3) The History of Economics as ‘Experimental Economics’
The important feature of Pareto’s sociology is that economic theories are observable facts, and the history of economics provides for the analysis of ‘facts’ from which hypothetical propositions can be developed in relation to the ‘sociological part’ of the economic phenomenon. That is, the history of economics provides the basis for developing theories pertaining to the sociological part of the economic phenomenon for which axiomatic modelling produces poor (unacceptable) approximations:
“… theories are facts. This remark is very true, and the history of those facts is the foundation of a theory about them, that is, a theory about theories.” (Pareto  1980 , p. 742).
To appreciate the importance of Pareto’s theory of theories to experimental economics, it is first necessary to consider the distinction between the intrinsic and extrinsic elements of theory, as outlined in his Sociologia. The intrinsicaspect concerns the soundness of theory in terms of its relationship to facts and the extrinsic5aspect concerns the relationship between a theory and society collectively without regard to the theory’s intrinsic scientific merit. Intrinsic merit is considered with reference to the closeness of the relationship to theory and fact pertaining to the primary phenomenon under investigation, and the extrinsic considers the merits or demerits of action advocated by theories for various members of society (those who advance a theory or those who adopt a theory) or for society more generally.
Pareto’s theory of theories, which he referred to as derivations, suggests that theories derive from subjective sentiments, which are linked to the extrinsic aspects. Derivations have been discussed thoroughly and extensively elsewhere (Parsons 1937, Finer 1966, Tarascio 1968, de Pietri-Tonelli and Bousquet 1994), so for the purpose of this study it is adequate to simply note that they combine sentiment and reason in a non-experimental manner. Importantly, the use of textual devices to persuade can be systematically analysed. Theories may utilise assertion, ranging from theoretical hypothesis to blunt statement of apparently self evident truths, make reference to the authority of authors (distinguished scholars, schools or traditions), make reference to sentiment and metaphysics (vague social welfare notions associating benefit to those who advance an idea to the benefit of the collective) or utilise analogies. If these devices are adopted in experimental theory, they are being used to support the development of intrinsic science by illustrating theory or persuade others of its scientific merit. However, the main use of such devices, which was of most interest to Pareto, was non-experimental: to persuade and influence others by exaggerating the extrinsic benefits of proposed action (policy).
Pareto’s textual exegesis of the history of economic theories revolves around the relationship between both the intrinsic and the extrinsic: “Both methods, if used exclusively, are equally incomplete.” (Pareto  1935, pp.503‑04). When the intrinsic aspects of theory dominate, science is experimental. Analysis of the extrinsic was designed to provide insight into human action when the individual’s subjective intent is interdependent with objective events, with the extrinsic aspects of social doctrines and theories significant sources variation in subjective forces through ‘persuasion’ and preference modification.
One important implication of the intrinsic-extrinsic distinction for the history of economics is the necessity of considering whether progress in economic theory is ‘asymptotic’, in which case scientific knowledge is cumulative but imperfect, or not asymptotic, in which case knowledge is neither cumulative nor perfect. It is asymptotic when the intrinsic aspect of theory continues to grow, permitting intellectual historians to speak of progress in particular classes of economic theory. If it is not asymptotic, the intrinsic aspects merely fluctuate as new extrinsic elements are introduced in response to extrinsic considerations, in which case historians cannot speak of progress in the remaining classes of economic theory.
Interestingly, in History of Economic Analysis, Joseph Schumpeter (1954) faced similar interpretive issues, and his classification of economics texts bear some relation to Pareto’s generic intrinsic-extrinsic distinction. Schumpeter classed economics texts in term of whether they dealt with: systems of political economy; economic thought; or economic analysis. The history of systems of political economy was characterised as dealing with opinions and desires about the economic system and the history of economic thought was characterised as dealing with ideas about public policy and economics. These two classes of text may be considered broad examples of what Pareto associated with the extrinsic elements of theory. Schumpeter’s notion of the history of economic analysis, which is the main aspect of his interest in intellectual history, is concerned with scientific progress and is closely related to Pareto’s general notion of intrinsic aspects of science. However, the similarity must not be overstated: while both masters stressed the sociology of knowledge, to Pareto this field was the primarily consideration for the extrinsic aspects of science, whereas to Schumpeter ‘vision’ and ideology were treated as important motivating factors in scientific progress. The similarity concerns their understanding of the interpretive problem, but their objectives for the history of economics were different. As a consequence, their solutions to this interpretive problem were diverse, and the focus of their historical research, were quite different.