Tuesday, August 13, 2013

(Hypothetical) Crimes Against Neural Art

We would expect that if there was any moral outrage to have over the treatment of cultured neural tissue, it would occur in an art gallery. Something about an art gallery sensitizes us to the well-being of critters we might not usually care about - as in the case of Garnet Hertz's Cockroach Controlled Mobile Robot (a three wheeled robot about half the size of R2D2, driven by a Madagascar hissing cockroach) - and to cry out over events that we might otherwise willfully ignore or even accept as routine - as in Guillermo Vargas's infamous “You Are What You Read,” (where a starving dog was taken off the street and brought into a gallery) [1].  Instead, when neural tissue is given a robotic body and placed on display (sometimes remotely) in an art gallery, most responses seem to focus on the ambiguous nature of the works.  Artist Stephane Dumas wrote, referring to MEArt (a drawing robot controlled by a culture of rat brain cells), that “the public can experience the drawing activity and at the same time sense the presence of its remote initiator, the brain [2],” implying a felt mental presence associated with the biological components of the work.  However, Dr. Stuart Bunt, one of the scientists who worked on Fish and Chips (a precursor to MEArt that used tissue taken from fish rather than rats), wrote that “many viewers of Fish and Chips embodied it with impossible sentience and feared it unnecessarily [3],” indicating that the attributed mental life (and implied moral obligations towards it) was an illusion constructed by the framing of the piece. This contradiction between the audience and creator's interpretation of these pieces is reflected in Dumas's assertion that embodied neural bioart (here referring to Silent Barrage, which featured a distributed robotic body that audience members could walk through) “is a work in progress that raises more questions about the relationship between neural mechanisms and creative consciousness than it answers [2].”  This ambiguity is even praised by artist Paul Vanouse, who states that “MEArt's creators have cleverly designed their thought-provoking apparatus to maximize cognitive dissonance [4],” while Emma McRae describes MEArt as an example of one of “an infinite multiplicity of agencies [5]” that don't fit into well established categories, which  humans must learn to share the world with [6].

If even cockroaches become objects of empathy in an art gallery, what would it take for us to feel sorry for neural culture?  Above photos by Sharmanka and Douglas Repetto, from here.

If Fish and Chips, MEArt, and Silent Barrage didn't raise any overt ethical alarms, what would be required for such an embodied neural artwork to be 'wronged'?  While moral transgressions are certainly possible on a multitude of grounds (e.g. affronts to the dignity of the cultures, or more likely the dignity of the animals they were derived from), for the moment let us narrow ourselves down to the possibility of causing morally relevant pain in such a system.  Previously on this blog I've discussed several different ways of looking at the possibility of pain embodied neural cultures.  Here I'd like to present my own hypotheses for what might be the minimal requirements for creating morally relevant pain under these different perspectives.

Both behavioral and anatomical perspectives on pain have serious problems with identifying the presence of morally relevant pain in neural culture.  This is as these perspectives both require the existence of reference points that both clearly demonstrate pain, as well as defining that pain based on qualities can be shared with neural culture.  From the behavioral perspective, such pain might seem possible if one created a robotic body that constantly whimpered or otherwise generated pain-associated behaviors.  However, we wouldn't trust that system to have any sort of 'authentic' pain; the signal produced could just as easily be giggling, and there wouldn't be any change from the perspective of the culture.  The relationship between bodily activity and morally relevant mental states, which while not fixed does possess evolved biological structure in 'full' animals, is arbitrary in the case of embodied neural culture.  From the anatomical perspective, pain also might seem possible at first.  Works like Silent Barrage and MEArt share some cellular similarities to some of the most morally relevant parts of the pain system - the cortical regions that appear necessary for caring about pain.  However, these isolated neural tissues lack the connections to other parts of the brain and body that usually interpret what the cortex does in such a way that pain is produced.

What would it take to 'wrong' a culture of dissociated rat neurons?  Image by Dr. Steve Potter, from here.
 A mathematical perspective on pain gets around the reference issues of the behavioral and anatomical perspectives by holding that the internal structure of a system determines its ability to feel different states, including pain.  Thus, this perspective can be applied to any system, whether naturally or artificially constructed - the only thing that matters are the (mathematical) rules that govern how the system changes over time.  Such a definition hasn't been built yet, but Integrated Information Theory of Consciousness seems poised to construct such a definition.  In the mean time, we might hypothesize that such a description of pain would include essential qualities such as aversion, negative reinforcement, redirection of attention, and sensitization to other 'painful' stimuli.  The described mathematical structure would need to be rich enough that it was convincingly authentic, such that even if the system under investigation was 'wired up incorrectly' - as in the case of the neural culture that could just as easily be made to laugh as to cry - there would be some latent structure present in the culture's actions that was clearly identifiable as pain.

