Fritz Machlup’s Joint Review of Three Works on Social-Science Method (1955)

Fritz Machlup’s joint book review examines John Madge’s The Tools of Social Science, Arnold M. Rose’s Theory and Method in the Social Sciences, and Hans L. Zetterberg’s On Theory and Verification in Sociology. Its organizing question is what makes social inquiry explanatory: the use of privileged observational techniques, or the construction and testing of coherent theoretical propositions. Machlup sharply criticizes Madge’s empiricism, finds a corrective in Rose’s methodological pluralism, and endorses Zetterberg’s account of theory and verification. His central distinction is between scientific inquiry and scientism—the elevation of particular methods into standards that constrain which problems can be studied.

The review begins by challenging Madge’s claim that social science is only a century old and has yet to achieve victory. For Machlup, this chronology wrongly identifies social science with positivist sociology, excluding a much longer history of inquiry. The metaphor of victory is equally misleading:

I do not want to be less honest than Mr. Madge is (p. 290) but I should like to submit that some twenty-four hundred years of known social science have left us where we are, and that we shall always be short of victory because there is no end to the battle for knowledge, social or natural.

This opening establishes both the historical breadth and the provisional character of knowledge. Neither social nor natural science reaches a final state of completion. Madge’s restricted genealogy anticipates the methodological restrictions that Machlup subsequently attacks.

Madge organizes his book around documents, observation, interviews, and experiments, devoting the largest share to interviewing and treating experiment as the culmination of scientific method. Machlup objects less to these techniques themselves than to their hierarchy and the subordinate position assigned to theory. In his reading, Madge tolerates theoretical explanation chiefly because it stimulates empirical research, while treating logical order and ideal assumptions with suspicion. Knowledge is confined to what can be seen and heard; mental experiments remain rehearsals for supposedly more authoritative actual experiments.

Machlup then identifies an additional difficulty: Madge combines this restrictive empiricism with epistemological relativism. His appeal to Mannheim’s perspectivism makes truth dependent on shared experience and agreed social aims, while casting doubt on the possibility and desirability of objective social science. Machlup’s alternative is not a single universal procedure but a plurality of methods suited to different social sciences:

If, instead of proclaiming the coexistence of competing truths, he came to recognize the coexistence of different methods and of several social sciences, he would be less myopic also in other matters and less liable to misguide his readers.

The distinction between multiple truths and multiple methods is a decisive conceptual move. Methodological diversity need not entail the abandonment of objective inquiry. Conversely, insistence on one scientific method can obscure the legitimate differences among disciplines and research problems.

Rose supplies the review’s constructive counterweight. Machlup praises his opening discussion of theory and hypothesis testing, then emphasizes Chapter 14’s contrast between problem orientation and method orientation. Rose’s approving use of Maslow challenges quantification pursued for its own sake, hierarchies of scientific prestige, and orthodoxies that exclude problems or discourage methodological innovation. His criticism of the division between behavioral and documentary sciences similarly resists defining disciplines by their favored techniques.

Chapter 15 develops the positive alternative: social science should become an integrated system of propositions rather than an accumulation of disconnected findings. Machlup reproduces Rose’s formulation at length:

"An adequate system of general propositions is based on an internally consistent set of assumptions and definitions, such that each proposition is capable of being logically deduced from the assumptions and definitions (in the manner of a theorem). The building up of an adequate theory—the shorthand term for the framework of assumptions and definitions as well as the propositions deducible from them—requires a constant interplay between theory and data in which the theory is regularly modified and new deductions are derived from it and once more tested by crucial facts" (p. 263).

Logical consistency and empirical responsiveness are complementary here. Theory gives findings explanatory relations; evidence prompts revision and renewed deduction. Machlup’s defense of theory therefore does not imply indifference to testing. It rejects the idea that collecting observations alone constitutes an adequate science.

Rose’s chapters addressed to economists make this argument especially relevant to Machlup’s own discipline. His discussion of the deductive ideal-type method distinguishes testing whether one event follows another under specified conditions from identifying the assumptions under which that connection follows necessarily. Empirical inquiry then examines whether those assumptions apply to observable phenomena. This assigns deduction a legitimate task without treating logical necessity as proof of empirical relevance.

Zetterberg extends the constructive position through a program favoring theories over classificatory frameworks, verification over description, and limited theories over an all-inclusive system. Yet these smaller theories must connect with one another:

These "miniature theories," however, should be integrated, not isolated.

Their value lies in explanation, not merely in naming phenomena or recording distributions. Zetterberg’s treatment of operationalism completes the review’s argument: measurement procedures matter only insofar as they bear on the concepts and hypotheses being tested. Evaluating the relation between operational and nominal definitions is consequently part of verification itself. Machlup closes by recommending Zetterberg to economists, making the review a cross-disciplinary defense of theoretically organized, empirically accountable inquiry against the primacy of technique.