International Encyclopedia of Systems and Cybernetics

1. "Any explicit move to obtain some quantitative knowledge of a system, or part or process of a system" (J.L. DESTOUCHES, 1990, p.479).

2."The assignment of numerals (and numbers) to represent properties" (N.R. CAMPBELL, 1957, p.267).

(According to CAMPBELL, a numeral is a material or quasi-material symbol, while a number is a mathematical concept. While numbers can be added or multiplied, numerals cannot be).

Measurements can be obtained using some instrument or device, but also by organizing a statistical inquiry or even using a more approximative way of evaluation.

J.L. DESTOUCHES states: "Any measurement is imprecise. It implies the interaction during a certain time of the studied system and another specific one, the measurement device (endowed with some observable characteristic, which will have changed after the measurement). Thus, any measurement implies a whole process with complex transmission devices of the obtained information and for showing the results. This introduces limitations to the knowledge of the system. One may conceive that some measurements cannot be made simultaneously because both measurement instruments cannot be used at the same moment, as it is for example, the case in quantical physics".

"While any positive knowledge about a system must be obtained by measurement, any measurement presuppose a theoretical concept, or at least a model. This is a dialectical situation. The model includes constants, parameters, variables, and relations between these quantities. Measurements are to be of these quantities: if the model is adequate and reliable, the quantities correspond to properties of the concrete system…

"A measurement is a concrete action, which is not at the same level as the model, but at the level of the system, while on the contrary its interpretation is at the level of the model" (p.430).

DESTOUCHES adds: "… the word 'measurement' must be understood in a broad meaning, it denotes any manipulation and any explicitely formulated method designed in order to provide a positive empirical or experimental knowledge about the studied system… using as well an apparatus, a device, like a questionary, an inquest or a measure of samples, any process explicitely formulated, giving a knowledge, i.e. a result about the system" (1980, p.197).

D. GERNERT refined still more these views, insisting that: "A scientist decides what to measure (or to observe) and how to do it (decision phase)". The results must finally be interpreted by this initiator, who in most cases depends on devices and intermediaries (1994).