Elementary symmetric polynomial¶

The degree-k sum of one product for every k-element subset of a finite variable list.

Version: v1 · 2026-10-17 · History
Domain-specific #: 14138
Domain group: Formal Sciences
Origin domain: Mathematics
Subdomains: Commutative Algebra, Symmetric Polynomial Theory → Mathematics
Aliases: Elementary symmetric function

Core Idea¶

For a finite list of formal variables x_1,...,x_n, the kth elementary symmetric polynomial e_k adds the product of the variables at each k-element subset of the index set. Each subset is used once. For three variables, e_2=x_1x_2+x_1x_3+x_2x_3: the three possible pairs, and no square terms. This precise selection rule, rather than permutation invariance alone, identifies the abstraction.^[1]^[2]

The usual family includes e_0=1, the empty-product case, and ends at e_n=x_1...x_n; no k-subset exists when k>n. Its members fit together in a single coefficient identity: ∏_(i=1)^n (1+x_i t)=∑_(k=0)^n e_k t^k. Picking either 1 or x_i t from each factor explains why the coefficient of t^k contains exactly the products from k selected positions.^[3] The variables may be formal indeterminates over a commutative coefficient ring; evaluating them at particular values is a later operation.^[2]

Every e_k is a symmetric polynomial, since permuting variable positions merely permutes the included subsets. The converse is false: x_1^2+...+x_n^2 is symmetric but is not one e_k. The theorem that general symmetric polynomials can be expressed using the family of elementary members describes their expressive reach; it does not widen the definition of one member.^[1]^[3]

Structural Signature¶

Finite indexed carrier: n named variable positions in a commutative polynomial ring. The positions matter even when two variables are later assigned the same numerical value; commutativity makes reordering a selected product harmless.^[2]
Degree choice: an integer k with 0≤k≤n selects subsets of exactly k distinct index positions. It fixes both the degree and which products qualify.^[1]
Exhaustive squarefree sum: multiply the variables in each eligible subset, then add every such product exactly once. Omitting a subset, repeating a variable within a product, or changing a term's coefficient changes the formal polynomial.^[1]^[3]
Family encoding, derived rather than required for one member: the product ∏(1+x_i t) collects all e_k as coefficients. It supplies a useful computation and explanation of the family's relations, while any single e_k is already defined by its subset sum.^[3]

What It Is Not¶

The entry is narrower than symmetric polynomial. A permutation-invariant polynomial can be a power sum such as p_2=∑x_i^2, or a polynomial such as e_1^2 built from elementary members. Neither expression is a single e_k, because its formal terms do not have the one-squarefree-product-per-k-subset support.^[1]^[3] A permutation-invariant quantity need not even be a polynomial, so the candidate seed's unqualified “any invariant quantity” claim is rejected.

It is also narrower than a formula or theorem using the family. Vieta's coefficient formulas apply e_k to roots; the fundamental theorem of symmetric polynomials represents other symmetric polynomials in the generators e_1,...,e_n. Neither result defines a different elementary member.^[1] Maclaurin's inequality is a separate theorem about values of the family. A list of selected pair products that omits one possible pair is simply an incomplete sum, even if it resembles the beginning of e_2.

Scope of Application¶

In commutative algebra, the rule constructs a formal polynomial from finitely many variables over a commutative coefficient ring. It makes sense without roots, matrices, positivity, or an ordering of numerical values.^[2] In symmetric-polynomial theory, the family supplies generators for permutation-invariant polynomials; in the field setting of the MIT notes, every symmetric polynomial has a unique expression in those generators.^[1] “Generator” refers to the whole family, not to every symmetric polynomial being one elementary member.

For a monic polynomial whose roots are supplied in an algebraic closure, the signed values of e_k at those roots are coefficients of the expanded root product.^[1] For a finite-dimensional complex operator, the eigenvalues counted with algebraic multiplicity fill the indexed-variable roles, and the characteristic polynomial is another root product; e_1 gives the sum of eigenvalues and e_n their product, the trace and determinant respectively.^[4]^[3] These are literal evaluations of the same formal construction, not a license to import a numerical result without its ring, dimension, and multiplicity assumptions.

The generating identity also admits a formal weighted-subset reading: treat each x_i as a commuting item weight, and the coefficient of t^k sums product weights of all k-item selections. This is a direct interpretation of the printed identity, not a separately reported application in that source.^[3]

Clarity¶

The subset test resolves three questions that the adjective “symmetric” leaves open: which variable positions may occur together, how many are chosen, and how many times each product is counted. For n=3, compare e_2=x_1x_2+x_1x_3+x_2x_3 with p_2=x_1^2+x_2^2+x_3^2. Both are degree two and invariant under permutations; only the first has the elementary degree-two support.^[1]^[3]

Formal identity must also be separated from numerical evaluation. At x_1=x_2=x_3=1, several different symmetric polynomials may yield the same number. That coincidence cannot establish that their formal monomial supports match. In an operator example, repeated eigenvalues still occupy distinct positions according to algebraic multiplicity; index subsets, not distinct numerical values, are what the rule enumerates.^[4]

Manages Complexity¶

An expanded e_k has one formal term for each of the binom(n,k) index subsets. The product identity records all degrees in one compact expression, and multiplying one further factor updates the coefficient list instead of separately rebuilding every subset sum.^[3] This is a structural compression: the coefficient of t^k retains the exhaustive subset rule even when the product is left factored.

