Some Notes on Truths and Comprehension
 1.4k Downloads
 1 Citations
Abstract
In this paper we study several translations that map models and formulae of the language of secondorder arithmetic to models and formulae of the language of truth. These translations are useful because they allow us to exploit results from the extensive literature on arithmetic to study the notion of truth. Our purpose is to present these connections in a systematic way, generalize some wellknown results in this area, and to provide a number of new results. Sections 3 and 4 contain some recursion and prooftheoretic results about Kripkestyle fixedpoint theories of truth. Section 5 shows how to derive full secondorder arithmetic from principles of truth. Section 6 investigates the prooftheoretic strength of disquotation without an arithmetical base theory.
Keywords
Truth Secondorder arithmetic Relative interpretations Definability1 Introduction
In this paper we study several translations, essentially known to experts, that map models and formulae of the language of secondorder arithmetic to models and formulae of the language of Peano arithmetic augmented by a truth predicate. These translations are useful because they allow us to exploit results from the extensiveliterature on arithmetic in order to study the notion of truth. Our purpose is to present these connections in a systematic way, to generalize some wellknown results in this area, and to provide a number of new results.
While this paper is essentially a technical one, many of its results should be of great philosophical interest. Deflationists such as Horwich [18] claim that truth and satisfaction (truth of) are merely expressive devices. This idea can be spelled out in various ways. According to Parsons [26], the notion of truth enables us to generalize sentence places while the notion of satisfaction allows us to quantify into predicate position. The direct method of generalizing sentence and predicate places is by introducing quantifiable variables that can occupy sentence and predicate position; that is, by introducing secondorder variables. A natural way, then, of understanding the idea that truth and satisfaction enable us to generalize sentence respectively predicate places is to say that a theory of truth must be able to emulate or interpret some form of secondorder quantification. The translations presented in this paper may help us to make this idea precise.
Secondly, interpretations of (subsystems of) secondorder arithmetic into theories of truth can be seen as ontological reductions. In secondorder arithmetic, the secondorder quantifiers are usually taken to range over sets of numbers. For some sets X we can find a predicate P such that n ∈ X if and only if P is true of n. Since we can identify predicates with their Gödel numbers, talk of certain sets of numbers can be reduced to or replaced by talk of numbers and satisfaction. Wellknown examples in this area are the reduction of the system of arithmetical comprehension to the typed compositional theory of truth and the reduction of ramified analysis to the Tarskian hierarchy of truth. An early philosophical discussion of both results can be found in Parsons [25], who also sketches a proof of both results. These examples show that we may identify arithmetical respectively hyperarithmetical sets with the predicates (or their Gödel codes) of which they are the extension. (For technical details I refer the reader to Halbach [10, 11]. For further discussion, see Halbach [12].) The results in this paper show that we can reduce even stronger subsystems of secondorder arithmetic to theories of truth.
There are some further philosophical issues on which our results may shed some light, such as the question of what constitues a good axiomatization of a semantic truth theory (Fischer et al. [8]), or the question of the ‘unsubstantiality’ of truth (Horsten [17], Shapiro [31], Ketland [19], Halbach [13]). However, I won’t attempt to draw any substantial philosophical conclusions in this paper and leave a proper assessment for another occasion.
This paper is structured as follows. In Section 2 we set up the overall framework of the paper. The ideas and results presented there are more or less known to experts. We fix a translation function from the language of secondorder arithmetic to the language of truth. The main idea is to translate a formula of the form t ∈ Y as ‘the result of substituting t for the free variable in the formula y is true’ or ‘y is true of t’. With every interpretation S for the truth predicate we canonically associate a Henkin structure \(\mathcal {M}_{S}\) for the secondorder language. It is then shown that a secondorder sentence is true in \(\mathcal {M}_{S}\) if and only if its translation is true when S is assigned as the extension of the truth predicate.
In Section 3, we apply the translation lemma to obtain some complexity results for Kripkestyle fixedpoint theories of truth [20]. We show that whenever a valuation scheme satisfies some minimal criteria, then all inductive sets are weakly definable in the least fixed point of that valuation scheme, while the sets that are strongly definable are exactly the hyperarithmetical ones. The proof presented here is more general than those currently available in the literature as it applies to the Strong Kleene scheme and the supervaluation schemes as well as to Leitgeb’s theory [21]. It also applies to a valuation scheme strictly weaker than the Strong Kleene. We also give a very short proof that all inductive sets are definable in Herzberger’s revision theory [16].
In Section 4, we draw some prooftheoretic consequences. It is shown that the least fixed points of the valuation schemes satisfying the criteria set out in the previous section validate (the translations of) the theories I D _{1} and \({\Delta ^{1}_{1}}\mathsf {CA}_{0}\). The axioms of the latter could therefore be taken into account for axiomatizations of the semantic theories. We also present some sufficient conditions under which all fixed points of a valuation scheme satisfy the theory of positive disquotation.
In Section 5 it is shown how to turn comprehension axioms into truththeoretic axioms. We present a disquotational theory of truth that relatively interprets full secondorder arithmetic, and show it to be consistent. This strengthens a result of Schindler [30] that parameterfree secondorder arithmetic is reducible to principles of truth.
Section 6 investigates the prooftheoretic strength of uniform disquotation without its base theory. This is done by recovering weak set theories from the Tbiconditionals. It is shown that typed uniform disquotation, over logic alone, interprets Tarski’s theory R while the theory of positive uniform disquotation, over logic alone, interprets Robinson arithmetic Q.
In Section 7 we conclude with some final remarks.
Technical Preliminaries
The language of Peano arithmetic, \(\mathcal {L}_{\text {PA}}\), is a firstorder language that contains a denumerably infinite set of individual variables v _{0}, v _{1}, v _{2},…, the connectives ¬,∨ and ∧, the quantifiers ∀,∃ and the identity symbol =. We assume that all other connectives are defined in the usual way. The sole nonlogical symbols are the individual constant \(\overline {0}\), the unary function symbol S for the successor function, the binary function symbols + and ⋅ for addition and multiplication, respectively, and function symbols for certain primitive recursive (p.r.) functions that we are going to specify in the course of the paper. If h is such a p.r. function, we write ḥ (with a subdot) for the corresponding function symbol. The language \(\mathcal {L}_{\text {T}}\) is obtained from \(\mathcal {L}_{\text {PA}}\) by augmenting the latter with the unary predicate symbol T.
If n is a number, we write \(\overline {n}\) for its numeral, i.e., the term that is obtained by applying the successor symbol S nmany times to the constant \(\overline {0}\). We assume some effective Gödelcoding of the expressions of \(\mathcal {L}_{\text {T}}\). If σ is some expression, we write # σ for its code and \(\ulcorner {\sigma }\urcorner \) for the numeral of its code. We occasionally identify expressions with their codes.
For more details on this notation, I refer the reader to Cantini [6] or Halbach [15].
Standard models of \(\mathcal {L}_{\text {T}}\) have the form \((\mathbb {N}, S)\), where \(\mathbb {N}\) is the standard model of P A and S ⊆ ω interprets the truth predicate T.
Let us now turn to the language of secondorder arithmetic. The language \(\mathcal {L}_{2}\) of secondorder arithmetic is obtained from \(\mathcal {L}_{\text {PA}}\) by adding the binary relation symbol ∈ plus secondorder or set variables X _{0}, X _{1}, X _{2},… (Let us call v _{0}, v _{1},…number variables.) This gives us new formulae of the form t ∈ X and ∀X φ.
\(\mathcal {L}_{2}\) is a twosorted firstorder language with usual (firstorder) rules for both set and number quantifiers. Intended Henkin models for \(\mathcal {L}_{2}\) have the form \((\mathbb {N}, \mathcal {M})\), where \(\mathcal {M}\subseteq \wp (\omega )\) and the set variables X _{ i } range over the elements of \(\mathcal {M}\). We call such models ωmodels.
A formula \(\varphi \in \mathcal {L}_{2}\) is called arithmetical iff it contains no secondorder quantifiers. Note that such a formula might contain free secondorder variables. (Sometimes we also refer to the formulae of \(\mathcal {L}_{\text {PA}}\) as arithmetical. It should always be clear from the context which sense is intended.)
A formula \(\varphi \in \mathcal {L}_{2}\) is \({\Pi ^{1}_{n}} \ ({\Sigma ^{1}_{n}})\) iff its has the form Q _{1} X _{1}…Q _{ n } X _{ n } ψ, where ψ is arithmetical, Q _{1}…Q _{ n } is a string of alternating secondorder quantifiers, and Q _{1} is universal (existential).
