Fact-checked by Grok 2 weeks ago

Axiom of extensionality

The is a foundational in axiomatic , asserting that two sets are equal precisely when they have exactly the same elements, thereby defining sets by their membership relations alone. Formally, it can be stated as: for all sets x and y, if for every z, z \in x z \in y, then x = y. This principle ensures the uniqueness of sets based on their contents, preventing distinct sets from being indistinguishable by their elements. Introduced by in his 1908 axiomatization of , which addressed foundational issues arising from paradoxes like , the establishes a clear for set identity. In Zermelo's original formulation, it was phrased as: if M \subset N and N \subset M, then M = N, emphasizing extensional equality through mutual relations. This forms the first principle in Zermelo's system and remains the initial in the modern Zermelo-Fraenkel (ZF), where it underpins the development of all subsequent axioms. The axiom's implications extend to proving key properties, such as the uniqueness of the , as any two sets with no elements must be identical under (with provided by other axioms). It also reinforces the idea that sets lack inherent "labels" or internal structure beyond membership, distinguishing from other mathematical structures like multisets or tagged collections. In the full Zermelo-Fraenkel-Choice (ZFC) system, interacts with axioms like separation and to enable rigorous constructions of mathematical objects, from natural numbers to uncountable infinities.

Introduction

Formal Definition

The axiom of extensionality asserts that sets are equal they have precisely the same elements, embodying the principle of extensional equality whereby the identity of a set is determined solely by its membership rather than by any intensional description or mode of presentation. This captures the intuitive notion that sets are collections defined purely by what belongs to them, without regard to order, repetition, or external labels. In , the axiom is formally stated as \forall x \forall y \left( \forall z (z \in x \leftrightarrow z \in y) \to x = y \right), where the universal quantifiers range over all sets, the biconditional \leftrightarrow specifies that every element of x is an element of y and vice versa, and the \to concludes that x equals y under this condition. In pure , this is a single rather than an infinite , as it applies uniformly to all sets without requiring separate instances for different predicates. For example, the sets \{1, 2\} and \{2, 1\} are equal because they share exactly the same elements, regardless of the order in which they are listed. This principle distinguishes sets from multisets, where order or multiplicity might affect equality.

Basic Implications

The axiom of extensionality ensures that sets are uniquely determined by their , implying that for any given collection of , there is at most one set that contains precisely those and no others. This uniqueness follows directly from the axiom's statement, as any two sets sharing identical membership relations must be identical. Consequently, this principle eliminates the possibility of distinct sets that are indistinguishable based on contents, providing a foundational criterion for set in extensional set theories. Under the axiom, sets serve as pure containers defined solely by membership, devoid of any additional inherent properties or identities beyond what their elements confer. This extensional view treats sets as abstractions where is exhausted by extension, reinforcing the idea that sets lack "internal " independent of their constituents. Such a conceptualization supports extensional reasoning across mathematical , ensuring consistency in how collections are identified and manipulated. The axiom is typically stated as \forall x \forall y \left( \forall z (z \in x \leftrightarrow z \in y) \to x = y \right), capturing one direction of the biconditional for set equality. The converse, \forall x \forall y \left( x = y \to \forall z (z \in x \leftrightarrow z \in y) \right), derives as a from the standard properties of equality in , particularly the rule. To sketch the proof: assume x = y; then for arbitrary z, if z \in x, substitution yields z \in y (since membership is an atomic ), and symmetrically if z \in y then z \in x; thus the membership biconditional holds universally. This logical derivation complements the , yielding the full equivalence that sets are equal they have the same elements. A concrete illustration of uniqueness arises with the , denoted \emptyset, which contains no elements at all. Suppose another collection Y also has no elements; then for every z, z \notin \emptyset and z \notin Y, so \forall z (z \in \emptyset \leftrightarrow z \in Y), implying \emptyset = Y by . Thus, the is singular, with no alternative empty collections possible to distinguish it.

