Per Wikipedia, Walter Murch "is an American film editor and sound designer. With a career stretching back to 1969, including work on Apocalypse Now, The Godfather I, II, and III, American Graffiti, The Conversation, and The English Patient, with three Academy Award wins (from nine nominations: six for picture editing and three for sound mixing), he has been referred to as "the most respected film editor and sound designer in the modern cinema."" He is also an amateur astrophysicist, and this book is the story of him trying to get professional astrophysicists interested in revisiting the mostly discredited 18th century Titius-Bode's Law, which predicts the positions of planets' orbits in the solar system. Bode's Law largely was assigned to the dustbin of history when Neptune was discovered in an orbit completely at odds with the law.
Murch seems like a very bright and creative guy. He helped develop the now ubiquitous Dolby 5.1 sound system with George Lucas. He carries a knapsack filled with "fortune cookie fortunes". 9 are given, this was my favorite:
Music is the pleasure the human soul experiences from counting without being aware that it is counting. — GOTTFRIED LEIBNIZThere is a a lot of interesting material on ancient ideas including the Music of the Spheres. I had never heard of the quadrivium.
From Pythagorean antiquity through the Middle Ages and well into the Renaissance, all learned gentlemen (and they were all mainly gentlemen) were steeped in the fourfold classical curriculum known as the quadrivium, which is to say, arithmetic (pure number), geometry (number in space), music (number in time), and astronomy (number in space and time).I liked this new word: apophenia - "“The widespread tendency of human beings to see patterns where there are no patterns.”" Ah, Wikipedia says our old friend Confirmation Bias "is a variation of apophenia", and lists another old friend, pareidolia, as its 1st example of apophenia.
There has always been opposition to the acceptance of Bode's Law. Carl Friederich Gauss, the 18th century "Mozart of Mathematics", raised 3 objections, the last 2 being
Gauss observed that, furthermore, Titius and Bode were dealing with just too few “planets” (eight) from which to derive a reliable law— any chance arrangement of a small number of objects (coins tossed on a table, birds singing on a tree branch), he argued, could be accounted for if you allowed yourself, as Bode and Titius had, a sufficient number of arbitrary constants.But Murch still forges ahead. He feels like his expertise with sound gave him insight into something that maybe was a resonance-based phenomenon, like musical harmony is. He applies Bode's law to Jupiter and Saturn's moons, and to some of the data on the new non-solar planetary systems being discovered by observatories like the Kepler satellite.Finally, there didn’t appear to be any physical explanation for why this clunky formula might be playing out in the actual physical world.
But, there are so many exceptions that the acrobatics required to try to make things work reminded me of the epicycles of earth-centric theories of the solar system. Munch theorizes the existence of some form of standing wave around gravitational bodies. This is evocative, but as 1 of the astrophysicists who reluctantly engages with Munch points out, if there were such standing waves in the metric, they would have to throw off all satellite communications, both nearby such as GPS, and far away, such as our spacecraft exploring the outer solar system. But then Munch finds an article saying that there are some discrepancies in GPS signals, so off we go again.
Several of the astrophysicists with whom Munch does succeed in engaging seem sympathetic, but, in the end, rightfully dismissive. Here is the conclusion of Lee Smolin of the Perimeter Institute of Toronto, who seemed to be very sympathetic to Munch and his efforts.
My sense is that Walter has done just about as well as any lay person could do, without the tools and discipline instilled by succeeding in a Ph.D program. The main thing a professional life in science teaches is that almost every idea is wrong. [my bold] All of us who work in science have seen most or all of our cherished, beautiful ideas fail. This is the tragic element of a life in science. Few are immune. What a life in science teaches one is that science is really, really, really hard to do right.Note that Smolin has been critical of science becoming too inbred. In 2006 he published a book titled "The Trouble With Physics", which was critical of string theory. Weschler references these objections, and recalls how at the start of the Enlightenment, a lot of scientists were amateurs.The main mistake Walter is making is to stay in love with his first idea. Only when you have seen your first 20 or 50 ideas die do you begin to appreciate what it means to have a good scientific idea. This is why laypeople like Walter never succeed in contributing to science— they cannot give up their first ideas.
The second mistake Walter is making is to underestimate the likelihood that a flexible theory can be invented and adjusted to fit random data, even when it is based on wrong ideas. This, unfortunately, happens all the time in medical science. If one has 100 factors to try to correlate with a disease, and one requires 95% probability for a match to be taken as a result, then at least five of them will appear to fit, just by chance. The literature of science and medicine is full of ideas that had some success at a 95% or even 99% level, but on further examination were found to be wrong.
Every year we see “discoveries” of new particles in experiments at the 95% or 99% level, which is to say better than Bode’s law does, which go away when larger data sets are taken. This is why the standard for discovery in particle physics is five sigma (one chance in 3 million of arising by chance). Even so, last year there was a five sigma “discovery” in cosmology that got lots of attention before it was shown that the effect could be explained by reflection off of dust in our own galaxy.
Bode’s law has if I understand right three free parameters. There were initially six known planets. The right question to ask is how likely is it someone could have invented a rule with three free parameters that fit six numbers. The answer is that given that there are a vast number of simple patterns with three parameters, the probability that something like Bode’s law could have been found to match random planetary orbits is close to one. And given Neptune, and the 50% success rate with some systems and the 5% success rate with other systems, the law has done about as poorly as could be expected were it an accidental fit, fine tuned to a small data set.
When a hypothesis works in half the new cases, or when it has a five percent success rate against a new data set, the right conclusion to draw is that the hypothesis is wrong. We must throw such ideas away if science is to be a source of reliable knowledge, and progress.
But, that was 250 years ago, we have moved beyond that now. Towards the end of the book, I found myself getting a little annoyed with the degree to which Weschler seems to have drunk Munch's kool-aid.
This brings me to the title of this post. Maybe we need something new to call what Munch is doing. It's not science - he simply doesn't have the training or math. But it is interesting and pretty. Who doesn't love a good pattern, even if there is nothing particularly fundamental underlying it?
So, maybe "science-art"? Googling "science as art" brings up shows and competitions for the prettiest pictures generated by science via microscopes, etc. "Science-fiction" and "science-fantasy" are already spoken for. There must be a good word for this. Maybe Aimee Mann can inspire us ...
A quick and enjoyable read, with several FFTKAT. Thanks Erica!
