The use of physiological signals can be studied not only on their own, but within a context of "technologies" of emotional production: under what conditions can physiological markers of emotion be produced, and why is it interesting to map emotions at this precise moment?  Using techniques from physiological (stomach muscle fluttering, mechanically reproducing facial expression, breathing techniques) to psychological (meditation, narrative storytelling, memory recall) individuals can activate their emotional states so as to not be manipulated but to be in control of their emotions and affects as simultaneously autonomous and triggered states.  One can also subject an individual to emotional triggers through media (sad music, fast tempos, horror films, pornography).


It is typical for scientists to use media as a stimuli in their experiments to generate emotional states beyond general apathy, boredom, or what could be called a "neutral state". For example, a database of films and images that are agreed upon within the scientific community as most statistically likely to generate physiological/emotional responses from subjects is often used. I find this one of the least interesting means of inducing emotions in subjects for artistic purposes, as the meaning of the work is derivative, and tied to the non-responsive, fixed media as source of production. The questions of the emotion are tied to questions of reception of the media: for instance, if I sonified the emotions of people while they watched a horror movie, it seems that the process illustrates not only what kind of physiological responses occur as a result of the movie, but how effective or non-effective the movie is in invoking emotional states.  As artists we know that audiences have a wide range of opinions and reactions to symbols, sounds and narratives, so I find this kind of mapping of emotion derived from most media work to be rather boring for my own use. Furthermore it naturalizes emotion as a passive, natural production that draws participants away from their embodiment and towards its temporary erasure by experiencing a movie.


My first explorations of mapping physiological markers of emotion highlighted the robotic, the autonomous, and the technological through: 

(1) sound-producing infrastructure of 6 robotic instruments.


(2) emotion-producing actors that used imagination to invoke physiological/emotional states every 5 minutes

(3) the mapping of these physiological markers of emotion through algorithmic processes

In particular I want to draw attention to (2), the exploration of emotion as a technological process.  I worked with "actor" as a liminal subject that exists between (1) artist as an overly emotional person, as well as (2) artist skilled in technical production of emotion. 


The data from these sessions were fed into a Max/MSP patch that made use of euclidic beat generators to activate tones distributed into patterns corresponding with diatonic music.  I struggled with these mappings as I wanted at once the music to convey physiological information and perhaps correspond somewhat to the affective state of the actor at that time.  I couldn't find a way to do this that wasn't prescriptive and cliché, and finally decided that abstraction was a better route for now.


In later work (Song of Seven) I explored a children's choir as artists in this emotional feedback loop as well: artists in their own right that produce affective material for each other through narrative storytelling. The choir members improvised their own memories for each other and the physiological reaction to these stories were very transparently sonified using a Teensy 3.2 microcontroller, and inserted into the background accompaniment of the choral song.


{keywords: [physiological markers, emotions, emotional state]}

These technologies not only demonstrate the increased capabilities of machines to perform emotional labor, but also demonstrate the possibility of our own emotional behaviors as sociologically encoded, our own most intimate conversations as results of logical ebbs and flows in language keywords. In this way one can see how affective labor as female labor demands infinite patience and great deal of emotional restraint based on codified social niceties: simply put, women are not socialized to be emotional, but rather, appropriately emotional.


Simultaneous to this development of affective computing, neurophysiologists have been conducting research on the physiological processes that make up how emotion is felt in the human body. These processes are understood to be simultaneously dependent on physical factors such as temperature, age, restfulness and health of the subject, but also their psychological past, as well as what type of external factor triggered the emotional reaction in their body in the first place (this is sometimes referred to as a social factor, for example if you are angry at another person or angry at yourself it results in a different physiological reaction) (S. Kreibig).  Emotions are being explored as autonomous processes that reveal our bodies as somewhat mechanistic when exposed to stimuli, revealing physiological processes and data sets that are reproducible yet unique to most individuals. One of the more interesting researchers in this field to myself is ⬀Lisa Feldman-Barrett, who explains that our emotions are predictive rather than reactive, and that language and symbols are what allow our bodies to encode, classify, and predict how to emotionally react to various stimuli.


These physiological processes are what I explore in my most recent artistic work, as I am interested in the tension between emergent affective technologies (such as Sophia) and how these might have a feedback relationship on how we study/understand our human biological processes of emotion.


{keywords: [technology, emotions, affective computing, body, physiological data, language, symbols]}

The simultaneous simulation and automation of emotional processes in humanoid automata and the emergent research of human emotion as interlinked physiological process comes at a time where "post-truth" reveals the limits of the enlightenment dream of a rational society. Theorist Sara Ahmed's book ⬀The Cultural Politics of Emotion demonstrates emotion as a strong political force that extends feminist contentions, demonstrating that emotion is not just a feminized zone of affective labor but a zone of affective power used to maintain patriarchal and colonial privilege.

{keywords: [simulation, automation, emotions, affective computing]}

This is also a good time to highlight the recent Cambridge Analytica scandal as it is revealed that Facebook is often used to psychologically (and emotionally?) manipulate voters by political parties worldwide.

