Skip to content

Boolean conditional-event algebra constructions

A Boolean conditional-event construction represents conditionals as elements of a Boolean structure with a probability interpretation under stated conditions.

Version
v1 · 2026-10-31 · History
Domain-specific #
14469
Domain group
Formal Sciences
Origin domain
Mathematics
Subdomains
Probability Theory, Mathematical Logic → Mathematics

Core Idea

A Boolean conditional-event construction represents expressions such as “a given b” as elements of a Boolean structure where they can be combined. A suitable measure can assign a basic conditional probability P(a∧b)/P(b), under the construction's stated assumptions and when P(b)>0. Booleanity is a feature of these constructions, not of every conditional-event approach.[1]

How would you explain it like I'm…

 

No faithful explanation at this level. All three generators judged a five-year-old picture would collapse 'a given b' into an ordinary event or plain if-then rule, erasing that it is a new element of a Boolean structure with its own construction-specific probability measure.

'Given' Statements You Can Combine

Mathematicians who study chance like to talk about things like 'it rains, given that it's cloudy.' Normally you can combine plain events with 'and', 'or', and 'not', but it's tricky to do that with 'given' statements. A Boolean conditional-event algebra construction is a way to build a special system where 'given' statements become things you can combine with 'and', 'or', and 'not'. With the right way of assigning chances, the chance of 'rain given cloudy' comes out the usual way: out of the cloudy times, how often it also rains. There is more than one way to build such a system, and not every approach to 'given' statements works like this.

Boolean Algebra of Conditionals

A conditional probability P(a|b) asks how likely a is when we restrict attention to cases where b holds, and it equals P(a and b) / P(b) when P(b) > 0. A Boolean conditional-event algebra construction turns the expression 'a given b' itself into an element of a Boolean algebra — a structure where 'and', 'or', and 'not' work with the usual laws — so conditionals can be combined with each other. Then a probability measure designed for that construction makes the probability of each basic conditional equal P(a∧b)/P(b), under the stated conditions. The algebraic structure and the probability assignment are two separate design decisions. Other theories of conditional events use different operations, and some aren't Boolean at all.

 

A Boolean conditional-event algebra construction embeds basic conditional expressions 'a given b' (often written a|b) as elements of a Boolean algebra, so that conjunction, disjunction, and complement are defined on conditionals and obey Boolean laws. This allows compound expressions such as (a|b) ∧ (c|d) to be formed and manipulated. A probability measure specific to the construction is then defined on the resulting algebra so that, for basic conditionals, the assigned probability agrees with the ratio P(a∧b)/P(b) whenever P(b)>0 and the construction's hypotheses are satisfied. Two choices are involved and should be kept separate: the algebraic structure that represents conditionals, and the probability assignment on it. Conditional-event theories differ in their operations, and not all of them yield a Boolean structure, so 'conditional-event algebra' is not synonymous with this Boolean family.

Examples

In the Goodman–Nguyen–van Fraassen product-space account described by Flaminio and colleagues, a finite base gives rise to countably infinite sequences of atoms. A basic conditional is represented by paths whose first occurrence of b also satisfies a; the product measure gives the ratio when P(b)>0. The 1994 original remains a library follow-up.[1][2]

Flaminio, Godo and Hosni construct a quotient of a free Boolean algebra on conditional symbols. Their four-element base example yields a four-element quotient; their finite-base theorem gives the basic conditional ratio for a positive probability on the base. Nonzero b and positive P(b) are distinct requirements.[1]

Boundaries

An arbitrary compound need not reduce to a single basic conditional. The displayed ratio does not define conditioning on a zero-probability event. Other conditional calculi may use non-Boolean operations, and these mathematical constructions do not themselves establish an empirical expert-system application.[1]

Neighborhood in Abstraction Space

Boolean conditional-event algebra constructions sits in a moderately populated region (56th percentile for distinctiveness): it has near-neighbors but no dense thicket of look-alikes.

Family — Formal Models & Logical Foundations (37 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-10-31

References

[1] Tommaso Flaminio, Lluís Godo and Hykel Hosni, “Boolean algebras of conditionals, probability and logic,” Artificial Intelligence 286 (2020), 103347, §§3, 6.2, 8.1–8.2. https://doi.org/10.1016/j.artint.2020.103347; author text: https://arxiv.org/abs/2006.04673 registry ↩a ↩b ↩c ↩d

[2] I. R. Goodman and H. T. Nguyen, “A Theory of Conditional Information for Probabilistic Inference in Intelligent Systems II Product Space Approach,” Information Sciences 76 (1994), 13–42, publisher abstract. DOI: 10.1016/0020-0255(94)90065-5. https://www.sciencedirect.com/science/article/abs/pii/0020025594900655 registry ↩