Stephen Edelston Toulmin, in opposition to logical positivism, argues that scientific theories operate as a system of rules that prescribe how inferences should be made. The instrumentalist conception is clearly relativistic and incompatible with the idea of cumulative development. In this sense they would not be true or false but simply ways of representing phenomena, such as a diagram or an image that can be more or less useful. And so, you will distinguish between:
1.Ideals (express expected natural behaviors)
2.Laws (refer to established models in which phenomena show deviation from the ideal)
3.Hypothesis (assumptions not yet studied that express natural regularities)
In a similar vein, Norwood Hanson, will observe that logical positivismlimits its interest to the finished product of scientific activity (i.e. theories) and neglects the rational procedure by which hypotheses, laws and theories are proposed for the first time. This author will study the methodology relevant to the context of each discovery. He will say that observation of the facts implies a contextual organization that prevents absolutely neutral language.
Paul Karl Feyerabend, will question the assumptions of logical positivism:
1.Principle of deductibility: for contemporary empiricism, reduction and explanation are achieved by deduction in the strict logical sense. Feyerabend will doubt this outright rationalism.
The consequence of this perspective is that at least some of the observational terms can be expected to be replaced in statements that are subjected to empirical contrast. The result is that successive theories cannot be compared to each other, because they are immeasurable, so that one cannot expect to deduce from each other. This idea of immeasurability will be shared by Thomas Khun, one of the most important representatives of the new epistemology.