HI Don,
The distinction between aesthetic and emotion in your post is pretty clear to me, but, being one of those thick ignorant you mentioned in your post, I was wondering: would the picture of a abnormously  anus, showing the first part of a man's colon trigger visceral responses to a specialist of colon medicine? I suspect he would have the same reaction I had when I was studying architecture and I was tremendously attracted by the bolt of a metal bridge (I remember watching all of them, one by one, on the Sydney arbour bridge, and my wife could not agree with my pleasure). So, if the colon specialist is not disgusted by the same picture could we say that his context is reducing or just changing his visceral response to a picture like that? My personal objection to the software, besides the jokes is that it was very unclear to me what criteria where used for the "unbiased assessment of the aesthetic quality" a scientific approach to that software should make those criteria very clear in the first place. But I am ignorant, I may be wrong...


Nicola Morelli, PhD
Associate Professor, School of Architecture and Design
Aalborg University, Aalborg, Denmark
http://servicedesign.wikispaces.com/
Blog http://nicomorelli.wordpress.com/
skype: nicomorelli


-----Original Message-----
From: PhD-Design - This list is for discussion of PhD studies and related research in Design [mailto:[log in to unmask]] On Behalf Of Don Norman
Sent: 16. maj 2009 19:23
To: [log in to unmask]
Subject: Emotional Theory Re: Online judgement of aesthetics

I am dismayed by the lack of knowledge of Emotion displayed in a lot of the posts about the computer program ACQUINE.

Science knows a lot about emotion, and all these posts seem completely ignorant of the tremendous amount of progress that has been made in the last decade. There are respectable studies of aesthetics and emotion, and ignoring them makes analyses of this ACQUINE program rather naïve.

Emotion is a very complex topic. Aesthetics (which is not the same thing, of course, but related) is even more complex. 

The Penn State group is a typical group of know-it-all engineers who think that they can solve any problem. They are completely unaware of the complexities of emotion and they seem to equate aesthetics and emotion. They also call some things "emotions" that no self-respecting emotion theorist would ever do. After reading their papers it is clear they do not understand emotion, let alone aesthetics.  Nonetheless, what they are attempting is actually pretty reasonable.

This program seeks simply to mimic human responses to the surface features of an image. The program knows nothing. Personally, this is a project doomed to failure, but there are components that might actually work in the hands of more skilled investigators.   But remember my words: it does not actually judge aesthetics or emotion: it tries to give the same responses to images that people give.

So there is a lot to criticize. But the basic idea is sound. Today there are many forms of pattern recognition systems (broadly defined) that use learning algorithms, Bayesian Networks, hidden Markov markers, or neural networks that when shown a large number of exemplars can set up their internal weights and structures so as to do an excellent job of mimicking human judgment.   This program fits there. Much of this work is very good and very valuable.

Note that this is why placement of an image matters. But then again, artists have always thought that placement matters, so the complaint that a person's face gets different ratings when seen large versus small is not legitimate: it should get different ratings. The judgement is not about the face -- it is about the total image.

I violently disagree with my good friend Chris Rust who said: "affect, our visceral response to an experience, is dependent on a context which is constantly changing and subject to an enormous number of factors."

Gee Chris, I wrote a book plus several peer-reviewed technical papers, some jointly  with a world-authority on emotion, to argue that Visceral responses were NOT context dependent and were pretty universal across the world.  

The catch here is the word "visceral."  Chris was using it in a very loose sense. I am using it in a very precise, scientific sense. In fact, in my design works (the book Emotional Design" I use the terms Visceral, Behavioral, and Reflective to describe three very different levels of emotional responses. In my scientific work, I (and my co-authors) use the more traditional labels of Reactive, routine, and reflective for these very same three levels..)

In a paper that Andrew Ortony (a well known emotion researcher) and I presented at the now-defunct  Ivrea Interactions Design School, we argued that Visceral responses were solely dependent upon perceptual factors. Relatively context free.  

We said that Visceral (reactive) affects are subconscious and are driven by innate, biological systems that lead to fear of heights, darkness, crowds. To dislike of bitter tastes and dissonant sounds. To dislike of scolding sounds. And to the like of warm, medium-populated spaces, to medium levels of light and sound, to praising voices and sweet tastes. Low-level anxiety or relaxation. Disgust fits here. These innate responses are pretty context independent. They can be modified slowly by classical condition to sensitize or desensitize them. 

Behavioral (routine) affective responses are also sub-conscious and consist of such things as hope and fear: these are all expectation based emotions. 

Reflective levels are conscious, highly learned and culturally sensitive. These can change with time. These are very context dependent.  Here is where blame (agency) is assigned (guilt, blame, pride, praise).

This computer program, if it does anything at all, looks at pure perceptual features: mostly visceral.  But it also is sensitive to placement of the object )the authors talk of common photographic rules such as the rule of thirds.

=================================

It is a lot of fun to joke about these things. But if we want to make progress as a respectable, academic field, we need to be more careful. We need to respect scientific knowledge, even when discussing art. We need to interpret computer programs such as this one according to its real goals

It is true that the authors are incredibly naïve and have the typical hubris of engineers. I like to tell them that Engineers go boldly where scientists fear to tread.   They use stupid descriptions of their goals.

But the work itself is solid.  Someday, research of this sort, will be very useful

So remember, emotions are complex. To understand them properly means to realize that the many different forms of emotional (affective) response have different brain structures, different causes, and require different analyses.  You can't lump them together.

As for aesthetics: that's even more messy. I believe that we should probably do the same with aesthetics. I joke that when a professional artist or designer sees something where their immediate, subconscious response (visceral!) is of liking, they will then consciously immediately denounce it as shallow and unworthy of consideration. "That's the sort of thing Alessi would do."  What this means is that the automatic visceral response says "good." The deeper reflective responses says, if I like it so easily, it is obviously shallow, unsophisticated, and devoid of content.  

Yup. T'is true. Visceral responses are indeed surface ones. Reflective ones are deep and reasoned. So don't confuse one with the other.

Relevant Papers

Norman, D. A., & Ortony, A. (2006). Designers and users: Two perspectives on emotion and design. In S. Bagnara & G. Crampton-Smith (Eds.), Theories and Practice in Interaction Design (pp. 91-103). Mahwah, NJ: Lawrence Erlbaum Associates. 

Norman, D. A., Ortony, A., & Russell, D. M. (2003). Affect and machine design: Lessons for the development of autonomous machines. IBM Systems Journal, 42 (1), 38-44. http://www.research.ibm.com/journal/sj/421/norman.pdf

Ortony, A., Clore, G. L., & Collins, A. (1988). The cognitive structure of emotions. Cambridge (UK): Cambridge University Press.
 
Ortony, A., Norman, D. A., & Revelle, W. (2005). The role of affect and proto-affect in effective functioning. In J.-M. Fellous & M. A. Arbib (Eds.), Who Needs Emotions? The Brain Meets the Robot (pp. 173-202). New York: Oxford University Press.

Respectfully

Don 



Don Norman
Nielsen Norman Group
Breed Professor of Design, Northwestern University
Visiting Distinguished Professor. KAIST, Daejeon, Korea
[log in to unmask]
www.jnd.org/