogy useless. It tells you how to use it. Keep the relations that clarify the question at hand: flow, driving difference, restriction. Set aside the details that the comparison cannot carry: the physical nature of charge, the way energy is transmitted, and the exact meaning of voltage. When you need those, you need a more precise model.

A good analogy is therefore a temporary scaffold, not a second name for the thing being built. It helps you reach a level from which you can see the actual structure more clearly. Then inspect the scaffold: Which beams supported the idea? Which were only convenient props? If you cannot say where an analogy breaks, you may still be borrowing its familiar imagery rather than understanding the unfamiliar system.

The same discipline applies beyond science. Calling inflation a “fever” can suggest that rising prices signal strain in an economy, but it may also imply a single illness with a single cause and a simple cure. Calling a computer’s memory a “desk” helps explain quick access to active information, but a computer does not remember as a person does. Each comparison highlights some relations and hides others.

So when an analogy clicks, don’t stop at the click. Ask what it maps, what it leaves out, and what mistaken conclusion it might tempt you to draw. The resemblance opens the door. Knowing its edge keeps you from walking through the wrong one.