A set X ⊆ ω is \({\Pi ^{1}_{n}} \ ({\Sigma ^{1}_{n}})\) iff \(X=\{n\mid (\mathbb {N}, \wp (\omega ))\models \varphi (\overline n)\}\), where φ(x) is a \({\Pi ^{1}_{n}} ({\Sigma ^{1}_{n}})\) formula with exactly x free. (That is, a set is \({\Pi ^{1}_{n}}\) iff it is definable by a \({\Pi ^{1}_{n}}\)formula in the standard model of secondorder arithmetic.) A set X ⊆ ω is \({\Delta ^{1}_{n}}\) iff X is both \({\Pi ^{1}_{n}}\) and \({\Sigma ^{1}_{n}}\). For more information on the analytic hierarchy I refer the reader to Shoenfield [32].
Another notion that we will use frequently in this paper is that of a relative interpretation. Giving a general definition of that notion is a tricky business (cf. Visser [36]), but for our purposes the following should suffice. Roughly, a theory T _{1} in \(\mathcal {L}_{T_{1}}\) is relatively interpretable in a theory T _{2} in \(\mathcal {L}_{T_{2}}\) if and only if there is a definitional extension of T _{2} in an expansion of \(\mathcal {L}_{T_{2}}\) and a translation I from \(\mathcal {L}_{T_{1}}\) to the expansion of \(\mathcal {L}_{T_{2}}\) such that (i) I replaces every nonlogical expression of \(\mathcal {L}_{T_{1}}\) by one of the same kind and arity (but not necessarily the same sort), (ii) I preserves logical structure, possibly relativizing quantifiers, and (iii) I preserves theoremhood. Here, the notion of a definitional extension is the standard one as in e.g. Monk [22, p. 208] ; for instance, we assume the usual uniqueness and existence conditions for function symbols. In practice, we will often specify a translation by specifying a formula with n free variables as translation of an nary predicate symbol instead of specifying a predicate symbol in an expansion as required by the definition.
We use bold face letters to indicate strings of expressions. For example, x denotes x _{1},…,x _{ n } and \({\ulcorner \varphi \urcorner }\) denotes \(\ulcorner {\varphi _{1}}\urcorner , \ldots , \ulcorner {\varphi _{n}}\urcorner \). It should always be clear from the context what the length of the string is.
2 TruthSets and SecondOrder Structures
In this section we introduce our first translation from \(\mathcal {L}_{2}\) to \(\mathcal {L}_{\text {T}}\). The ideas and results presented here are more or less known to experts, although we are not aware of a systematic presentation of them in the literature. The purpose of this section is mainly to set up the overall framework of this paper. Recall that standard models of \(\mathcal {L}_{\text {T}}\) have the form \((\mathbb {N}, S)\), where \(\mathbb {N}\) interprets the arithmetical vocabulary and S ⊆ ω interprets the truth predicate T. Let us call S a truthset. Any truthset S ⊆ ω encodes or determines a Henkin structure \((\mathbb {N}, \mathcal {M}_{S})\) for \(\mathcal {L}_{2}\) as follows:
Definition 1
 1.
\(S_{\varphi } = \lbrace n\mid \#\varphi (\overline n)\in S\rbrace \subseteq \omega \)
 2.
\(\mathcal {M}_{S} = \left \lbrace S_{\varphi } \mid \varphi \in \mathit {Form}^{1}_{T} \right \rbrace \subseteq \wp (\omega )\)
In the terminology of Cantini [5, 6], \(\mathcal {M}_{S}\) is the envelope of S; we will call it the set of sets encoded by S. Note that by Tarski’s undefinability theorem, in general we do not have \(S_{\varphi }=\{n\mid (\mathbb {N}, S)\models \varphi (\overline n)\}\). Hence \(\mathcal {M}_{S}\) does not coincide with the set of sets that are definable in the classical structure \((\mathbb {N}, S)\). However, we will later see (Section 3) that \(\mathcal {M}_{S}\) may coincide with the set of sets that are definable in S in some nonclassical logic.
As mentioned in the introduction, we occasionally identify expressions with their codes. Accordingly, we also write S _{ k } for S _{ φ }, provided that k = # φ. Occassionally, we also denote S _{ φ } by S _{ t } when t is a term denoting # φ.
Now consider the following translation function from the language \(\mathcal {L}_{2}\) to the truth language \(\mathcal {L}_{\text {T}}\).
Definition 2
If h is a variable assignment for \((\mathbb {N}, \mathcal {M}_{S})\), define the assignment h ^{∗} for \((\mathbb {N}, S)\) by h ^{∗}(v _{2i }): = h(v _{ i }) and h ^{∗}(v _{2i+1}):= min\(\{ k \in \mathit {Form}^{1}_{T} \ \mid \ S_{k} = h(X_{i})\}\). It is easily seen that our translation preserves the denotation of arithmetical terms.
Proposition 1
Let h be an assignment for \((\mathbb {N}, \mathcal {M}_{S})\) . Then \(t^{(\mathbb {N}, \mathcal {M}_{S}), h}=t^{*(\mathbb {N}, S), h^{*}}\) for all number terms t of \(\mathcal {L}_{\text {PA}}\).
The following central proposition shows that a secondorder sentence is true in the set of sets encoded by S if and only if its translation into the language of truth is true in the truthset S (i.e., if the truth predicate is interpreted by S). Roughly, the gist of this result is that we can convert or translate questions regarding truthsets into questions regarding secondorder structures which are often better understood. We will see several applications of this in subsequent sections.
Proposition 2 (Translation Lemma)
Proof
By induction on the complexity of formulae.
The case t _{1} = t _{2} follows from Proposition 1.
The cases ¬ψ, ψ ∧ χ, ψ ∨ χ, ∃ x ψ and ∀x ψ follow easily from the I.H.
The case ∃X ψ is proved in a similar way. □
As a corollary we get:
Proposition 3
Proposition 4
Sometimes it is convenient to assume that the language \(\mathcal {L}_{2}\) additionally contains nary relation variables \({X^{n}_{1}}, {X^{n}_{2}}, \ldots \) for every n > 1. It is straightforward to extend our apparatus to cover this more general situation and we will occasionally make use of this in what follows.
For example, on the syntactic side, a formula of the form Y ^{2}(x _{1}, x _{2}) translates into a formula of the form T ṣ(ṣ \(_{2}^{2}\)(y, x _{2}), x _{1}). A formula of the form ∀Y ^{ n } φ translates into a formula of the form \(\forall y(F{m^{n}_{T}}(y)\rightarrow \varphi ^{*})\), where \(F{m^{n}_{T}}(x)\) represents the set of \(\mathcal {L}_{\text {T}}\)formulae with exactly n free variables. On the semantic side, if φ(x, y) is an \(\mathcal {L}_{\text {T}}\)formula containing exactly x, y free, and S ⊆ ω, we let \(S_{\varphi }=\{\langle n,m \rangle \mid \#\varphi (\overline n, \overline m)\in S\}\). It should be obvious that Propositions 1 to 4 can be extended to the more general case.
Of course, since there is a primitive recursive coding machinery in the language of arithmetic, we could simply do with the unary case. However, I believe that the presentation of some of the later results will be more perspicuous if we avoid additional coding.
3 Definability in FixedPoint Theories
The Translation Lemma is useful because it allows us to transfer results about arithmetic to theories of truth. Let us illustrate this by proving some definability results for Kripkestyle fixedpoint theories of truth. These results are wellknown for particular choices of a valuation scheme, but the Translation Lemma allows us to generalize these results to a wide range of valuation schemes (in particular, it also applies to Leitgeb’s [21] theory of truth and to a scheme strictly weaker than the Strong Kleene, as well as to the revision theory of truth) and prove them in a uniform way. We assume that the reader is familiar with Kripke’s theory (cf. Kripke [20]) and introduce the following definitions only to fix some terminology.

(V1) The arithmetical vocabulary is interpreted in the standard way and every \(\mathcal {L}_{\text {PA}}\)sentence receives the same truth value under V as it receives in the standard model \(\mathbb {N}\).

(V2) A sentence of the form T t receives the value 1 iff \(t^{\mathbb {N}}\in E\) and receives the value 0 iff \(t^{\mathbb {N}}\in A\).

(V3) A conjunction is true under V iff both conjuncts are true under V.

