ABSTRACT

But, if computer programs will evolve and approach the human ways of reasoning (independently of whether they will reach this or not), their softwares will need to make inferences involving various forms of reasoning, other than mere deductive ones and, in particular, those involving vague assertions (or propositions), as we, humans, usually do. For instance, we usually also 'make inferences' from vague propositions, to which we in general cannot attribute, with certainty, one of the two truth-values true or false. Frequently, our reasoning is performed by attributing only some degree of confidence as either the involved propositions are true or not. This is what may happen, for instance, when we are visiting a foreign country and someone gives us (vague) information about the location of a certain place (perhaps because the native is also not sure about the right geography of the city), and we 'believe' in the information with a certain degree of confidence and 'decide' the way to be taken.