Alphabeta Math
Pipeline-generated
How statement and proof provenance work

The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.

  • Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
  • AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
  • AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.

These labels describe origin, not correctness: citations and verification chips remain separate evidence.

Formal Set-Theoretic Syntax, Structures, and Satisfaction

1 · Prerequisites

2 · Summary

This page constructs finite syntax for arbitrary set signatures, then defines satisfaction for nonempty set structures. Parsing and structural recursion support the free-variable, substitution and isomorphism arguments. Relativization to a set agrees with its induced membership structure, without a transitivity assumption. For a proper class, relativization remains a separate formula for each input; it does not provide a uniform truth predicate for the universe. The metatheory is ZF. No countability or effective natural-number coding of an arbitrary signature is assumed.

3 · Logical flowchart

4 · Definitions, theorems and proofs

DefinitionDefinition: AI-adaptedProof: Not applicablejudge pass (gpt-5.6-terra)audited 2026-09-07Open item page →

Set signatures and finite syntax strings

Definition

A set signature L=(C,F,P,a) has pairwise disjoint sets of constants, function symbols and relation symbols, with a:FPω{0}. Its alphabet ΣL is a tagged disjoint union of these sets, variables vi for iω, constructor tags var,const,fun,rel,eq,neg,and,exists, and punctuation (,),,. Equality is a logical symbol.

A word is a function w:nΣL for a natural number n, its length. The set WL of words is obtained by Separation inside P(ω×ΣL). The empty word has domain 0. Concatenation of u:mΣL and w:nΣL is the function on m+n whose value at i<m is u(i) and whose value at m+j is w(j) for j<n. Signatures may be empty or uncountable; finite words here are set codes, not necessarily natural-number codes.

Conventions and prerequisites: The natural numbers N (von Neumann).

DefinitionDefinition: AI-adaptedProof: Not applicablejudge pass (gpt-5.6-terra)audited 2026-09-07Open item page →

Terms and formulas as finite set codes

Definition

For an alphabet token x, let [x] denote the one-letter word. If each ui is either a token or a word, define the fully delimited serialization

enc(u0,,uk1)=[(]uˉ0[,][,]uˉk1[)],

where uˉi=[ui] for a token field and uˉi=ui for a word field. Concatenation is the set function defined in Set signatures and finite syntax strings, so every displayed constructor below is an actual member of WL.

A leaf term is enc(var,vi) or enc(const,c). A composite term is enc(fun,f,t1,,tn) with a(f)=n. Atomic formulas are enc(eq,s,t) and enc(rel,R,t1,,tn) with a(R)=n. Formula constructors are enc(neg,ϕ), enc(and,ϕ,ψ), and enc(exists,vi,ϕ). The punctuation, tag, and symbol fields are tokens; an immediate subexpression field is inserted by concatenating its complete word.

Formally, start with the leaf terms and no formulas, repeatedly apply these serialization maps to the current pair of subsets of WL, and take the two unions over ω. The one-stage map is a function on the set P(WL)2, so natural recursion applies. Each finite list of arguments appears by a common finite stage; hence the union is closed. Induction on stages puts it inside any constructor-closed pair containing the leaves. These are the least term and formula sets, denoted TermL and FormL.

We print the familiar abbreviations f(t1,,tn), s=t, R(t1,,tn), ¬ϕ, ϕψ and viϕ. Define ϕψ:=¬(¬ϕ¬ψ), ϕψ:=¬(ϕ¬ψ) and viϕ:=¬vi¬ϕ. In the pure membership signature there is just one binary relation symbol and every term is a variable.

Conventions and prerequisites: Set signatures and finite syntax strings, The recursion theorem.

LemmaStatement: AI-adaptedProof: AI-generatedjudge pass (gpt-5.6-terra)audited 2026-09-07Open item page →

Unique parsing of finite syntax

Statement

Every term and every formula in the delimited coding has exactly one outer constructor, with uniquely determined symbol fields and immediate subexpressions. Each immediate subexpression is shorter than its parent.

Facts & Assumptions

Given: Work in ZF unless the statement explicitly weakens or supplements it; fix the objects and hypotheses of the statement.

[F1]

Terms and formulas as finite set codes defines every constructor by an explicit concatenation map enc: outer parentheses and balance-one commas delimit token fields and complete child words. Terms and formulas are the least subsets of WL closed under those maps.

