Purpose: The purpose of this study is to investigate the possibility of soft science measurement of motivation under strict hard science criteria from observations of individual animals and to suggest the conditions under which an observation can be classified as a measurement. Design/methodology/approach – The methodology starts from reconciling second-order cybernetics/ radical constructivism (SOC/RC) understanding of the central role of the observer with physical measurements, which accepts the existence of a mind-independent reality. As a result of the reconciliation, parallels were identified between the SOC/RC experiences of as_is and as_if, on the one hand, and the measurement concepts of accuracy and resolution, on the other hand. The scales of physical measurement are defined by criteria of varying strictness, and the scales that meet the strict criterion of concatenation are generally considered hard science and lead to well-defined accuracy and precision. The similarity between SOC/RC and physical measurement suggests that if accuracy and precision can be computed from observations, then the observations can be classified as measurements in a strict hard science fashion; otherwise, the observations are just observations. Findings: A nonlinear dynamic model of motivation is reintroduced as an example for reference in measurements of motivation. If there was an agreement on its use among observers (Ethologists), which in reality is not the case, then empirical data may be collected, and the averages and spreads of parameter estimations will define a reference for an animal species. Later, observers with their own data will calibrate with the reference model, so that new observers will have calculated values of accuracy and precision for their data. Research limitations/implications – Unlike hard science whose scales of measurement are practically unambiguous, measuring the purpose of behaviour of an animal has inherent ambiguity according to the reintroduced model. The ambiguity cannot be resolved from instantaneous readings. The necessary existence of ambiguity renders the criticism of hard science invalid, that of expecting to measure motivation with a static scale as if it were temperature. Practical implications: Human observers can be treated as measuring devices of motivation from observing behaviour. Each observer can have characteristic accuracy/precision, or validity/reliability, calculated from empirical data. Social implications – This is an inductive, rather than deductive, study of individual animal behaviour; the author believes it is extensible to individual human behaviour and personality studies. However, group behaviour studies are beyond its scope. Originality/value – The author believes that the suggestion of ambiguity of scales of animal motivation is original, and the suggested link between SOC/RC and a mainstream hard science is new.