Changes

Barseghyan explains that this indicates that is because "we do expect confirmed novel predictions but only in very special circumstances. There was one common characteristic in all those episodes [...] episodes… they all altered our views on the structural elements of the world".[[CiteRef::Barseghyan (2015)|pp. 146]] For instance, in our key examples, Newton’s proposal of unobservable entities, such as gravity and absolute space, challenged the ''accepted ontology'' of the time, while Clark’s simply accounted for the data already available.

Barseghyan presents his historical hypothesis that this specific requirement for CNP has been employed in natural science since the 18th century. Assuming he is correct (for the sake of argument), he continues: "The ''third law'' stipulates that the requirement of confirmed novel predictions could become employed only if it was a deductive consequence of the accepted theories and other employed methods of the time. So a question arises: what theories and methods does this requirement follow from?".[[CiteRef::Barseghyan (2015)|pp. 147-8148]]

Barseghyan answers the question with two principles. For one, there is a principle, implicit in our contemporary mosaic and accepted since the eighteenth century, that states: "the world is more complex than it appears in observations, that there is more to the world than meets the eye".[[CiteRef::Barseghyan (2015)|pp. 148]] Thus, observations may not tell the whole story, as what we observe may an effect of an unobservable. Secondly, "it has been accepted since the early eighteenth century that, in principle, any phenomenon can be produced by an infinite number of different underlying mechanisms".[[CiteRef::Barseghyan (2015)|pp. 148]] "This leads us to the thesis of underdetermination that, in principle, any finite body of evidence can be explained in an infinite number of ways".[[CiteRef::Barseghyan (2015)|pp. 148]] Therefore:

<blockquote> The abstract requirement that follows from these two principles is that whenever we assess a theory that introduces some new internal mechanisms (new types of sub-stances, particles, forces, fields, interaction, processes etc.) we must take into account that this hypothesized internal mechanism may turn out to be fictitious even if it manages to predict the known phenomena with utmost precision. In other words, we ddo not tolerate "fiddling" with the ''accepted ontology;'' if a theory attemptes to modify the accepted ontology, it must show that it is not cooked-up.</blockquote> [[CiteRef::Barseghyan (2015)|pp. 148]]</blockquote>

Thus, in utilizing the Third Lawthird law, one can discover both when certain criteria become an implicit rule and under what conditions they are necessary.