[F2]

Let P be a property of naturals such that for every nN, if P(m) holds for all m<n then P(n). Then P(n) holds for all nN. (At n=0 the hypothesis is vacuous, so P(0) is forced.) (Strong (complete) induction)

Proof

1.1

Induct on construction stages to verify that every expression starts with an opening parenthesis, ends with its matching closing parenthesis, and has positive parenthesis balance at every nonempty proper prefix. A parent contributes one to the balance while its child words each have nonnegative balance and total balance zero; leaf interiors have no parentheses. Thus these assertions hold for leaves and are preserved by all constructors.

F1
2.1

For a word known to be an expression, the second token is its unique constructor tag. Scan from its opening parenthesis: commas at balance one separate exactly its outer fields, because child interiors have balance at least two. The first return to balance zero is the end of the expression. These boundaries depend only on the word, not on a proposed construction.

step 1.1F1
3.1

The tag prescribes the sorts of the fields, and any function or relation symbol prescribes their number through its arity. The scan therefore recovers every child word uniquely. Existence follows from membership in the constructor closure. Every child omits at least the outer opening, tag and closing tokens, so is shorter; strong induction on length also recovers all descendant parses.

step 2.1F2
TheoremStatement: AI-adaptedProof: AI-generatedjudge pass (gpt-5.6-terra)audited 2026-09-07Open item page →

Structural induction and recursion on syntax

Statement

Constructor induction is valid for the term and formula sets: a property true of leaves and preserved by each licensed constructor holds of every expression. For a specified set A, supplied leaf values in A, and supplied operations into A for all constructors on the appropriate finite powers of A, there is a unique evaluation E:TermLFormLA obeying those rules. The construction is uniform in the supplied parameters. Operations may also depend on the constructor's symbol fields.

Facts & Assumptions

Given: Work in ZF unless the statement explicitly weakens or supplements it; fix the objects and hypotheses of the statement.

[F1]

Every term and every formula in the delimited coding has exactly one outer constructor, with uniquely determined symbol fields and immediate subexpressions. Each immediate subexpression is shorter than its parent. (Unique parsing of finite syntax)

[F2]

Let (N,0,σ) be a Peano system (def-peano-system), in particular the natural numbers N (def-natural-numbers). For any set A, any element aA, and any function f:AA, there is a unique function g:NA such that g(0)=a and g(σ(n))=f(g(n)) for all nN. (The recursion theorem)

[F3]

Let P be a property of naturals such that for every nN, if P(m) holds for all m<n then P(n). Then P(n) holds for all nN. (At n=0 the hypothesis is vacuous, so P(0) is forced.) (Strong (complete) induction)

Proof

1.1

Immediate subexpressions are uniquely determined and shorter. Strong induction on word length therefore proves constructor induction, including the vacuous shorter-expression hypothesis at the first leaves.

F1F3
1.2

Let D be the expression set and use the state set B=ω×P(D×A). Start with (0,). For a state (n,h) whose second component is the correct evaluation of expressions of length at most n, extend h to all expressions of length n+1 using the prescribed operations and shorter child values. Leaves use their supplied values. On any other state set the successor state to (n+1,), making the rule a total set function BB. Parsing gives a unique value at every newly added expression.

F1construct
2.1

Natural recursion in B gives the state sequence. Induction on n shows only correct states are reached and their graphs extend one another. Their union is a set function with domain D obeying the rules. Any other evaluation agrees on leaves and then on a constructor whenever it agrees on its children, so step 1.1 proves uniqueness. The state definition uses only the given parameters, proving uniformity.

F2step 1.1step 1.2
DefinitionDefinition: AI-adaptedProof: Not applicablejudge pass (gpt-5.6-terra)audited 2026-09-07Open item page →

Free variables and free-for substitution

Definition

An occurrence is a token position in the parsed finite word. A variable occurrence is free when it is in a term field and no ancestor quantifier binds that variable. The variable field of a quantifier is a binder, not a free occurrence. Write FV(e) for the finite set of variables with free occurrences in e, and Var(e) for all variables appearing anywhere. A sentence is a formula with empty FV.

The recursive rules are FV(v)={v}, FV(c)=, union over the arguments for function and atomic relation/equality expressions, unchanged under negation, union for conjunction, and FV(yψ)=FV(ψ){y}. Structural recursion justifies these set-valued definitions, with targets P(ω) after identifying variables with their indices.