Historical Development

Etymology

The of "" in originates from the distinction between the "extension" of a —the full set of objects to which it applies—and its "," the inherent qualities or properties that define it. This opposition, rooted in 19th-century logical analysis, underscores how entities like sets or classes are identified solely by their members rather than their descriptive attributes. Gottlob Frege employed the term in this sense in his 1893 work Grundgesetze der Arithmetik, where he defined the equality of concepts through their shared extensions, laying foundational groundwork for treating logical objects extensionally. , in his 1908 axiomatization of , referred to the corresponding principle as the "Axiom der Bestimmtheit" (axiom of definiteness), emphasizing the unique determination of sets by their elements. The English term "extensionality" gained traction in mathematical and logical discourse during the and , as formal axiomatic systems proliferated in works by and Paul Bernays, who adopted "Extensionalitätsaxiom" in their Grundlagen der Mathematik. By the mid-20th century, through translations and standardizations in texts like Jean van Heijenoort's anthology, "Axiom of Extensionality" emerged as the conventional designation in English-language . This shift marked a broader of German logical terminology into international mathematical practice, contrasting with intensional approaches that prioritize meaning over mere membership.

Origins and Early Formulations

The roots of the axiom of extensionality trace back to the late , embedded in the foundational concepts of developed by . In his seminal 1895 paper Beiträge zur Begründung der transfiniten Mengenlehre, Cantor defined a set as "any collection M into a whole of definite and separate objects m of our or our thought." This characterization implicitly embodies by treating sets as indistinguishable if they share the same elements, disregarding order or multiplicity, as membership equality determines the collection's identity. Building on such ideas, advanced a more formal version in his 1893 work Grundgesetze der Arithmetik. There, Basic Law V equates the extensions of two functions f and g precisely when they agree on all arguments: the course-of-values of f equals that of g if and only if f(x) = g(x) for every x. This principle of for concept extensions served as a cornerstone for deriving arithmetic from logic but ultimately contributed to Bertrand Russell's 1901 paradox, exposing inconsistencies in unrestricted comprehension. Ernst Zermelo explicitly formulated the axiom in 1908 to establish a paradox-free axiomatic foundation for . In his paper Untersuchungen über die Grundlagen der Mengenlehre I, the axiom appears as the first postulate, known as the Axiom der Bestimmtheit (Axiom of Definiteness): if every element of set M belongs to N and every element of N belongs to M, then M = N. Zermelo's inclusion resolved ambiguities in set equality arising from the paradoxes, using the term Bestimmtheit to denote this determinacy by membership. In the ensuing decades, the axiom gained widespread acceptance and refinement in emerging systems. By the 1920s, John von Neumann integrated an extensionality for sets and classes into his 1925 axiomatization, emphasizing equality via identical elements. This approach was expanded in the von Neumann–Bernays–Gödel (NBG) framework through collaborations in the 1930s and 1940s, where extensionality extended to classes, reinforcing its status as a core principle in axiomatic .

Role in Zermelo-Fraenkel Set Theory

Statement in ZF

In Zermelo-Fraenkel set theory (ZF), the axiom of extensionality is the first axiom in the standard list and is formally stated as \forall x \forall y \, [\forall z (z \in x \leftrightarrow z \in y) \rightarrow x = y]. This asserts that two sets are equal they have precisely the same elements, with the converse direction following from the definition of equality in . Within ZF, which is a of pure sets without urelements, the ensures that every set is fully determined by its elements, preventing distinctions or accidental equalities based on different constructions or notations. This extensional characterization underpins the uniqueness of set constructions throughout the . The is compatible with other ZF axioms, such as the of the , which posits the existence of a set with no elements (\exists x \, \neg \exists y (y \in x)); combined with , it guarantees that there is a unique empty set \varnothing. Similarly, the (\forall x \forall y \exists z \forall w (w \in z \leftrightarrow (w = x \lor w = y))) ensures the existence of a set containing exactly x and y, and makes this pair set \{x, y\} unique regardless of how it is formed. For instance, consider the axiom (\forall x \exists y \forall z (z \in y \leftrightarrow z \subseteq x)), which asserts the existence of a set containing all subsets of x; proves the of this \mathscr{P}(x), as any two such collections must have and thus be equal.

