---
title: Lattice of Theories (LOT)
url: https://www.emergentmind.com/topics/lattice-of-theories-lot
type: topic
---

# Lattice of Theories (LOT)

Searching arXiv for the cited LOT papers to ground the article in the listed literature.
arxiv_search.query({"search_query":"id:2004.00989 OR id:1811.02041 OR id:1811.01144 OR id:2511.17374 OR id:2601.20759 OR id:1905.11748","max_results":10,"sort_by":"submittedDate","sort_order":"descending"})
arXiv search returned the following relevant records:
- 2601.20759 — "The Latent Space of Equational Theories"
- 2511.17374 — "Characterizing Sets of Theories That Can Be Disjointly Combined"
- 2004.00989 — "Lattices of Intermediate Theories via Ruitenburg's Theorem"
- 1905.11748 — "Modelling competing theories"
- 1811.02041 — "Conceptua: Institutions in a Topos"
- 1811.01144 — "The IFF Approach to the Lattice of Theories"
Lattice-of-Theories (LOT) denotes a family of order-theoretic constructions in which theories are organized by entailment, model classes, or closure under a semantic relation. In current arXiv literature, LOT is not a single invariant object but a recurrent pattern: it appears as the concept lattice of theories obtained from Tarski satisfaction, as a lattice of $\chi$-variants of intermediate logics generated by fixed-point axioms, as a complete lattice of theory-combination properties for decidable first-order theories, as a lattice of axiomatic and empirical theories in many-valued modal semantics, and as an empirical geometric approximation to the classical equational LOT [1811.02041] [1811.01144] [2004.00989] [2511.17374] [1905.11748] [2601.20759].

## 1. Order-theoretic core

At its most general, a LOT is a partially ordered collection of theories equipped with meet and join operations. What varies across frameworks is the choice of elements and the direction of order. In the institution/FCA tradition, the elements are closed intents or formal concepts, and the order on theories is reverse inclusion of intents, equivalently inclusion of extents. In the intermediate-logic setting, the elements are intermediate logics or their $\chi$-variants ordered by inclusion of theories. In the disjoint-combination setting, the elements are closed sets of decidable theories ordered by inclusion. In the equational and lattice-based modal settings, theories are again ordered by entailment or inclusion of axioms, with lattice operations defined by closure under consequence [1811.02041] [1811.01144] [2004.00989] [2511.17374] [1905.11748] [2601.20759].

The resulting meet and join operations are correspondingly presentation-dependent. In the truth-concept-lattice formulation, join on intents is intersection of theories and meet is semantic closure of their union. In the classical equational LOT, by contrast, $T_1 \wedge T_2 = Cn(T_1) \cap Cn(T_2)$ and $T_1 \vee T_2 = Cn(T_1 \cup T_2)$. In the lattice of closed theory-combination properties, $\bigwedge_i X_i = \bigcap_i X_i$ and $\bigvee_i X_i = Cl(\bigcup_i X_i)$. This comparison shows that LOT is best understood as a structural schema linking consequence, models, and closure, rather than as a single fixed lattice presentation.

## 2. Satisfaction, concept lattices, and institutional LOT

A canonical LOT construction arises from Tarski satisfaction. In an institution $I = (\mathrm{Sig}, \mathrm{Sen}, \mathrm{Mod}, \models)$, each signature $\Sigma$ has a sentence set $\mathrm{Sen}(\Sigma)$, a model category $\mathrm{Mod}(\Sigma)$, and a satisfaction relation $\models_\Sigma$. For a signature morphism $\sigma : \Sigma \to \Sigma'$, the satisfaction condition is
$$
M' \models_{\Sigma'} \mathrm{Sen}(\sigma)(\varphi) \iff \mathrm{Mod}(\sigma)(M') \models_{\Sigma} \varphi.
$$
Fixing $\Sigma$, one obtains the truth classification with tokens $\mathrm{Mod}(\Sigma)$, types $\mathrm{Sen}(\Sigma)$, and incidence $\models_\Sigma$. FCA then supplies the derivation Galois connection
$$
A^{\uparrow} = \{\varphi \in \mathrm{Sen}(\Sigma) \mid \forall M \in A: M \models \varphi\},
$$
$$
\Phi^{\downarrow} = \{M \in \mathrm{Mod}(\Sigma) \mid \forall \varphi \in \Phi: M \models \varphi\},
$$
with
$$
A \subseteq \Phi^{\downarrow} \iff \Phi \subseteq A^{\uparrow}.
$$
Formal concepts are pairs $(A,\Phi)$ with $A = \Phi^{\downarrow}$ and $\Phi = A^{\uparrow}$, and the intents $\Phi$ are precisely theories closed under logical consequence relative to $\models_\Sigma$ [1811.02041].

The LOT is the concept lattice of these formal concepts, ordered by
$$
(A_1,\Phi_1) \le (A_2,\Phi_2) \iff A_1 \subseteq A_2 \iff \Phi_2 \subseteq \Phi_1.
$$
Hence, on theories, join is intersection of theories and meet is the closure of the union. In the IFF formulation, the same construction appears as the truth concept lattice of the truth classification $C = (\mathrm{Tok}, \mathrm{Typ}, \vDash) = (\mathrm{mod}(L), \mathrm{expr}(L), \vDash)$. Closed theories are intents $T = \mathrm{Th}(E)$ paired with extents $E = \mathrm{Mod}(T)$, and entailment is represented by closure:
$$
T \vDash \varphi \iff \varphi \in \mathrm{Th}(\mathrm{Mod}(T)) = T''.
$$
Top is $\mathrm{Th}(\mathrm{Tok})$, the set of logical truths, and bottom is $\mathrm{Th}(\emptyset)$, the inconsistent closed theory [1811.01144].

The categorical refinement in the institutional setting is that derivation factors through the concept lattice of theories. The derivation adjunction between powersets factors as a reflection and a coreflection through an axis $N(\models_\Sigma)$, and this axis is exactly $\mathrm{LOT}(\Sigma)$. The paper states this as the factorization of abstract truth through LOT by an extent reflection and an intent coreflection, and gives the restricted equivalence
$$
\mathrm{Clsn}(B) \simeq \mathrm{Clg}(B).
$$
LOT is also functorial under signature morphisms, yielding induced adjunctions between signature-indexed theory lattices, and the Grothendieck construction assembles these fibers into a category $\mathrm{Theory} = \mathrm{Gr}(\mathrm{the})^{op}$; under mild cocompleteness assumptions on $\mathrm{Sig}$, this category is cocomplete, so semantic integration of ontologies is represented by colimits in $\mathrm{Theory}$ [1811.02041].

## 3. Intermediate logics and fixed-point lattices

A different LOT construction is developed for intermediate logics. An intermediate logic $L$ is any set of formulas between intuitionistic propositional calculus $\mathrm{IPC}$ and classical propositional calculus $\mathrm{CPC}$, closed under modus ponens and uniform substitution. The class $\mathrm{IL}$ of intermediate logics forms a bounded lattice under inclusion, with meet given by intersection and join by least intermediate logic containing both [2004.00989].

Fix a univariate formula $\chi(p)$ and write $\chi^2(p) = \chi(\chi(p))$. The central construction is the $\chi$-logic generated by a set $\Gamma$: the smallest set $\Gamma^\chi$ containing $\mathrm{IPC}$, all substitution instances of formulas in $\Gamma$, the fixed-point axiom scheme $\chi^2(p) \leftrightarrow p$ for every atom $p$, and closed under modus ponens. If $L$ is the intermediate logic generated by $\Gamma$, then $\Gamma^\chi = L^\chi$. For each fixed $\chi$, the collection $\mathrm{IL}^\chi$ of $\chi$-variants forms a bounded lattice, and the map $L \mapsto L^\chi$ is a lattice homomorphism satisfying
$$
L^\chi \wedge M^\chi = (L \wedge M)^\chi, \qquad
L^\chi \vee M^\chi = (L \vee M)^\chi.
$$

The key input is Ruitenburg’s theorem: for a univariate $\chi$, there exists $N_\chi$ such that
$$
\mathrm{IPC} \vdash \chi^{n+2}(p) \leftrightarrow \chi^n(p) \quad \text{for all } n \ge N_\chi.
$$
This makes fixed points of $\chi^2$ syntactically tractable and yields the characterization
$$
L^\chi = \{\, \varphi(\#p) \mid \varphi[\chi^n(\#p)/\#p] \in L \,\}.
$$
Up to IPC-equivalence, univariate formulas have only six possible fixed-point shapes:
$$
\bot,\ p,\ \neg p,\ \neg\neg p,\ p \vee \neg p,\ \top,
$$
and since $\mathrm{IL}^{\neg p} = \mathrm{IL}^{\neg\neg p}$ there are exactly five distinct lattices $\mathrm{IL}^\chi$.

These five cases have sharply different behavior. $\mathrm{IL}^p = \mathrm{IL}$ because the fixed-point axiom $p \leftrightarrow p$ is trivial. $\mathrm{IL}^{\bot}$ and $\mathrm{IL}^{\top}$ are singleton lattices: the axioms $p \leftrightarrow \bot$ and $p \leftrightarrow \top$ force all atoms to denote $0_H$ and $1_H$, respectively, and $\mathrm{IPC}^{\bot} \ne \mathrm{IPC}^{\top}$ as theories. The negative-variant case $\mathrm{IL}^{\neg p} = \mathrm{IL}^{\neg\neg p}$ imposes $p \leftrightarrow \neg\neg p$ and yields the negative variants, or DNA-logics; an explicit example is
$$
KP^{\neg p} = ND^{\neg p} = ML^{\neg p} = InqB.
$$
Here the $\chi$-core consists of regular elements, forming a Boolean algebra in the signature $\{1,0,\wedge,\to\}$, so for any $L$, the logic $L^{\neg p}$ makes the $\{1,0,\wedge,\to\}$-fragment classical, while disjunction may remain non-classical. Moreover,
$$
L^{\neg p} = CPC \iff L \supseteq WEM,
$$
where $WEM = IPC + (\neg p \vee \neg\neg p)$. The dense-element variant $\mathrm{IL}^{p \vee \neg p}$ imposes $p \leftrightarrow p \vee \neg p$; algebraically, the $\chi$-core is the set of dense elements, and
$$
L^{p \vee \neg p} = CPC \iff L = CPC.
$$

The construction has an algebraic dual. If $\bar\chi : H \to H$ is the unary polynomial induced by $\chi$ on a Heyting algebra $H$, the $\chi$-core is
$$
H^\chi := \bar\chi^n[H]
$$
for some sufficiently large $n$, equivalently the set of fixed points of $\bar\chi^2$. Restricting valuations to $H^\chi$ yields $\chi$-semantics, and
$$
H \models^\chi \varphi \iff H \models \varphi[\chi^n(\#p)/\#p].
$$
On the algebraic side, $\chi$-varieties are exactly the classes of Heyting algebras closed under subalgebras, homomorphic images, products, and core superalgebras, and the paper proves
$$
\mathrm{IL}^\chi \cong^{op} \mathrm{HA}^\chi.
$$
This produces a full duality between the logical and algebraic LOTs [2004.00989].

## 4. LOT for disjoint theory combination

In the theory-combination setting, LOT is not a lattice of individual first-order theories but a lattice of theory-combination properties. Let $\mathfrak{T}$ be the class of decidable first-order theories over countable signatures, and for $X \subseteq \mathfrak{T}$ define
$$
G(X) := \{\, T \in \mathfrak{T} : \text{for all } T' \in X,\ T \sqcup T' \text{ is decidable} \,\}.
$$
Here $T_1$ and $T_2$ are combinable when their disjoint union $T_1 \sqcup T_2$ is decidable. The operator $G$ is antitone and satisfies the antitone Galois connection
$$
X \subseteq G(Y) \iff Y \subseteq G(X).
$$
Therefore $Cl := G \circ G$ is a closure operator, and the class $\mathcal{C}$ of closed subsets of $\mathfrak{T}$ forms a complete lattice under inclusion with
$$
\bigwedge_i X_i = \bigcap_i X_i, \qquad
\bigvee_i X_i = Cl\Big(\bigcup_i X_i\Big).
$$
On closed sets, $G$ is an antitone involution, so the lattice is isomorphic to its dual [2511.17374].

