A formality obstruction that sees beyond ordinary Massey products
The Bianchi–Massey tensor captures how several cup-product relations fit together at chain level. In a precise connectivity and dimension range it completes the cohomology algebra as rational-homotopy data and detects non-formality that ordinary Massey products can miss.
This guide belongs to The Gold Standard of Topology, within The Gold Standard of Education and the verified How X Works hub.
Choose a reading route
Chapter contents
Open all twelve chapters
- 1. Formality asks whether cohomology tells the whole rational story
- 2. Vanishing Massey products do not always prove formality
- 3. Begin with the kernel of cup product
- 4. The Bianchi domain cancels the obvious fourth-order product
- 5. Worked bookkeeping: from exact products to a cohomology class
- 6. A ledger prevents a symbolic calculation from overclaiming
- 7. The Crowley–Nordström theorem has a precise range
- 8. Non-formal manifolds can hide behind zero Massey products
- 9. Relation to Sullivan minimal models
- 10. Changed context: sphere bundles add hypotheses
- 11. Evidence checks for a formality claim
- 12. The Gold Standard conclusion
1. Formality asks whether cohomology tells the whole rational story
CHAPTER 1 OF 12 · Orient
A differential graded algebra is formal when it is connected to its cohomology algebra, equipped with zero differential, by quasi-isomorphisms. In that case rational homotopy calculations can discard many chain-level choices without losing the rational homotopy type.
Non-formality means that multiplication in cohomology has missed higher coherence data. Massey products are familiar witnesses, but their definitions depend on vanishing products and choices. The Bianchi–Massey tensor packages a wider, choice-independent obstruction in a controlled range.
2. Vanishing Massey products do not always prove formality
CHAPTER 2 OF 12 · Build
Every formal space has vanishing Massey products whenever they are defined. The converse fails: there are spaces and manifolds for which all ordinary Massey products vanish while the space is still non-formal.
That logical direction matters. A non-zero Massey product is evidence against formality; a list of zero products may only show that the chosen probes missed the obstruction. The Bianchi–Massey tensor was designed to see some of this blind spot.
3. Begin with the kernel of cup product
CHAPTER 3 OF 12 · Build
Let H be the graded cohomology algebra and let μ multiply pairs of classes. The kernel E of the graded-symmetric multiplication map contains formal linear combinations of products that vanish in cohomology.
Each element of E therefore admits a chain-level null-homotopy after choosing cocycle representatives. Those choices are precisely where higher information can hide: the product is zero as a cohomology class, but the way it becomes exact can interact with other exact products.
4. The Bianchi domain cancels the obvious fourth-order product
CHAPTER 4 OF 12 · Build
The tensor is not evaluated on every pair of elements of E. Its natural domain is a subspace of graded-symmetric combinations whose image in the fourth symmetric power of H cancels. This Bianchi-style symmetry removes the part already determined by ordinary cup products.
What remains is a place where chosen primitives can combine to form closed cochains. Their cohomology classes provide the tensor value, while the domain restriction makes that value independent of the auxiliary choices.
5. Worked bookkeeping: from exact products to a cohomology class
CHAPTER 5 OF 12 · Work
Suppose e₁ and e₂ are two independent relations in E. Choose cocycle representatives for the underlying cohomology classes and primitives γ₁, γ₂ satisfying dγᵢ equal to the represented product relation. A valid Bianchi combination b built from e₁ and e₂ has its obvious fourfold product cancel.
The construction then combines represented products with the opposite primitives, with graded signs, to produce a closed cochain. Passing to cohomology gives F(b). Changing cocycle representatives or primitives changes the cochain by an exact term, so the resulting class is an invariant.
| Stage | Object | Verification |
|---|---|---|
| Cohomology | Graded algebra H | Degrees and cup products |
| Relations | E = kernel of multiplication | Each relation maps to zero |
| Domain | Bianchi-symmetric b | Fourfold product cancels |
| Chain choice | Primitive γ for each relation | dγ equals represented relation |
| Output | Closed cochain and class F(b) | Choice changes are exact |
6. A ledger prevents a symbolic calculation from overclaiming
CHAPTER 6 OF 12 · Work
A calculation should record degrees, cup-product relations, a basis for E, the Bianchi-domain relations, the chosen primitives and the final cohomology class. If the candidate expression is not closed, a sign or domain condition is wrong.
If the tensor evaluates to zero on a selected element, only that element has been tested. Proving the tensor vanishes requires evaluating a spanning set of its domain or using a theorem that forces vanishing.