Raw substitution e[t/x] replaces just the free occurrences of variable x by term t. On terms it replaces x by t, keeps other variables and constants, and acts on each argument. It commutes with atoms and Boolean constructors. At yψ it leaves the whole expression unchanged if y=x; otherwise it gives y(ψ[t/x]).

The term t is free for x in e if, at every replaced occurrence, the path to the root crosses no binder for a member of FV(t). Thus at yψ with yx and xFV(ψ) it requires both yFV(t) and that t be free for x in ψ. If no free x occurs, the condition is vacuous. Simultaneous substitution replaces the original free occurrences once; it does not perform substitutions inside inserted terms. It need not equal sequential substitution.

Conventions and prerequisites: Structural induction and recursion on syntax.

DefinitionDefinition: AI-adaptedProof: Not applicablejudge pass (gpt-5.6-terra)audited 2026-09-07Open item page →

Structures and variable assignments

Definition

An L-structure M has a nonempty set carrier M, values cMM, total functions fM:Ma(f)M, and relations RMMa(R), supplied as set-indexed interpretation data. Logical equality is literal equality on M. An assignment is a function s:ωM, also written s(vi)=s(i). The update s[x:=b] changes only the value at variable x to bM.

For structures in the same signature a homomorphism h:MN preserves constants and function values and takes related tuples to related tuples. It is strong if it also reflects every nonlogical relation. An embedding is an injective strong homomorphism; an isomorphism is a surjective embedding. A substructure has a nonempty subcarrier containing all constants, closed under all functions, with functions and relations restricted from the larger structure. A reduct keeps the carrier and only the interpretations for a smaller signature; an expansion is the reverse relationship. All finite powers and the assignment set Mω are sets.

Conventions and prerequisites: Set signatures and finite syntax strings.

DefinitionDefinition: AI-adaptedProof: Not applicablejudge pass (gpt-5.6-terra)audited 2026-09-07Open item page →

Term denotation

Definition

For an L-structure M and assignment sMω, define vis=s(i), cs=cM and

f(t1,,tn)s=fM(t1s,,tns).

This is a simultaneous definition for all assignments: recurse on terms into the set M(Mω) of functions from assignments to M. Variable leaves are projections, constant leaves are constant maps, and each function operation acts pointwise. Structural recursion gives a unique evaluation. The operations are total because the interpreted functions are total. Superscripts M on brackets will distinguish structures when needed.

Conventions and prerequisites: Structures and variable assignments, Structural induction and recursion on syntax.

TheoremStatement: AI-adaptedProof: AI-generatedjudge pass (gpt-5.6-terra)audited 2026-09-07Open item page →

Existence and uniqueness of set satisfaction

Statement

For every set signature L and L-structure M there is a unique set SatMFormL×Mω. Writing M,sϕ for membership, its clauses are

M,su=v    us=vs, M,sR(t1,,tn)    (t1s,,tns)RM, M,s¬ϕ    not M,sϕ, M,sϕψ    (M,sϕ and M,sψ), M,sxϕ    bM (M,s[x:=b]ϕ).

The set is uniformly first-order definable from L and the structure data.

Facts & Assumptions

Given: Work in ZF unless the statement explicitly weakens or supplements it; fix the objects and hypotheses of the statement.

[F1]

For an L-structure M and assignment sMω, define vis=s(i), cs=cM and f(t1,,tn)s=fM(t1s,,tns). This is a simultaneous definition for all assignments: recurse on terms into the set M(Mω) of functions from assignments to M. Variable leaves are projections, constant leaves are constant maps, and each function operation acts pointwise. Structural recursion gives a unique evaluation. The operations are total because the interpreted functions are total. Superscripts M on brackets will distinguish structures when needed. Conventions and prerequisites: def-set-structures-and-variable-assignments, thm-structural-recursion-on-set-coded-syntax. (Term denotation)

Proof

1.1

Put S=Mω. Term denotations give each atomic formula its truth subset of S by Separation: equality compares the two values and a relation tests its interpreted tuple.

F1
2.1

On truth subsets use complement in S, intersection, and for binder x the operation Qx(B)={sS:bM s[x:=b]B}. These are total operations on the set P(S). The structural recursion used to define denotation also applies to formulas with atoms as leaves, giving truth sets Bϕ.