(V4) A universal sentence is true under V iff all its instances are true under V.

(V5) A sentence receives value 1 under V iff its negation receives value 0 (and vice versa).
A set X ⊆ ω is called \({\Pi ^{1}_{1}}\)hard iff every \({\Pi ^{1}_{1}}\)set Y is manyone reducible to X, i.e., there is a recursive function f such that for all n ∈ ω, n ∈ Y iff f(n) ∈ X. A set X is called \({\Pi ^{1}_{1}}\)complete iff X is \({\Pi ^{1}_{1}}\)hard and X is a \({\Pi ^{1}_{1}}\)set.
It is wellknown that E ^{ ∞ } is a \({\Pi ^{1}_{1}}\)complete set of integers for several choices of V. This was proved for the Strong Kleene scheme by Burgess [3] and Cantini [5] (the result was already announced by Kripke), for the van Fraassen scheme by Burgess [3], for the Cantini scheme by Cantini [6],^{1} and for Leitgeb’s [21] theory (which is based on his notion of semantic dependence) by Welch [40]. The proof in [40] was generalized by Fischer et al. [8] to cover a range of further supervaluational schemes (for example, supervaluations based on maximal consistent extensions). Leitgeb’s theory is not defined in the Kripkean way, but in Beringer and Schindler [1] it is shown that Leitgeb’s theory coincides with the least fixed point of a particular 3valued valuation scheme that we dubbed the Leitgeb valuation scheme, V _{ L }. Thus, Leitgeb’s theory can be considered as a special case of Kripke’s.^{2} (Thus, whenever we talk about Kripke’s theory of truth in what follows, Leitgeb’s theory is always included.) We observe the following:
Proposition 5
If \(\mathcal {M}_{E^{\infty }}\) contains all \({\Pi ^{1}_{1}}\) sets, then E ^{ ∞ } is \({\Pi ^{1}_{1}}\) hard.
Proof
For every \({\Pi ^{1}_{1}}\)set P we need to find a recursive function such that n ∈ P iff f(n) ∈ E ^{ ∞ }. By assumption, \(P=(E^{\infty })_{\varphi }\in \mathcal {M}_{E^{\infty }}\) for some φ(x). Let \(f(n):=\#\varphi (\overline n)\). □
Using the Translation Lemma, we can give a short and uniform proof that E ^{ ∞ } is \({\Pi ^{1}_{1}}\)hard for a wide range of valuation schemes, including all of the schemes mentioned above. (We call the class of these valuation schemes nice.) There is no general argument by which one could show that E ^{ ∞ } is itself a \({\Pi ^{1}_{1}}\)set, but for the valuation schemes that are usually discussed in the truththeoretic literature such arguments are straightforward.
The general strategy of the proof is as follows. Any \({\Pi ^{1}_{1}}\)set P can be seen as the slice or projection of the fixed point I of a particular operator. This operator is given by a particular formula φ of \(\mathcal {L}_{2}\). Whenever V is nice, we can utilize our translation function to find a formula φ ^{′} of \(\mathcal {L}_{\text {T}}\) such that \((E^{\infty })_{\varphi ^{\prime }}=I\). The set P can be easily recovered from \((E^{\infty })_{\varphi ^{\prime }}\). This shows that \(\mathcal {M}_{E^{\infty }}\) contains all \({\Pi ^{1}_{1}}\)sets and we can apply Proposition 5.

\(I^{0}_{\varphi }=\emptyset \)

\(I^{\alpha +1}_{\varphi }=\Gamma _{\varphi }(I^{\alpha }_{\varphi })\)

\(I^{\gamma }_{\varphi }=\bigcup _{\alpha <\gamma }I^{\alpha }_{\varphi }\), when γ is a limit ordinal.
Our goal is to find a formula φ ^{′} of \(\mathcal {L}_{\text {T}}\) such that \((E^{\infty })_{\varphi ^{\prime }}=I_{\varphi }\). In fact, we will show that \((E^{\alpha })_{\varphi ^{\prime }}=I^{\alpha }_{\varphi }\) for every α.
The displayed formula is our desired formula φ ^{′}.
There are some valuation schemes such that E ^{ ∞ } is not \({\Pi ^{1}_{1}}\)hard (for instance, Cain and Damnjanovic [4] have shown that the Weak Kleene scheme does not necessarily generate a \({\Pi ^{1}_{1}}\)hard fixed point, depending on the chosen Gödel coding).^{4} However, we will show that whenever V is nice in the sense of the following definition, E ^{ ∞ } will indeed be \({\Pi ^{1}_{1}}\)hard.
Definition 3
 (N1)
if \(\psi \in \mathcal {L}_{PA}\) and \(\mathbb {N}\models \psi \) then \((\mathbb {N}, E, A)\models _{V} \psi \vee \varphi \)
 (N2)
if \((\mathbb {N}, E, A)\models _{V} \varphi \) and \(\psi \in \mathcal {L}_{PA}\) then \((\mathbb {N}, E, A)\models _{V} \psi \vee \varphi \)
 (N3)
if a disjunction ψ ∨ φ is true under V and \(\psi \in \mathcal {L}_{PA}\), then (at least) one of ψ, φ is true under V
In other words, the conditions require that if \(\psi \in \mathcal {L}_{PA}\), then ψ ∨ φ is true under V iff ψ or φ is true under V. The Strong Kleene scheme, the Leitgeb scheme, the van Fraassen scheme and the Cantini scheme are all nice. Moreover, it is easily seen that any supervaluation scheme is nice.^{5} The Weak Kleene scheme, however, is not nice, because it does not satisfy property (N1). There is a nice valuation scheme strictly lying between the Weak and Strong Kleene scheme. It is obtained by adopting the Weak Kleene rules for the existential and universal quantifier but the Strong Kleene rules for conjunction and disjunction.
Properties (N1)(N3) ensure (together with (V1)(V4)) that the formula \(\varphi ^{*}(u, y, I^{*}_{\varphi })\) is satisfied in a partial model \((\mathbb {N}, E, A)\) if and only if it is satisfied in its classical closeoff \((\mathbb {N}, E)\):
Proposition 6
Proof
Combining the above proposition with the Translation Lemma we get:
Proposition 7
In other words, the set \(I^{\alpha }_{\varphi }\) build up by the positive operator Γ _{ φ } at stage α is identical with the set weakly defined by the formula \(\varphi ^{*}(u, y, I^{*}_{\varphi })\) in the partial model obtained in the Kripke construction at stage α (starting with the empty set as extension and antiextension at stage 0).
Proof
By transfinite induction on α.
α = 0: Since E ^{0} = ∅, we get \((E^{0})_{I^{*}_{\varphi }}=\emptyset =I^{0}_{\varphi }\).
α = β+1: Let \(\langle m, n\rangle \in (E^{\beta +1})_{I^{*}_{\varphi }}\), whence by definition \(\#\varphi ^{*}(\overline m, \overline n, I^{*}_{\varphi })\in E^{\beta +1}\). Since (at successor levels) the extension of the truth predicate comprises precisely those sentences that are true at the preceding level, we get \((\mathbb {N}, E^{\beta }, A^{\beta })\models _{V} \varphi ^{*}(\overline m, \overline n, I^{*}_{\varphi })\). It follows from Proposition 6 that the formula also holds in the classical closeoff, that is \((\mathbb {N}, E^{\beta })\models \varphi ^{*}(\overline m, \overline n, I^{*}_{\varphi })\). By I.H., \((E^{\beta })_{I^{*}_{\varphi }}=I^{\beta }_{\varphi }\), so Proposition 4 yields \(\mathbb {N}\models \varphi \left (\overline m, \overline n, \overline {I^{\beta }_{\varphi }}\right )\), whence by definition \(\langle m, n\rangle \in I^{\beta +1}_{\varphi }\).
For the other direction, assume that \(\langle m, n\rangle \in I^{\beta +1}_{\varphi }\). So \(\mathbb {N}\models \varphi \left (\overline m, \overline n, \overline {I^{\beta }_{\varphi }}\right )\), whence by I.H. and Proposition 4, \((\mathbb {N}, E^{\beta })\models \varphi ^{*}(\overline m, \overline n, I^{*}_{\varphi })\). Hence, by Proposition 6, \((\mathbb {N}, E^{\beta }, A^{\beta })\models _{V}\varphi ^{*}(\overline m, \overline n, I^{*}_{\varphi })\) and therefore \(\#\varphi ^{*}(\overline m, \overline n, I^{*}_{\varphi })\in E^{\beta +1}\) by definition of Kripke jump, which means that \(\langle m, n\rangle \in (E^{\beta +1})_{I^{*}_{\varphi }}\).