The full family also compresses general symmetric-polynomial reasoning. Once the coefficient ring and variable list are fixed, a symmetric polynomial can be expressed through elementary generators under the standard hypotheses, replacing an arbitrary permutation-invariant expression with a polynomial in e_1,...,e_n.^[1]^[2] The reduction does not make the resulting expression itself a single elementary polynomial. Keeping that distinction prevents a compact basis representation from erasing the class boundary.

Abstract Reasoning¶

The construction licenses a direct recognition test. Declare n and k; enumerate or characterize every k-subset of the index positions; compare its product with the proposed formal polynomial's monomials and coefficients. A missing pair rejects the candidate. A square term rejects e_2, even if the overall polynomial remains symmetric. For k=0, the empty subset contributes 1; for k=n, the only subset contributes the product of all variables.^[1]^[3]

The product encoding then licenses coefficient reasoning. Expanding ∏(z-α_i) shows that choosing k negative root terms and n-k copies of z gives coefficient (-1)^k e_k(α_1,...,α_n) at z^(n-k).^[1] This is an inference from the selection mechanism, so it preserves multiplicity and signs rather than treating “root sum” as an informal shortcut. It does not require the roots to be distinct. The same selection logic explains the characteristic-polynomial use without relying on a damaged extracted coefficient equation in the linear algebra source.^[4]

Knowledge Transfer¶

The pattern transfers literally between formal symmetric-polynomial calculations, supplied polynomial roots, and finite operator spectra: there is always a finite indexed variable list, a fixed k, and one product per eligible index subset. Roots and eigenvalues change what fills the variable positions; they do not change the subset rule.^[1]^[4]

The weighted-subset interpretation preserves those roles as well, but is a derived use of the generating formula, not evidence that a source tested an external inventory method.^[3] Outside commutative polynomial arithmetic, “combine every k things” may be a useful analogy, but it is not an elementary symmetric polynomial unless multiplication, addition, the variable carrier, and the exact formal rule are present.

Examples¶

Canonical — three-variable pair sum¶

For formal variables a,b,c, the second member is e_2(a,b,c)=ab+ac+bc.^[1] Choosing k=2 produces the index pairs {1,2}, {1,3}, and {2,3}. Each gives one degree-two product, and all three appear once. a^2+b^2+c^2 would still be symmetric, but it selects repeated positions instead of pairs.

Mapped back: commutative polynomial ring in a,b,c → pair-index subsets → ab, ac, bc exactly once → coefficient of t^2 in (1+at)(1+bt)(1+ct).

Applied — roots and a diagonal operator¶

Suppose a monic cubic has roots 2,3,5. Its pair sum is e_2=2·3+2·5+3·5=31, and direct expansion gives (z-2)(z-3)(z-5)=z^3-10z^2+31z-30.^[1] A diagonal complex operator with eigenvalues 2,3,5 fills the same positions; its characteristic polynomial has those coefficients, with e_1=10 as trace and e_3=30 as determinant.^[4]^[3] The polynomial-root and operator cases have different starting objects but preserve the selected-variable roles.

Mapped back: three roots or three eigenvalue positions → choose two positions → add each pair product to obtain 31 → read 31 as the degree-one coefficient of the monic cubic; keep the alternating signs for the other degrees.

Derived reading — weighted selections¶

Assign formal weights u,v,w to three labeled items. The coefficient of t^2 in (1+ut)(1+vt)(1+wt) is uv+uw+vw: the sum of product weights of all two-item selections.^[3] This illustrates the source's generating identity; it is not a claim that the source documented an external inventory use.

Mapped back: three commuting weights → all two-item index subsets → one product per subset → coefficient of t^2.

Structural Tensions¶

T1 — Catalog breadth vs. identity precision. The broad symmetric-polynomial class covers power sums and arbitrary expressions in elementary generators, while one e_k has the specific squarefree, coefficient-one subset test.^[1]^[3] Treating e_k only under that broad label saves a catalog distinction but cannot by itself license an e_k coefficient or subset calculation. Giving the elementary family its own entry preserves that diagnostic, at the cost of a finer distinction whose relation to the broad class must stay explicit. Diagnostic: Does the task need only permutation invariance, or must it certify one product for every k-subset?

T2 — Formal support vs. evaluated value as evidence. The formal sum exposes which index subsets contribute. Evaluation at specific values is useful for root coefficients or a matrix invariant, but equal or zero values can collapse distinct contributions. Both operations can be performed; the tradeoff arises when one numerical value is used instead of formal support as evidence of membership. Diagnostic: Is the claim about the polynomial's term structure, or only its value at one input?