F1step 1.1construct
3.1

Separate {(ϕ,s):sBϕ} inside FormL×S. The displayed clauses follow directly from the operations. Any rival satisfies the same atomic and constructor rules, so structural induction forces equality of all truth sets. All operations and the assertion of existence of the unique recursive evaluation are first-order formulas on sets with parameters L,M; this yields uniform definability.

step 1.1step 2.1
LemmaStatement: AI-adaptedProof: AI-generatedjudge pass (gpt-5.6-terra)audited 2026-09-07Open item page →

Coincidence for term values and satisfaction

Statement

Assignments agreeing on FV(e) give equal denotations when e is a term and the same truth value when e is a formula. Also, passage to a reduct preserves denotations and satisfaction for expressions in the smaller signature. In particular sentence truth is independent of assignment; the truth of a formula can be specified by any tuple assigning all its free variables.

Facts & Assumptions

Given: Work in ZF unless the statement explicitly weakens or supplements it; fix the objects and hypotheses of the statement.

[F1]

For every set signature L and L-structure M there is a unique set SatMFormL×Mω. Writing M,sϕ for membership, its clauses are M,su=v    us=vs, M,sR(t1,,tn)    (t1s,,tns)RM, M,s¬ϕ    not M,sϕ, M,sϕψ    (M,sϕ and M,sψ), M,sxϕ    bM (M,s[x:=b]ϕ). The set is uniformly first-order definable from L and the structure data. (Existence and uniqueness of set satisfaction)

[F2]

An occurrence is a token position in the parsed finite word. A variable occurrence is free when it is in a term field and no ancestor quantifier binds that variable. The variable field of a quantifier is a binder, not a free occurrence. Write FV(e) for the finite set of variables with free occurrences in e, and Var(e) for all variables appearing anywhere. A sentence is a formula with empty FV. The recursive rules are FV(v)={v}, FV(c)=, union over the arguments for function and atomic relation/equality expressions, unchanged under negation, union for conjunction, and FV(yψ)=FV(ψ){y}. Structural recursion justifies these set-valued definitions, with targets P(ω) after identifying variables with their indices. Raw substitution e[t/x] replaces just the free occurrences of variable x by term t. On terms it replaces x by t, keeps other variables and constants, and acts on each argument. It commutes with atoms and Boolean constructors. At yψ it leaves the whole expression unchanged if y=x; otherwise it gives y(ψ[t/x]). The term t is free for x in e if, at every replaced occurrence, the path to the root crosses no binder for a member of FV(t). Thus at yψ with yx and xFV(ψ) it requires both yFV(t) and that t be free for x in ψ. If no free x occurs, the condition is vacuous. Simultaneous substitution replaces the original free occurrences once; it does not perform substitutions inside inserted terms. It need not equal sequential substitution. Conventions and prerequisites: thm-structural-recursion-on-set-coded-syntax. (Free variables and free-for substitution)

Proof

1.1

For terms, agreement at the variables gives the same leaf values, constants are fixed, and equal argument values give equal function values. This proves the term assertion by constructor induction and the free-variable union rule. Reducts have exactly the same relevant leaf interpretations and function operations.

F1F2
2.1

For equality and relation atoms use step 1.1; negation and conjunction preserve equality of truth values by their clauses. At xψ, if s,s agree on FV(ψ){x}, then for each bM their updates at x agree on FV(ψ). The induction hypothesis identifies truth for each witness in both directions.

F1F2step 1.1
3.1

Reducts have the same carrier, update assignments and witness range, so the identical atom, Boolean and existential argument proves reduct invariance. For a sentence the agreement condition is empty. A finite free-variable assignment extends to all variables by one fixed element of the nonempty carrier, and any two extensions agree on the free variables.

F1step 2.1
LemmaStatement: AI-adaptedProof: AI-generatedjudge pass (gpt-5.6-terra)audited 2026-09-07Open item page →

Term substitution commutes with evaluation

Statement

For any terms u,t, variable x, structure M and assignment s, put b=ts. Then u[t/x]s=us[x:=b].

Facts & Assumptions

Given: Work in ZF unless the statement explicitly weakens or supplements it; fix the objects and hypotheses of the statement.

