[This is a pre-publication version of a thesis to be published in the near future]
This thesis investigates the relationship between natural languages and those of formal
logic. It focusses on mismatches, which primarily concern semantic differences
between logical constants in formal systems and their designated counterparts in natural
language. These mismatches are directly relevant to the goals of philosophical
logic, which primarily aims to provide tools for assessing reasoning, the latter w…
Read more[This is a pre-publication version of a thesis to be published in the near future]
This thesis investigates the relationship between natural languages and those of formal
logic. It focusses on mismatches, which primarily concern semantic differences
between logical constants in formal systems and their designated counterparts in natural
language. These mismatches are directly relevant to the goals of philosophical
logic, which primarily aims to provide tools for assessing reasoning, the latter which
is most often carried out in the vernacular. Furthermore, they are important to various
debates in the philosophy of language and the philosophy of logic, not to mention
to linguistics and neighbouring fields. Despite this relevance, mismatches themselves
are rarely discussed in the literature. This thesis aims to fill this lacuna. Chapter
1 establishes the methodological background, which is that of so-called ‘connective
analysis’. It also introduces the main ideas behind use theories of meaning, the central
approach to linguistic meaning discussed in this book. Chapter 2 further delves
into the topic of meaning by discussing inferentialism. It finds its main variations
wanting, and develops a novel version that adopts a middle ground position, thereby
combining the advantages of its predecessors. Chapter 3 defends this version against
objections from the principle of compositionality. It also discusses the subject matter
of formal semantics, thereby clarifying the relationship between use theories and the
former. Chapter 4 deals with the mismatches themselves, surveying their different
forms as discussed since modern logic’s inception. It argues that there are no ‘qualitative’
mismatches between the two types of languages, and that natural language
contains a relation of logical consequence. This vindicates a conception of logic in
which the prior analyses of the vernacular under use-theoretic lights is methodologically
central. This approach is then applied starting in Chapter 5, which studies the
uses of modal expressions and quantifiers in natural language, their interactions in
particular. This will lead to the introduction of the so-called ‘quantified argument
calculus’ (‘Quarc’), a logic of quantification arguably closer to natural language than
the predicate calculus. Afterwards, Chapter 6 introduces two-valued modal extensions
of Quarc, and proves soundness and completeness theorems for a family of
associated Gentzen-style natural deduction systems. Chapter 7 applies one of these
systems to a specific type of categorical reasoning, and discusses an associated polygon
of opposition. Finally, Chapter 8 considers three-valued modal extensions of
Quarc, motivated by phenomena related to presupposition-failure. This will lead to
the introduction of a tableaux system, whose soundness and completeness will also
be proven. Thus, overall, this thesis vindicates the use of natural language as tool
to study logic, and of logic as a tool to study natural language.