research29 min studyLesson 543 in Mathematics

Projective, injective, and flat modules

Distinguish lifting, extension, and tensor exactness through concrete module tests.

On this page

What you’ll understand

  • Translate projectivity and injectivity into lifting and extension diagrams.
  • Test a failure of flatness using an injection.
  • State the hypotheses under which flatness implies finite projectivity.

Three useful classes of modules repair three different failures of exactness. Projectives let maps lift through surjections; injectives let maps extend across injections; flat modules preserve injections when tensoring. These properties answer different questions, even when some implications connect them. We first allow a general unital ring and specify left or right modules where necessary, then use commutative rings for the numerical examples so that both tensor factors have an unambiguous compatible action.

Projectivity repairs lifting

A left module PP is projective if every map from PP into the target of a surjection lifts to its source. Equivalently, Hom⁡R(P,−)\operatorname{Hom}_R(P,-) is exact, rather than merely left exact. Free modules have this property because images of basis elements can be lifted individually. A direct summand of a free module remains projective. Conversely, choose a free module surjecting onto PP and lift its identity; the resulting section exhibits PP as a direct summand.

This characterization does not require a finite basis. It also explains why projective is weaker than free over a general ring: a summand can fail to admit a global basis. A short exact sequence ending in a projective module splits by lifting that module's identity. The quantifier matters. Splitting one convenient sequence is not the definition; projectivity concerns every surjection and every compatible map, or equivalently every short exact sequence having that module as quotient.

P projective  ⟺  Hom⁡R(P,−) exact  ⟺  P⊕Q≅R(I) for some Q,IP\text{ projective}\iff\operatorname{Hom}_R(P,-)\text{ exact}\iff P\oplus Q\cong R^{(I)}\text{ for some }Q,I
The direct sum can be infinite; finite projectivity requires finite generation as well.↔ Scroll wide equations horizontally.

Injectivity repairs extension

A left module JJ is injective if a map from a submodule into JJ extends to the containing module. The contravariant functor Hom⁡R(−,J)\operatorname{Hom}_R(-,J) is then exact. Every short exact sequence beginning with JJ splits, by extending its identity to a retraction. Reversing arrows is essential: a projective condition at the quotient cannot be substituted for an injective condition at the submodule simply because both statements mention a splitting.

Over integers, injective modules are exactly divisible abelian groups. The rationals and the rationals modulo integers are standard examples; division is solvable inside these groups, although uniqueness is unnecessary. This criterion is specific to the integer ring, not a characterization of injectives over arbitrary rings. Categories of left modules have enough injectives, so every module embeds in one. An arbitrary abelian category need not have enough injectives or enough projectives; resolution constructions must explicitly assume the relevant existence property.

J injective  ⟺  Hom⁡R(−,J) exact,F right-flat  ⟺  F⊗R− exactJ\text{ injective}\iff\operatorname{Hom}_R(-,J)\text{ exact},\qquad F\text{ right-flat}\iff F\otimes_R-\text{ exact}
Extension uses a contravariant Hom functor; right-module flatness uses tensoring left modules.↔ Scroll wide equations horizontally.

Flatness is exactness after tensoring

Tensor product is always right exact. For a right module FF, flatness says that F⊗R−F\otimes_R- also preserves injections of left modules, hence all short exact sequences. Every right projective module is flat because it is a summand of a free module, and tensoring with a free module forms a direct sum. Flatness does not generally supply lifts of maps; it only controls how relations change after tensoring, which is weaker than the projective lifting condition.

For a commutative ring, a finitely presented flat module is projective. Omitting finite presentation changes the statement. The integer module Q\mathbb Q is flat because it is a localization, but it is not projective: projective integer modules are free, and a nonzero free abelian group is not divisible. Thus localizing can remove torsion without creating a basis suitable for arbitrary lifting. When selecting a resolution, decide whether the calculation needs exact tensoring, exact Hom, or both before choosing its building blocks.