The semantic basis is Fontaine’s lemma: for decidable theories $T_1$ and $T_2$, $T_1 \sqcup T_2$ is decidable iff there is an algorithm that, given cubes $\varphi_1$ and $\varphi_2$, decides whether the spectra intersect:
$$
\sigma_{T_1}(\varphi_1) \cap \sigma_{T_2}(\varphi_2) = \emptyset.
$$
The LOT then organizes classical combination properties by exact Galois duality. Stable infiniteness is a closed symmetric node:
$$
G(\{SI\}) = \{SI\}.
$$
Shiny theories, which are smooth, have the finite model property, and have a computable minimal model function, are exactly the largest class combinable with all decidable theories:
$$
G(\mathfrak{T}) = \{shiny\}.
$$
The paper also proves
$$
G(\{polite\}) = \{SI\},
$$
showing that politeness does not extend indiscriminate combination beyond stable infiniteness.

Several additional closed nodes are paired by $G$. Gentleness and computable finite spectra satisfy
$$
G(\{gentle\}) = \{CFS\}, \qquad G(\{CFS\}) = \{gentle\}.
$$
Smoothness plus computable spectra and infinite decidability satisfy
$$
G(\{SM{+}CS\}) = \{ID\}, \qquad G(\{ID\}) = \{SM{+}CS\}.
$$
Computable spectra is a symmetric fixed point:
$$
G(\{CS\}) = \{CS\},
$$
and the paper presents this as a new sharp symmetric combination theorem generated by lattice operations, with
$$
\{CS\} = \{CFS\} \wedge \{ID\}.
$$
There are parallel dualities for $n$-shiny and $n$-decidable classes, and the free-filter families $\mathcal{F}$-QG and co-$\mathcal{F}$-QG yield chains and antichains of size $2^{\aleph_0}$. The resulting LOT answers several long-standing open questions negatively: no weaker symmetric property below stable infiniteness suffices for a Nelson-Oppen-style method applying to all pairs in that class, politeness is unsalvageable in the indiscriminate sense, and there is no single $U \in \{CFS\}$ whose combinability would characterize gentleness for all decidable theories [2511.17374].

## 5. Many-valued modal LOT and competing theories

In lattice-based modal logic, LOT arises in three intertwined forms. The paper on competing theories states that LOT is not named explicitly there, but its lattice-theoretic machinery induces a logical LOT of axiomatic extensions, a variable-set LOT of empirical theories, and a concept lattice associated with a reflexive many-valued graph [1905.11748].

The logical LOT is standard. If $Ax(T)$ is the set of axioms or theorems of a theory $T$ extending the basic normal lattice-based modal logic $L$, then
$$
T_1 \le T_2 \iff Ax(T_1) \subseteq Ax(T_2),
$$
with
$$
T_1 \wedge T_2 = Cn(Ax(T_1) \cap Ax(T_2)), \qquad
T_1 \vee T_2 = Cn(Ax(T_1) \cup Ax(T_2)).
$$
The paper isolates two special extensions:
$$
L_0 := L + (\Box \bot \vdash \bot), \qquad
L_1 := L + (\Box p \vdash p),
$$
with $L_1$ extending $L_0$. Semantics is given on reflexive many-valued graphs over a complete, frame-distributive and dually frame-distributive commutative residuated lattice
$$
A = (D,1,0,\vee,\wedge,\otimes,\to).
$$
For a graph-based $A$-frame $G = (X,R_\Diamond,R_\Box)$, the complex algebra
$$
G^+ = (X^+, [R_\Box], \langle R_\Diamond \rangle)
$$
is a complete normal lattice expansion in which $[R_\Box]$ is completely meet-preserving and $\langle R_\Diamond \rangle$ is completely join-preserving. The main completeness theorem states that, for $i \in \{0,1\}$, $L_i$ is sound and complete with respect to the corresponding class of graph-based $A$-frames.

