"The first and last thing required of genius is the love of truth."
-Johann Wolfgang von Goethe

Sunday, October 18, 2009

7. Sun's Day & Bus #51



Like an East-born extrovert with round face all alit on his Name Day, the Chicago Sun showed itself at all Times this morning. Paradoxically, it reminded me of my first night in town. I sat alone at a bar self-spoon-feeding ice-cream and watching the Yankees game when a friendly, forthright, forty-year-old blonde began to mock me. "How's your sundae?" she teased.

Shining . . . shining. 

Today's morning lecture, titled "Moving in an Uncertain World," was delivered by Daniel Wolpert, from the University of Cambridge. He applied Bayesian probability models to explain how the brain predicts and prescribes action. The inter-disciplinary nature of this conference continues to fascinate me. Creative scholars like Wolpert cross-pollinate multiple fields with potent data and information. Wolpert's research shows that the brain processes prior knowledge [e.g. memory] in addition to data [e.g. sensory input] in order to produce "beliefs" or, in mathematical language, "probabilities." Bayes's Rule is:

P(A|B) = P(B|A)P(A)
               __________
                     P(B)

P( ) means "probability of" and | means "given."

Then, in a deft, symbolic switcheroo, he plugged in the appropriate neuroscientific concepts, where:

A=state
B=sensory input

Now substituted:

P(state|sensory) = P(sensory|state)P(state)
                             ___________________
                                       P(sensory)

The expressions translate thusly:

The posterior of the equation, P(state|sensory), is the belief after, the brain's product after synthesizing information. P(sensory|state), the likelihood, represents the noise in our own senses, the information that makes the process unnecessarily complicated, I guess. P(state) is prior belief, the statistics of the task. In other words, if I are preparing to hit a tennis ball, my brain uses the information from every tennis ball that has ever been hit to me by that opponent. In predicting what will occur, Wolpert proves that we ought to shift towards the mean of prior, to compensate for the "noise" of likelihood. If a good tennis player usually hits the ball to the baseline, I ought to shift there despite the fact that the naked probability dictates that it will arrive towards the center of the court. "Noise" is a fundamental limit on performance. Don DeLillo's novel "White Noise" comes to mind. His protagonist leads a difficult life, while the television at times spews voluminous nonsense in the background. Good thing that is not the case today . . . ha! Again, good art intuitively understands human phenomena. The languages are just different. One is true, one is fiction. So what? Brian Boyd's argument, through which I am slowly making my way, convincingly asserts that art is an adaptive trait. I will get to that all once I finish the book.


Anyway, I tried to pay excellent attention to Wolpert's lecture, and it exhausted me. SO, I decided to skip some lectures [Bueller? Bueller?] and take a cab to the Art Institute of Chicago for A Sunday Afternoon on Michigan Avenue. After a day and a half of arid, acronymic description, I sought some rich artistic expression to revitalize my soul. Monet and Picasso's interpretations of the human being in the universe are just as valuable as scientific discovery; one cannot exist without the other. As an Adidas commercial tautologically proclaimed: "Only greatness equals greatness." Greatness understands it, with a paint brush or a microscope. Greatness transcends. Greatness is never forgotten? Not always . . .

Toward the end of my her ninety-five-year-long life, my beloved Bubbie developed dementia. It was a slow process; she began forgetting in her mid-eighties. But at that time, early on, there was one seemingly insignificant fact she always knew. When she traveled to Israel [in the early 1960s], she rode Bus #51 to her cousins's house. We were all amazed. #51, she said each time we asked. I chalked it up to her devotion to detail. But it was hard not to wonder: How could she remember this, but forget something monumental such as her husband's premature death?

Well, now I know why, because science does. Richard Morris explained in his lecture "Brain Systems of Learning and Memory." I am sure that only a handful of people in the whole auditorium understood the entire presentation. But I was able to hook on, most likely because of my own memory, to this one hold. It had long been thought that long-term memory of inconsequential events did not exist. But Morris challenged this with two undermining words: not always. OK, here goes my attempt at an explanation. In experiment, strongly-tetanized neural pathways in the presence of a low concentration of something called KN-93 fail to stabilize. In other words, "important" events are prevented from being encoded by the brain. When, immediately following this, a second pathway is weakly-tetanized, its receptors bind to proteins left-over from the failed first attempt. Morris used the word "paradoxical" to describe this. Just think: a failed memory of the love-of-your-life could contribute to the indelible image of an unromantic fact of daily-life. The former is lost forever while the latter lingers on. Because, after the receptors from the weakly-tetanized pathway bind to the first pathway's proteins, long-term potentiation [LTP] stabilizes. Memory is formed. My Bubbie's memory, now mine, of Bus #51 immediately struck me as a perfect example of this. Who knows what important information, forever lost, produced proteins that paved the way for the Bus. Cajal says that there is no primary and secondary in Nature; every link in the chain proves equally valuable. Morris's findings on memory support this egalitarian theory.

A former SUNday school teacher, my Bubbie was my first teacher, supervising penmanship drills when I was young. I am proud to say that I still pen elegant characters, even in the notes I wrote during a lecture that reminded me of her. I loved her and am thankful for my memories, all that remains from nearly a century of life. I will not soon forget her; she keeps springing spritely up off of my trampoline hippocampus.

Plus, her daughter is my mother and I am my mother's SON, who likes to play with words.

Thanks for reading.

2 comments:

  1. Richard Morris may have demonstrated "not always," but did he explain "why sometimes?" Why are some "important" events "prevented" from being encoded? What does it mean to say "prevented?" By what and for what reason? It doesn't seem to be the case that there is some ranking of importance with respect to the events themselves, so what is behind this selectivity?

    ReplyDelete
  2. He did explain "why sometimes," I just cannot. It is nearly impossible for a novice to grasp. It has to do with KN-93, an inhibitor [there are two types of postsynaptic potentials: E(xcitatory)PSPs and I(nhibitory)PSPs. Coming off of an initial synapse, they determines the nature of the next cellular process, because they cause the cell membrane to either depolarize or hyperpolarize]. Those basics are all I know. I am not even sure they are relevant. I can't answer your specific questions. Maybe investigate the experimental procedure "strongly-tetanized" vs. "weakly-tetanized" pathways? That seems to be the essential difference between the two memory processes.

    I wish I could do better.

    ReplyDelete