Lastly, we might trade these views on moral pain that focus on the neural culture in isolation for a perspective that focuses on the interactions between the neural culture and its environment - what I have previously termed a 'social' perspective.  This perspective looks for grounds on which audience members might interpret the activity of a neural culture the same way as they might interpret morally relevant pain in animals - as a signal that the audience is obliged in some way to help the culture deal with its imminent destruction, whether real or perceived.  As the neural culture is not any living animal that the audience is familiar with, the social perspective does not attempt to interpret the neural culture as such. (Though certainly, such similarity if it did exist would be reasonable grounds for a different interpretation.  We have the option of treating a slice of ACC as if it still existed with the rest of a rat, just as we have the option of treating a deposed dictator with the same respect they commanded while in power.)  It is useful to note here that by treating the neural culture as a sign open for interpretation, the social view actually encompasses the other views examined above, each as methods of interpretation in their own right, and perhaps appropriate in their own sets of situations.

A torn image created by MEArt.  In early shows the control system for the robotic arms was still being developed.  Was the destruction of one of MEArt's products, the drawing, a moral failing of the artists who created the piece?  Could such destruction be considered analogous to the painful experience of the destruction of one's own body?  Image from here
Without an evolutionary 'narrative' to tell us how to behave (as we do when interpreting the facial expressions and vocalizations of social mammals, for instance), we are left up to the artists and scientists who created the work to provide some sort of moral structure.  From this perspective, pain might be something as simple as a LED that the culture could trigger if the pH of its media deviated to far from homeostatic limits.  All the necessary components are there - the audience can receive the signal, the signal signifies the pending doom of the culture, and the signal is in some way generated by the culture itself (that is, the signal would stop if the culture actually did die - as might be the case if we 'euthanize' the culture).  However, while there is enough structure in this scenario for some audience members to interpret the signal as pain, the lack of interaction between the audience and the culture makes this pain seem flat and robotic.  A more 'authentic' pain might require a richer relationship between the audience and the culture.  For instance, perhaps the audience has the ability to feed the culture and thus 'relieve' its pain (by bringing the media back to a safe pH limit) - or at least a way to request the researchers to feed the culture.  Such a moral relationship could be strengthened by adding repercussions for the audience's actions [6] - the culture could respond to feeding with a neutral sound that was described by the creators as a 'thank you,' the culture could blink its LED with greater frequency if it found that 'polite' requests for feeding were ignored, or human protesters could be staged outside of the gallery crying for 'tissue culture liberation' and preaching that only a philosopher could ignore pain in neural culture.  The key bit here isn't so much the qualities of the culture itself, but the interactions between the culture and the audience that might generate a truly moral pain.

Want to cite this post?

Zeller-Townson, RT. (2013). (Hypothetical) Crimes Against Neural Art. The Neuroethics Blog. Retrieved on , from http://www.theneuroethicsblog.com/2013/07/hypothetical-crimes-against-neural-art.html

[1] Other examples of this trend include  Gregor Schneider's “Death Room” (a room intended specifically for someone to die in, with plans to allow a volunteer to pass away publicly, of natural causes), or Huang Yong Ping's “Theatre of the World” (where a variety of reptiles, amphibians, insects, and arachnids were allowed to hunt and eat one another).  Perhaps in an art context, with its overtones of control and even frivolity, prevents us from excusing events such as starvation and natural death as being unavoidable, and predation as being necessary for some greater good.

[2] Dumas, Stephane. "Creation as Secretion. An externalist model in esthetics."  Situated aesthetics: art beyond the skin. Ed. Riccardo Manzotti.  Imprint Academic, 2011.

[3] Bunt, Stuart.  "Cybernetics and the Interaction Between Pure and Applied Sciences and the Humanties."  Proceedings of The 17th International Symposium on Electronic Art, Istanbul, September 14-21 2011. retrieved from <http://isea2011.sabanciuniv.edu/paper/cybernetics-and-interaction-between-pure-and-applied-sciences-and-humanities> on July 19th, 2013

[4] Vanouse, Paul. "Contemplating MEArt- the semi-living artist"  retreived from <http://www.paulvanouse.com/MEART_PV_essay.pdf> on July 19th, 2013

[5] McRae, Emma. "A report on the practice of SymbioticA Research Group in their creation of MEART-the semi living artist." retrieved from <http://www.fishandchips.uwa.edu.au/project/emma_text.pdf> on July 19th, 2013

[6] McRae does go on to point out how such works, performed in a university setting and using animal tissue, required ethical approval by animal ethics committees/Institutional animal care and use committees.  She notes that such committees often focus on the scientific merits of these works to justify them, even refusing to comment on works where the value is primarily artistic.  These committees are primarily concerned with the use of the animals from which these tissues will be derived, however, so they don't reflect ethical concerns with neural cultures themselves so much as the processes that produce them.  

No comments: