Plotbot   Science Fact-ion by Randall Hayes

Read more by Randall Hayes

All this mean I by Love, that my feeling
Astonishes with its wondrous working
So fiercely that when I on love do think
I know not well whether I float or sink.
-Geoffrey Chaucer, inventor of Valentine’s Day

There's an old BBC show about the history of science and technology called Connections, hosted by James Burke, which had a big influence on me during grad school. He would start by standing in a pasture somewhere and claiming something outrageous, like “Sheep led directly to computers.” Having defined the endpoints of his argument, he would then step through documented historical events, building a completely plausible narrative to support that argument, commenting on the individual events and the emerging pattern with a characteristically dry British wit. My roommates were just as enchanted as I was, and we made it a habit of meeting for supper and watching it together.

Some scientists really like this detailed, step-by-step understanding of the world, which I call “Rube Goldberg science,” after the famous cartoonist of the early 20^th century, who specialized in drawing complicated machines that performed simple tasks. There are molecular biologists who have spent decades working out one particular biochemical pathway or signaling cascade. Those diagrams often remind me (and others) of Rube Goldberg machines. One might ask why they need to be so complicated, and there are a couple of surprisingly straightforward answers:

1. the more steps there are, the more control points there are, and the more finely the level of the final product can be controlled (I would tell my music-obsessed students to think about a single volume knob vs. separate bass and treble knobs); and

2. speed. Each step in a cascade can serve as a multiplier, allowing the cell to make a lot of the final product really quickly. In engineering terms, three amplifiers with gain three, arranged in a row, are equivalent to a single amplifier with gain 27 (3x3x3), which might be a lot harder to build.

Neither of those reasons addresses the process by which biological networks are built, namely random mutation and environmental selection. The most common way to get a new enzyme is to duplicate and then mutate the gene that codes for it. That process is an important constraint on the evolving networks, but really a topic for another day.

Other scientists, constrained by current data and technology, or simply by the way their own imaginations work, prefer to think at a more abstract level. They might say, “Sheep to computers—right! That makes perfect sense!” and start working out the implications of that relationship. BF Skinner used the engineering phrase black box to describe this situation. For those personalities, as long as the input → output relationships work experimentally, who cares what's going on inside the box?

Fantasists are more often of this type than SF-writing engineers. They have long described social relationships between people as lines of light. Two random examples from my own reading would be Paul de Filippo’s “Lennon Spex,” where connections are color-coded by type and strength, and Jack Chalker’s Changewinds series, where the thin red line of a love spell connecting two of the characters could be followed like a scent trail by those with magical vision.

For a visual species like humans, this is a compelling metaphor, and a useful one. Connections can have different thicknesses, or weights, to use the term neural network researchers prefer. They can be one-way connections, as the emotional “ties” between characters in the Australian swashbuckling RPG Lace & Steel, or they can be two-way connections, the way most online social networks are currently described through graph theory. The most appropriate abstraction depends entirely on what specific behaviors the modeler is attempting to capture, and what will most easily allow comparison to what previous modelers have done.

One could attempt to break social connections down to every light ray that bounces off the partners' bodies, every ripple of sound that passes between them, every chemical messenger that forms a physical vapor trail connecting the skin of one partner to the mucous membranes inside the nose of the other. That would be difficult, time-consuming, and in many cases might be overkill. Correlating the observable behaviors of the two partners, Skinner-style, might be good enough in most cases. Again, Gottman's group can predict divorce with over 90% success from recording a few minutes of contentious interaction.

The reason I'm laying all this out is to introduce a pair of self-help books I have found very useful, both of which address classic science fiction/fantasy character dynamics. The first, Controlling People, lays out a fairly abstract model in which a Controller, alienated from his own emotions, behaviorally broadcasts imaginary constructs about how humans should act rather than receiving information about how humans really do act (the most common one I experience is random strangers telling me to smile). The author, Patricia Evans, calls these assaults on the psychic boundaries of others “backward connections.” Now, any and all human perception is a compromise between a predicted reality and the sensed reality, but the Controller does not revise his hypotheses about people in the face of contrary evidence. Instead he insists on the truth of his own view to the exclusion of all others. Vulnerable personalities, such as children or traumatized adults, can in turn be disconnected from their own emotions—from reality itself—and be replaced by the Controller's faulty model, which the author calls a Pretend Person. The Controller relies on this Pretend Person the way a child relies on a security blanket, or a stuffed animal.

