Classes of valuations closed under operations Galois-dual to Boolean sentence connectives (Q1922666)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Classes of valuations closed under operations Galois-dual to Boolean sentence connectives
scientific article

    Statements

    Classes of valuations closed under operations Galois-dual to Boolean sentence connectives (English)
    0 references
    0 references
    26 October 1997
    0 references
    Where \(U\) is a language and \(\mathcal V\) is a class of valuations (truth-value assignments to the formulas of \(U\)), compare the hypotheses \[ \begin{aligned}\forall x_1, x_2 &\in U \exists x_3\in U \forall v\in{\mathcal V}: v(x_3)= T\text{ iff }v(x_1)= T\text{ or }v(x_2)=T,\tag{\(\dag\)}\\ \forall v_1, v_2 &\in{\mathcal V}\exists v_3\in{\mathcal V}\forall x\in U: v_3(x)= T\text{ iff }v_1(x)= T\text{ or }v_2(x)= T.\tag{\(\dag\dag\)}\end{aligned} \] We describe a Galois connexion between \(U\) and \(\mathcal V\) in terms of which \((\dag)\) and \((\dag\dag)\) are Galois-dual statements; the first saying that there are disjunctive combinations on the left (`\(U\)') side of this connexion, and the second saying that there are disjunctive combinations of the right (`\({\mathcal V}\)') side. Traditionally, sentential logic explores the effect of such assumption as \((\dag)\) on the consequence relation (on \(U\)) determined by (= sound and complete w.r.t.) \(\mathcal V\); for example, writing \(x_1\vee x_2\) for the \(x_3\) promised by \((\dag)\), we have, for all \(x,y,z\in U:x\vee y\lvdash y\vee x\); if \(x\lvdash z\) and \(y\lvdash z\) then \(x\vee y\lvdash z\); and so on. We shall open up for exploration in this paper the effect of such assumptions as \((\dag\dag)\) on the consequence relation \(\lvdash\) (on \(U\)) determined by \(\mathcal V\). These effects include, for example, the principle: If \(x,y\lvdash z\) then either \(x\lvdash z\) or \(y\lvdash z\). We investigate, that is, the logical repercussions of closure assumptions like \((\dag\dag)\) on the right, rather than of closure assumptions like \((\dag)\) on the left. In reasoning about such matters we notate the disjunctive combination of \(v_1\) and \(v_2\) (the \(v_3\) promised by \((\dag\dag)\)) as \(v_1\nabla v_2\), describing \(\nabla\) as that operation on valuations which is Galois-dual to the sentence connective \(\vee\). Clearly any Boolean sentence connective gives rise similarly to a Galois-dual operation on valuations. Actually the above description is oversimplified in numerous respects, one of which being that we treat the case in which \(U\) is a class of formulas as a special case, reserving talk of `logic' for this case, and instead of talking of consequence relations, which is too suggestive of the special case, we will speak of `closure relations' on an arbitrary underlying set \(U\); the need for this extra generality will be evident from 0.4 and 1.1. The latter passage introduces an example raising a problem in terms of which we have chosen to organize much of the discussion, namely the operation Galois-dual to material equivalence (so that we have the `or' in \((\dag\dag)\) replaced by `if and only if'); this problem arises naturally from reflection on the philosophical concept of supervenience or on the database-theoretic concept of functional dependency.
    0 references
    closure under operations
    0 references
    Boolean connectives
    0 references
    truth-value assignments
    0 references
    valuations
    0 references
    Galois connexion
    0 references
    sentential logic
    0 references
    consequence relation
    0 references
    supervenience
    0 references
    functional dependency
    0 references

    Identifiers

    0 references
    0 references
    0 references
    0 references
    0 references
    0 references