Orthodox Semigroup¶

A regular semigroup whose idempotents are closed under multiplication, equivalently one in which chosen inverses of two factors compose in reverse order to an inverse of their product.

Version: v1 · 2026-08-30 · History
Domain-specific #: 2438
Origin domain: semigroup theory
Subdomain: regular semigroups
Aliases: Regular E-semigroup

Core Idea¶

An orthodox semigroup is a semigroup \(S\) satisfying two linked requirements. First, it is regular: for every \(a\in S\), there is some \(x\in S\) such that \(axa=a\) and \(xax=x\). Such an \(x\) is an inverse of \(a\), although it need not be unique. Second, its idempotents are closed under multiplication. Writing

\[ E(S)=\{e\in S:e^2=e\}, \]

the second condition says that \(e,f\in E(S)\) implies \(ef\in E(S)\). Because the ambient operation is already associative, \(E(S)\) is then a subsemigroup—and specifically a band, a semigroup all of whose elements are idempotent.^[1]^[2]

This definition earns an autonomous mathematical identity through its consequences. If \(V(a)\) denotes all inverses of \(a\), then a regular semigroup is orthodox exactly when

\[ x\in V(a),\ y\in V(b)\quad\Longrightarrow\quad yx\in V(ab). \]

Thus the group law \((ab)^{-1}=b^{-1}a^{-1}\) survives in set-valued form even when inverses are nonunique. Equivalent tests say that every inverse of an idempotent is idempotent, or that intersecting inverse sets are equal. Hall's foundational 1969 paper used these properties to study the least inverse-semigroup congruence, ideal extensions, and principal factors.^[1] The class is therefore more than the phrase “regular semigroup with one extra closure condition”: it is the stable setting in which nonunique inverses and a closed band of idempotents remain mutually controlled.

Structural Signature¶

The recurring structure is:

associative carrier + elementwise generalized inverses + distinguished idempotent set + closure of that set under multiplication → a regular semigroup with coherent reverse-order inverse calculus.

Its mandatory roles are:

Semigroup carrier \(S\) — a set with one total associative binary operation.
Regularity witness — each \(a\) has at least one \(x\) with \(axa=a\) and \(xax=x\).
Inverse-set map \(V\) — \(V(a)\) records every such inverse, not a chosen global inverse function.
Idempotent band \(E(S)\) — every \(e\in E(S)\) satisfies \(e^2=e\), and products of idempotents stay idempotent.
Reverse-order invariant — any inverse of \(a\) followed in reverse order by any inverse of \(b\) yields an inverse of \(ab\).

The recognition test is conjunctive but not arbitrary: verify regularity for every element and multiplicative closure of \(E(S)\). When regularity is already known, any of the equivalent inverse-set properties can replace the closure test.^[1]^[3] The identity fails if even one element has no inverse, or if two idempotents have a non-idempotent product.

What It Is Not¶

It is not a synonym for regular semigroup. Regularity guarantees nonempty \(V(a)\), but idempotents in a regular semigroup need not multiply to idempotents. It is also not merely an E-semigroup: that term asks for idempotent closure but does not, by itself, impose regularity. “Regular E-semigroup” is the exact modern descriptive synonym.

It is not an inverse semigroup. An inverse semigroup gives every element exactly one inverse; an orthodox semigroup may give an element many. Inverse semigroups form a proper subclass because their idempotents commute and hence are closed under multiplication. Nor is an orthodox semigroup necessarily a group, monoid, band, completely regular semigroup, or commutative semigroup. Those impose different or additional conditions.

Finally, the object is not the moral or religious sense of “orthodox.” Hall explicitly introduced the adjective as a short name for the class defined in his title.^[1] The term carries no evaluative claim.

Scope of Application¶

The construct belongs to regular-semigroup structure theory. It is used when generalized inverses exist but uniqueness is too restrictive, and when the behavior of idempotents must still be algebraically tractable. Its scope includes congruences on regular semigroups, decompositions by principal factors, ideal extensions, bands of idempotents, covers, and comparisons with inverse semigroups.