[F1]

For an L-structure M and assignment sMω, define vis=s(i), cs=cM and f(t1,,tn)s=fM(t1s,,tns). This is a simultaneous definition for all assignments: recurse on terms into the set M(Mω) of functions from assignments to M. Variable leaves are projections, constant leaves are constant maps, and each function operation acts pointwise. Structural recursion gives a unique evaluation. The operations are total because the interpreted functions are total. Superscripts M on brackets will distinguish structures when needed. Conventions and prerequisites: def-set-structures-and-variable-assignments, thm-structural-recursion-on-set-coded-syntax. (Term denotation)

[F2]

An occurrence is a token position in the parsed finite word. A variable occurrence is free when it is in a term field and no ancestor quantifier binds that variable. The variable field of a quantifier is a binder, not a free occurrence. Write FV(e) for the finite set of variables with free occurrences in e, and Var(e) for all variables appearing anywhere. A sentence is a formula with empty FV. The recursive rules are FV(v)={v}, FV(c)=, union over the arguments for function and atomic relation/equality expressions, unchanged under negation, union for conjunction, and FV(yψ)=FV(ψ){y}. Structural recursion justifies these set-valued definitions, with targets P(ω) after identifying variables with their indices. Raw substitution e[t/x] replaces just the free occurrences of variable x by term t. On terms it replaces x by t, keeps other variables and constants, and acts on each argument. It commutes with atoms and Boolean constructors. At yψ it leaves the whole expression unchanged if y=x; otherwise it gives y(ψ[t/x]). The term t is free for x in e if, at every replaced occurrence, the path to the root crosses no binder for a member of FV(t). Thus at yψ with yx and xFV(ψ) it requires both yFV(t) and that t be free for x in ψ. If no free x occurs, the condition is vacuous. Simultaneous substitution replaces the original free occurrences once; it does not perform substitutions inside inserted terms. It need not equal sequential substitution. Conventions and prerequisites: thm-structural-recursion-on-set-coded-syntax. (Free variables and free-for substitution)

Proof

1.1

If u=x, both sides equal b. If u is another variable, the update leaves its value unchanged. If u is a constant, both sides are its interpretation. These cover all leaves, including when t itself contains x.

F1F2
2.1

At u=f(u1,,un), substitution acts on its arguments. The induction hypothesis identifies every argument value on the two sides, and applying the same total function fM gives the displayed equality. Constructor induction completes the proof.

F1F2step 1.1
LemmaStatement: AI-adaptedProof: AI-generatedjudge pass (gpt-5.6-terra)audited 2026-09-07Open item page →

Free-for substitution commutes with satisfaction

Statement

If t is free for x in ϕ, then for every structure M and assignment s,

M,sϕ[t/x]    M,s[x:=ts]ϕ.

Facts & Assumptions

Given: Work in ZF unless the statement explicitly weakens or supplements it; fix the objects and hypotheses of the statement.

[F1]

For any terms u,t, variable x, structure M and assignment s, put b=ts. Then u[t/x]s=us[x:=b]. (Term substitution commutes with evaluation)

[F2]

Assignments agreeing on FV(e) give equal denotations when e is a term and the same truth value when e is a formula. Also, passage to a reduct preserves denotations and satisfaction for expressions in the smaller signature. In particular sentence truth is independent of assignment; the truth of a formula can be specified by any tuple assigning all its free variables. (Coincidence for term values and satisfaction)

Proof

1.1

Write b=ts. The term substitution identity proves the assertion for both kinds of atoms. Negation and conjunction transport the inductive equivalences through their truth clauses.

F1
1.2

At yψ, if y=x, raw substitution leaves this quantified formula unchanged and updating its bound variable externally cannot affect its truth by coincidence. If xFV(ψ), substitution again changes nothing and coincidence gives the result.

F2
1.3

Otherwise yx and xFV(ψ). The free-for hypothesis gives yFV(t) and the free-for condition in ψ. For each aM, coincidence makes ts[y:=a]=b. The two distinct-variable updates commute: s[y:=a][x:=b]=s[x:=b][y:=a]. Applying the induction hypothesis to ψ at s[y:=a] therefore gives the same truth at each witness a on both sides.

F1F2
2.1

Existential quantification over these pointwise equivalent assertions proves both directions for yψ. The cases above exhaust its possibilities, so structural induction gives the result for every formula.

step 1.1step 1.2step 1.3
LemmaStatement: AI-adaptedProof: AI-generatedjudge pass (gpt-5.6-terra)audited 2026-09-07Open item page →

Fresh bound-variable renaming preserves truth

Statement

If zy is absent from Var(ψ), then z is free for y in ψ, and yψ and z(ψ[z/y]) have the same truth in every structure and assignment. Here raw substitution replaces only occurrences free in ψ, hence only those bound by the displayed outer binder. A least fresh variable exists outside any specified finite set of variables.