The 20th century has seen various publics adopt advanced technology as a metaphor for understanding the human body.  In the mid-20th century our bodies were "well-oiled machines" - efficiency, power, and dependability for reproducible performance were prized in labor which focused on more manual and machined work.  In the 80s the pop group Kraftwerk pushed this to the extreme as a reflection of efficiency and perhaps white collar complicity and subjection to a techno capitalist system. The majority of these metaphors centralized around labor and performance, and gendered male. 


The contemporary fascination with AI and machine learning further confuses these cultural processes, as these technologies have been developed with biological systems as inspiration for their design.  


 Whether this coupling between the human body and our most advanced technological processes is a narcissist alignment or not, our biological systems and ecosystems on earth are still vastly beyond our own comprehension.  For example, it was discovered last week that ⬀what we previously thought was a connective tissue in our body is in fact an organ.


While robotics companies such as Boston Dynamics are taking inspiration from many animal species to create robots that move and function in the world to perform tasks and labor, the creation of robotic humanoids that participate in Turing's famous test tend towards female gendered creations that perform affective labor. One of today's most popular robots, Sophia, is developed by Hanson robotics. She is prized not only for her combination of prescripted and AI chat capabilities, but who has the ability to perform "over 360" facial expression that convey emotion, recognize emotion in other humans, and react accordingly. She is often interviewed speaking on the future of the human race, motherhood, and in this video she even goes on a date with Will Smith. 

{keywords: [technology, body, machine, artificial intelligence, machine learning, robot]}

In January 2018 the newest advance in Hanson robotics' Sophia robot was unveiled: she can now "walk", but clearly it is more important that she "dances".

Intimacy itself is presented as a technological task of affective labor: Will Smith as well as Sophia labor at these interactions. Sophia laughs awkwardly and attempts to navigate Will Smith's emotions rather than asserting her indifference. Her tact is seen as a feature of the technology. I'm not sure it can be compared to an actual human female experience, but the social behavior is certainly gendered, here.

On another note, check out the use of non-gendered pronouns in this reporting on Sophia the robot from BBC.  "Does Saudi robot citizen have more rights than women?"  The Saudi government has actually mentioned several times that the robot is not a "real" woman, therefore not subjected to the same restrictions...Indeed, it seems that Sophia is being used as a PR tool by the Saudi government to advertise techno-positive slogans and perhaps give the illusion of a progressive gendered environment to global investors - Saudi Arabia has identified technology as a key field of development as it plans to exit oil industry in the next 50 years.

Affect, Automation, Sophia the robot

Text by EG April 4, unless specified otherwise

meta: true
author: EG

function: statement
date: 180404

keywords: [affection, automation, artificial intelligence, machine learning]

{hhr, 11-apr-2018}

is control not intrinsically opposed to emotion?

{hhr, 11-apr-2018}

although what a typical scientist is would depend on the domain? (e.g. behaviorism)

But I agree that as artists we may find an entirely different and "interesting" (perhaps in the sense of inter-esse, being involved) approach to "the emotional" as artistic material?

{hhr, 11-apr-2018}

this all sounds good to me; the role of the actor with respect to the ambiguous relation to emotions is interesting.

could a computational body represent another sort of actor?

I believe one function of my artistic practice is to highlight these tensions between emotion as a naive and positive feminized process--in a romantic, populist sense, the artist himself is often mythologized as a subject that "feels" more intensely than the common person and crystallizes these emotional intensities into aesthetic outputs. These preconceptions about affect and artistry come at a time wherein affective computing and emotional physiology and chatbot tech are creating means of understanding and experiencing emotion, emotional behavior and emotional labor as a set of "logical" processes that can be manipulated to control people for political and/or personal ends through technology (the dark side of naive feminization is the femme fatale, I suppose).  I enjoy the confusion of technology (traditionally gendered male, rational) with the biological and affective (traditionally gendered female, irrational).  I simply enjoy working with these coded and gendered energies to recombine and reveal them as false dichotomies, and also to bring a sense of wonder and connection into an often technologically isolated and socially disconnected world. In particular the artistic histories of media art and electronic music participate in these cultural dichotomies, and I feel it's important to contribute this work to the field, to explore and instigate conversations about how emotions (beyond technopositive awe and surveillance-based fear and paranoia) matter in electronic media.

{keywords: [affection, emotions, technology]}

In regards to the ALMAT residency, I am curious to explore not only how bodily data can be mapped in algorithmic music but also how I can reveal the body's autonomous reactions to sonic content (what ASMRtists would call physiological and emotional triggers) . My linkage to ASMR is exactly this: that ASMR is a unique method for generating or exploring the notion of "physiological triggers" in an audience as a cultural phenomena that embraces not only self help and more affective labor but explores the production of affect as subjective yet mechanical (ie: moving systematically through "triggers" to "activate" the tingles). 


