This book traces how such a seemingly immutable idea as measurement proved so malleable when it collided with the subject matter of psychology. It locates philosophical and social influences reshaping the concept and, at the core of this reshaping, identifies a fundamental problem: the issue of whether psychological attributes really are quantitative. It argues that the idea of measurement now endorsed within psychology actually subverts attempts to establish a genuinely quantitative science and it urges a new direction. It relates views (...) on measurement by thinkers such as Holder, Russell, Campbell and Nagel to earlier views, like those of Euclid and Oresme. Within the history of psychology, it considers contributions by Fechner, Cattell, Thorndike, Stevens and Suppes, among others. It also contains a non-technical exposition of conjoint measurement theory and recent foundational work by leading measurement theorist R. Duncan Luce. (shrink)

It has become customary to locate the origins of modern measurement theory in the works of Helmholtz and Hölder. If by ‘modern measurement theory’ is meant the representational theory, then this may not be an accurate assessment. Both Helmholtz and Hölder present theories of measurement which are closely related to the classical conception of measurement. Indeed, Hölder can be interpreted as bringing this conception to fulfilment in a synthesis of Euclid, Newton, and Dedekind. The first explicitly representational theory appears to (...) have been Russell's. He rejected the traditional concept of quantity and its connection to number, making the use of number in quantitative science problematical. He solved the problem via representationalism. (shrink)

According to the realist interpretation, measurement commits us not just to the logically independent existence of things in space and time, but also to the existence of quantitatively structured properties and relations, and to the existence of real numbers, understood as relations of ratio between specific levels of such attributes. Measurement is defined as the estimation of numerical relations (or ratios) between magnitudes of a quantitative attribute and a unit. The history of scientific measurement, from antiquity to the present may (...) be interpreted as revealing a progressive deepening in the understanding of this position. First, the concept of ratio was broadened to include ratios between incommensurable magnitudes; second, the concept of a quantitative attribute was broadened to include non-extensive quantities; third, quantitative structure and its relations to ratios and real numbers were elaborated; and finally, the issue of empirically distinguishing between quantitative and non-quantitative structures was addressed. This interpretation of measurement understands it in a way that is continuous with scientific investigation in general, i.e., as an attempt to discover independently existing facts. (shrink)

The thesis that numbers are ratios of quantities has recently been advanced by a number of philosophers. While adequate as a definition of the natural numbers, it is not clear that this view suffices for our understanding of the reals. These require continuous quantity and relative to any such quantity an infinite number of additive relations exist. Hence, for any two magnitudes of a continuous quantity there exists no unique ratio. This problem is overcome by defining ratios, and hence real (...) numbers, as binary relations between infinite standard sequences. This definition leads smoothly into a new definition of measurement consonant with the traditional view of measurement as the discovery or estimation of numerical relations. The traditional view is further strengthened by allowing that the additive relations internal to a quantity are distinct from relations observed in the behaviour of objects manifesting quantities. In this way the traditional theory can accommodate the theory of conjoint measurement. This is worth doing because the traditional theory has one great strength lacked by its rivals: measurement statements and quantitative laws are able to be understood literally. 1 This paper was completed in 1990-1. while the author was a visiting scholar at the Irvine Research Unit in Mathematical Behavioral Sciences. University of California. Irvine. The author wishes to thank the Director. Professor R. D. Luce, for the generous provision of space and facilities within the Unit and for his critical comments upon some of the ideas expressed herein: Professor L. Narens. for his trenchant criticisms: and the University of Sydney, for granting Special Study Leave and financial assistance to make the visit possible. (shrink)

As an aspiring science in the nineteenth and early twentieth centuries, psychology pursued quantification. A problem was that degrees of psychological attributes were experienced only as greater than, less than, or equal to one another. They were categorised as intensive magnitudes. The meaning of this concept was shifting, from that of an attribute possessing underlying quantitative structure to that of a merely ordinal attribute . This fluidity allowed psychologists to claim that their attributes were intensive magnitudes and measurable . This (...) claim was supported by an argument that order entails quantity. As adapted by psychometricians, the argument was that if an attribute is ordered, then the differences between its degrees are quantitative and, therefore, measurable. However, in a paper ignored in psychology for six decades, the issue was resolved mathematically and the resolution implies that the psychometricians’ argument was fallacious. (shrink)

In his review of Charles Spearman’s The Nature of ‘Intelligence’ (1923), R. G. Collingwood launched an attack upon psychometrics that was expanded in his Essay on Metaphysics (1940). Although underrated by friend and foe alike, Collingwood’s critique identified a number of defects in the thinking of psychometricians that subsequently became entrenched. However, his main complaint was that psychology generally (and, by implication, psychometrics) was a ‘fashionable scientific fraud’. This charge was inspired by his more general views on logic and metaphysics, (...) which, however, as I argue, are logically unsustainable. Ironically, other elements of his philosophy – his ‘fallacy of calculation’ and concept of ‘scale of forms’ – are relevant to psychometrics and tip the scales in favour of his otherwise unwarranted charge. (shrink)

The transition from the traditional to the representational theory of measurement around the turn of the century was accompanied by little sustained criticism of the former. The most forceful critique was Bertrand Russell''s 1897 Mind paper, On the relations of number and quantity. The traditional theory has it that real numbers unfold from the concept of continuous quantity. Russell''s critique identified two serious problems for this theory: (1) can magnitudes of a continuous quantity be defined without infinite regress; and (2) (...) can additive relations between such magnitudes exist if magnitudes are not divisible? The present paper shows how the traditional theory answers these questions and compares the traditional and representational theories as contributions to our understanding of the logic of application. (shrink)