Structural–Framed Character¶

The construction is strongly structural inside algebra: index positions, fixed cardinality, subset products, and a formal sum define it independently of a named theorem or a preferred application. Its evaluative weight is low: calling a polynomial elementary records a formal identity, not an endorsement or a preferred social outcome. Its definition does not depend on a human practice or institution, although teaching and notation transmit the name. Its vocabulary travels literally among algebra, root calculations, and operator spectra because each has an indexed list of commuting values; outside those settings, “subset” and “aggregate” alone do not carry the polynomial test.

The pattern is recognized by checking a candidate's formal terms, not imported as a rhetorical frame onto a setting with no polynomial ring. A broad selection metaphor can suggest a comparison, but it cannot decide membership. Its character: structurally precise and weakly framed by human convention, yet domain-bound by commutative polynomial operations and the one-per-k-subset diagnostic.

Structural Core vs. Domain Accent¶

The portable skeleton selects all fixed-size subsets and aggregates a contribution from each. The identity-bearing core is more exact: the carrier is a commutative polynomial ring, each contribution is the product of variables at distinct index positions, the sum includes every eligible subset once, and the result is a formal degree-k polynomial. These are mathematical conditions, not decorative terminology.^[2]^[3]

The live Prime Symmetry captures permutation invariance but does not by itself express this entire select-and-aggregate skeleton. A distinct future Prime for that skeleton would need two unlike nonalgebraic systems in which all fixed-size subsets are selected, each produces the same typed contribution, and aggregation preserves a diagnostic that supports new reasoning. It would fail that test if the examples merely share a “choose some things” metaphor, use different selection rules, or gain no invariant beyond a loose analogy. No such broader identity is asserted here.

One can use the skeleton to think about weighted selections in other settings. Calling every such aggregation an elementary symmetric polynomial would be unwarranted when its “weights” do not multiply in a commutative coefficient ring, when the sum is incomplete, or when the object is merely a numerical summary with no declared polynomial. The exact formal rule explains both why the construction travels among algebraic settings and why it remains domain-specific.

This entry is a kind of Symmetric polynomial.

Permuting variable positions leaves e_k unchanged, a literal instance of the live Symmetry pattern of invariance under transformation.^[1] That invariant alone is insufficient to identify an elementary member: p_2 passes it and fails the subset test.^[3] The broader Symmetric polynomial class contains both; Polynomial names a still broader formal carrier.

Relationships to Other Abstractions¶

Current abstraction Elementary symmetric polynomial Domain-specific

Parents (1) — more general patterns this builds on

Elementary symmetric polynomial is a kind of Symmetric polynomial Domain-specific

Every elementary k-subset polynomial is permutation-invariant, with exhaustive squarefree support as its differentia.

Hierarchy path (1) — routes to 1 parentless root

Elementary symmetric polynomial → Symmetric polynomial → Symmetry

Neighborhood in Abstraction Space¶

Elementary symmetric polynomial sits in a sparse region of the domain-specific corpus (69^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Field Theory & Polynomial Structures (26 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-10-17

Not to Be Confused With¶

Symmetric polynomial: the whole class fixed by variable permutations; an elementary member is a special case with exhaustive squarefree degree-k support.^[1]
Power sum at degree two or higher: p_2=∑x_i^2 repeats one variable in each degree-two term instead of selecting two distinct positions. At degree one, the constructions coincide: p_1=e_1=∑x_i.^[1]^[3]
General symmetric expression in elementary generators: a polynomial in the e_k family need not itself equal one e_k.^[1]
Vieta's formulas and characteristic-polynomial coefficients: these evaluate the family on roots or spectra; they are uses of the object.^[1]^[4]
Maclaurin's inequality: a separate theorem about elementary symmetric values, rather than the formal polynomial object.

References¶

[1] Linus Setiabrata, Math 4370, Computational Algebra, notes for Karola Mészáros, last compiled 2020, Example 7.20.1, Definition 7.20.4, Observations 7.20.5–7.20.6, and Theorem 7.20.7, printed pp. 65–66 (one-based PDF pp. 66–67). registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p ↩q ↩r ↩s ↩t ↩u

[2] George M. Bergman, An Invitation to General Algebra and Universal Constructions, second edition, Springer, author PDF dated April 2, 2015, Exercise 5.3:5(i)–(iv), one-based PDF p. 137. The exercise states the commutative-ring setting and the elementary subset definition; its generation proof is assigned as an exercise. registry ↩a ↩b ↩c ↩d ↩e ↩f

[3] J. M. Landsberg, Geometry and Complexity Theory, author prepublication text, §6.1.4, equations (6.1.1)–(6.1.3), printed p. 153 (one-based PDF p. 163), and §6.1.5, printed p. 154 (one-based PDF p. 164). registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p ↩q

[4] Yitzhak Katznelson and Yonatan R. Katznelson, A (Terse) Introduction to Linear Algebra, draft, Appendix A.6.7, one-based PDF p. 178. The extracted coefficient equation at the page break is corrupt; the cited subset definition, eigenvalue list, and characteristic-polynomial linkage are legible. registry ↩a ↩b ↩c ↩d ↩e ↩f