1.10.1 Tensor Algebra Notation

Tensor Algebra Notation is the overall system of symbols, formats, and conventions used in practice to write tensors and the operations performed on them, encompassing several distinct notational styles, index notation, direct notation, and graphical notation among them, each suited to different purposes, and each translatable into the others once the underlying tensor is properly understood. Where the foundational notational conventions establish the basic rules, index placement, summation, the metric, this broader notation covers the full range of concrete writing systems that those rules support, along with the practical considerations that determine which notational style is appropriate in a given setting.

Tensor algebra notation exists because a single tensor can be expressed in multiple equivalent ways depending on whether the emphasis is on its components in a specific basis, its identity as a basis-independent object, or the pattern of operations connecting several tensors together. Fluency in tensor algebra requires the ability to move between these notational styles as needed, recognizing that a change in notation never changes the underlying mathematical object being described.

Index Notation

Component-Level Expression

Index notation, sometimes called Ricci calculus, represents a tensor through its components relative to a chosen basis, using superscripts for contravariant indices and subscripts for covariant indices, together with the Einstein summation convention for repeated indices. This is the most computationally direct notation, since every operation, contraction, tensor product, index raising, corresponds to an explicit and mechanical manipulation of indices.

Strengths and Limitations

Index notation excels at making explicit precisely which indices are summed, which survive, and how the type of a tensor changes under an operation, which is invaluable for verifying that a lengthy computation is well-formed. Its main limitation is verbosity: expressions involving many tensors and many indices can become difficult to read, and the notation ties the expression to a specific coordinate system, obscuring the basis-independent nature of the underlying tensor.

Direct (Coordinate-Free) Notation

Emphasizing the Tensor as an Object

Direct notation, also called symbolic or invariant notation, refers to tensors and their operations without reference to indices or a chosen basis at all, writing expressions such as T(u, v) for a bilinear form evaluated on two vectors, or A ⊗ B for a tensor product, or tr(T) for a trace. This style emphasizes that a tensor is fundamentally a basis-independent object, a multilinear map or an element of an abstract tensor space, and that its components are merely one particular representation of it.

When Direct Notation Is Preferred

Direct notation is typically preferred when stating general theorems, definitions, or structural relationships that hold regardless of the basis chosen, since it avoids the risk of a reader mistaking a coordinate-dependent detail for an intrinsic property. It is less convenient, however, for carrying out explicit numerical computations, which is why index notation remains dominant in applied calculations.

Matrix and Dyadic Notation

Matrices as a Special-Case Notation

For rank-2 tensors specifically, matrix notation offers a compact rectangular arrangement of components, together with the familiar operations of matrix multiplication, transposition, and inversion. This notation is a special case that applies cleanly only to two-index tensors and does not extend gracefully to tensors of rank three or higher, which is one of the main motivations for adopting general index or direct notation once tensors beyond rank two are involved.

Dyadic Notation

Dyadic notation writes a rank-2 tensor as a formal sum of outer products of vectors, such as T = Σ e_i ⊗ e_j T^ij, blending features of both index and direct notation: it retains explicit basis vectors while treating the outer product as a formal algebraic operation rather than a purely numerical one. This notation is common in physics and engineering contexts where rank-2 tensors, such as stress or inertia tensors, are described in terms of physical basis directions.

Graphical (Diagrammatic) Notation

Representing Tensors as Nodes

Graphical tensor notation, associated with Penrose diagrams, represents each tensor as a shape, often a box or circle, with a number of lines emerging from it equal to the tensor's rank: one line per index. Contraction between two tensors is drawn by connecting a line from one shape to a line from another, visually representing the summation over the shared index. Free indices are drawn as lines that remain unconnected, extending outward from the diagram.

Practical Value of Diagrams

Graphical notation is particularly effective for tracking complex networks of contractions among many tensors simultaneously, since the pattern of connecting lines makes the overall structure of a computation visually apparent in a way that a long string of indices often does not. It is less suited to expressing algebraic identities symbolically or to precise written derivations, where index or direct notation remains standard.

Translating Between Notational Styles

A Single Underlying Object

Regardless of which notation is used, index, direct, matrix, dyadic, or graphical, the object being described is the same tensor, and any correct manipulation performed in one notation corresponds to an equivalent, correct manipulation in every other notation. Learning to translate confidently between these styles is what allows a computation carried out efficiently in index notation to be checked structurally against a diagram, or stated compactly in direct notation for inclusion in a general theorem.

Choosing a Notation for the Task

In practice, the choice of notation is governed by the task at hand: index notation for explicit computation and verification, direct notation for stating general results, matrix or dyadic notation for low-rank tensors tied to a physical basis, and graphical notation for visualizing complex contraction patterns. Tensor algebra notation, taken as a whole, is this entire toolkit together with the judgment needed to select the right tool for a given purpose.