□
Now we can easily see that all \({\Pi ^{1}_{1}}\)sets are weakly definable in the minimal fixed points of a nice valuation scheme.
Proposition 8
 1.
\(\mathcal {M}_{E^{\infty }}\supseteq \left \{P\mid P\ \text {is}\ {\Pi ^{1}_{1}}\right \}\)
 2.
E ^{ ∞ } is \({\Pi ^{1}_{1}}\)hard
Proof
Ad 1. If P is a \({\Pi ^{1}_{1}}\)set, then for appropriate φ we have n ∈ P iff 〈〈〉,n〉 ∈ I _{ φ } for every n. Let \(\varphi ^{\prime }:=\varphi ^{*}(\langle \rangle , x, I^{*}_{\varphi })\). It follows from the previous theorem that \(P=(E^{\infty })_{\varphi ^{\prime }}\).
Ad 2. This follows from the first item and Proposition 5. □
Note also the following:
Proposition 9
If V is nice and E ^{ ∞ } is itself \({\Pi ^{1}_{1}}\) , then E ^{ ∞ } is \({\Pi ^{1}_{1}}\) complete and \(\mathcal {M}_{E^{\infty }}=\left \{P\mid P\ \text {is}\ {\Pi ^{1}_{1}} \right \}\).
Proof
It suffices to show that \(\mathcal {M}_{E^{\infty }}\subseteq \left \{P\mid P\ \text {is}\ {\Pi ^{1}_{1}}\right \}\). Now if \((E^{\infty })_{\varphi }\in \mathcal {M}_{E^{\infty }}\), then (E ^{ ∞ })_{ φ } is elementary definable in the structure \((\mathbb {N}, E^{\infty })\) by the formula T ṣ(\(\ulcorner \varphi \urcorner \), x) and thus is \({\Pi ^{1}_{1}}\), because (by assumption) E ^{ ∞ } is \({\Pi ^{1}_{1}}\). □
Note that in order to prove the above complexity results it is sufficient to assume that (N1)(N3) hold merely for those partial models that arise in the construction of the least fixed point.
Let us say that a set X ⊆ ω is strongly definable in the partial model \((\mathbb {N}, E, A)\) iff \(X=\{n\mid (\mathbb {N}, E, A)\models _{V}\varphi (\overline n)\}\) for some Etotal formula φ. It is known for the Strong Kleene and Cantini scheme that the sets strongly definable in the least Kripke fixed point are exactly the hyperarithmetical sets. (See Cantini [5, 6].) The proposition below extends this result to all valuation schemes that are nice.

For each e,(0,e) is an Hindex.

If e is an Hindex, then (1,e) is an Hindex.

If every number in W _{ e } is an Hindex, then (2,e) is an Hindex.

If i = (0,e), then J _{ i } = W _{ e }.

If i = (1,e) and e ∈ H, then J _{ i } = (ω∖J _{ e }).

If i = (2,e) and W _{ e } ⊆ H, then \(J_{i}=\bigcup _{k\in W_{e}}J_{k}\).
Thus, the hyperarithmetical sets are obtained by starting with the recursively enumerable sets and closing them under complements and certain countable unions. Let H Y P denote the collection of hyperarithmetical sets.
Proposition 10
Proof
“ ⊆”: With every set J _{ i } we will associate an E ^{ ∞ }total \(\mathcal {L}_{\text {T}}\)formula φ _{ i } such that \(J_{i}=(E^{\infty })_{\varphi _{i}}\). Notice first that for every W _{ e } there is an \(\mathcal {L}_{\text {PA}}\)formula ψ _{ e }(x) that elementary defines W _{ e } in \(\mathbb {N}\). Now, if i = (0,e) let φ _{ i } = ψ _{ e }(x). If i = (1,e) then let φ _{ i } = ¬φ _{ e }(x). If i = (2,e) then let φ _{ i } = ∀y(¬ψ _{ e }(y)∨T ṣ(y, x)). Using the Translation Lemma and the properties of a nice valuation scheme, one proves by induction on the Hindices that every φ _{ i } is E ^{ ∞ }total and \((E^{\infty })_{\varphi _{i}}=J_{i}\). If i = (0,e) this follows from (V1). If i = (1,e) this follows from the I.H. and the definition of E ^{ ∞ }totality. If i = (2,e) this follows from the I.H. and (N1) and (V4).
“ ⊇”: Assume that φ is E ^{ ∞ }total. We have to show that (E ^{ ∞ })_{ φ } is \({\Delta ^{1}_{1}}\). Under the assumptions, Proposition 9 implies that both (E ^{ ∞ })_{ φ } and (E ^{ ∞ })_{¬φ } are \({\Pi ^{1}_{1}}\). Now by (V5), ¬φ is E ^{ ∞ }total too, and (E ^{ ∞ })_{¬φ } is the complement of (E ^{ ∞ })_{ φ }. This means that (E ^{ ∞ })_{ φ } must be \({\Sigma ^{1}_{1}}\). Thus (E ^{ ∞ })_{ φ } is both \({\Sigma ^{1}_{1}}\) and \({\Pi ^{1}_{1}}\), hence it is \({\Delta ^{1}_{1}}\). □
Notice that the only use of (V5) that we made in this paper is in the ⊇direction of the proof of Proposition 10. All previous results (and most results in the next section) can be proved without assuming (V5).
At the beginning of this Section 1 briefly mentioned that it is also possible to show that every \({\Pi ^{1}_{1}}\)set is weakly definable in the revision theory of truth. (Of course, it is wellknown that the revision theory defines much more sets than that. See Welch [39].) In fact, the proof is even simpler than for Kripke’s theory. Let us briefly consider the case of Herzberger’s theory in [16].
Proposition 11
Let Γ _{ φ } be an elementary positive operator given by the formula φ(x _{1} ,…,x _{ n } ,X ^{ n } ). Let \(I^{*}_{\varphi }=\ulcorner {\varphi ^{*}(\textbf {x}, I^{*}_{\varphi })}\urcorner \) . Then \((R_{\alpha })_{I^{*}_{\varphi }}=I^{\alpha }_{\varphi }\) for every α ≤ κ (where κ is the closure ordinal of the operator Γ _{ φ } ).
Notice that Proposition 11 is more general than Proposition 7: here, φ can be any elementary X ^{ n }positive formula, in contrast to Proposition 7 which only applies to formulae of the form Eq. 7. In the next section, we will see that this strengthening also holds for Kripkestyle theories that satisfy stronger requirements than (N1)(N3).
Proof
Finally, for limits γ, if \(\textbf {m}\in (R_{\gamma })_{I^{*}_{\varphi }}\) then by definition there must be a β such that \(\varphi ^{*}(\textbf {m}, I^{*}_{\varphi })\) is true in all models \((\mathbb {N}, R^{\alpha })\) where β ≤ α < γ. Thus, by I.H. and Proposition 4, \(\mathbb {N}\models \varphi (\overline {\mathbf {m}}, I_{\varphi }^{\alpha })\) for β ≤ α < γ, which implies \(\textbf {m}\in I^{\gamma }_{\varphi }\).
Conversely, if \(\textbf {m}\in I^{\gamma }_{\varphi }\) then there is some β such that for all α with β ≤ α < γ, \(\textbf {m}\in I^{\alpha }_{\varphi }\). By I.H. and Proposition 4 we get \(\textbf {m}\in (R_{\alpha })_{I^{*}_{\varphi }}\) for all α with β ≤ α < γ and thus \(\textbf {m}\in (R_{\gamma })_{I^{*}_{\varphi }}\). Therefore \((R_{\gamma })_{I^{*}_{\varphi }}=I^{\gamma }_{\varphi }\). □
4 Axioms
The correspondence between truthsets and secondorder models that we introduced in Section 2 seems to be a good way to measure the amount of secondorder quantification that a semantic theory of truth is able to mimick. The theorems of the previous section indicate, I believe, a certain lower bound on the prooftheoretic strength that we should expect from a good axiomatization of the minimal Kripke fixed points. For example, a semantic theory that encodes all inductive sets should be axiomatized by a theory that formalizes or interprets the theory of inductive definitions, I D _{1}. A similar idea is suggested in Fischer et al. [8, section 3.2] .
