Exploration of positivist theories of knowledge led to the understanding that our theories about the world must be evaluated as a whole. It is not possible to separate an individual sentence X of a given theory and ask whether it is true or false, analytic or synthetic etc. The interpretation of the sentence is only possible within the context of the theory as a whole. As a result, any analysis of the sentence is always conditional on the assumption of the validity of the background theory to which it belongs. This makes it impossible to distinguish sentences with empirical contents from those without. Several examples exist in the literature of definitions which are motivated by empirical regularities. On the surface, the definition is an analytic truth. Deeper examination shows that it has empirical content, since the definition was made to crystallize an empirical regularity, and summarize a pattern of observations. This is one reason why the dogma of the analytic/synthetic distinction does not survive a close examination.
It is widely agreed that epistemic and esthetic values are inevitably involved in the process of selection of scientific theories. The idea that scientific theories must be judged as a whole means that these values are reflected to some extent even in apparently purely observational sentences of the theory. To avoid this objection, Carnap tried to systematize the process of theory selection so as to avoid this problem, but could not succeed. A concrete example in the context of economic theory may be helpful in clarifying this issue.
The Pareto principle is widely accepted and regarded as a scientific and ethically neutral way of making welfare comparisons by economists. On the other hand, going further to recommend redistributions requires “unscientific” value judgments. This is a faithful representation of Locke’s theories of property: the initial property endowments must not be called into question, even if they leave some segments of the society starving, while others have far beyond their need. This leads to the paradoxical position that it is scientific and objective to support property rights over the basic needs of the poor, while it is unscientific and value-laden to advocate basic needs over property rights. The problem arises because the Pareto principle, which appears perfectly objective, reflects background commitments made elsewhere in the theory. As we have already seen, changing these commitments leads to equally objective alternatives, such as the Pareto-Style Longevity principle.