Reinventing personality from a Volterra nonlinear model of motivation

Abstract

Multidimensional Laplace Transforms (MLT) is a generalisation of the Laplace transforms widely used in the analysis of linear control systems. The use of MLT allows the response of continuous nonlinear systems to be expressed as a power series of the stimulus. MLT was used in modelling animal motivation. The two distinct motivational mechanisms of priming and homeostasis are combined in a single mathematically rigorous nonlinear dynamic model. Projecting the model to simulate human motivation led to an age-dependent cybernetic personality model. The model comprises four complex dimensions, where sentences are classified to reflect states of motivation. The Age Trend Classification (ATC) is based on empirical evaluation of funniness scores over adult age: a sentence is classified according to changes in its average scores, either as falling, rising, peaking or constant. In contrast, psychometric personality models assume stable human traits; age dependence in adulthood is often considered insignificant. In this respect the age dependent artificial personality is a re-invention of and radically different from current psychological models. This paper explains the nonlinear roots of the motivation model, how the complex motivational dimensions reconcile with psychological traits and suggests a cybernetic hypothesis of human aging. Statistical analysis compares the consistency of Big Five personality with ATC funniness scores, the analysis shows that Cronbach's Alpha coefficients of raw funniness scores are higher in value than Big Five traits, and that the correlation coefficients between raw ATC funniness and Big Five scores show that nearly all ATC correlations with Extroversion and Neuroticism are closely distributed. The high consistency of ATC scores supports the reliability of funniness score data as a source of personality measures, the nearly uniform correlation with two of five personality traits suggests the reliance on such empirical measures as a criterion to select sentences with strong links to personality.