Propositional Measure Logic

Francisco Aragão · 2025

Read Propositional Measure Logic by Francisco Aragão (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-14693.txt.

Introduction

Classical logic calls every statement true or false, but real reasoning often deals in degrees of confidence. Francisco Aragão, a solo author, presents a logic in 2025 where each formula carries a measure between zero and one as its degree of truth. Deduction keeps its classical shape, while meaning becomes probabilistic from the start. He proves the system sound, so its deductions respect the measures. Aragão then applies the logic to a stubborn problem in Bayesian networks, and sketches extensions for reasoning under uncertainty.

Catalog record

CREATORFrancisco Aragão
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE26 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-14693

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Propositional Measure Logic Francisco Aragão Abstract We present a propositional logic with fundamental probabilistic semantics, in which each formula is given a real measure in the interval [ 0 , 1 ] [0,1] that represents its degree of truth. This semantics replaces the binarity of classical logic, while preserving its deductive structure. We demonstrate the soundness theorem, establishing that the proposed system is sound and suitable for reasoning under uncertainty. We discuss potential applications and avenues for future extensions of the theory. We apply probabilistic logic to a still refractory problem in Bayesian Networks. Keywords : Artificial intelligence, Probabilistic Logic, Bayesian Networks, Imprecise Information. 1 Introduction Classical formal logic underpins mathematics, computer science and formal inference. However, to deal with ambiguity and partial information, new approaches have emerged - examples of which are fuzzy logic, probabilistic modal logic, Bayesian networks and belief-based systems. Even though progress has been made, these approaches generally have a limitation: the probability or degree of belief, in general, being kept out of the logical semantics, remaining at another level of interpretation on a deterministic model. In other words, maintaining the binary characteristic of truth - true or false, with uncertainty being treated as associated with models, rather than a property of logical language in itself. The proposed logic will be used to solve the problem of tackling certain types of uncertainty and imprecision with Bayesian Networks. The aim is to take advantage of the conceptual and practical benefits of this system in practical situations that have not yet been adequately explored. We introduce an extension to classical propositional logic which follows Hilbert’s axiomatic model, which is correct and quasi-complete. The syntactic aspect of classical propositional language remains intact, though the semantics is given by beliefs , e.g. a degree of truth, to which each formula is assigned, that is, a measure in the real interval [0, 1]. This approach shows ability to deal with applications from theory of knowledge to artificial intelligence. We develop a unified and principled logic for reasoning under uncertainty, including formalism, as well as expressiveness and interpretability. We have proven that the proposed probabilistic semantics is sound in relation to classical logic: every classically valid formula has positive probability. 2 The Propositional Measure Logic - PML In this section, we introduce a probabilistic logic that aims to work on logical reasoning and probabilistic reasoning, called PML. It is defined semantically in the Definition 4 , based on a generalization of the concept of truth valuation, the well-formed sequence of truth valuations, Definition 2 . The union of Logic and Probability can be traced back to the 17th century. A comprehensive list of pioneering approaches can be found in [ 12 ] , [ 13 ] and [ 14 ] . We introduce the Probabilistic Epistemic Logic whose language differentiates between ‘safe’ knowledge and ‘inaccurate’ knowledge. The latter may be an affirmative testimony, while the former can be a truth accepted on the basis of solid, consistent, argumentation. Probabilistic Epistemic Logic differs from traditional logic, such as Nilsson’s [Nil86] and Hailperin’s [Hail84] probabilistic logic because it has extended the truth function concept. As an ultimate motivation, we aim an appropriated probabilistic logic to fusion imprecise belief. The intuition behind Probabilistic Epistemic Logic is to measure the Probability of a sentence, you need a semantic concept whose structure is as sophisticated as the sentence structure itself. With a probabilistic semantics, it will be possible to make decisions in situations where purely logical systems, classical or non-classical, do not. We can handle contradiction without trivialization by doing logical reasoning, and also we can make decisions in a probabilistic-like way. The proposed system accommodates reasons and measurements for imprecise information. We proceed by introducing the basic notion of a formula: Definition 1 . [Formula] [ Van04 ] : Given a set of propositional letters, we define recursively a Well Formed Formula (WFF) as • \textrm{Propositional letters are formulae}; • \textrm{If }\alpha\ \textrm{ and }\ \beta\textrm{ are formulae, then }(\alpha\vee\beta)\ and\ (\alpha\rightarrow\beta)\textrm{ are formulae} • \textrm{If }\alpha\textrm{ is a formula, then }(\neg\alpha)\textrm{ is also a formula} . A truth valuation is a map from the set of all formulae from a logical language into the set \{true,false\} . In order to elaborate a probabilistic-like semantics to the logical language presented in Definition 1 , we need a structured extension of truth valuations that proves to be isomorphic to the structure of those formulas. Section 2.1 formalizes the Well Formed Sequences of truth valuations : 2.1 Well Formed Sequences of truth valuations The set of all possible truth valuations determining the truth of a sentence does not have the necessary structure to define a measure over a language. We are obliged to define a new truth object called well-formed sequence . Traditionally, logical truth is established through a meta-linguistic relationship between a simple object (a function) and a complex object (a proposition). We propose to replace this by the confrontation of two complex objects: a proposition, as before, but instead of the simple truth valuation, we would have a structured sequence of truth valuations. To define a mathematical measure over a language, it is necessary to capture some idea of quantity, as Nilsson and Hailperin among others have tried. See Section 5 . Each structured sequence represents a different way of being true for a logical formula, thus we define: Definition 2 . [Well-Formed Sequence]: Let {C} be a collection of truth valuations. We say that s is a Well Formed Sequence (WFS) generated from a collection {C} of truth valuations when • If v is a truth valuation then \big<v\big> is a well-formed sequence. • If s_{i} and s_{j} are well-formed sequences, then \big<s_{i},s_{j}\big> is also a well-formed sequence. • Nothing else is WFS. We call WFS ( {C} ) the set of all well-formed sequences generated by the collection {C} of truth valuations. Each language formula partitions the universe of well-formed sequences by two, those WFS associated to the formula and those not associated: Definition 3 . [Association]: Let \varphi be a WFF and s be a WFS. We say that s is associated to \varphi , (notated \varphi_{s} ) when if \varphi=A and s=\big<v\big> , then \varphi is associated to s ; if \beta is associated with s , then (\neg\beta) is also associated with s , if \alpha is associated with s_{i} and if \beta is associated with s_{j} , then \varphi=(\alpha\vee\beta) is associated with s=\big<s_{i},s_{j}\big> . if \alpha is associated with s_{i} and if \beta is associated with s_{j} , then \varphi=(\alpha\rightarrow\beta) is associated with s=\big<s_{i},s_{j}\big> . We call {A}_{\varphi} the set of all WFS associated to \varphi . Sequences associated with an atom A are of the form v , where v is a truth valuation. Formulae constructed by binary connective as (\alpha\vee\beta) are associated with sequences of type \big<s_{1},s_{2}\big> , where s_{1} and s_{2} are well-formed sequences. Negated formulae as (\neg\alpha) are specially associated to the same type of sequences as those associated to \alpha . Next, we develop a probability measure that will work as a logical semantics. 2.2 Probabilistic semantics - Mathematical Measures as Logical Semantics We introduce the justification between a sequence and an associated formula. Some

Read the complete text