The empirical LOT is indexed by sets of relevant variables. Each theory $X \subseteq Var$ determines a relation $R_X : Z \times Z \to A$ encoding similarity relative to $X$, and the powerset of variables forms a complete lattice ordered by inclusion. The modalities $\Box_X$ and $\Diamond_X$ induced by $R_X$ act on the concept lattice and preserve meets or joins, respectively. In the paper’s case study, the theories $A$, $M$, and $H$ correspond to ancient, modern calories, and hormonal or macronutrient response. Worlds are databases, $E$ records local similarity, and modal evaluation captures how a hypothesis changes when viewed through the lens of another theory. This yields a formal semantics of competing theories in which partial compatibility and theory-ladenness are represented by truth degrees in the underlying residuated lattice [1905.11748].

## 6. Latent and empirical approximations of LOT

A recent line of work constructs a geometric approximation to the classical LOT for equational logic. In the underlying exact order, an equational theory $T$ over a binary-operation signature is ordered by entailment:
$$
T_1 \le T_2 \iff Th(T_1) \subseteq Th(T_2),
$$
equivalently $Mod(T_2) \subseteq Mod(T_1)$. The classical lattice operations are
$$
T_1 \wedge T_2 = Cn(T_1) \cap Cn(T_2), \qquad
T_1 \vee T_2 = Cn(T_1 \cup T_2).
$$
The paper then embeds theories into a latent space using Stone pairings over random finite magmas. For a finite magma $A$ and a formula $\varphi(\bar x)$,
$$
\llbracket \varphi \rrbracket_A = \frac{1}{|A|^n}\#\{\bar a \in A^n : A \models \varphi(\bar a)\}.
$$
Using the ET dataset of $K = 4694$ magma equations and a sample of $n$ random magmas, one forms the Stone matrix $R \in [0,1]^{K \times n}$, quotients by coordinate permutations through Stone spectra, and performs PCA on the centered row covariance to obtain a three-dimensional latent embedding [2601.20759].

The resulting geometry correlates strongly with proof-theoretic structure. The ET implication graph contains $8{,}178{,}279$ implications among $4{,}694$ equations; quotienting by provable equivalence yields $1{,}415$ vertices and $4{,}824$ arrows. Reversible edges are much shorter in the latent space than strict implication edges: the reported mean lengths are approximately $0.69$ for reversible edges, $5.29$ for strict atomic edges, and $7.17$ for general strict edges. The first principal component $X$ is strongly correlated with expectation, $Y$ with variance, and the $Z$-axis separates conjugate equations by an emergent reflection symmetry around $Z=0$, with self-conjugate laws near $Z=0$. A Hasse-like diagram is obtained by contracting each reversible clique to its center of mass and drawing atomic arrows between clique centers.

The framework also proposes sample-based surrogates for logical operations. For composite theories represented by binary satisfaction vectors $s(T)$ over the sampled magmas,
$$
s(T_1 \vee T_2) = s(T_1) \wedge s(T_2),
$$
while
$$
s(T_1 \wedge T_2) \ge s(T_1) \vee s(T_2)
$$
in the sense that any sampled model of $T_1$ or $T_2$ satisfies the meet. A proposed statistical test for implication estimates conditional satisfaction by
$$
\hat P(T_2 \mid T_1) =
\frac{\sum_i s_i(T_1)s_i(T_2)}{\sum_i s_i(T_1)}
$$
and uses a Wilson score interval to decide whether the lower confidence bound exceeds a threshold. The paper is explicit that this geometry is only an empirical approximation: finite-random-magma statistics can diverge from general validity, and the current experiments rely on proof-theoretic ground truth rather than statistically inferred edges [2601.20759].

Across these settings, LOT functions as a unifying device for relating theories to one another through order, closure, and duality. In some cases it is exact and algebraic, as with concept lattices, $\chi$-logics, and Galois lattices of theory-combination properties; in others it is semantic and many-valued; in still others it is empirical and geometric. This suggests that the enduring content of LOT lies in the systematic organization of theory space, together with explicit operations that connect syntactic strength, semantic extent, and transformations between theories.

Source: https://www.emergentmind.com/topics/lattice-of-theories-lot