Manifolds 26 | Ricci Calculus [dark version]

Ricci calculus—often called tensor calculus—is presented as a coordinate-based shorthand for doing differential geometry on manifolds. The central payoff is practical: by working in local charts and tracking indices, calculations become shorter and more systematic, which is especially useful in physics. The tradeoff is conceptual: the index machinery can make it easier to compute while risking a weaker connection to the underlying geometric ideas.

The setup starts with a manifold M and a local chart (U, h), where h maps points in U to R^n. In Ricci calculus, attention shifts from the full map to its coordinate component functions, written as x^1, …, x^n (or equivalently as component functions of the chart). This coordinate viewpoint aligns with the tangent space T_pM: the standard basis of T_pM is built by pushing forward the canonical basis of R^n using the parameterization. Derivatives with respect to these coordinates then naturally appear as the coordinate vector fields, denoted using an operator ∂ with indices.

A tangent vector at p is expressed in that coordinate basis as a linear combination of basis vectors. The notation becomes compact through the Einstein summation convention: whenever an index appears once upstairs (superscript) and once downstairs (subscript), summation over that index is automatic, so explicit Σ signs are omitted. This single convention drives much of the brevity—and it also forces careful bookkeeping about where indices sit.

That bookkeeping is tied to how vectors transform. Components with superscripts are labeled contravariant vectors, while components with subscripts are covariant vectors. The distinction matters because these two types behave differently under coordinate changes, even though the calculus itself is carried out through component computations.

To show how geometry is encoded in indices, the transcript uses the inner product on a tangent space T_pM. In abstract form, an inner product takes two tangent vectors and returns a real number, capturing lengths and angles. In Ricci notation, the same information is stored in a matrix G (the metric tensor), and the inner product becomes a component expression where indices are placed downstairs so that the Einstein convention produces the required double sum over indices.

Finally, the discussion moves to dual objects. A covector (dual to a contravariant tangent vector) has components with subscripts, and its natural basis pairs with tangent vectors using the Einstein convention. The key example is dx^j: a one-form, interpreted as a linear map from the tangent space to R. When dx^j is evaluated on the coordinate basis vector ∂_k, it returns 1 if j = k and 0 otherwise—captured by the Kronecker delta δ^j_k (with equivalent placements of indices depending on convention). The transcript closes by noting that one-forms will be formalized next as linear maps from tangent spaces to R, setting up the integration machinery to come.