Why Your White Sneakers Prove That All Ravens Are Black
The Copper Coin and the Question
Imagine you hold a copper coin over a candle flame. Seconds later, the coin barely fits through a slot it slid through before. You ask, “Why?” Someone says, “Because copper expands when heated, and this coin is copper.” That answer feels satisfying — it pulls in a general law about copper and applies it to this specific coin. In 1948, the philosopher Carl Hempel (1905–1997) argued that this is exactly what a scientific explanation looks like: an argument that links a single event to a general law.
Hempel was born near Berlin and trained in physics and mathematics before turning to philosophy. He joined a group called the Vienna Circle, who wanted to make philosophy as clear as a mathematical proof. They believed that any statement that couldn’t be verified by observation was meaningless. But Hempel, more than anyone, would discover cracks in that simple picture — and his attempts to fix them produced some of the strangest puzzles in the philosophy of science.
The Raven Puzzle: How a White Shoe Confirms a Black Bird
Let’s start with a puzzle. You want to check the truth of the statement “All ravens are black.” You go outside, see a raven, and it’s black — that confirms the idea. According to what’s called Nicod’s criterion, an example that fits the “if” part and the “then” part (a black raven) confirms the rule; an example that fits the “if” part but not the “then” part (a white raven) would disprove it. Things that don’t fit the “if” part at all (non‑ravens) are simply neutral. That sounds obvious.
Hempel noticed a twist. The sentence “All ravens are black” is logically equivalent to “All non‑black things are non‑ravens.” If you test that second claim using Nicod’s rule, a non-black thing that is also a non-raven — like a white sneaker — counts as confirming evidence. Since the two sentences say the same thing, Hempel argued that whatever confirms one must confirm the other. So seeing your white shoes actually makes “All ravens are black” slightly better supported. Hempel published this paradox of confirmation in 1945, and philosophers have argued about it ever since.
Hempel insisted the result was logically correct, even if it felt absurd. Some critics said that because there are far more non‑black things than black things, a single white shoe provides only a minuscule boost compared to a black raven. Hempel conceded that a quantitative approach might make the paradox feel less shocking. But he never backed away from the idea that confirmation depends on logical structure, not just our habits of thought.
Explanation as Argument: The Covering‑Law Model
Hempel and his co‑author Paul Oppenheim proposed that scientific explanations are a type of argument. They called this the deductive‑nomological (DN) model, or the covering‑law model. In its simplest form, you have:
- One or more general laws (the “nomological” part).
- One or more statements about initial conditions — the situation before the event you want to explain.
- The explanandum — the statement describing the event to be explained.
The explanation works if the event logically follows from the laws and the initial conditions. Hempel set out four conditions: (1) the event must be a logical consequence of the premises; (2) the premises must include at least one general law; (3) the premises must have empirical content — you can test them against the world; and (4) all the premises must be true. For example, explaining why a flagpole casts a shadow of a certain length requires a law about light traveling in straight lines, plus the sun’s angle and the pole’s height. The shadow’s length then follows necessarily.
Hempel noticed something else: if you had known the laws and conditions ahead of time, you could have predicted the event. He called this the symmetry thesis: every adequate explanation could have served as a prediction, and every scientific prediction could, in principle, serve as an explanation. This idea would soon be tested.
When Prediction Isn’t Explanation: The Flagpole Fights Back
The symmetry thesis faced clever counterexamples. Suppose you measure the length of a flagpole’s shadow and, using the laws of optics and the sun’s angle, deduce the pole’s height. According to Hempel’s original conditions, that deduction qualifies as a scientific explanation. Yet it seems obvious that the shadow’s length does not explain why the flagpole has that height — the pole’s height was chosen by the builders, not forced by the shadow.
Another trouble came from the modus tollens form of argument. If you know that all ravens are black, and you spot a bird that is definitely not black, you can conclude it isn’t a raven. That’s a valid deduction, and it satisfies Hempel’s conditions. But does the bird’s non‑blackness really explain why it isn’t a raven? Hardly. Its being a raven is about its species and DNA, not about its color. So even valid deductions from true laws can fail to explain.
These puzzles nudged philosophers toward the idea that explanation requires more than logical shape — it needs causal relevance. A property is causally relevant if changing whether it holds makes a difference to the chance of the outcome. The flagpole’s height causes the shadow’s length, not the other way around. Modus tollens deductions, while logically airtight, can run the causal arrow backward.
The Hidden “Unless” That Every Law Carries
Toward the end of his career, Hempel spotted another crack. Take the simple law “Copper expands when heated.” It assumes nothing else is interfering — no super‑cold blasts, no weird electromagnetic fields. In real life, other forces almost always lurk. Hempel called these hidden assumptions provisos: you must assume that no factors outside the theory are at play.
This means that testing a single law in isolation is trickier than it looks. If a coin fails to expand, it might not mean the law is wrong — it could mean a proviso was violated. Some critics, like Nancy Cartwright, went further and argued that the “simple” laws in textbooks are never exactly true in the messy world. Hempel’s proviso problem shows that applying a theory to the real world always demands extra checking — it’s never a purely logical plug‑and‑play operation.
Why Your Curiosity Needs Hempel’s Checklist
When you ask why a coin expanded, you’re doing what scientists do every day: hunting for the general rules that cover the case. Hempel gave us a checklist for what makes an answer genuinely explanatory, even though his original checklist needed revisions. He showed that verification isn’t enough — a single observation can sneakily support a whole theory, just like the white sneaker. He showed that explanation and prediction aren’t mirror images; a good explanation tracks what actually makes things happen.
These ideas matter far beyond philosophy classrooms. When a news headline says “study shows X causes Y,” you can ask: Did the scientists show a law‑like connection, or just a correlation? Could the prediction be true without explaining anything? Hempel’s puzzles remind us that science is not just a pile of facts — it’s an ongoing argument about the invisible laws that make the universe tick. And every time you feel that satisfying click when a “why” question finally gets a good answer, you’re feeling the force of a covering law.
Think about it
- If a scientist can perfectly predict what time the sun will rise every day for the next century, does that prediction explain why the sun rises? What would a satisfying explanation need to add?
- Hempel said that seeing a white shoe confirms “All ravens are black.” If you were to design a bird survey, would you count non‑black non‑ravens as evidence? Why or why not?
- Think of a time you asked “why” and got an answer like “that’s just how it is.” What was missing — a law, a cause, or something else?