Hall proved that homomorphic images of orthodox semigroups are orthodox and that a semigroup with ideal \(I\) is orthodox exactly when both \(I\) and the Rees quotient \(S/I\) are orthodox. For finite semigroups this supports a principal-factor criterion.^[1] Later work on semigroups whose idempotents form a subsemigroup established E-unitary covers preserving the relevant E-density or orthodoxy conditions, showing that the class continues to organize structure and representation questions rather than naming a one-off example.^[4]

The scope remains algebraic. A database operation, workflow, or distributed reduction may instantiate the parent Semigroup, but it is not an orthodox semigroup unless its carrier, generalized inverse relation, and idempotent band are literally defined and satisfy the equations.

Clarity¶

A reader can decide membership through the following procedure.

Confirm one associative operation on \(S\).
For every \(a\), show \(V(a)\neq\varnothing\); equivalently supply an inverse \(x\) satisfying both sandwich equations.
Find \(E(S)\) and check \(ef\in E(S)\) for every pair \(e,f\in E(S)\).

For a finite semigroup this is a direct table calculation. For a presented or represented semigroup, an equivalent inverse-set condition may be easier. In particular, once regularity is known, test whether every inverse of every idempotent is itself idempotent, or whether arbitrary \(x\in V(a)\) and \(y\in V(b)\) always give \(yx\in V(ab)\). These are equivalent, not optional strengthenings.^[1]^[3]

The most common false positive is to verify that each element has a generalized inverse and stop. The most common false negative is to demand uniqueness of the inverse. Regularity alone is too weak; uniqueness is too strong. Orthodoxy occupies the precise middle position.

Manages Complexity¶

Regular semigroups allow many inverses per element, so an argument about products could require tracking choices in \(V(a)\) and \(V(b)\) independently. Orthodoxy compresses that choice explosion: every allowed pair composes in reverse order to an allowed inverse of the product. The closed band \(E(S)\) similarly turns a distinguished subset that is generally not closed in a regular semigroup into an algebraic object with its own band structure.

This compression has structural payoffs. Equality of inverse sets defines the key relation used by Hall to obtain the least inverse-semigroup congruence; quotients can therefore expose an inverse-semigroup skeleton while retaining information in the noncommutative band of idempotents.^[1] Closure under homomorphic images and ideal extensions supports modular reasoning: one can study an ideal and quotient separately, or reduce finite questions to principal factors, without leaving the class.

The mechanism does not make inverse choices unique and does not force idempotents to commute. It manages nonuniqueness by making it coherent, rather than deleting it.

Abstract Reasoning¶

Diagnostic inference. If \(e\) is idempotent in a regular semigroup and some \(x\in V(e)\) is not idempotent, the semigroup is not orthodox. This supplies a single-witness refutation. Dually, proving the reverse-order inverse condition for all factors certifies orthodoxy without separately enumerating all idempotent products.

Predictive inference. In an orthodox semigroup, a choice \(x\in V(a)\), \(y\in V(b)\) predicts \(yx\in V(ab)\). This is the exact set-valued residue of reversing factors under group inversion. It also predicts that the relation of “having the same inverse set” behaves as the least inverse-semigroup congruence in Hall's construction.^[1]

Decomposition inference. If an ideal and its Rees quotient are both orthodox, then their extension is orthodox. In a finite semigroup, principal factors can therefore be used as local units of analysis. The move is not merely classificatory: it identifies which smaller components must be checked to establish a global property.

Boundary inference. If the idempotents commute, then their product is idempotent, so a regular semigroup with commuting idempotents is inverse and therefore orthodox. The converse fails: a band can be noncommutative and still be orthodox. This locates precisely what orthodoxy preserves from inverse semigroups and what it relaxes.

Knowledge Transfer¶

The full mechanism transfers literally among subareas that use semigroups: abstract algebra, transformation semigroups, regular-semigroup congruence theory, and finite-semigroup structure. The carrier and representation may change, but \(V(a)\), \(E(S)\), idempotent closure, and the reverse-order condition remain exact.