7. The Crowley–Nordström theorem has a precise range
CHAPTER 7 OF 12 · Connect
For a closed (n−1)-connected manifold of dimension at most 5n−3, with n greater than one, Crowley and Nordström show that the rational homotopy type is determined by the cohomology algebra together with the Bianchi–Massey tensor. In that range the manifold is formal exactly when the tensor vanishes.
Outside the stated connectivity and dimension range, the tensor remains meaningful but need not be the only missing invariant. Quoting the equivalence without the hypotheses turns a sharp theorem into a false universal test.
8. Non-formal manifolds can hide behind zero Massey products
CHAPTER 8 OF 12 · Transfer
The same work constructs many highly connected manifolds in dimensions 4n−1 that are non-formal even though they have no non-zero Massey products. This is the central transfer lesson: the absence of familiar higher products does not close the case.
The tensor combines relations collectively. An obstruction may live in how several null products fit together even when no single ordinary Massey bracket exposes it.
9. Relation to Sullivan minimal models
CHAPTER 9 OF 12 · Connect
A Sullivan minimal model records generators and differentials that build the rational homotopy type. Formality asks whether this model can collapse, up to quasi-isomorphism, to cohomology with zero differential.
The Bianchi–Massey tensor can be read as structured information carried by the first relevant higher multiplication. It complements, rather than replaces, the minimal-model method: one supplies a computable invariant in a range; the other remains the general language.
10. Changed context: sphere bundles add hypotheses
CHAPTER 10 OF 12 · Transfer
Recent work on orientable sphere bundles shows that, when the base is formal, vanishing of an appropriate Bianchi–Massey tensor can characterise formality of the total space under the theorem’s conditions. Euler classes and the cohomology of the base control the calculation.
This is not permission to test any fibration with one tensor value. Check orientability, compactness, base formality, degree conditions and the precise statement being applied.
11. Evidence checks for a formality claim
CHAPTER 11 OF 12 · Audit
State the coefficient field, connectivity, dimension, Poincaré duality assumptions and the algebraic model. Give the cup-product kernel and the tensor domain, not only a verbal claim that the tensor was computed.
For a vanishing result, show why every domain element is covered. For a non-vanishing result, exhibit one element with a non-zero target class and verify independence from the chosen representatives.
| Claim | Minimum evidence | Common overreach |
|---|---|---|
| A Massey product vanishes | Defined bracket and indeterminacy | All higher data vanish |
| Tensor vanishes | Spanning domain checked | One element evaluated |
| Manifold is formal | Vanishing plus theorem hypotheses | Ignoring dimension range |
| Spaces share rational type | Cohomology and tensor in range | Matching Betti numbers only |
12. The Gold Standard conclusion
CHAPTER 12 OF 12 · Use
Use ordinary Massey products as useful but incomplete probes. When the theorem’s range applies, compute the Bianchi–Massey tensor to close the blind spot and combine it with the cohomology algebra.
The disciplined conclusion has three parts: what invariant was computed, which theorem turns it into a formality statement, and which hypotheses limit the transfer.
A changed-context transfer check
A simply connected seven-manifold has all computed triple Massey products zero. That evidence does not by itself prove formality. In the relevant highly connected low-dimensional setting, compute the Bianchi–Massey tensor or use a theorem that forces its vanishing.
If the manifold falls outside the connectivity or dimension range, a zero tensor is still information but may not be complete. Return to the minimal model and check whether further higher operations can survive.
Frequently asked questions
Does a formal space have zero Bianchi–Massey tensor?
Yes. Formality removes the higher chain-level obstruction measured by the tensor.
Do zero Massey products imply a zero tensor?
Not in general. Crowley and Nordström exhibit non-formal manifolds with no non-zero ordinary Massey products.
Is the tensor determined only by Betti numbers?
No. It uses the graded cup-product structure and chain-level null-homotopy information, not only dimensions of cohomology groups.
Can the low-dimensional theorem be applied to every space?
No. Its connectivity, closed-manifold and dimension hypotheses are essential; outside them further invariants may be needed.
Authoritative sources and evidence route
- Crowley and Nordström: the rational homotopy type of highly connected manifolds and the Bianchi–Massey tensor
- University of Bath research record for the Journal of Topology publication
- Zhou: Formality of sphere bundles, Algebraic & Geometric Topology
- Massey products, defining systems and indeterminacy
- Formality, minimal models and vanishing Massey products