Facts & Assumptions

Given: Work in ZF unless the statement explicitly weakens or supplements it; fix the objects and hypotheses of the statement.

[F1]

If t is free for x in ϕ, then for every structure M and assignment s, M,sϕ[t/x]    M,s[x:=ts]ϕ. (Free-for substitution commutes with satisfaction)

Proof

1.1

No binder in ψ uses z, so substitution introduces no capture. At an assignment s[z:=a], the free-for identity gives truth of ψ[z/y] exactly when ψ is true at s[z:=a][y:=a].

F1
2.1

Since z is absent from ψ, coincidence (as used in the free-for lemma) removes the update at z when evaluating ψ. The last assignment therefore gives the same truth as s[y:=a]. Quantifying over the same aM proves the claimed equivalence. Inner binders for y were untouched by the raw substitution rule.

F1step 1.1
3.1

A finite set of variable indices is bounded in ω, by induction over its finite listing and taking successive maxima. There is an index outside it, and the least such index is uniquely specified by the natural order. This proves the fresh-variable assertion without choosing from a family of sets.

given
PropositionStatement: AI-adaptedProof: AI-generatedjudge pass (gpt-5.6-terra)audited 2026-09-07Open item page →

Canonical capture-avoiding substitution

Statement

There is a canonical total capture-avoiding substitution ϕt/x on formulas, obtained by renaming obstructing binders using least unused variable indices. It replaces the original free occurrences of x and satisfies

M,sϕt/x    M,s[x:=ts]ϕ.

Canonicity refers to the specified coding and traversal, not to literal invariance under other fresh-variable conventions.

Facts & Assumptions

Given: Work in ZF unless the statement explicitly weakens or supplements it; fix the objects and hypotheses of the statement.

[F1]

If zy is absent from Var(ψ), then z is free for y in ψ, and yψ and z(ψ[z/y]) have the same truth in every structure and assignment. Here raw substitution replaces only occurrences free in ψ, hence only those bound by the displayed outer binder. A least fresh variable exists outside any specified finite set of variables. (Fresh bound-variable renaming preserves truth)

[F2]

If t is free for x in ψ, then for every structure M and assignment s, M,sψ[t/x]    M,s[x:=ts]ψ. (Free-for substitution commutes with satisfaction)

Proof

1.1

Traverse the original parse tree in preorder, visiting the left child before the right child. Begin with the finite forbidden set Var(ϕ)Var(t){x}. Rename every binder whose variable belongs to FV(t){x} to the least index outside the current forbidden set, add that variable to the forbidden set, and rename exactly the occurrences governed by that binder. At each operation the new name is absent from the entire current formula, so the fresh-renaming lemma applies.

F1construct
2.1

There are only finitely many original binder nodes. Renaming changes names but neither the tree nor the number of nodes; hence the traversal terminates and produces a uniquely specified formula ϕ. Repeated fresh-renaming equivalence gives the same truth as ϕ at every assignment. Original free occurrences are unchanged; all remaining binders avoid x and FV(t).

F1step 1.1
3.1

Set ϕt/x=ϕ[t/x]. Now t is free for x in ϕ, so F2 identifies its truth at s with the truth of ϕ at s[x:=ts]. Step 2.1 changes that last formula back to ϕ. This proves both directions and totality.

F2step 2.1
CorollaryStatement: AI-adaptedProof: AI-generatedjudge pass (gpt-5.6-terra)audited 2026-09-07Open item page →

Semantic equivalence is preserved by substitution

Statement

If ϕ and ψ have the same truth at every assignment in every L-structure, then so do ϕt/x and ψt/x. Moreover, in any fixed structure and assignment, terms t,u of equal value give the same truth for ϕt/x and ϕu/x.

Facts & Assumptions

Given: Work in ZF unless the statement explicitly weakens or supplements it; fix the objects and hypotheses of the statement.

[F1]

There is a canonical total capture-avoiding substitution ϕt/x on formulas, obtained by renaming obstructing binders using least unused variable indices. It replaces the original free occurrences of x and satisfies M,sϕt/x    M,s[x:=ts]ϕ. Canonicity refers to the specified coding and traversal, not to literal invariance under other fresh-variable conventions. (Canonical capture-avoiding substitution)

Proof

1.1

Fix a structure and assignment s and put b=ts. The two substituted formulas have the truth values of ϕ and ψ at the same assignment s[x:=b]. The hypothesis applies at this assignment, so they agree.