Relation to Other Axioms

The in Zermelo-Fraenkel (ZF) relies on the underlying logic's properties of for its converse implication, which states that equal sets have the same ; this follows from the substitution property of , allowing the axiom to focus solely on the direction that sets with are equal. Without this logical foundation, the axiom would require a biconditional , but in ZF, it integrates seamlessly with to define set identity purely in terms of membership. In interaction with the axiom schema of separation (the restricted comprehension in ZF), ensures the of sets defined by properties. ZF's restricted limits definitions to of existing sets, preventing the formation of paradoxical entities that arise in unrestricted comprehension schemes of , while guarantees that such subsets are distinct only if their elements differ. For instance, the , obtained via separation as the subset with no elements, is unique due to , avoiding multiple indistinguishable empties that could undermine set construction. Extensionality works with the axiom schema of separation to derive unique subsets from given sets based on formulas, ensuring that the resulting collection is a set identical to any other with the same elements, which is essential for iterative definitions in the cumulative hierarchy. Similarly, in conjunction with the , it guarantees that the image of a set under a definable function is a unique set, as extensionality identifies any two such images with matching elements; this interplay supports the transfinite needed to build higher ranks in the hierarchy. The axiom of extensionality is independent of the remaining ZF axioms, as demonstrated by models where the other axioms hold but fails, such as partial models with multiple empty sets; this , established through proofs, underscores extensionality's primitive status in ZF, though it is often assumed as foundational for practical set-theoretic reasoning.

Formulations in Alternative Set Theories

In (NF)

In Quine's () set theory, the axiom of extensionality is explicitly stated as a core postulate, formulated as \forall x \forall y \, (x = y \leftrightarrow \forall z (z \in x \leftrightarrow z \in y)), which asserts that two sets are identical they have precisely the same members. This formulation, introduced by in his 1937 paper "New Foundations for Mathematical Logic," serves as one of only two fundamental principles in NF, alongside the of stratified . Unlike in Zermelo-Fraenkel (ZF), where extensionality is paired with multiple other axioms to enforce well-foundedness, in NF it is integrated directly with stratified comprehension, which restricts set formation to formulas that respect a type-like to avoid paradoxes like Russell's. A key difference from ZF arises in NF's treatment of equality and membership, where extensionality applies universally within a single, stratified universe that includes a total set V comprising all sets. This allows the membership relation \in to define equality extensionally across the entire system, enabling the construction of sets that would be "proper classes" in ZF, such as the Russell class \{x \mid x \nsubseteq x\}, without leading to contradiction due to the stratification constraint. Quine's 1937 framework thus positions extensionality as the foundational assumption for identifying sets solely by their elements, providing an alternative logical basis for mathematics that avoids the iterative hierarchy of ZF while maintaining extensional uniqueness. The implications of in NF extend to supporting non-well-founded structures, such as hypersets—sets that can contain themselves indirectly through infinite descending membership chains—and circular sets, which permit cyclic memberships like x \in y \in x. This contrasts sharply with ZF's , which enforces well-foundedness and prohibits such constructions; in NF, extensionality's integration with allows these features without , opening avenues for modeling recursive or self-referential entities in .

In Zermelo Set Theory with Urelements (ZU)

In with Urelements (ZU), the axiom of extensionality is a primitive axiom, typically formulated to apply only to sets: \forall x \forall y \, (\mathrm{Set}(x) \land \mathrm{Set}(y) \land \forall z (z \in x \leftrightarrow z \in y) \to x = y), where \mathrm{Set} distinguishes sets from urelements. Zermelo's original axiomatization incorporated the possibility of urelements—non-set atoms—as part of the domain, alongside axioms enabling the formation of sets via separation (Aussonderung), which extracts subsets defined by definite properties, and collection (implicit in and operations), which ensures the existence of sets whose elements satisfy given definite conditions. These mechanisms, together with , guarantee that any two sets built iteratively from urelements, sharing identical membership, must coincide, as their construction uniquely determines their identity based on elements. Urelements play a crucial role by serving as foundational atoms that are neither sets nor composed of elements, distinguished formally by the property that no object belongs to them (\forall u (U(u) \to \neg \exists v (v \in u)), where U denotes urelements). thus applies post-construction: while urelements themselves may be indistinguishable except by primitive identity, any sets formed from them via separation and collection inherit uniqueness from their elements, preventing duplicate representations in the . This distinction maintains the purity of set formation atop an atomic base. In his 1908 paper, Zermelo introduced as the first , stating that two sets are equal if they have the same elements, while accommodating urelements optionally to resolve foundational paradoxes. Later developments have explored set theories with urelements in various non-standard contexts, such as anti-foundational approaches, while retaining as a primitive postulate.

Extensions and Modifications

With Urelements

In set theories incorporating urelements, also known as atoms, the axiom of requires modification to accommodate these primitive objects, which lack elements and internal set structure. The standard formulation of extensionality would equate all urelements to the , as they share the property of having no members, thereby collapsing distinct urelements into a single entity. To preserve the distinctness of urelements while maintaining extensionality for constructed sets, the axiom is restricted to apply only to sets (excluding urelements). The modified axiom states:
\forall x \forall y \bigl( \mathrm{set}(x) \land \mathrm{set}(y) \land \forall z (z \in x \leftrightarrow z \in y) \to x = y \bigr),
where \mathrm{set}(x) indicates that x is a set, not a . This ensures that any two sets with identical members are equal, while allowing multiple urelements to coexist as distinct non-set objects separate from the unique \emptyset. In this setup, urelements serve as foundational building blocks without violating the identity principle for sets formed via the other s. The remains the unique set with no elements, as the applies to it.
The rationale for this adjustment stems from the nature of urelements, which are intended to model indivisible entities without membership relations, such as points in geometry or individuals in permutation models. Applying full over all objects would undermine their utility by forcing all such objects to be indistinguishable and equivalent to \emptyset, rendering them ineffective for applications like proofs. By restricting to sets, the axiom upholds the extensional characterization of sets built from urelements, ensuring based on membership while permitting a potentially or class-sized collection of distinct urelements. In theories such as ZFA (Zermelo-Fraenkel set theory with atoms), this modified extensionality allows for the construction of unique sets containing urelements. For instance, given distinct urelements a and b, the sets \{a\} and \{b\} are distinct because their sole members differ, and \{a, b\} is uniquely determined by its contents. This framework supports the formation of hereditarily symmetric sets over urelements, where higher-rank sets inherit uniqueness from their elements via the modified axiom. Historically, this adaptation emerged in the 1920s to facilitate models for investigating the of the . Abraham introduced urelements in 1922 to construct models where choice fails, using permutations of the urelements to define symmetric subsets, with adjusted to preserve distinctness. In the 1930s, Andrzej Mostowski refined these models, solidifying the role of restricted in handling atoms. Forcing techniques, developed by in the 1960s for ZFC, were later extended to ZFA contexts, enabling broader results by incorporating urelements into generic extensions while maintaining the modified .

Non-Extensional Approaches

In theory, the axiom of is rejected to accommodate multiplicity, allowing collections such as {a, a} to differ from {a} despite sharing the same distinct elements. This weakened form of distinguishes multisets by the number of occurrences of each element, enabling applications in areas like and data structures where repetition matters. Wayne D. Blizard formalized this approach in 1989, emphasizing that traditional fails for such structures since equality depends not only on presence but on of repetitions. Structural set theories, such as those incorporating Peter Aczel's anti-foundation axiom, permit hypersets—non-well-founded sets that may contain themselves or form cycles—while maintaining through graph-based definitions of . In this framework, sets are uniquely determined by accessible pointed graphs (APGs), where arises from structural rather than solely from membership relations, allowing for infinite descending chains without violating extensional principles. Aczel's 1988 formulation ensures every APG has a unique , providing a bisimulation-based that extends beyond well-founded membership to handle circular definitions. This structural approach contrasts with pure by emphasizing relational decorations, yet preserves the core idea that sets with identical elements are equal. Vopěnka's alternative set theory (AST), introduced in 1979, explicitly adopts a non-extensional stance by incorporating semisets—subcollections that behave like sets but allow distinctions based on intensions or construction rather than pure extensions. In AST, while standard sets may adhere to extensionality, proper semisets and classes can differ despite having the same elements, addressing philosophical concerns about vagueness and infinite hierarchies without relying on the full axiom. This theory motivates non-extensionality through axioms like selection and supertransitivity, enabling a "limit universe" where natural numbers lack a maximal infinite element, thus critiquing Cantor's extensional foundations for overlooking semantical nuances. Fuzzy set theory, pioneered by in , further weakens extensionality by replacing binary membership with degrees in [0,1], violating the strict equality condition of classical sets since two fuzzy sets may have identical supports but differ in membership gradations. Zadeh characterized fuzzy sets via membership functions, allowing partial belongings that capture in real-world phenomena, such as linguistic terms like "tall," where extensional fails due to graded overlaps. This philosophical shift prioritizes imprecision over rigid extensional equivalence, influencing fields beyond . In modern computer science, non-extensional and weakened extensional approaches, particularly via hypersets and non-well-founded structures, enable modeling of recursive data types, pointers, and infinite processes that standard extensional foundations cannot capture without ad hoc fixes. For instance, Aczel's framework supports semantic models for concurrent systems like Milner's Calculus of Communicating Systems (CCS), using bisimulations to resolve circularities in process equivalences. These applications extend to coalgebraic methods for coinductive definitions, contrasting extensional well-foundedness by accommodating non-terminating computations in programming language semantics.

References

  1. [1]
    [PDF] The Axioms of Set Theory ZFC
    We will start with Zermelo's first axiomatisation of Set Theory and will show how basic mathematics can be developed within this system. Then we will introduce ...
  2. [2]
    [PDF] fundamentals of zermelo-fraenkel set theory - UChicago Math
    Aug 23, 2011 · Axiom of Extensionality. If every element of X is an element of Y and every element of Y is an element of X, then X = Y .
  3. [3]
    [PDF] Dr. Bob's Axiom of Choice Centennial Lecture Fall 2008 (updated ...
    Sep 26, 2020 · Zermelo states 7 axioms (to paraphrase):. Page 3. Axiom of Choice. 3. 1. Axiom of Extensionality. If M ⊂ N and N ⊂ M, then M = N. 2. Axiom of ...
  4. [4]
    [PDF] Math 310 October 30, 2020 Lecture
    Oct 30, 2020 · The Axiom of Extensionality mandates that two sets are equal precisely when they have the same elements. In our treatment in Section 3.1, we ...
  5. [5]
    [PDF] AN INTRODUCTION TO SET THEORY
    We begin. We have the following axioms: The Axiom of Equality. ∀x ∀y [x = y → ∀z (x ∈ z ↔ y ∈ z)]. The Axiom of Extensionality. ∀x ∀y ... This is true by Axiom of ...
  6. [6]
    [PDF] Axioms of Set Theory
    The Axioms of Zermelo–Fraenkel Set Theory. In 1905, Zermelo began to axiomatise Set Theory and in 1908 he published his first axiomatic system consisting of ...
  7. [7]
    [PDF] Set Theory
    ... Set Theory. Page 11. 1. Axioms of Set Theory. Axioms of Zermelo-Fraenkel. 1.1. Axiom of Extensionality. If X and Y have the same elements, then. X = Y . 1.2 ...
  8. [8]
    [PDF] 1 Elementary Set Theory - Penn Math
    Axiom of Extensionality Let A, B be sets. If (∀x)x ∈ A iff x ∈ B then A ... ∀x∀y∀z(x = y ⇒ (y = z ⇒ x = z)). Proof. We have y = x ⇒ (y = z ⇒ x ...Missing: textbook | Show results with:textbook
  9. [9]
  10. [10]
    [PDF] mat.1 The Theory of Sets - Open Logic Project Builds
    (Note that the axiom of extensionality implies that there is only one empty set.) Whenever we want to talk about ∅ in the language of set theory, we would ...
  11. [11]
    Frege's Theorem and Foundations for Arithmetic
    Jun 10, 1998 · Principle of Extensionality:​​ Frege was made aware of the inconsistency by Bertrand Russell, who sent him a letter formulating 'Russell's ...Frege's Theory of Extensions... · Frege's Analysis of Cardinal... · Frege's Theorem
  12. [12]
    Gottlob Frege (1848—1925) - Internet Encyclopedia of Philosophy
    Gottlob Frege was a German logician, mathematician and philosopher who played a crucial role in the emergence of modern logic and analytic philosophy.
  13. [13]
    Zermelo's Axiomatization of Set Theory
    Jul 2, 2013 · The first axiomatisation of set theory was given by Zermelo in his 1908 paper “Untersuchungen über die Grundlagen der Mengenlehre, I”The Axioms · The Background to Zermelo's... · The Major Problems with...
  14. [14]
    [PDF] Foundations of Mathematics Vol. 1 (1934) David Hilbert/Paul Bernays
    To make it easier the first axioms are accompanied by a linguistic version. The demarcation of the axioms does not correspond completely to that in. Hilbert's “ ...
