Nenad Smokrovic

Dealing with deductive reasoning, performed by ‘real-life’ reasoners and expressed in natural language, the paper confronts Harman’s denying of normative relevance of logic to reasoning with a logicist thesis, a principle that is supposed to contribute for solving the problem of incongruence between descriptive nature of logic and normativity of reasoning. The paper discusses in detail John MacFarlane’s (2004) and Hartry Field’s (2009) variants of “bridge principle”. Taking both variants of bridge principles as its starting point, the paper proceeds arguing that there is more than one logical formalism that can be normatively suitable for deductive reasoning, due to the fact that reasoning can assume different forms that are guided by different goals. A particular reasoning processing can be modelled by specific formalism that can be shown to be actually used by a real human agent in a real reasoning context.

In this paper I am developing the theses that argumentation is a means for extending knowledge. The theses are founded on two focal points:1. Reasoning is designed for argumentation, and 2. Argumentation process is an exceptionally successful media that provokes usage of methods reliable for the extension of knowledge. The first point relies on Sperber’s and Mercier’s evolutionary psychological approach to argumentation which I consider the most convincing theory in the field. Taking this ground as a departing point, the goal of the paper is to broaden this approach with epistemological insights that I base on Williamson’s safety theory of knowledge.

In traditional semantic theory the meaning of a word or operator (logical constant) is permeated with normativity. It is held that if one grasps the meaning of a word (or logical constant), one ought to behave in a certain way. This view is labelled as normativism here. Normativists hold that meaning is intrinsically and irreducibly normative. The naturalistic approach to semantics, on the other hand, which tries to reconcile the traditional approach to meaning with a naturalistic world-view, has to naturalise the normative character ofmeaning. Naturalists employ several strategies of argumentation, two of which I deem to be particularly significant. These two strategies are exemplified by P. Horwich’s revisionistic and C. Peacocke’s reductionistic approaches. This paper elucidates and critiques the former. My criticism tries to show that Horwich’s theory does not offer a successful answer to the normativist challenge.

In traditional semantic theory the meaning of a word or operator (logical constant) is permeated with normativity. It is held that if one grasps the meaning of a word (or logical constant), one ought to behave in a certain way. This view is labelled as normativism here. Normativists hold that meaning is intrinsically and irreducibly normative. The naturalistic approach to semantics, on the other hand, which tries to reconcile the traditional approach to meaning with a naturalistic world-view, has to naturalise the normative character ofmeaning. Naturalists employ several strategies of argumentation, two of which I deem to be particularly significant. These two strategies are exemplified by P. Horwich’s revisionistic and C. Peacocke’s reductionistic approaches. This paper elucidates and critiques the former. My criticism tries to show that Horwich’s theory does not offer a successful answer to the normativist challenge.