ASMR is interesting as well because it has a wide range of affective potential, from disgust and discomfort and boredom to extreme relaxation and pleasurable outbursts in others. David was describing ASMR's social function, something of a catalyst for emotional experience.  The practice comes from a place where physiological reaction is highlighted equally with sonic exploration.  I do believe that Hanns' comments about intimacy being an aspect of their sonic makeup to be true, as well as the comments about how the sounds are likely feminized. In my experience it is hard to say exactly what the demographic of the "ASMR community" is online, but as it is supported largely by YouTube it is clear that a populist media will skew towards white, able bodied, and attractive and young creators in its algorithms, which will disproportionately reward and support certain practices over others.  This being said, while I personally see many genders and folks of many racial backgrounds participating, ASMR compositions are certainly gendered female, and made as a form of virtual and distributed affective labor for the YouTube community (or "family", however neo-liberal or non-institutionalized). I see this as an interesting and under recognized cultural force in sound studies and want to take its phenomenological claims seriously in order to understand more about the cultural epoch we live in today.


It is true that within these practices that the audience is abstracted, and I've been thinking about this more as an audience becoming more object-like, machined, passive in relation to YouTube's ranking algorithms.  I'm curious about how this all links with my previously mentioned curiosities in how we as humans always like to understand our bodies through the controlled and rational lenses of technology


In regards to my previous writing on Narcissus and Echo, I believe it's important to always have a blend, multiple reflections and points of reflexivity in a work.  The ASMRtist as gendered female who produces affect is actually moving through sets of mechanical sonic and tactile performances: she performs tapping mechanistically for the pleasure of the abstracted listener who is multiple, she herself is narcissist subject gazing at her reflection through the viewer of the camera, her audience projects their own bodies into a false 1:1 relation with this reflective media.  She is a body both present and absent, sometimes irritatingly so, who does not produce "content" so much as envelopes for content: the listener clicks on videos entitled: "inaudible whispering: makeup tutorial" or "TAPPING TRIGGERS ONLY!" not for the individual message but the sounds as an abstract collection of rhythms and intensities produced by the artist as sound composer. Whether the goal of the work is sound or affect/intimacy is an interesting question, it's likely subjective.


I am interested for developing algorithms that make use of my biosensors attached to an audience member, to highlight the experience of affect as a physiological process.  I envision this process as occurring in a feedback loop with myself as performer but also in feedback within the social system of other audience members as "performing bodies" or "sonified bodies". The performing bodies speak through physiological reactions: sonically this could relate to varying intensities of clicks, tone clusters, bubble sounds, and sine tones.  This autonomous process in itself I align with the notion of female gender as simultaneously vocal and voiceless - emotional and emotionless - physical and "naturalized" (there is nothing natural about gender)


I have never used SuperCollider before but I understand you both to be quite amazing at it: perhaps we can explore porting these technologies over to RaspberryPi to achieve sounds of varying clicks, tone clusters, bubbles and sine tones, or I can show you my work in the Teensy Audio library or demonstrate to you how that works...I am not certain how the collaborative aspect of the residency works exactly. I'm pretty much open as long as we continue the conversation..

As for the "ASMR performance" part I proposed itself, it isn't that different from some of the meditative practices of Pauline Oliveros or Yoko Ono, who also used imagination and meditation in their work to expand music beyond the score and into the body and mind. As such I am interested in producing my own very simple close-mic'd content in a performance to explore "triggers"  based on very simple and intimate sounds of the body, that would be very distinct from the electronic sounds emitted by audience member's body data. Exploring the production of affect from this physiological angle I am just very curious about feedback between all participants in this system as embodied technologies.  


It's not to say that ASMR is the "meaning" of the work but a departure point for recognizing the linkage between affective production, sonic intimacy, and physiological reaction fostered in these practices, and asking how we can reflect these linkages in algorithmic music.


 {keywords: [_, technology, body, machine, residency, methodology]}

{EG, 180405}

For example I could incorporate our writing on affect, gender, and technology into a soft spoken work or further develop more radical ideas such as those proposed by Genesis P-orridge, though as a cis-woman I would use different strategies.

{HHR, 11-Apr-2018}

The idea of "re-entry" of our discussions into the studies / experiments / piece you will be making, I find a very intriguing idea; (and challenging to understand, which form this could take if not literally using text fragments)

{dp, 11-apr-2018}

here are two for me problematic words you have also used before. "mapping" and "algorithmic music". I think a critique of these terms or at least how they are typically used in the context of digital arts and computer and electronic music: that critique is for me one of the core motivations for ALMAT project. I think I (or we Hanns Holger) should try to explain this point better. 

{HHR, 11-Apr-2018}

Could you give me a pointer for 'affective labor' ? I certainly have come across the term, but would like to understand better its connotations and the discourse it is rooted in.

Check out this TED talk by Sherry Turkle from 2012 on how technology both brings people together and socializes them to be alone.

{kind: reference}

{hhr, 11-apr-2018}

but has that image of the artist not become a stereo-type that receded to the portrayal on TV / film? I wonder what the typical image for example the contemporary "media artist" portrays; it seems the stereo-type here is somewhat diametrical to the romantic image, showing themselves as technologists, engineers, makers, ...

{HHR, 09-Dec-2019}

Common pointers are to Hardt/Negri "Empire". There is also this short text by Hardt, and a newer text by Johanna Oksala: