Towards an integrative design-oriented theory of sleep-onset and insomnolence from which a new cognitive treatment for insomnolence (serial diverse kinesthetic imagining, a form of cognitive shuffling) is proposed for experimentally testing this

We present progress towards an integrative design-oriented (IDO) theory of sleep onset and insomnolence: the somnolent information-processing theory (SIPt; Beaudoin, 2013, 2014). We define “insomnolence” as difficulty falling (back) asleep — a key feature of insomnia (DSM-V).We argue that theories of human sleep onset and propensity require an IDO approach. By “design-oriented” we mean adopting the design stance (Dennett, 1982; Poggio, 2012; McCarthy, 2008) which is universally known in theoretical Artificial Intelligence and cognitive science but unused in theories of sleep onset and insomnolence, SIPt aside. Like other cognitive science, IDO involves interdisciplinary information-processing theories; but it is also integrative, aiming to specify how requirements of autonomous agency (competence) are realized by the interaction of diverse component processes (subsuming motivational, cognitive, executive and ancillary functions). The IDO approach requires that any appeal to key psychological constructs (“consciousness”, “arousal”, “emotion”, “attention”, “goals”, “intention”, etc.) be grounded in specific IDO theories. This approach is meant to contribute to a paradigm shift in research in insomnia, “emotion” and psychology more generally, in response to what Beaudoin, Hyniewska & Hudlicka (2017) and Muthukrishna & Henrich (2019) identified as the root of psychology’s replication crisis: lack of rigorous, ambitious, progressive, evolutionarily grounded theoretical integration. We claim control of human somnolence posed a significant evolutionary challenge particularly due to their abundant cortex.Leading theories of insomnia tend to explain insomnolence in terms of cognitive and/or physiological activity (Perlis, 2011) or “arousal” (Harvey, 2005). Cognitive theories of insomnia assume that attention, intention and effort to sleep are insomnolent (e.g., inhibiting “de-arousal”, Espie, 2006). Rejecting these assumptions, we argue that arousal is a problematically polymorphic concept unsuitable for IDO explanations of somnolence.In contrast, SIPt grounds its major concepts in specific IDO theories. In accordance with Moors' (2017) skepticism, SIPt replaces “emotion” with computational architectures of motivation. More precisely, we leverage the H-CogAff (Sloman, 2003) and LIDA (Franklin et al, 2013) architectures. We replace the concept of “emotion” and “arousal” with IDO concepts of perturbance and alarms. Perturbance is an emergent state in which an insistent motivator tends to control executive functions (Beaudoin, 1994; Wright, Sloman & Beaudoin, 1996). Perturbance is theoretical grounding for repetitive thought (Watkins, 2008). Alarms (Oatley, 1992; Sloman, 2003; Baars & Franklin, 2009) are urgent global control signals which, we claim, also underlie the alarm reaction (Selye 1936).SIPt postulates that (1) chronobiological processes (Borbély, 2016) are the principal contributors to somnolence; (2) sleep-onset-like information-processing is pro-somnolent (increases sleep propensity); (3) perturbance is insomnolent; (4) alarms are insomnolent; (5) some perceptual states affect sleep propensity: sensing supineness, rocking (Bayer et al, 2011) or skin temperature, Romeijn et al (2011).We describe an effortful form of cognitive shuffling, serial diverse kinesthetic imagining (SDKI). It is suitable for an experiment pitting SIPt against other theories (eg, Espie, 2006 and Havey, 2005) since only according to SIPt should SDKI be both counter-insomnolent (per postulates 3 and 4) and pro-somnolent (per postulates 2 and 5).Yet more theoretical work is required towards an IDO theory of somnolence.