Returned from Munich now write!

That that I am back after my weeklong stint in Munich for my Birthday (yes, yesterday!) I now back at the office and need to get my head around a few ideas that have been on my mind.

Firstly its that of the role of anthroporphism in my project and more specifically in the synthetic psychology (SP). Surely SP is an anthropomorphic view of robotics and dynamic systems… Braitenberg uses psychology terminology to describe the behaviour of his vehicles.

The question is whether this is common in ethology and other fields of study of behaviour. The alife connection to ethology is strong, but in traditional robotics attributing notions of “feelings”, “taste”, “experience” etc. are surely less highly regarded (correct me if I am wrong).

The example that came to mind regarding my project is a comparison between a human and a dog. If a human is disgusted, the common facial expression is that we pinch our eyes and close our mouth as to not let any of discusting substance enter our body. We have come to associate different “practical” reactions to stimuli such as fear, anger, disgust, love boredom etc. and take them for granted. Similarly we have learnt to understand the body language of some of the animals closest to us such as dogs, cats and horses.

The interesting bit is where we attribute an emotional state to a specific behaviour. A dog wagging his tail is more likely to be excited or happy, while pinching his tail between its legs is a sure sign of discontent or fear. Dogs don’t have the same facial expressions as we do, yet we attribute our emotions to their behavior. This behaviour is not similar to the behaviour exhibited by humans, it is grounded in the personal experience that this species of creature has accumilated throughout its evolution. Of course, and we know this from humans especially, imitation plays a role – such that some reactions have been abstracted over time and have become less connected to the original reaction (Im guessing laughter is one here).

The crux with this is that if we were to apply the human reaction to an emotional stimuli, such as an angry or happy face, to a robot, we would be making the mistake of failing to “ground” that matching reaction to situation in the previous experience of the “species”. Thus a reaction would not be specific to that species and seem uncanny.

We shouldn’t attribute to much HUMAN emotion to our creatures, instead we should strive to find new emotions from the CREATURES perspective. We can still call them love, hate, disgust, joy or whatever afterwards…

Previous Research Blog Posts

This is my research blog. I started my PhD at the University of Brighton in October 2007.

Topics in my project include videogames, interaction-gameplay/flow, artificial life and animation.

18. March 2008

Finally getting to grips with honing in on my thesis title/question. The problem was the wide
range of fields that effect my topic – ranging from evolution/biology over cognitive psychology
to AI and computer graphics.

I am currently reading papers by Steve Grand (developed
“Creatures” game series), Dave Cliff (Uo Bristol, Uo Southhampton)
and Pattie Maes (MIT Media Lab) on integrating Alife agents into
interactive entertainment.

I am also still keeping Valentino Braitenberg’s (Max-Planck Institute Thueringen) Vehicle designs in mind.
A summary of Alife papers (more on the definition of the field) collected by Margaret Boden (Uo Sussex) provides a
sound background.

I have found several papers about emotional modelling, interactive narrative (novel gameplay) and actor performance
by Michael Mateas (Uo California) and Andrews Stern (Creator of the “dogz” desktop
pet), who both made large strides in terms of emotional agents, even before their game Facade.