It does not clear the prime bar. Outside semigroup theory, “orthodox” often means conventional, and “idempotent” may be used for operations without a regular-semigroup inverse calculus. Those lexical echoes are not instances. The portable residue is already captured by the live prime Semigroup, together with its parent properties Associativity and Closure. Orthodox Semigroup adds a specialized regularity/idempotent package whose diagnostic equations remain native to algebra.

Examples¶

Groups. Every group is orthodox. Each \(a\) has the unique inverse \(a^{-1}\), and \(E(G)=\{1\}\), which is closed. The example shows inclusion, but not the distinctive nonunique-inverse behavior.

Inverse semigroups. The symmetric inverse semigroup of all partial bijections of a set under composition is orthodox. Each partial bijection has its unique relational inverse. Its idempotents are partial identity maps; composing two produces the partial identity on the intersection of their domains. Hence the idempotents form a semilattice and the semigroup is orthodox.

Bands. Every band is orthodox. Every element is idempotent, so \(E(S)=S\), which is automatically closed; each element is its own inverse. A rectangular band supplies a noncommutative orthodox example and demonstrates that idempotents need not commute.

A regular nonexample. Let \(T_3\) be the full transformation monoid on \(\{1,2,3\}\), with \((fg)(x)=f(g(x))\). Every transformation of a finite set is regular: select one preimage for every point in its image to construct an inverse. Define idempotents

\[ e=(1\mapsto1,2\mapsto1,3\mapsto3),\qquad f=(1\mapsto1,2\mapsto2,3\mapsto2). \]

Both square to themselves. But \(fe=(1\mapsto1,2\mapsto1,3\mapsto2)\), while \((fe)^2\) sends \(3\) to \(1\), so \((fe)^2\ne fe\). Thus \(T_3\) is regular but not orthodox. This boundary case isolates the missing condition: idempotent closure.

Structural Tensions¶

Regularity versus coherence. Regularity maximizes availability—every element has an inverse—but permits many incompatible choices. Orthodoxy restricts that freedom enough to make reverse-order composition reliable, without collapsing it to uniqueness. Diagnostic: are inverse sets merely nonempty, or are arbitrary choices closed under reverse-order multiplication?

Closed idempotents versus commuting idempotents. Closure makes \(E(S)\) a band, which is enough for orthodoxy. Commutativity would make it a semilattice and pushes a regular semigroup into the narrower inverse-semigroup class. Diagnostic: does the argument need only \(ef\) to remain idempotent, or does it silently swap \(ef\) and \(fe\)?

Definitional conjunction versus autonomous class. The definition looks like Regular Semigroup plus E-semigroup, inviting composite closure. But equivalent inverse-set tests, least inverse congruences, extension theorems, and a continuing literature give the conjunction stable inferential work. Diagnostic: would replacing the name by two generic properties preserve the established theorems and recognition tests without rebuilding the class? Here it would not.

Broad closure versus local failure. Orthodoxy is preserved by homomorphic images and ideal extensions, but it can fail inside the broad class of regular transformation semigroups through a single product of idempotents. Diagnostic: use closure theorems when decomposing known orthodox objects, but use an explicit idempotent-product witness when testing a new regular object.

Structural–Framed Character¶

Orthodox Semigroup is structural. Its membership is exhausted by equations and closure properties. No evaluator, institution, purpose, or preferred outcome makes a semigroup orthodox. The historical choice of the adjective is conventional, but the algebraic class it labels can be recognized independently of that framing.

On the structural–framed criteria, evaluative weight, institutional origin, human-practice dependence, and import-versus-recognize framing are all absent. Vocabulary travel is limited only because the technical word “orthodox” is domain-bound, not because the structure depends on a frame.

Structural Core vs. Domain Accent¶

The structural core is a semigroup with an inverse relation and a multiplication-closed idempotent subset. The core is completely formal, but its indispensable objects—generalized semigroup inverses, idempotents, bands, congruences, and principal factors—belong to semigroup theory. The exact reverse-order formula travels wherever this algebraic structure is instantiated, including transformation and partial-transformation semigroups.

The domain accent is therefore not historical decoration; it supplies the mathematical typing that makes the equations meaningful. A generic system with reversible actions is not enough, and a generic idempotent operation is not enough. Removing the semigroup carrier and its sandwich inverse equations leaves only analogies already handled by Semigroup, Associativity, Closure, or Inverse.

Orthodox Semigroup specializes the live prime semigroup: every orthodox semigroup is a semigroup, while most semigroups are not regular and regular semigroups need not have closed idempotents. This is the single minimal prospective DAG parent.

It also depends internally on associativity and closure, but those are already inherited through Semigroup and need no redundant direct edge. group, monoid, and inverse are related algebraic neighbors or subclasses/superclasses in ordinary mathematical discourse, not minimal direct parents for this catalog placement.

Relationships to Other Abstractions¶

Current abstraction Orthodox Semigroup Domain-specific

Parents (1) — more general patterns this builds on

Orthodox Semigroup is a kind of Semigroup Prime

Orthodox Semigroup specializes the live prime semigroup: every orthodox semigroup is a semigroup, while most semigroups are not regular and regular semigroups need not have closed idempotents.

Hierarchy paths (4) — routes to 4 parentless roots

Orthodox Semigroup → Semigroup → Set and Membership

Show alternative paths (3)

Neighborhood in Abstraction Space¶

Orthodox Semigroup sits in a sparse region of the domain-specific corpus (75^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Unclustered & Miscellaneous (1565 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-09-08

Not to Be Confused With¶

Semigroup. Requires only an associative binary operation. Orthodoxy additionally requires elementwise generalized inverses and a closed idempotent set.
Regular semigroup. Requires an inverse for each element but does not require products of idempotents to remain idempotent.
E-semigroup. Requires idempotents to form a subsemigroup but need not be regular. A regular E-semigroup is precisely orthodox.
Inverse semigroup. Gives each element a unique inverse; equivalently, in the regular setting, idempotents commute. It is a proper subclass of orthodox semigroups.
Band. Makes every element idempotent. Bands are orthodox examples, not synonyms for the whole class.
Completely regular semigroup. Is a union of groups. Complete regularity and orthodoxy are independent conditions in general; their conjunction is often called an orthogroup.
Orthodox in ordinary language. Means conventional or doctrinally established and has no algebraic connection.

References¶

[1] T. E. Hall, “On Regular Semigroups Whose Idempotents Form a Subsemigroup,” Bulletin of the Australian Mathematical Society 1, no. 2 (1969): 195–208. https://doi.org/10.1017/S0004972700041447 registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h

[2] K. S. S. Nambooripad, Theory of Regular Semigroups (Sayahna Foundation, 2018), especially Theorem 2.43 and the discussion of orthodox and inverse semigroups. https://books.sayahna.org/en/pdf/TheoryOfRegularSemigroups.pdf registry ↩

[3] Ana Catarina Cristino Monteiro, Classes of Semigroups, Congruences and Languages (M.Sc. dissertation, Universidade de Lisboa, 2024), Definition 1.1.1 and Proposition 1.2.1. https://repositorio.ulisboa.pt/server/api/core/bitstreams/b56c8d99-4313-4fe4-b560-2ac140dfa287/content registry ↩a ↩b

[4] J. Almeida, J.-É. Pin, and P. Weil, “Semigroups Whose Idempotents Form a Subsemigroup,” Mathematical Proceedings of the Cambridge Philosophical Society 111, no. 2 (1992): 241–253. https://doi.org/10.1017/S0305004100075332 registry ↩

[5] P. H. H. Fantham, “On the Classification of a Certain Type of Semigroup,” Proceedings of the London Mathematical Society s3-10, no. 1 (1960): 409–427. https://doi.org/10.1112/plms/s3-10.1.409 registry