F1
2.1

If ts=us, the updated assignments are literally equal. Apply the substitution identity once for each term to obtain equality of the two truth values.

F1
DefinitionDefinition: AI-adaptedProof: Not applicablejudge pass (gpt-5.6-terra)audited 2026-09-07Open item page →

Theories, models and semantic consequence

Definition

An L-theory T is a set of L-sentences, without any requirement that it be deductively closed. A structure M is a model of T if it satisfies every sentence of T. By coincidence this does not depend on the chosen assignment. Write Tϕ if for every set L-structure M which is a model of T and every assignment s, one has M,sϕ. Thus if T has no models the consequence condition is vacuous.

Validity means truth in every structure at every assignment; semantic equivalence means equality of truth under all such choices. The universal closure of ϕ binds its finitely many free variables in increasing index order (and is ϕ if none occur). Repeated use of the universal truth clause shows that a structure satisfies this closure exactly when ϕ holds at every assignment. Quantification over all set structures is a first-order class description, not a set of all structures.

Conventions and prerequisites: Coincidence for term values and satisfaction.

PropositionStatement: AI-adaptedProof: AI-generatedjudge pass (gpt-5.6-terra)audited 2026-09-07Open item page →

Isomorphisms preserve satisfaction

Statement

For any homomorphism h:MN, term t and assignment s, thsN=h(tsM). If h is a surjective strong homomorphism, then M,sϕ iff N,hsϕ for every equality-free formula ϕ. If h is an isomorphism, the equivalence holds for all formulas, including equality.

Facts & Assumptions

Given: Work in ZF unless the statement explicitly weakens or supplements it; fix the objects and hypotheses of the statement.

[F1]

For every set signature L and L-structure M there is a unique set SatMFormL×Mω. Writing M,sϕ for membership, its clauses are M,su=v    us=vs, M,sR(t1,,tn)    (t1s,,tns)RM, M,s¬ϕ    not M,sϕ, M,sϕψ    (M,sϕ and M,sψ), M,sxϕ    bM (M,s[x:=b]ϕ). The set is uniformly first-order definable from L and the structure data. (Existence and uniqueness of set satisfaction)

[F2]

An L-structure M has a nonempty set carrier M, values cMM, total functions fM:Ma(f)M, and relations RMMa(R), supplied as set-indexed interpretation data. Logical equality is literal equality on M. An assignment is a function s:ωM, also written s(vi)=s(i). The update s[x:=b] changes only the value at variable x to bM. For structures in the same signature a homomorphism h:MN preserves constants and function values and takes related tuples to related tuples. It is strong if it also reflects every nonlogical relation. An embedding is an injective strong homomorphism; an isomorphism is a surjective embedding. A substructure has a nonempty subcarrier containing all constants, closed under all functions, with functions and relations restricted from the larger structure. A reduct keeps the carrier and only the interpretations for a smaller signature; an expansion is the reverse relationship. All finite powers and the assignment set Mω are sets. Conventions and prerequisites: def-set-signature-and-finite-syntax-strings. (Structures and variable assignments)

Proof

1.1

For variables the term identity is the definition of composition; for constants it is preservation of constants. For a function application, use the identities for its arguments and preservation of that function. Constructor induction proves the term claim.

F2
2.1

The term identity and strong preservation give equivalence for relation atoms. For equality atoms, the term identity gives the forward direction; injectivity gives the reverse direction when h is an isomorphism. Negation and conjunction preserve these equivalences.

F1F2step 1.1
3.1

A witness aM is sent to h(a)N, and h(s[x:=a])=(hs)[x:=h(a)]. Conversely, for a witness bN, surjectivity supplies one a with h(a)=b. The induction hypothesis on the quantified body then transfers witnesses in both directions. Thus formula induction proves both asserted scopes, with injectivity needed precisely for equality.

F1F2step 2.1
DefinitionDefinition: AI-adaptedProof: Not applicablejudge pass (gpt-5.6-terra)audited 2026-09-07Open item page →

Relativization to sets and definable classes

Definition

Fix a pure membership formula δ(z,p) defining a class C={z:δ(z,p)}. This is eliminable notation for a predicate, not a class object. For a fixed pure membership formula ϕ, first rename its binders away from the parameter variables p. Define ϕC by keeping atoms, commuting with Boolean constructors, and setting

(xψ)C=x(δ(x,p)ψC),(xψ)C=x(δ(x,p)ψC).

Copies of δ(x,p) are inserted by capture-avoiding substitution and fresh internal bound variables. In the set case C=M use the predicate zM, with a fresh parameter variable for M. This operation is meaningful even for an empty class. It does not make the empty class an admitted structure: carriers of structures remain nonempty. For proper classes, evaluation of ϕC means a separate ambient formula for each fixed ϕ, not a uniform universe satisfaction relation.

Conventions and prerequisites: Canonical capture-avoiding substitution.

TheoremStatement: AI-adaptedProof: AI-generatedjudge pass (gpt-5.6-terra)audited 2026-09-07Open item page →

Relativization agrees with induced set satisfaction

Statement

For every nonempty set M, let M=(M, ⁣M2) be its membership structure. For every fixed pure membership formula ϕ and assignment s:ωM, its set satisfaction agrees with the ambient evaluation of ϕM at those free-variable values and parameter M. No transitivity of M is required. For a definable proper class the corresponding relativized expression is only a formula-by-formula schema.

Facts & Assumptions

Given: Work in ZF unless the statement explicitly weakens or supplements it; fix the objects and hypotheses of the statement.

[F1]

Fix a pure membership formula δ(z,p) defining a class C={z:δ(z,p)}. This is eliminable notation for a predicate, not a class object. For a fixed pure membership formula ϕ, first rename its binders away from the parameter variables p. Define ϕC by keeping atoms, commuting with Boolean constructors, and setting (xψ)C=x(δ(x,p)ψC),(xψ)C=x(δ(x,p)ψC). Copies of δ(x,p) are inserted by capture-avoiding substitution and fresh internal bound variables. In the set case C=M use the predicate zM, with a fresh parameter variable for M. This operation is meaningful even for an empty class. It does not make the empty class an admitted structure: carriers of structures remain nonempty. For proper classes, evaluation of ϕC means a separate ambient formula for each fixed ϕ, not a uniform universe satisfaction relation. Conventions and prerequisites: prop-capture-avoiding-substitution. (Relativization to sets and definable classes)

[F2]

For every set signature L and L-structure M there is a unique set SatMFormL×Mω. Writing M,sϕ for membership, its clauses are M,su=v    us=vs, M,sR(t1,,tn)    (t1s,,tns)RM, M,s¬ϕ    not M,sϕ, M,sϕψ    (M,sϕ and M,sψ), M,sxϕ    bM (M,s[x:=b]ϕ). The set is uniformly first-order definable from L and the structure data. (Existence and uniqueness of set satisfaction)

Proof

1.1

All terms are variables. Equality and membership between two values in M are exactly the ambient equality and membership, because the structure restricts the ambient relation. Atoms are unchanged by relativization, giving the atomic equivalence.

F1F2
2.1

Negation and conjunction preserve the inductive equivalences. For xψ, the ambient guard xM ranges over exactly the same witnesses bM as the satisfaction clause, and the induction hypothesis applies at s[x:=b]. This proves both directions for quantifiers and hence all formulas. The parameter for M cannot be captured because the construction uses fresh binders.

F1F2step 1.1
3.1

The argument used only restriction of membership and equality, never transitivity. Replacing the set guard by a fixed defining formula for a class still makes a separate ambient formula at each syntactic step; it does not assemble a truth relation with proper-class assignment domain.

F1step 2.1
RemarkRemark: AI-adaptedProof: Not applicablejudge pass (gpt-5.6-terra)audited 2026-09-07Open item page →

Set truth and the Tarski interface

Remark

Set satisfaction and relativization must be distinguished. Set satisfaction is uniform in the code of a formula and the data of a set structure. Relativization to a definable proper class supplies an ambient formula separately for each fixed input formula. Writing Vϕ(a) in this latter sense merely abbreviates ϕ(a).

For the truth-with-parameters interface, a proposed pure membership formula T(w,x,y,z) and a set parameter t would have to satisfy, for every pure membership formula ψ with free variables among x,y and all sets r,s, the biconditional ψ(r,s)T(ψ,r,s,t). Here ψ is its finite set code. This is a scheme of requirements, not a single first-order assertion quantifying over ambient truths. The companion refutation tests this exact scheme by one formula built from the proposed T. No sentence-only arithmetized diagonal lemma or representability theorem is asserted here.

Conventions and prerequisites: Relativization agrees with induced set satisfaction.

5 · Examples, counterexamples and false statements

None yet.

Sources