About categorizing and interpreting physiological data [24-04-2018]


[DP] Categorization is a deeply incarnated way of thinking in our western society, that of dividing everything into parts, so that we can build complex things out of simpler units. This is a common approach in many fields, but I believe it's also interesting to think in other directions: maybe you don’t need to categorize the emotional changes. You know that there is a strong correlation between physiological changes and emotional state, isn’t it already a lot? If you gather people in a concert hall or performance space, these people usually have already a sort of expectation about what’s going to happen. But still you know that if you get the people to wear your sensors, or if you find any other way of 'sensing' them, you can observe their physiological variations, and how these change according to your actions. This is great, however you interpret it. It could be interesting to stay longer at this abstract level, instead of breaking these data down immediately to some categories. You have these data, you can use them as abstract material in very interesting ways, and at the same time you know that whatever you'll do will influence their responses. You are looking at a feedback loop where the microphones are the people, you are sensing the environment through the people. I believe in the moment you break some material down into categories, you lose something important.


[EG] The question for me is also that of stimulating the audience. I’m interested in building the shape of the music by stimulating the audience. It’s really difficult to think how you could stimulate emotional states without at least sort of going into these categories. 


[DP] But do you stimulate directly emotional states? I would say you are stimulating the perception of the people in order that they get into some emotional state, is that correct? Or can you directly stimulate an emotional state?


[EG] Well, I could for example scream into my own microphone and just make everyone really uncomfortable. Maybe that discomfort doesn’t have a valence, but I know that this is probably not going to bliss everyone out and relax them. And so it is something I need to be aware of obviously, there needs to be some form of understanding of these principles: if not a categorization, there has to be at least something.


[HHR] It occurs to me that, if you go into this kind of orthogonal space, what might happen physiologically from your 'microphones' is that you get a 'signal of attentive listening', or something similar. I’m now thinking that probably is going to be flattening out to a large degree if you go away from this kind of attempt to at least push people to some extremes. And then the question is how could you push people to these extremes, in a way which is not so iconic and that can account for the complexity in different responses of people. You need a strategy to figure out how to activate these kind of physiological responses, maybe more abstract than trying to refer to some specific states. But still, how to activate these physiological responses? It is a difficult question, because it implies experimentation and experimentation in this sense assumes a sort of reproducibility, because if you develop it with some probation or test persons, you may find some strategies to trigger specific things by playing a particular combination or musical structure. That may or may not produce the same response in somebody else in the audience. 


[EG] Emotion is so difficult to categorize. Our reactions to stimuli are just so all over the board in a way you can’t predict what people are going to do. For this iteration I am interested in the situation where I can’t predict what people will do: part of the structure of the work reside in this provocation, and it would be nice for the provocation to be able to be open-ended somehow, but I think that it’s going to be generally on one side or the other side of positive or negative valence. I think that categorization happens even in the heads of the audience, but if it was left kind of abstract, as to what you were supposed to be feeling, it could work.

meta: true
persons: [HHR, POZ, DP, EG]
place: experimental studio
date: 180424
kind: conversation
origin: conversation transcript
keywords: [categorisation, emotions, physiological data, audience]