One module fails two different tests

Show directly that M=Z/2ZM=\mathbb Z/2\mathbb Z is neither projective nor flat as an integer module. Identify the particular map that fails in each test.

  1. If MM were projective, the quotient q:Z→Mq:\mathbb Z\to M would admit a section ss. The element s(1‾)s(\overline1) would satisfy 2s(1‾)=02s(\overline1)=0 in the integers, forcing it to be zero.

  2. That forced value contradicts q(s(1‾))=1‾q(s(\overline1))=\overline1. This proves nonprojectivity using an actual lifting problem, rather than simply observing that the module has a finite presentation.

  3. Tensor the injection Z→2Z\mathbb Z\xrightarrow{2}\mathbb Z with MM. Both resulting groups identify with MM, but the resulting map is multiplication by two, hence zero. Its source is nonzero, so the map is not injective and MM is not flat.

The result

The obstruction to projectivity is a missing section, whereas the obstruction to flatness is a lost injection. They are related but distinct witnesses.

Explore the model

Interactive lab

Interactive lab · modular

Detect annihilated elements

Pause/Reset, then set both modulus to 8 and additive step to 2. Predict period four and an image of four residues. After four Step presses, compare the return to zero. Reset and set both modulus to 8 and step to 3; predict period eight. The missing residues in the first image correspond to a noninjective endomorphism of the finite group.

Detect annihilated elementsAddition modulo m starts at zero: rₙ = na mod m. Playback advances two additions per second. The orbit has period m/gcd(m,a), including period one for a = 0. This is not a cryptographic security test or a primality algorithm. Current readings: Residue r0: 0; Orbit period · m/gcd(m,a): 12; Distinct residues visited: 1 of 12. All parameters are adjustable with the labeled controls below.Repeated addition modulo 12: the current residue is 0, and the orbit period is 12.01234567891011+ 5mod 12An additive orbitn = 0rₙ = 0Period = 12Cycle starts at 0:0 → 5 → 10 → 3 →8 → 1 → 6 → 11 →4 → 9 → 2 → 7 → 0Exact residues · two additions per second
Additive orbitCurrent residue / next edge
Residue r0
0
Orbit period · m/gcd(m,a)
12
Distinct residues visited
1 of 12
t = 0 s

Addition modulo m starts at zero: rₙ = na mod m. Playback advances two additions per second. The orbit has period m/gcd(m,a), including period one for a = 0. This is not a cryptographic security test or a primality algorithm. Parameter changes pause playback and recompute the model at the current elapsed time. Reset restores both parameters and returns time to zero. Playback starts paused. Adjust either parameter, then choose Play or Step.

Look forFinite cyclic maps expose torsion relations, but testing one map is only a witness of failure, not a universal proof of flatness or projectivity.

Check your understanding

Which implication holds for every commutative ring without finiteness assumptions?

Try it yourself

Explain why the injection Z→5Z\mathbb Z\xrightarrow{5}\mathbb Z remains injective after tensoring with Q\mathbb Q. Then decide whether this single test proves that the rationals are projective. Give a correct reason for flatness and a reason projectivity fails.

Take a moment to make a prediction or work it out on paper. Open a hint when you need one.

Give me a hint

Separate invertibility of five after localization from the lifting property of a projective module.

Compare with the solution

The tensor map becomes multiplication by five on Q\mathbb Q, an isomorphism with inverse division by five. One successful test proves neither flatness nor projectivity. Flatness follows because localization of a commutative ring is flat. Projectivity fails because every projective integer module is free, while Q\mathbb Q is nonzero and divisible and no nonzero free abelian group is divisible. The three assertions therefore have different justifications.

Take it with you

  • Projectives lift maps; injectives extend maps.
  • Tensoring detects flatness through preservation of all injections.
  • Finite presentation is essential in the general flat-to-projective criterion.

Read further

Original sources and deeper explanations to follow at your own pace.

Optional progress, saved in this browser when available. Revisit any lesson at any time.

Lesson not marked complete.