Infinitesimals, Tangent Spaces, Derivations and Jets
Infinitesimal geometry can be encoded algebraically through functions modulo higher-order vanishing, derivations, jets and differential operators.
This handbook article treats Infinitesimals, Tangent Spaces, Derivations and Jets as a connected mathematical system rather than a list of isolated definitions. The source develops the subject through definitions, examples, structural correspondences, formulas and diagrams. The practical reading strategy is to identify the objects under discussion, state the permitted operations, separate assumptions from consequences, and then test every construction against the examples supplied in the source.
Core concepts
Functions modulo second-order terms
Near a point, functions that vanish to second order can be discarded when only first-order behaviour matters. The resulting quotient separates the value of a function from its linear infinitesimal part.
Maximal ideals and tangent directions
Functions vanishing at a point form a maximal ideal m. The quotient m/m² captures first-order vanishing and its dual gives an algebraic tangent space.
Derivations
A derivation is a linear map satisfying the Leibniz rule. Evaluation of directional derivatives at a point produces derivations, and derivations provide a coordinate-free algebraic description of tangent vectors.
Singular points
For an algebraic set defined by equations, the rank of first derivatives determines whether the expected tangent dimension is achieved. Singular points are precisely where first-order constraints lose independence.
Higher-order jets
Replacing m² by higher powers mᵏ⁺¹ retains Taylor information through order k. Jets package higher-order local behaviour without requiring an actual infinitesimal number.
Differential operators
Higher-order differential operators can be defined algebraically by how repeated commutators with multiplication operators reduce order, linking local algebra to differential calculus.
How the ideas fit together
Infinitesimal geometry can be encoded algebraically through functions modulo higher-order vanishing, derivations, jets and differential operators.
The source's recurring method is structural. It begins with a class of mathematical objects and specifies operations or maps, then asks what can be proved from those rules alone. This is why definitions matter more than notation: two apparently different systems can be treated together when they satisfy the same defining laws, while two expressions that look similar can behave differently if their ambient structures differ.
Within this topic, Functions modulo second-order terms provides the entry point. The later ideas—Maximal ideals and tangent directions, Derivations, Singular points, Higher-order jets, Differential operators—either refine that first structure, construct new objects from it, or describe information preserved by a suitable map. Read the topic as a sequence of dependencies rather than as independent vocabulary.
Whenever the source passes to a quotient, extension, decomposition or representation, keep two questions visible: what information is deliberately forgotten? and what information is preserved? Those questions explain why quotient objects, extension structures and invariant quantities appear repeatedly across algebra. They are mechanisms for changing the form of a problem without losing the relationships that the theory is designed to study.
The examples also serve as boundary tests. A finite example can prove that an unusual structure is possible; a function-ring example can reveal zero divisors; a geometric example can show how an abstract invariant recovers visible shape; and an operator example can show why multiplication may become noncommutative. The safest study practice is therefore to move in both directions: derive consequences from the definition and then use an example to test whether the consequences have been understood correctly.
A reliable way to reason through the topic
1. Identify the ambient structure. Before manipulating symbols, determine what kind of objects are present and which operations are actually defined. In this topic, the central ideas include Functions modulo second-order terms, Maximal ideals and tangent directions, Derivations. Results that are valid in one algebraic setting do not automatically transfer to another simply because the notation looks similar.
2. Track closure and compatibility. Algebraic definitions are built from operations that must remain inside the chosen structure and satisfy specified laws. When a map or construction is introduced, check which laws it preserves. This prevents a common error: using an operation that exists in a familiar number system but has not been established in the current setting.
3. Separate representation from structure. A matrix, polynomial, coordinate tuple, diagram or formula may represent an object without being the object itself. Isomorphism and other structure-preserving maps are important precisely because they allow different representations to express the same underlying algebraic organisation.
4. Use examples as tests, not universal rules. The source repeatedly uses finite systems, function spaces, geometric models and operator examples to expose what a definition permits. An example demonstrates possibility and mechanism; it does not by itself turn its numerical values or special properties into a general axiom.
5. Look for invariants and quotients. Once a structure and its maps are understood, the next question is what survives a change of coordinates, decomposition or identification. Dimensions, kernels, images, quotient objects, factor structures and equivalence classes are recurring devices for retaining essential information while removing representational detail.
Key symbolic relationships
The defining product rule for derivations.
First-order vanishing functions modulo second-order vanishing form the cotangent space in algebraic form.
The quotient retains information through order k around the chosen point.
Examples and what they demonstrate
| Example | Structural lesson |
|---|---|
| Smooth point of a plane curve | For F(x,y)=0, the linearised equation Fₓdx+Fᵧdy=0 describes the tangent direction when the gradient does not vanish. |
| Singular curve point | If both first partial derivatives vanish at a point, the first-order tangent condition degenerates and higher-order terms become necessary. |
| Jet of a function | Two functions have the same k-jet at a point when their difference vanishes to order k+1. |
How the source diagrams support the mathematics
- The source uses formulas, structural diagrams and worked examples to move from definitions to invariant properties.
- This article converts those visual and symbolic relationships into responsive cards, process sequences and formula panels rather than reproducing page images.
The web article expresses the purpose of these visuals with responsive HTML/CSS rather than embedding scanned source pages.
Common mistakes to avoid
- Treating a source example as if it were an additional axiom or a universal numerical requirement.
- Using familiar arithmetic operations before confirming that the current structure supports them.
- Confusing an object with one particular coordinate, matrix, polynomial or diagram used to represent it.
- Assuming that a property preserved by an isomorphism is also preserved by every map.
- Skipping the domain, codomain, coefficient field or scalar ring when interpreting a formula.
- Forgetting that quotient constructions identify whole equivalence classes rather than deleting inconvenient elements.
Verification questions
- Can you define the central objects in Infinitesimals, Tangent Spaces, Derivations and Jets without relying on a single example?
- Can you explain why Functions modulo second-order terms is structurally different from Differential operators?
- Can you state the role of each operation in the principal formulas and identify where it is defined?
- Can you distinguish an equality of objects from an isomorphism between differently represented objects?
- Can you reconstruct at least one source example from its defining rules rather than memorising the finished result?
- Can you identify which conclusions depend on extra hypotheses such as finiteness, irreducibility, commutativity or finite generation?