Why does the Kuratowski definition (a, b) = {{a}, {a, b}} successfully encode the ordered pair, while the naive definition {a, b} fails?
ABecause {a, b} = {b, a} by extensionality, so {a, b} cannot distinguish first from second coordinate; the Kuratowski encoding uses nesting depth to uniquely identify the first element
BBecause ZFC requires all sets to have exactly two elements, and Kuratowski satisfies this requirement
CBecause the singleton {a} is larger than {a, b}, making the ordering visible to set operations
DBecause Kuratowski ordered pairs are primitive objects in ZFC, unlike unordered pairs
The fundamental problem is that {a, b} = {b, a}: sets are unordered, so swapping elements gives the same set. The Kuratowski pair (a, b) = {{a}, {a, b}} fixes this by encoding the first coordinate a as the unique element of the singleton {{a}}. To recover a from {{a}, {a, b}}, you find the element that appears in a singleton member; to recover b, you take the non-a element of the two-element member. Proving (a,b) = (c,d) iff a=c and b=d requires this asymmetry. Order is encoded not by position (which doesn't exist in sets) but by nesting structure.
Question 2 Multiple Choice
Starting only from the Axiom of Pairing and Extensionality, which of the following sets is guaranteed to exist?
A{a, b, c} for any three sets a, b, c
B{a} (the singleton containing only a) for any set a
CThe empty set ∅
DThe infinite set {a, {a}, {{a}}, ...}
Pairing guarantees that for any a and b, the set {a, b} exists. Applying pairing to a and a gives {a, a}, and extensionality says {a, a} = {a} since the only element is a. So {a} (the singleton) is guaranteed. Option A requires three elements — you can build {a, b, c} by applying pairing to {a, b} and c and then union, but union is a separate axiom not assumed here. The empty set requires its own axiom (or the Axiom Schema of Separation). The infinite set requires the Axiom of Infinity. Pairing alone is constructively quite limited: it only moves from existing sets to two-element (or one-element) collections.
Question 3 True / False
The Axiom of Pairing can directly produce sets with more than two elements by applying it repeatedly to the results.
TTrue
FFalse
Answer: False
False. Each application of the Axiom of Pairing produces a set with at most two elements: {a, b} or {a} (from {a, a}). Repeatedly applying pairing only ever produces new two-element sets — for example, {{a}, {a, b}} or {{a, b}, {c}} — never a three-element set. To produce {a, b, c} as a flat three-element set, you need the Axiom of Union: form {a, b} and {c} via pairing, then {{a, b}, {c}} via pairing, then take the union to get {a, b, c}. Each axiom in ZFC has a specific constructive role; they cannot substitute for each other.
Question 4 True / False
The Kuratowski ordered pair (a, b) satisfies the characteristic property: (a, b) = (c, d) if and only if a = c and b = d.
TTrue
FFalse
Answer: True
True, and this is precisely what makes the Kuratowski definition the standard. The proof: suppose {{a}, {a,b}} = {{c}, {c,d}}. The singleton {a} must equal either {c} or {c,d}. If it equals {c,d}, then a is the only element, forcing c = d = a. In either case a = c. Once a = c is established, the two-element members {a,b} = {c,d} = {a,d} must give b = d. The converse is trivial. This property is all that ordered pairs need to do — capture 'first' and 'second' — and the Kuratowski encoding achieves it using only sets, with no primitive notion of order required.
Question 5 Short Answer
Explain why sets alone cannot represent ordered pairs, and how the Kuratowski definition solves this problem using only the resources of set theory.
Think about your answer, then reveal below.
Model answer: Sets are inherently unordered: {a, b} and {b, a} are identical by the Axiom of Extensionality. An ordered pair needs to distinguish which element comes first. The Kuratowski solution is to encode the first element via a singleton and the pair together: (a, b) = {{a}, {a, b}}. The first coordinate is recoverable as the unique element of the unique singleton in this set; the second is recovered from the two-element member. This uses only pairing and extensionality — no primitive notion of 'first' or 'second' is needed.
The deeper insight is that ZFC must derive all mathematical structure — including order — from the single primitive notion of set membership. This forces creative encodings: order cannot be assumed, so it must be constructed. The Kuratowski pair is the canonical solution because it satisfies the essential property (equality iff components match) using the simplest possible nesting. Alternative encodings exist (e.g., (a, b) = {a, {a, b}} in some formulations), but Kuratowski's is standard. Once ordered pairs exist as sets, the entire apparatus of relations, functions, sequences, and Cartesian products follows — all expressible as sets of ordered pairs.