Does Aristotelian logic describe human reasoning? Valid syllogisms and canonical models

Abstract

In the case of quantified propositions, the mental models theory distinguishes between canonical and noncanonical models. While people identify the canonical models in an immediate, rapid, and easy way, the noncanonical models cannot be detected without reflection and cognitive effort. In this paper, I try to show that all of the valid syllogisms in Aristotelian logic can be considered to be correct by resorting only to the canonical models of their sentences. In this way, I argue that this means that Aristotelian logic can be a useful criterion to explain, describe, and even predict people’s conclusions from quantified assertions.

Mood Baroco, Avicenna’s proof of it, and reasoning by means of models

“…All of the conclusions I will present in this last section are, for obvious reasons, very akin to conclusions raised in papers such as, for example, those of López-Astorga (2016a, 2016b. It is worth resorting to the mental models theory to analyze ancient logics because, as said, it has a strong empirical support.…”

“…However, returning to the beginning of this section and always in keeping with conclusions similar to the ones of López-Astorga (2016a, 2016b, it can also be said that activities of that type can be relevant for ancient logics as well. Such activities can allow discovering their real status and what they actually described.…”

“…On the one hand, the theory proposes that people usually think about sentences considering possible cases that can be consistent with them. On the other hand, such cases are not always detected without difficulties: some of them require minimal cognitive effort and others need more cognitive effort (see also, e.g., López-Astorga, 2016a). In this way, a set of cases requiring little cognitive effort is called by Khemlani et al (2015) a 'canonical model', and a set of cases needing further effort is called by them a 'noncanonical model', their main prediction in this way being that most of the individuals will tend to consider only canonical models when reasoning, since these last models are, as said, easier.…”

“…This theory has already been related to ancient logics in several works. For example, if we only consider Aristotelian logic, we can say that it has been tried to prove that, if we assume the aforementioned theory, we must also suppose that all of the moods of all of the figures in this last logic have to be necessarily easy for the human mind (López-Astorga, 2016a). Clearly, papers in this direction are interesting because they allow reviewing the predictive potential of a logic such as the Aristotelian one and, at the same time, checking some of the predictions of the mental models theory.…”