The framework gathers a small family of results under one idea: existence theorems allow us to draw conclusions about a function’s behavior on an interval without precisely locating that behavior. The Intermediate Value Theorem is the first of them, and that sentence is the whole shape of it. Something happens. The theorem declines to say where.

Stated in full: if \(f\) is continuous on the closed interval \([a,b]\) and \(d\) is a number between \(f(a)\) and \(f(b)\), then there is at least one number \(c\) between \(a\) and \(b\) with \(f(c) = d\).

Four things to notice, and the interactive below is built to make each of them visible.

The statement, one clause at a time

Both functions run from f(0) = 1 to f(4) = 5, so the bracket the theorem checks is identical in the two modes and only continuity differs. Drag the target. Where the theorem applies it promises at least one crossing, and the crossings actually present are marked and counted underneath — usually three, which is the gap between what is guaranteed and what is true. Switch to the jump and drag the target between 3 and 4: the bracket still holds, the theorem no longer applies, and there is genuinely no crossing to find.

The two hypotheses are checked separately because they fail separately. Drag the target above 5 or below 1 and the second one goes out while the first holds: the function is still continuous, the theorem has nothing to say, and the honest report is that it does not apply rather than that no crossing exists.

“At least one” is not a hedge

Leave the target at 3 on the continuous function and count the marked crossings. There are three, at \(x = 2-\sqrt{3}\), \(x = 2\), and \(x = 2+\sqrt{3}\), and the theorem promised one.

That gap is the point of an existence theorem. It is a lower bound on how much happens, obtained from almost no information — two endpoint values and continuity. It cannot be a count, because the same two endpoint values are consistent with one crossing or with a hundred.

So a conclusion that says “the function equals 3 exactly once” is not what the theorem gives you, and neither is “the function equals 3 at \(x = 2\).” Both are true here. Neither follows from the IVT.

Continuity is not decoration

Switch to the jump and set the target to 3.5. The endpoints have not moved: \(f(0) = 1\) and \(f(4) = 5\), and 3.5 sits squarely between them. The second hypothesis is satisfied and the first is not, and the readout reports what that costs — no crossing anywhere on the interval.

The left piece climbs from 1 to just under 3; the right piece starts at 4 and climbs to 5. Everything in \([3, 4)\) is stepped straight over. The bracket is a genuine bracket and it guarantees nothing, because the guarantee was never about the endpoints alone.

This is also why “continuous on the closed interval” is written the way it is. The endpoints are where \(f(a)\) and \(f(b)\) are read, so continuity has to reach them.

The sentence that earns the point

The suggested skill attached to this topic is providing reasons or rationales for a conclusion, which is a fair warning that the writing is the assessed part. A complete invocation has four moves, in this order:

  1. Name the function and assert continuity, with a reason. Polynomials are continuous everywhere; a quotient is continuous on its domain; a function given as continuous in the stem is continuous because you were told.
  2. Evaluate at both endpoints and state the two values.
  3. Observe that the target lies between them.
  4. Conclude, in the theorem’s own words, that there is at least one \(c\) in the open interval with \(f(c) = d\), and name the theorem.

Written out for the function above:

\(f\) is a polynomial, so it is continuous on \([0,4]\). Since \(f(0) = 1\) and \(f(4) = 5\), and \(1 < 3 < 5\), the Intermediate Value Theorem guarantees there is at least one \(c\) in \((0,4)\) with \(f(c) = 3\).

The two failures worth naming are skipping step 1, which is the one graders are actually checking, and overclaiming in step 4 — writing “exactly one” or naming a location. The theorem is being invoked precisely because you cannot locate anything, so a conclusion that locates something is not a conclusion the theorem supports.

A self-test at the jump: drag the target slowly from 2 up to 5 and watch the crossing count. It reads one, then none, then one again, while the verdict never moves — the theorem has said “does not apply” for the whole drag. Account for that in one sentence. A theorem that is silent across the entire range is still telling you something true at every point of it, and saying what is the difference between having learned the statement and being able to use it.