Congruence and similarity

In brief

Two puzzle pieces cut from the same template fit into the same slot; a photograph enlarged to a poster shows the same picture, just bigger. The first is an example of congruence, the second of similarity, and both are precise mathematical relationships, not vague resemblance. Congruent figures are identical in size and shape, related by a move, a slide, a turn, or a flip, that carries every point of one exactly onto the corresponding point of the other. Similar figures share the same shape but not necessarily the same size, related by that same kind of move plus a uniform stretch or shrink. The power of both ideas is that you can prove two figures stand in one of these relationships from a short list of matching measurements, sides, angles, without ever needing to physically lay one on top of the other.

The full treatment

First look: stencils and shadows

Trace around a stencil twice on the same sheet of paper, once directly and once after sliding the stencil to a new spot. The two traced shapes are congruent: they have exactly the same size and shape, and if you cut one out you could lay it exactly on the other by sliding, turning, or flipping it, with every edge and corner matching. Now shine a light on the same stencil held at two different distances from a wall, casting two shadows. The shadows are similar, not congruent: the closer shadow is smaller, the farther one is larger, but both have the identical proportions, the same angles at every corner and the same ratio between any two of their sides. Congruence is "same size and shape, possibly moved." Similarity is "same shape, possibly resized (and possibly also moved)."

Building the idea: what has to match, and what does not

For congruence, the defining test is whether a rigid motion, some combination of sliding (translation), turning (rotation), and flipping (reflection), carries every point of one figure exactly onto the corresponding point of the other. Rigid motions are exactly the moves that preserve every distance: if two points on the original figure are 3 units apart, the corresponding points after any rigid motion are still exactly 3 units apart. Because distances between all points are preserved, so are all angles. Two figures are congruent exactly when some rigid motion maps one onto the other.

For similarity, allow one further operation: a uniform scaling, stretching or shrinking every distance in the figure by the same scale factor in every direction. A uniform scaling does not preserve individual distances (unless the scale factor is exactly 1), but it does preserve every ratio of one distance to another within the figure, and preserves every angle exactly, because stretching all directions by the same factor leaves the relative directions, and therefore the angles between them, unchanged. Two figures are similar exactly when some combination of a rigid motion and a uniform scaling maps one onto the other. Congruence is the special case of similarity where the scale factor happens to be exactly 1.

The formal model: proving the relationship from a handful of parts

The practical power of these definitions is that you rarely need to exhibit the actual rigid motion or scaling to prove two figures stand in one of these relationships; instead, a short list of matching measurements is provably sufficient. For triangles, three classical sufficient conditions for congruence are: side-side-side (all three corresponding sides equal in length, since three side lengths pin down a triangle's shape completely), side-angle-side (two corresponding sides and the angle directly between them equal), and angle-side-angle (two corresponding angles and the side directly between them equal). Each can be proved sufficient by construction: given the stated matching parts, attempt to build a triangle satisfying them, and show the third side or angle is thereby forced to a single possible value, with no remaining freedom for a different-shaped triangle to also satisfy the same conditions.

For similarity, the corresponding sufficient condition for triangles is angle-angle: if two angles of one triangle equal two corresponding angles of another, the triangles are similar, with no condition on side lengths at all. This holds because a triangle's third angle is always forced by the first two (all three interior angles must sum to 180 degrees), so matching two angles already forces all three to match, and a classical construction argument confirms that matching angles alone, with no constraint on absolute size, is enough to force the same shape. This is exactly why all right triangles containing a given acute angle, regardless of size, are similar to one another, a fact with direct consequences once ratios between their sides are studied as fixed numbers depending only on the angle.

Why matching a short list is as good as checking everything

The deeper reason these shortcuts work is that a triangle has only limited freedom: three points in a plane, once you fix the three pairwise distances between them, have no remaining freedom to form a different-shaped triangle, the whole figure, all three angles included, is forced. Checking three well-chosen measurements is therefore logically equivalent to checking every measurement, because those measurements already use up all of the triangle's degrees of freedom. This is what separates a genuine proof of congruence or similarity from a visual judgment that two figures merely look alike.

Lineage

Euclid's Elements treats congruence in Book I largely through an informal appeal to superposition, the idea that one figure can be picked up and laid exactly on another, and treats similar figures and proportional sides formally in Book VI, building on the theory of proportion developed in Book V. The superposition argument for congruence, while intuitively clear, was recognized over the following centuries as logically informal, since "picking up and moving a figure" was never itself defined from Euclid's stated postulates. David Hilbert's Foundations of Geometry, 1899, resolved this by making congruence a primitive relation governed by its own explicit axioms. Later, Felix Klein's 1872 Erlangen program recast the whole picture in the language of transformation groups: congruence is the study of properties invariant under rigid motions, similarity the study of properties invariant under rigid motions plus uniform scaling, a reframing that unified both as members of a broader family of geometries, each defined by which transformations it allows.

The strongest case for it

The congruence and similarity criteria are among the most heavily used tools in practical and theoretical geometry precisely because they let a handful of measured or given facts stand in for an entire figure's shape. Surveying, engineering drawing, and manufacturing all depend on similarity to relate a scale drawing or model to a full-size object, confident that every angle and proportion carries over exactly. Congruence criteria are the backbone of geometric proof generally, since showing two triangles congruent is the standard route to proving two segments or angles elsewhere in a larger figure are equal, a technique used continuously from Euclid's original proofs through modern geometric reasoning. The angle-angle similarity criterion in particular is the single fact that makes trigonometric ratios possible at all, since it guarantees that the ratio between two sides of a right triangle depends only on one of its acute angles, not on the triangle's size.

The strongest case against it

The criteria are sharp but narrow, and misapplying them is a common source of error: side-side-angle, unlike side-angle-side, is not in general a valid congruence criterion, because two triangles can share two sides and a non-included angle while still differing in shape, an ambiguous case that trips up learners who assume any three matching parts are automatically sufficient. Similarity also has a boundary worth stating honestly: matching angles forces the same shape only in flat, Euclidean geometry; on a curved surface, two triangles can share all three angles while having different sizes, since the angle sum of a triangle on a sphere depends on its size. A further limitation is practical rather than logical: real physical objects are never perfectly rigid or perfectly measured, so real congruence and similarity claims about manufactured parts are always approximate, holding within some stated tolerance rather than exactly. A common misconception is treating "similar-looking" as a synonym for the formal relation; two figures can look roughly alike while failing every one of the precise angle and ratio conditions similarity actually requires.

Where it stands now

Congruence and similarity, and the classical sufficient conditions for proving them (side-side-side, side-angle-side, angle-side-angle for congruence; angle-angle for similarity), are settled, exactly proved facts of Euclidean plane geometry, unchanged since their essential content was established in Euclid's Elements and made fully rigorous by Hilbert's later axiomatization. What has changed is the conceptual frame: Klein's transformation-group perspective is now the standard lens through which mathematicians understand congruence and similarity as two instances of a general pattern, invariance under a specified family of transformations, that recurs throughout modern geometry and physics.

Test yourself

You are given two triangular metal brackets cut for an assembly. Bracket A has sides measuring 6 centimeters, 8 centimeters, and 10 centimeters. Bracket B has sides measuring 9 centimeters, 12 centimeters, and 15 centimeters. Prove, citing the specific criterion you use, whether the two brackets are similar, and if so, state the exact scale factor relating B to A. Then a third bracket, C, has two angles measuring 37 degrees and 53 degrees, matching two of the angles you can compute for bracket A. Prove whether C must be similar to A, and explain precisely what additional single measurement, if any, you would need in order to instead prove C congruent to some specific one of the three bracket sizes given here.