Some non-standard interpretations of the axiomatic basis of Leśniewski's Ontology

Rafal UrbaniakAustralas. J. Log 2006 · DOI

TL;DR

This paper challenges the common assumption that the axioms of Leśniewski's Ontology uniquely determine the meaning of its primitive constant, typically denoted by 'ɛ'. The authors demonstrate that multiple, non-standard interpretations of 'ɛ' can be constructed that still satisfy the system's axioms, thereby showing that the axiomatic basis alone does not fix the meaning of the primitive constant.

What Problem Does This Paper Address?

In formal logic and philosophy, a key goal of axiomatization is to provide a precise and unambiguous definition of the fundamental concepts (primitive constants) within a system. The prevailing view regarding Leśniewski's Ontology is that its axioms, particularly Axiom 1, are sufficient to uniquely determine the meaning of its primary primitive constant, 'ɛ'. This constant is often interpreted as a form of "is" or "is a member of." The authors of this paper question this widely held belief, proposing that the axiomatic basis might not be as restrictive as commonly assumed. They aim to explore whether alternative interpretations of 'ɛ' can exist that are consistent with the axioms, thereby challenging the notion of "axiomatic determination" of meaning.

Methods

The paper employs a model-theoretic approach to analyze Leśniewski's Ontology. They first define a simplified basic language (BL) for the ontology, including name variables, logical connectives, and a universal quantifier, with a primitive functor 'f' (later specified as 'ɛ'). They then introduce a standard semantics for this language, defining notions of "satisfaction" within models (OBJ).

The core of their method involves constructing and analyzing "non-standard" interpretations of the primitive functor 'ɛ'. Instead of adhering to the traditional interpretation, they explore various ways 'ɛ' can be assigned a semantic meaning (e.g., as identity, as a falsum functor, or based on set-theoretic properties like emptiness or cardinality). For each non-standard interpretation, they rigorously prove whether the axioms of Ontology remain valid under that interpretation. This involves demonstrating that if the axioms are satisfied in the standard sense, they are also satisfied in the non-standard interpretation, and vice-versa. They also examine the validity-preserving nature of key rules of inference, such as substitution and extensionality, under these non-standard interpretations.

Datasets

Not discussed in the paper.

Key Insights and Ideas

  • Axioms Don't Uniquely Determine Meaning: The central finding is that the axioms of Leśniewski's Ontology do not uniquely determine the meaning of the primitive constant 'ɛ'. This directly challenges a common philosophical assumption about axiomatic systems.
  • Multiple Valid Interpretations: The authors demonstrate the existence of several distinct, non-standard interpretations of 'ɛ' that are fully compatible with the standard axioms of Ontology. This suggests that the "meaning" of 'ɛ' is more flexible than previously thought.
  • The Role of "Standard Semantics": The paper highlights that what is often considered the "standard" interpretation of 'ɛ' is just one among many possible interpretations that satisfy the axioms. This raises questions about the criteria used to select a "standard" interpretation.
  • Distinction Between Validity and "Determined Meaning": The work implicitly distinguishes between an axiom being valid under an interpretation and that interpretation being "determined" by the axiom. The paper shows that validity can be achieved in multiple ways, undermining the idea of unique determination.
  • Importance of Rules of Inference: Beyond just axioms, the paper also considers whether rules of inference remain validity-preserving under these non-standard interpretations, adding another layer to the analysis of axiomatic systems.

Key Results

The paper successfully demonstrates that several axioms of Leśniewski's Ontology (specifically, Axioms 1-7, and others related to numerical identity) remain valid under various non-standard interpretations of the primitive functor 'ɛ'. For instance:

  • Axiom 1 is shown to be valid under an interpretation where 'ɛ' signifies that its arguments are empty names.
  • Axiom 5 is shown to be valid under interpretations where 'ɛ' signifies that the first argument is empty and the second is non-empty, or where it signifies identity of non-empty extensions.
  • Axiom 7 is shown to be valid under an interpretation where 'C' (defined in terms of 'ɛ') acts as a verum functor (always true).

These results collectively support the main thesis that the axiomatic basis does not fix a unique meaning for 'ɛ'.

Limitations and Future Work

  • Focus on Axioms: While the paper proves that rules of inference are validity-preserving for some interpretations, the primary focus is on the axioms. The authors acknowledge that a deeper dive into the interaction between rules and non-standard interpretations is warranted.
  • Completeness: The paper notes that it does not require completeness with respect to a given non-standard interpretation to be a necessary condition for "being a theory of the functor." This suggests that further work could explore the implications of requiring completeness.
  • Rule of Definition: The paper touches upon complexities with the rule of definition in non-standard interpretations, particularly concerning whether definitions truly "determine" semantics. This is identified as a complex problem that the authors hope to address in future work.
  • Broader Interpretations: The authors mention that the list of interpretations presented is not exhaustive and that there are many more possible non-standard interpretations, particularly those related to numerical identity. They suggest that their method can be generalized to these cases.

Summary updated

BibTeX
@article{urbaniak2006some,
  author = {Rafal Urbaniak},
  title = {{Some non-standard interpretations of the axiomatic basis of Leśniewski's Ontology}},
  journal = {Australas. J. Log.},
  year = {2006},
  doi = {10.26686/ajl.v4i0.1778},
}

Loading related content...

Figure