"Teddy, like a real teddy bear, doesn’t leave, is as comforting as one can imagine, and could be male or female, child or adult, and could even be split into a number of imaginary people.”
-p95

It's a darker, almost parasitic version of the process I described in my very first column, where Douglas Hofstadter kept his wife “alive” in his head as a detailed mental model that he could interact with, consciously and conscientiously, after her death. Evans instead stresses that the Control Connection established in her model is unconscious, a childish defense mechanism against a painful reality. She describes these interactions at an abstract, fantasy/magic level, but there's now growing evidence from the hybrid field of neuropsychoanalysis that the bizarre delusions of certain brain-damaged patients, such as Capgras Syndrome, could fit this pattern. There are also many papers detailing models of schizophrenic delusions. We might eventually be able to work up wiring diagrams for delusion more generally. That would certainly please the Rube Goldbergs of the scientific community.

The other book, The Gaslight Effect, is more a practical guide to ending a Controller's reality-warping behaviors. Being male, I found it mildly annoying and distracting that, following the example of the film that inspired the name, practically all the book's examples of Controllers were also male. But my psychic boundaries are intact, guarded by my vigilant inner Fianna, and I'm pretty sure I'll get over it.

Randall Hayes, Ph.D., read Iron John during graduate school, and while he has never bought a drum, he did grow a crappy goatee. :^{| >

REFERENCES

http://www.imdb.com/title/tt0183790/

Kind of like GRR Martin’s “Hedge Knight,” but with a posse—including a drunken, lecherous Chaucer.

Chaucer citation train
https://en.wikipedia.org/wiki/Saint_Valentine#Saint_Valentine's_Day
https://en.wikipedia.org/wiki/Parlement_of_Foules
http://www.poetryintranslation.com/PITBR/English/Fowls.php

https://topdocumentaryfilms.com/james-burke-connections/

Burke went on to write a long-running column of the same name for Scientific American.

http://www.rube-goldberg.com

https://pdfs.semanticscholar.org/2694/f84262846b04dde3a3012cdb15af582b0cda.pdf

Lays out a “maximum intermediate steps” or “incompetent design” model of evolution, quite different from the normal assumption that efficiency rules. I may have to do a whole column about this at some point.

https://www.quora.com/What-does-weight-mean-in-terms-of-neural-networks

https://www.rpg.net/reviews/archive/classic/rev_3867.phtml

One of my favorites. More about games as writing tools in an upcoming column.

https://www.psychologytoday.com/articles/201509/the-einstein-love

http://www.slate.com/articles/double_x/doublex/2010/03/can_you_really_predict_the_success_of_a_marriage_in_15_minutes.html

Or maybe only 43%, depending on how you do the math. . .

https://www.psychologytoday.com/blog/life-art/200807/how-thin-are-your-boundaries

http://nautil.us/issue/19/illusions/how-your-brain-decides-without-you

http://www.intergalacticmedicineshow.com/cgi-bin/mag.cgi?do=columns&vol=randall_hayes&article=001

https://www.theatlantic.com/health/archive/2015/08/neuroscience-psychoanalysis-casey-schwartz-mind-fields/401999/

An excerpt of her book, which describes her experience in a hybrid master’s program, the first of its kind.

https://en.wikipedia.org/wiki/Capgras_delusion

https://www.frontiersin.org/articles/10.3389/fnhum.2016.00303/full

Stern, Robin (2007). The Gaslight Effect: How to Spot and Survive the Hidden Manipulation Others Use to Control Your Life. Harmony Books, New York, NY.

https://en.wikipedia.org/wiki/Fianna

Hey, I’m Irish, partly.

Bly, Robert (1990). Iron John: A Book about Men. 25^th anniversary edition by Da Capo Press, a member of the Perseus Books Group, Philadelphia, PA.