The language \(\mathcal {L}_{\text {ID}_{1}}\) extends the language \(\mathcal {L}_{\text {PA}}\) by a predicate constant \(\overline I_{\varphi }\) for every arithmetical \(\mathcal {L}_{2}\)formula φ(x, X) (with all free variables displayed) in which X occurs only positively. On the intended interpretation, the constant \(\overline {I_{\varphi }}\) is interpreted by the least fixed point I _{ φ } of the operator Γ _{ φ }.
Since every fixed point I _{ φ } of an elementary positive operator Γ _{ φ } is \({\Pi ^{1}_{1}}\), we can find (via Proposition 8) a formula \(\varphi ^{\prime }\in \mathcal {L}_{T}\) such that \(I_{\varphi }=(E^{\infty })_{\varphi ^{\prime }}\) for every nice valuation V. (Note, however, that while φ is an arbitrary Xpositive formula, φ ^{′} will be of the form Eq. 7.) In order to translate I D _{1} into \(\mathcal {L}_{\text {T}}\), we may stipulate that \(\overline {I_{\varphi }}\) is translated as \(\ulcorner {\varphi ^{\prime }}\urcorner \) and \(\overline {I_{\varphi }}(x)\) as \(T\oalign {{\textit {s}}\cr \hfil .\hfil }(\ulcorner {\varphi ^{\prime }}\urcorner , x)\). Call the resulting translation ^{ # }. (On the arithmetical vocabulary, ^{ # } is just the identity function.) As I will point out below, the translation ^{ # } has a certain disadvantage, but it is enough for the following result, which is a corollary of Proposition 8:
Proposition 12
If V is nice then \((\mathbb {N}, E^{\infty })\models (\mathsf {ID}_{1})^{\#}\)
For example, if V is the Strong Kleene scheme, then V satisfies the antecedent of the above proposition. The KripkeFeferman theory K F (see Feferman [7]) can be seen as an axiomatization of the Strong Kleene fixed points, but it doesn’t count as a satisfactory axiomatization of the minimal fixed point if one adheres to the prooftheoretic criterion suggested at the beginning of this section. Burgess has given a variant of K F, called K F B (the acronym stands for ‘KripkeFefermanBurgess’) that does have the same strength as I D _{1}. The additional strength over K F is obtained by adding a minimality axiom scheme which basically says that if φ(x) satisfies the K F axioms, then all true sentences fall under the extension of φ(x) (cf. Halbach [15, chap. 17] for details). Another way to strengthen K F, which has already been pointed out by Cantini [5, p. 105] , is to add the (translation of the) second axiom scheme of I D _{1} to K F. Proposition 12 above shows that the resulting system is still sound with respect to the minimal Strong Kleene fixed point.
The theory P U T B (the acronym stands for positive uniform Tbiconditionals) is a subtheory of K F and has been investigated by Cantini [5] and Halbach [14]. In the following paragraphs, we prove some simple facts about models of P U T B. In Section 6 we return to this theory, where we will investigate the prooftheoretic strength of positive disquotation without the arithmetic base theory.
Definition 4
The theory P U T B interprets the theory \(\widehat {\mathsf {ID}}_{1}\), but the translation ^{ # } introduced above won’t do. Under the translation ^{ # }, a fixed point I _{ φ } gets mapped to a formula φ ^{′} with the appropriate extension, but the formula φ ^{′} does not ‘preserve’ the syntactic shape of the formula φ. (To repeat, the formula φ ^{′} is an instance of Eq. 7, while φ can be an arbitrary Xpositive formula.) In order to get a formula with the right syntactic properties, we apply Cantini’s trick again. Thus, where φ(x, X) is a Xpositive formula, let \(I^{*}_{\varphi }\) be a closed term of \(\mathcal {L}_{\text {T}}\) such that \(I^{*}_{\varphi }=\ulcorner {\varphi ^{*}(x, I^{*}_{\varphi })}\urcorner \). Now expand the translation ^{∗} (of Section 2) by translating \(\overline {I_{\varphi }}\) as \(I^{*}_{\varphi }\).
Proposition 13 (Cantini [5])
\(\mathsf {PUTB}\vdash (\widehat {\mathsf {ID}}_{1})^{*}\)
Proof
Notice that the proof doesn’t work if we use the translation ^{ # } instead of ^{∗}, because \(\ulcorner {\varphi ^{\prime }}\urcorner \neq \ulcorner {\varphi ^{*}(x, \ulcorner {\varphi ^{\prime }}\urcorner )}\urcorner \).
We have already seen that the closeoffs of minimal fixed points of nice valuation schemes satisfy (I D _{1})^{ # } and therefore \((\widehat {\mathsf {ID}}_{1})^{\#}\). But do all of them also satisfy P U T B and therefore \((\widehat {\mathsf {ID}}_{1})^{*}\)? If V is strong in the sense of the definition given below, the answer is ‘yes’. In that case, the answer applies not only to the minimal fixed points but to arbitrary fixed points as well. Moreover, if V is strong in the sense of the following definition then the minimal fixed point also satisfies (I D _{1})^{∗}.
In order to state the definition, we need to fix some terminology. A formula of \(\mathcal {L}_{T}\) is in negation normal form if it is build up from atomic and negated atomic formulas using ∧,∨,∀ and ∃ only, without further use of ¬. For every formula \(\varphi \in \mathcal {L}_{T}\), let φ ^{nf} be a unique formula in negation normal form that is logically equivalent (in classical logic) to φ. To fix things, we let φ ^{nf} be the unique formula that is obtained from φ by applying the transformation rules in [2, chap. 19] .
Definition 5
 (S1)
if φ is true under V, then φ ∨ ψ and ψ ∨ φ are true under V
 (S2)
if for some t, φ(t) is true under V, then ∃x φ is true under V
 (S3)
if \((\mathbb {N}, E, A)\models _{V}\varphi \) then \((\mathbb {N}, E)\models \varphi \)
 (S4)
\((\mathbb {N}, E, A)\models _{V}\varphi \) iff \((\mathbb {N}, E, A)\models _{V}\varphi ^{\text {nf}}\)
Note that (S1) implies (N1) and (N2). Condition (S3) expresses that V is classically sound: if a sentence is true in a partial model, then it is true in its classical closeoff. The Strong Kleene, the van Fraassen and the Cantini scheme are strong. The Weak Kleene and the Leitgeb scheme are not strong, however, because both violate condition (S1).^{6} In Beringer and Schindler [1] it is shown that the Cantini scheme is the strongest valuation scheme that is classically sound. Thus, no supervaluational scheme stronger than Cantini’s satisfies condition (S3). Properties (S1)(S4) enable us to extend Proposition 6 to every Tpositive formula:
Proposition 14
Proof
The lefttoright direction follows from the classical soundness of V, (S3). The righttoleft direction is proved by induction on the buildup of the Tpositive formula φ. We assume that φ is in negation normal form. This is justified by (S4). If φ is of the form s = t, s≠t, or T t, this is trivial. Since ¬T t is not Tpositive, we can ignore it. If φ is a disjunction, the claim follows from the induction hypothesis and property (S1) of a strong valuation. Similarly, if φ is a conjunction, a universal or existential statement, the claim follows from the induction hypotheses and properties (V3), (V4) and (S2), respectively. □
Proposition 15
If V is strong and \((\mathbb {N}, E, A)\) is any fixed point of \(\mathcal J_{V}\) then \((\mathbb {N}, E)\models \mathsf {PUTB}\).
Proof
Let φ be a Tpositive sentence and assume that \((\mathbb {N}, E)\models \varphi \). Then Proposition 14 shows that \((\mathbb {N}, E, A)\models _{V}\varphi \), which implies by the fixed point property that \((\mathbb {N}, E, A)\models _{V} T\ulcorner \varphi \urcorner \). This means # φ ∈ E, so also \((\mathbb {N}, E)\models T\ulcorner \varphi \urcorner \). Now assume that \((\mathbb {N}, E)\models \neg \varphi \). Since V is classically sound (S4), φ has value 0 or u in the partial model \((\mathbb {N}, E, A)\) and consequently, # φ∉E. So \((\mathbb {N}, E)\models \neg T\ulcorner \varphi \urcorner \). So \((\mathbb {N}, E)\models \varphi \leftrightarrow T\ulcorner \varphi \urcorner \) for all Tpositive φ, whence by standardness \((\mathbb {N}, E)\models \mathsf {PUTB}\). □
I do not know the answer to the following:
Question 1
If \((\mathbb {N}, E, A)\) is the minimal (any) fixed point of the Leitgeb valuation scheme, do we have \((\mathbb {N}, E)\models \mathsf {PUTB}\)?
A further consequence of Proposition 14 is the following:
Proposition 16
Suppose that V is strong. Let φ(x _{1} ,…,x _{ n } ,X ^{ n } ) be an arithmetical \(\mathcal {L}_{2}\) formula (with all free variables displayed) in which X ^{ n } occurs only positively, and \(I^{*}\varphi =\ulcorner {\varphi ^{*}(\textbf {x}, I^{*}_{\varphi })}\urcorner \) . Then \((E^{\alpha })_{I^{*}_{\varphi }}=I^{\alpha }_{\varphi }\) for all α∈ON.
The difference to Proposition 7 is that here φ can be an arbitrary arithmetical X ^{ n }positive formula while Proposition 7 only applied to formulae of the form Eq. 7.
Proof
If α is a limit ordinal, the claim follows easily from the I.H. □
As a corollary to Proposition 16, we get:
Proposition 17
If V is strong then \((\mathbb {N}, E^{\infty })\models (\mathsf {ID}_{1})^{*}\)
In the previous section, we have seen that the sets that are strongly definable in the least fixed points of nice valuation schemes are exactly the hyperarithmetical sets. A little modification of the Translation Lemma shows that the system \({\Delta ^{1}_{1}}\mathsf {CA}_{0}\) is true in the classical closeoffs of those fixed points.
The minimal ωmodel of the system \({\Delta ^{1}_{1}}\text {}\mathsf {CA}_{0}\) is the structure \((\mathbb {N}, \mathit {HYP})\). (For details, I refer the reader to Simpson [33].)
Let \(\mathcal {M}_{S}^{tot}=\{S_{\varphi }\mid \varphi \in \mathcal {L}_{T}, \varphi \ \text {is} \ S\text {total}\}\). The following proposition states that whenever φ is a secondorder formula, then its translation under ^{∗∗} is true in S iff the original sentence is true in \(\mathcal {M}_{S}^{tot}\).
Proposition 18
Proof
Similar to the proof of Proposition 2. □
Proposition 19
Proof
Since \(\mathcal {M}^{tot}_{E^{\infty }}= \mathit {HYP}\) by Proposition 10 and \((\mathbb {N}, \mathit {HYP})\models {\Delta ^{1}_{1}}\mathsf {CA}_{0}\), the claim follows from Proposition 18. □ □
5 Reducing Comprehension to Disquotation
Unfortunately, this will lead to inconsistency. The reason is that we can instantiate the quantifier on ∀y to any \(\mathcal {L}_{\text {T}}\)formula, in particular to formulae that are not in the range of the translation function ^{∗}, such as the liar sentence.
Proposition 20
The schema (8) is inconsistent with P A.
Proof
There are at least two ways out of the problem. First, one could restrict the permissible instances of Eq. 8 to formulae that do not contain free set variables. The resulting system is ωconsistent and interprets \(\mathsf {Z}_{2}^{}\), that is secondorder arithmetic with comprehension for all \(\mathcal {L}_{2}\)formulae that do not contain free set variables. This was shown in Schindler [30]. The second option is to keep the parameters but restrict the quantifers in Eq. 8 to a suitable class of formulae. The obvious choice is to restrict the quantifiers to formulae that lie in the range of the translation function; thus, we may instantiate the quantifier on y above to the term \(\ulcorner \neg T\) ṣ(y, x)\(\urcorner \), for example, but not to \(\ulcorner \neg T\) ṣ(z, z)\(\urcorner \).
We will first show how this works for the case of arithmetical \(\mathcal {L}_{2}\)formulae because the general case involves some subtleties that obscure the main idea for the consistency proof. At the end of this section, we will sketch how to extend the method to formulae of arbitrary complexity. It should be possible to extend it to third and higherorder systems as well. Since we will only deal with classical theories in what follows, we assume for simplicity that the symbols ∨,∃ are no longer among the primitive vocabulary but defined.
We first expand the language of \(\mathcal {L}_{2}\) by adding abstraction terms. This is necessary to close the range of the translation function under instantiations of quantified formulae (see below). Let \({\mathcal {L}_{2}^{0}}\) be the result of adding to \(\mathcal {L}_{2}\) a term {x∣φ} for every \(\mathcal {L}_{\text {PA}}\)formula φ containing only the number variable x free. Let \(\mathcal {L}_{2}^{n+1}\) be the result of adding to \({\mathcal {L}_{2}^{n}}\) a term {x∣φ} for every arithmetical \(\mathcal {L}_{2}\)formula φ containing only the number variable x free. Thus, while φ must not contain any set variables at all, it may contain expressions of the form t ∈ {x∣ψ} as subformulae (where ψ is an arithmetical formula of \({\mathcal {L}_{2}^{n}}\)). Finally, let \(\mathcal {L}_{2}^{+}\) be the union of these languages.
Proposition 21
We abbreviate the formula Open image in new window —which expresses that x is in the range of the translation function τ—by T r s l(x).
The reason for adding the abstraction terms to the language \(\mathcal {L}_{2}\) is to ensure that the predicate T r s l(x) is closed under “instantiation” in the following sense: if φ(x _{2i+1}) is in the range of τ and \(t=\ulcorner {\tau (\psi )}\urcorner \) for some ψ in the domain of τ, then φ(t/x _{2i+1}) is also in the range of τ. Note that this is provable in P A (by “internal” induction on the complexity of φ). This is crucial for the proof of Proposition 22 below.
Definition 6
It is well known that A C A, which is nonconservative over P A, is interpretable in the compositional theory C T but not in the disquotational theory T B, which is a conservative extension of P A (for a proof, see Halbach [15]). The following proposition shows that A C A is interpretable in the disquotational theory U T B(A C A). While the latter (presumably) does not prove any (nontrivial) truththeoretic generalizations for its own truth predicate, it is able (via the interpretability result) to define a truth predicate for P A for which the compositional clauses are derivable.
Proposition 22
A C A is relatively interpretable in U T B(A C A).
Proof
This is the translation of the comprehension axiom for φ. □
In order to show the consistency of U T B(A C A), we will define a set S ⊆ ω such that \(\mathcal {M}_{S}\) is the collection A R I T H of arithmetically definable sets. Since \((\mathbb {N}, \mathit {ARITH})\) is a model of A C A, a variant of the Translation Lemma will then show that \((\mathbb {N}, S)\) is a model of U T B(A C A).
Let the term {x∣φ} denote the set \(\{n\mid (\mathbb {N}, \mathit {ARITH})\models \varphi (\overline n)\}\in \mathit {ARITH}\). This is welldefined, since A R I T H is closed under arithmetical definability.
 1.
\(S=\{\#\tau (\varphi )\mid \varphi \in \mathcal {L}_{2}^{+}, (\mathbb {N}, \mathit {ARITH})\models \varphi \}\)
 2.
\(S_{\tau (\varphi )}=\{n\mid \#\tau (\varphi (\overline n))\in S\}\)
 3.
\(\mathcal {M}_{S}=\left \{S_{\tau (\varphi )}\mid \varphi \in \mathcal {L}_{2}^{+}\right \}\)
It is not very hard to see that \(\mathcal {M}_{S}=\mathit {ARITH}\). By induction on the buildup of φ we prove:
Proposition 23
For every \(\varphi \in \mathcal {L}_{2}^{+}\) : \((\mathbb {N}, \mathit {ARITH}) \models \varphi \Leftrightarrow (\mathbb {N}, S) \models \tau (\varphi ) \)
Proposition 24
The theory U T B(A C A) has an ωmodel.
Proof
Now let us see how this method can be extended to \(\mathcal {L}_{2}\)formulae of arbitrary complexity. In order to prove the consistency of the resulting system, we will have to define a set S ⊆ ω such \((\mathbb {N}, \mathcal {M}_{S})\) is a model of secondorder arithmetic. Pick a countable set \(\mathcal A\subseteq \wp (\omega )\) such that \((\mathbb {N}, \mathcal {A})\) is a model of Z _{2}. Such models exist; we refer the interested reader to Putnam et al. [28]. That \((\mathbb {N}, \mathcal A)\) is a model of Z _{2} implies that \(\mathcal {A}\) is closed under secondorder definabilty with parameters from \(\mathcal A\).
Since our goal is to construct S such that \(\mathcal {M}_{S}=\mathcal {A}\), every set in \(\mathcal {A}\) must be coded by a formula (with one free variable) of \(\mathcal {L}_{\text {T}}\). In contrast to the arithmetical case, it is not clear that every member of \(\mathcal A\) can be defined by a formula of \(\mathcal {L}_{2}\). Thus we need to expand \(\mathcal {L}_{2}\) by countably many set constants P _{1}, P _{2},…; then we inductively add an abstraction term {x∣φ} for every formula of the expanded language that contains exactly one free number variable. Call the resulting language \(\mathcal {L}_{2}^{\dagger }\).
Now we define a translation τ ^{′} from \(\mathcal {L}_{2}^{\dagger }\) to \(\mathcal {L}_{\text {T}}\) as in Proposition 21. In order to translate the set constants, we pick for every i a term \(P^{*}_{i}\) with \(P^{*}_{i}=\ulcorner T\) ṣ(\(P_{i}^{*}\), v _{2i })\(\urcorner \) (via diagonalization). The translation function τ ^{′} is defined exactly like τ with the proviso that \(\tau ^{\prime }(P_{i})=P_{i}^{*}\) and τ ^{′}(t ∈ P _{ i }) = T ṣ(\(P_{i}^{*}, \tau ^{\prime }(t))\).
Definition 7
Similarly to Proposition 22, we can prove:
Proposition 25
Z _{2} is relatively interpretable in U T B(Z _{2}).
The inclusion of the terms \(P^{*}_{i}\) into the range of the translation function is entirely dispensable for the interpretability result, but they play a crucial role in the consistency proof.
Let us now indicate how to modify the proof of Proposition 24 to obtain a consistency proof for U T B(Z _{2}). Let e be a bijection between {P _{1}, P _{2},…} and \(\mathcal {A}\). Such a function exists since \(\mathcal {A}\) is countable. We let the constant P _{ i } denote the set e(P _{ i }). Moreover, we let the term {x∣φ} denote the set \(\{n\mid (\mathbb {N}, \mathcal {A})\models \varphi (\overline n)\}\). This is welldefined, since \(\mathcal {A}\) is closed under definability.
As above, we let S consist of the set of sentences that are translations of \(\mathcal {L}_{2}^{\dagger }\)sentences that hold in \((\mathbb {N}, \mathcal A)\).
Obviously, \(\mathcal {M}_{S}\subseteq \mathcal {A}\). Moreover, since every set in \(\mathcal A\) is definable by a formula of the form x ∈ P _{ i }, we can find an \(\mathcal {L}_{\text {T}}\)formula φ (namely, the translation of x ∈ P _{ i } under τ ^{′}) such that S _{ φ } = e(P _{ i }). This implies \(\mathcal {A}\subseteq \mathcal {M}_{S}\), and therefore \(\mathcal {A}=\mathcal {M}_{S}\). A variant of the Translation Lemma yields the desired consistency proof.
Proposition 26
The theory U T B(Z _{2}) has an ωmodel.
While U T B(Z _{2}) is by no means a very elegant theory, it shows that principles of truth need not be any weaker than our strongest arithmetical theories. The expressive power gained by secondorder quantification can be recovered in theories of truth. There is, however, a certain mismatch. While secondorder languages allows us to generalize over any predicate of the secondorder language, the truth predicate of U T B(Z _{2}) does not allow us to generalize over all predicates of the language of truth, but only over those predicates that can be translated back into the secondorder formalism.
6 Truth Without a Base Theory
In the last section, we saw how to derive comprehension axioms from uniform Tbiconditionals. The derivation made use of some of the axioms of Peano arithmetic. The use of those arithmetical axioms was essential, because we needed to show that the formula in question is in the range of the translation function. This, in turn, is crucial because, under our translation, we relativized the set quantifiers to formulae of a particular shape. If, on the other hand, we do not relativize the set quantifiers to a predicate when setting up the translation, no arithmetical axioms are needed for deriving comprehension axioms from uniform Tbiconditionals.
The fact that uniform disquotation sentences behave like instances of naïve comprehension has some interesting consequences, as we will show next. First, unlike the ordinary (i.e., nonuniform) Tschema, the uniform Tschema leads to inconsistency over logic alone. As Gupta [9] has shown, a theory can be classical, contain names for its own expressions, obey the ordinary Tschema, and be consistent at the same time. Only when in addition selfreferential liarlike statements can be formulated do we get inconsistency. In stark contrast, the uniform Tschema can lead to inconsistency even in the absence of selfreference. Second, restricted (and consistent) instances of the uniform Tschema, taken over logic alone, commit us to the existence of infinitely many objects and suffice to interpret certain amounts of arithmetic, so that the set of their consequences is essentially undecidable. In particular, typed uniform disquotation interprets R while uniform positive disquotation interprets Q. It goes without saying that the existence of infinitely many objects does not follow from the mere presence of infinitely many names in the language.
Proposition 27
The scheme \(\forall x(T\ulcorner \!\varphi (\dot {x})\!\urcorner \leftrightarrow \varphi (x))\) is inconsistent over classical firstorder logic.
Proof
By a simple application of Russell’s paradox. The uniform Tbiconditional for the formula ¬T ṣ(x, x) delivers, via the ‘translation’ provided at the beginning of this section, that ∃y∀x(x ∈ y⇔x∉x), which is logically inconsistent. □ □
Next, we show that the typed theory of uniform disquotation, taken over logic alone, interprets the arithmetical theory R introduced in Tarski et al. [35], which is essentially undecidable and has only infinite models. (Roughly, R is Q formulated with numerals. See also Monk [22, chap. 14] .)
Proposition 28
U TB^{−} relatively interprets R.
Proof
Proposition 29
P U TB^{−} relatively interprets Q.
Proof
The results in this section show that a good amount of arithmetic is already encoded into the uniform Tbiconditionals. As mentioned above, the use of the primitive function symbols ṣ \(_{n}^{n}\) (for n≥1) in the formulation of the uniform Tschema plays an essential role (because we need to ensure totality and functionality), although the fact that there are infinitely many is of no relevance. Instead of the infinitely many symbols ṣ \(_{n}^{n}\) we could use a single binary function symbol ḥ in the formulation of the uniform Tschema. (On the intended interpretation, ḥ denotes the function h such that \(h(m,n)=\#\varphi (\overline n/x_{j})\) if m = # φ codes a formula with some free variables and x _{ j } is the variable with the highest index in φ, and h(m, n) = 0 otherwise.)
7 Conclusions
In this paper we considered translations that map formulae of the language of secondorder arithmetic to formulae of the language of truth. From a technical point of view, such translations are useful because they allow us to transform questions about truth into questions about secondorder arithmetic and to exploit the extensive literature on the latter to answer questions about the former. We have seen several such applications in this paper. From a philosophical point of view, such translations indicate, I believe, that there is a close connection between truth and satisfaction on the one hand and secondorder quantification on the other hand. The results in this paper, and other similar results in the literature, support Parsons’ claim, mentioned in the introduction, that truth and satisfaction are a means to quantify into sentence respectively predicate position.
There are at least two directions into which the study of translations between secondorder arithmetic and theories of truth ought to be further developed. First, our attention in this paper was restricted to ωmodels and it might be interesting to generalize some of the results to nonstandard models. Second, our attention in this paper was restricted to translations from secondorder arithmetic into the language of truth. Of course, the reverse is also possible. It is well known that theories of truth can be interpreted in subsystems of secondorder arithmetic too. As an example (see [11]), one may take the interpretation of typed compositional theories of truth (e.g. the systems of ramified truth R T _{ α }) into systems of predicative comprehension (e.g. the systems of ramified analysis R A _{ α }). This raises the question of the composition of such interpretations. In [24], Nicolai has shown that the mutual embeddings between typed truth and predicative comprehension cannot be lifted to stricter notions of equivalence such as biinterpretability. For example, he shows that while for each n ∈ ω, R T _{ n } is an eretract of R A _{ n }, the converse does not hold. Further interesting work awaits us there.
Footnotes
 1.
 2.
We briefly sketch Leitgeb’s theory for those readers not familiar with it. A sentence φ depends on S ⊆ ω iff for all P ⊆ ω we have \((\mathbb {N}, P)\models \varphi \Leftrightarrow (\mathbb {N}, S\cap P)\models \varphi \). Let G _{0} = ∅ and let G _{ α+1} be the set of (codes of) sentences that depend on G _{ α }. If γ is a limit ordinal, let G _{ γ } be the union of all G _{ α } for α < γ. Next, let T _{0} = ∅ and let T _{ α+1} be the set of (codes of) sentences that are elements of G _{ α } and are true in the classical model \((\mathbb {N}, T_{\alpha })\). At limit stages, we take unions again. This hierarchy reaches a fixed point, T ^{ ∞ }, which Leitgeb proposes as interpretation for the truth predicate. The valuation scheme V _{ L } is defined as follows. Let \(V_{L}(\mathbb {N}, E, A)(\varphi )=1\) if φ depends on E∪A and \((\mathbb {N}, E)\models \varphi \), and =0 if φ depends on E∪A and \((\mathbb {N}, E)\nvDash \varphi \), and =u otherwise. Then E ^{ ∞ } = T ^{ ∞ }.
 3.Let and Open image in new window and Open image in new window Observe that
 4.
For additional results on the Weak Kleene scheme see Speranski [34], where it is shown that adding a symbol for proper substraction solves the problem.
 5.
Where a scheme is supervaluational iff it is given by a rule of the following form: \((\mathbb {N}, E, A)\models _{V}\varphi \) iff for all S⊇E, if Φ(S, E, A) then \((\mathbb {N}, S)\models \varphi \), where Φ is a condition such that for every partial model \((\mathbb {N}, E, A)\) there is an S⊇E with Φ(S, E, A).
 6.
This claim is obvious for the Weak Kleene scheme. For the Leitgeb scheme, a counterexample to (S1) can be found in Schindler [29].
 7.
Notice that in formulating the Tbiconditionals, we have also assumed a Gödelcoding in the metalanguage in order to pick a name for each formula. However, for the derivation (and the results below) it is entirely irrelevant which name is actually chosen, and we could simply pick names at random, with the proviso that no two formulae get the same name assigned.
Notes
Acknowledgments
For their help in preparing this paper, I want to thank Timo Beringer, Catrin CampbellMoore, Volker Halbach, Carlo Nicolai, Stanislav Speranski, and two anonymous referees, one of which spotted a flaw in an old proof of Proposition 14. A special thanks goes to Lavinia Picollo, in particular for her help with Proposition 21 .
References
 1.Beringer, T., & Schindler, T. (2017). A graphtheoretic analysis of the semantic paradoxes. Under review.Google Scholar
 2.Boolos, G., Burgess, J., & Jeffrey, R. (2003). Computability and logic, 4th edn. Cambridge: Cambridge University Press.Google Scholar
 3.Burgess, J.P. (1986). The truth is never simple. Journal of Symbolic Logic, 51, 663–681.CrossRefGoogle Scholar
 4.Cain, J., & Damnjanovic, Z. (1991). On the Weak Kleene scheme in Kripke’s theory of truth. Journal of Symbolic Logic, 56, 1452–1468.CrossRefGoogle Scholar
 5.Cantini, A. (1989). Notes on formal theories of truth. Zeitschr. f. Math. Logik und Grundlagen d Math, 35, 97–130.CrossRefGoogle Scholar
 6.Cantini, A. (1990). A theory of formal truth arithmetically equivalent to ID _{1}. Journal of Symbolic Logic, 55, 244–259.CrossRefGoogle Scholar
 7.Feferman, S. (1991). Reflecting on incompleteness. Journal of Symbolic Logic, 56, 1–49.CrossRefGoogle Scholar
 8.Fischer, M., Halbach, V., Kriener, J., & Stern, J. (2015). Axiomatizing semantic theories of truth? Review of Symbolic Logic, 8, 257–278.CrossRefGoogle Scholar
 9.Gupta, A. (1982). Truth and paradox. Journal of Philosphical Logic, 11, 1–60.CrossRefGoogle Scholar
 10.Halbach, V. (1995). Tarski hierarchies. Erkenntnis, 43, 339–367.CrossRefGoogle Scholar
 11.Halbach, V. (1996). Axiomatische Wahrheitstheorien Logica Nova. Berlin: Akademie Verlag.Google Scholar
 12.Halbach, V. (2000). Truth and reduction. Erkenntnis, 53, 97–126.CrossRefGoogle Scholar
 13.Halbach, V. (2001). How innocent is deflationism? Synthese, 126, 167–194.CrossRefGoogle Scholar
 14.Halbach, V. (2009). Reducing compositional to disquotational truth. Review of Symbolic Logic, 2, 786–798.CrossRefGoogle Scholar
 15.Halbach, V. (2011). Axiomatic theories of truth. Cambridge: Cambridge University Press.CrossRefGoogle Scholar
 16.Herzberger, H. (1982). Notes on naive semantics. Journal of Philosphical Logic, 11, 61–102.CrossRefGoogle Scholar
 17.Horsten, L. (1995). The semantical paradoxes, the neutrality of truth and the neutrality of the minimalist theory of truth. In P. Cortois (Ed.), The many problems of realism, vol. 3 of Studies in the general philosophy of science (pp. 173–187). Tilburg University Press.Google Scholar
 18.Horwich, P. (1998). Truth, 2nd edn. Oxford: Basil Blackwell.CrossRefGoogle Scholar
 19.Ketland, J. (1999). Deflationism and Tarski’s paradise. Mind, 108, 69–94.CrossRefGoogle Scholar
 20.Kripke, S. (1975). Outline of a theory of truth. Journal of Philosophy, 72, 690–716.CrossRefGoogle Scholar
 21.Leitgeb, H. (2005). What truth depends on. Journal of Philosphical Logic, 34, 155–192.CrossRefGoogle Scholar
 22.Monk, D.J. (1976). Mathematical logic. New York: Springer.CrossRefGoogle Scholar
 23.Moschovakis, Y.N. (1974). Elementary induction on abstract structures. Dover Publications.Google Scholar
 24.Nicolai, C. (2016). Equivalences for truth predicates. Review of Symbolic Logic. forthcoming.Google Scholar
 25.Parsons, C. (1983). Ontology and mathematics. In Mathematics in philosophy (pp. 37–62). Cornell University Press.Google Scholar
 26.Parsons, C. (1983). Sets and classes. In Mathematics in philosophy, (pp. 209–220). Cornell University Press.Google Scholar
 27.Pohlers, W. (2009). Proof theory the first step into impredicativity. Berlin Heidelberg: Springer.Google Scholar
 28.Putnam, H., Boyd, R., & Hensel, G. (1969). A recursion theoretic characterization of the ramified analytical hierarchy. Transactions of the American Mathematical Society, 141(/142), 37–62.Google Scholar
 29.Schindler, T. (2014). Axioms for grounded truth. Review of Symbolic Logic, 7, 73–83.CrossRefGoogle Scholar
 30.Schindler, T. (2015). A disquotational theory of truth as strong as Z\(^{}_{2}\). Journal of Philosophical Logic, 44, 395–410.CrossRefGoogle Scholar
 31.Shapiro, S. (1998). Proof and truth: Through thick and thin. Journal of Philosophy, 95, 493–521.Google Scholar
 32.Shoenfield, J.R. (1967). Mathematical logic. AddisonWesley.Google Scholar
 33.Simpson, S.G. (2009). Subsystems of second order arithmetic, 2nd edn. Cambridge: Cambridge University Press.CrossRefGoogle Scholar
 34.Speranski, S.O. (2017). Notes on the computational aspects of Kripke’s theory of truth. Studia Logica, 105, 407–429. https://doi.org/10.1007/s1122501696948.
 35.Tarski, A., Mostowski, A., & Robinson, R. (1953). Undecidable theories. North Holland, Amsterdam.Google Scholar
 36.Visser, A. (1997). An overview of interpretability logic. In Advances in modal logic, (pp. 307–359). Stanford: CSLI Publications.Google Scholar
 37.Visser, A. (2008). Pairs, sets and sequences in firstorder theories. Archive for Mathematical Logic, 47, 299–326.CrossRefGoogle Scholar
 38.Visser, A. (2009). Cardinal arithmetic in the style of Baron von Münchhausen. Review of Symbolic Logic, 2, 570–589.CrossRefGoogle Scholar
 39.Welch, P. (2001). On GuptaBelnap revision theories of truth, Kripkean fixed points, and the next stable set. Bulletin of Symbolic Logic, 7, 345–360.CrossRefGoogle Scholar
 40.Welch, P. (2015). The complexity of the dependence operator. Journal of Philosophical Logic, 44, 337–440.CrossRefGoogle Scholar
Copyright information
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.