Perturbance: Understanding Why Robots Will Have High-Order Emotions Matters to Psychology

Yesterday, Drs. Sylwia Hyniewska, Eva Hudlicka and I submitted a paper for the AISB-2017 Symposium on Computational Modelling of Emotion: Theory and Applications. The title of our paper is “Perturbance: Unifying research on emotion, intrusive mentation and other psychological phenomena with AI”.

We are trying to draw attention to the fact that some research in affective Artificial Intelligence (AI) is highly relevant to research in several realms of psychology. Our paper focuses mainly on perturbance, which is a class of emotions systematically explored by Aaron Sloman and colleagues, mostly at the University of Birmingham. Perturbance refers to the persistent disturbance of attention by highly insistent motivators. Some of the (overlapping) literatures to which perturbance is relevant are:

  • Emotions,
  • Motivation,
  • Attention,
  • Repetitive thought (worry, intrusive mentation, rumination, perseverative thinking, etc.),
  • Rationality,
  • anorexia,
  • cognitive productivity,
  • self-regulation, volition and self-help, and
  • Psychotherapy.

Take cognitive productivity for example. Most authors who discuss distraction from work tend to focus on external distractions. However focusing and distractibility ultimately need to be understood in terms of how the mind generates and processes internal ‘affective’ information. The H-CogAff theory is relevant to that.

In its discussion of emotions, our paper focuses mainly on “limerence”, which is the attraction phase of romantic love. (The term was coined by Dorothy Tennov in Love and Limerence.) Limerence is an emotion that has been neglected in general theories of emotion. Limerence is interesting for many reasons, some of which we explore in the paper. One of them is that is of tremendous evolutionary significance, as it is often leads to mating and attachment (a process that Helen Fisher has described). Also whereas emotion researchers disagree about the defining features of emotion, and most general theories of human emotion fail to recognize the extreme significance of perturbance in emotion, limerence researchers all agree that attention is a core property of limerence. Our paper supplies the concept of perturbance to the limerence community.

We also argue that researchers, of all people!, and the general public, ought not to cast limerence as a pathology. We write:

While it may be tempting to cast limerence as a pathological form of romantic love [31,32], this would distort the original and common academic conception of limerence [33]. This would also overlook the near universality and evolutionary significance of limerence. Like other obsessions and other emotional states, limerence lies on continua [34] and may or may not be pathological. We believe the distorted casting should be resisted by scholars, and other terms used to describe pathological limerence. We also recommend scientific literature on this phase of romantic love should converge on the term limerence and help shape folksonomy.

Don’t expect a lot of answers in our article. It is brief, dense, broad and mainly suggestive. But there are plenty of references from which you can learn more. (Just search the web with the parameter to find the papers on this theory, some of which are listed below. You can also check out my book, Cognitive Productivity.) I hope our paper will help awaken the “sleeping beauty” that is Aaron Sloman’s theory of mind, which you can think of as a potentially progressive research programme.

Here’s the abstract of the paper we submitted to AISB.

Intrusive mentation, rumination, obsession, and worry, referred to by Watkins [1] as “repetitive thought” (RT), are of great interest to psychology. This is partly because every typical adult is subject to “RT”. In particular, a critical feature of “RT” is also of transdiagnostic significance—for example obsessive compulsive disorder, insomnia and addictions involve unconstructive “RT”. We argue that “RT” cannot be understood in isolation but must rather be considered within models of whole minds. Researchers must adopt the designer stance in the tradition of Artificial Intelligence augmented by systematic conceptual analysis [2]. This means developing, exploring and implementing cognitive-affective architectures. Empirical research on “RT” needs to be driven by such theories, and theorizing about “RT” needs to consider such data. We draw attention to H-CogAff theory of mind (motive processing, emotion, etc.) and a class of emotions it posits called perturbance (or tertiary emotions) [3,4], as a foundation for the research programme we advocate. Briefly, a perturbance is a mental state in which motivators tend to disrupt executive processes. We argue that grief, limerence (the attraction phase of romantic love) and a host of other psychological phenomena involving “RT” should be conceptualized in terms of perturbance and related design-based constructs. We call for new taxonomies of “RT” in terms of information processing architectures such as H-CogAff. We claim general theories of emotion also need to recognize perturbance and other architecture-based aspects of emotion. Meanwhile “cognitive” architectures need to consider requirements of autonomous agency, leading to cognitive-affective architectures.

It is an honour and great pleasure for me to collaborate with Sylwia Hyniewska and Eva Hudlicka. Sylwia is an empirical psychologist whose research focuses mainly on emotion recognition. Eva is an AI researcher who focuses on emotions and psychotherapy. She is also a psychotherapist. Combining such areas of expertise is,in fact, one of our suggestions. AI and Psychology by themselves are not enough.

Sylwia, Dr. Célyne H. Bastien and I have also recently submitted an abstract for ISRE-2017 named “Towards an affective information-processing theory of sleep onset and insomnia”. More on that later.


Published by

Luc P. Beaudoin

Head of CogZest. Author of Cognitive Productivity books. Co-founder of CogSci Apps Corp. Adjunct Professor of Education, Simon Fraser University. Why, Where, and What I Write. See About Me for more information.

One thought on “Perturbance: Understanding Why Robots Will Have High-Order Emotions Matters to Psychology”

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.