  15. [15]
    Set Theory - Stanford Encyclopedia of Philosophy
    Oct 8, 2014 · The Reflection Principle encapsulates the essence of ZF set theory, for as shown by Levy (1960), the axioms of Extensionality, Separation ...
  16. [16]
    Beiträge zur Begründung der transfiniten Mengenlehre - EuDML
    Cantor, Georg. "Beiträge zur Begründung der transfiniten Mengenlehre." Mathematische Annalen 46 (1895): 481-512. <http://eudml.org/doc/157768>.
  17. [17]
    Grundgesetze der arithmetik : Frege, Gottlob, 1848-1925. [from old ...
    Nov 3, 2014 · 1893. Usage: Public Domain Mark 1.0 Creative Commons License ... PDF download · download 1 file · SINGLE PAGE PROCESSED JP2 ZIP download.
  18. [18]
    Untersuchungen über die Grundlagen der Mengenlehre. I
    Download PDF · Mathematische Annalen Aims and scope ... Zermelo, E. Untersuchungen über die Grundlagen der Mengenlehre. I. Math. Ann. 65, 261–281 (1908).
  19. [19]
    [PDF] PhilSci-Archive - CANTOR-VON NEUMANN SET-THEORY
    Next come the familiar axioms of. Extensionality, Pairing, Union, Power, Infinity, Separation and Regularity. (there is implicit universal quantification over ...
  20. [20]
    Zermelo-Fraenkel Set Theory (ZF)
    Axioms of ZF​​ This axiom asserts that when sets \(x\) and \(y\) have the same members, they are the same set. Since it is provable from this axiom and the ...
  21. [21]
    [PDF] SET THEORY
    This book provides an introduction to relative consistency proofs in axiomatic set theory, and is intended to be used as a text in beginning.
  22. [22]
    [PDF] Set Theory (MATH 6730)
    The Axioms of Set Theory and Some Basic Consequences. We will use the axiom ... Ext] expresses better. (more fully) what we want to require by the extensionality ...
  23. [23]
    Quine's New Foundations - Stanford Encyclopedia of Philosophy
    Jan 4, 2006 · Quine, W.V., 1937a, “New Foundations for Mathematical Logic,” American Mathematical Monthly, 44: 70–80. ... Preview the PDF version of this ...The Roots of Set Theory · Type Theory and the... · History · Implementation of...
  24. [24]
  25. [25]
    ZFA in nLab
    Oct 18, 2020 · Zermelo's original 1908 axiomatisation of set theory included atoms, but they were soon discarded as a foundational approach as they could be ...
  26. [26]
    Zu den Grundlagen der Cantor-Zermeloschen Mengenlehre
    Cite this article. Fraenkel, A. Zu den Grundlagen der Cantor-Zermeloschen Mengenlehre. Math. Ann. 86, 230–237 (1922). https://doi.org/10.1007/BF01457986.
  27. [27]
    Multiset theory.
    - **Author and Year**: Wayne D. Blizard, 1989
  28. [28]
    Fuzzy sets - ScienceDirect.com
    A fuzzy set is a class of objects with a continuum of grades of membership. Such a set is characterized by a membership (characteristic) function.Missing: extensionality | Show results with:extensionality
  29. [29]
    Non-wellfounded Set Theory - Stanford Encyclopedia of Philosophy
    Apr 16, 2008 · The set theoretic side of our story is connected to two axioms, the Foundation Axiom and the Anti-Foundation Axiom. We present them here ...