Wednesday, July 27, 2011

communication and the end of science in the States

I like reading. I read a lot. Recently I have been on a fiction reading jag, John D. MacDonald, Rex Stout, Lawrence Block, J.K. Rowling, Jack Higgins, Michael Crichton, Tom Clancy.

Since the Mid-1990's I focused on reading papers I was downloading from *.edu websites. Thesis and published papers that were inside of faculty members campus computer storage. Over the years I have watched the security change so that it has become more difficult to reach published papers without accessing through specific libraries or campus computer security systems. You can still find a lot of papers out there though.

There are a couple of things in common between fact based papers and fiction, one is the attempt to communicate an idea or group of ideas and the other is subjective and often inaccurate assumptions.

I worked for 6 years with a company that was developing a manufacturing process. The theoreticians pretty much ruled the roost and didn't understand the variability in the application. I would have arguments and be told things like "amplitude is constant" quite loudly and with very little respect for my expertise. After a year or two I would prove that process variability had a range. For example I claimed we were losing amplitude under force. It took about 18 months to fund the testing and prove my theory.

What I loved about that was once we discovered I was right and amplitude decreased as force increased people came up with a really neat linear equation to predict amplitude loss.

Just so people get this, every time you place force on an object you are essentially creating a spring. Grab a broom stick and put it between two chairs. Push down in the middle. The more force the more the stick will bend, BUT, as force increases the displacement per an amount of force will decrease. The first ten pounds of force will bend the broomstick more than the second ten pounds of force. This is just the way stuff works.

Now if you have ever graphed anything like this you know that the curve cannot be linear, it has to be curved.

It becomes even better, when I finally showed the variability in amplitude was skewed to the low side of the range some moron with a masters pulled out a statistics book which said "all distributions are normal distributions". If you know statistics you are probably laughing right now. It's no joke.

These are the kinds of assumptions that people writing papers make and the hard part in understanding them is identifying the assumptions.

One of Einstein's most famous incorrect assumptions was that the sun is a sphere. Duh, a plasma ball spinning deforms into a flattened out ellipsoid. If Einstein had thought about it he understood the issue and being unable to define the ellipsoid he reverted to a well defined sphere. When scientists could define the ellipsoid they did and discovered that the modification changed Einstein's results only slightly and not enough to invalidate the conclusions.

Did Einstein know his assumption was incorrect? Maybe. Maybe he didn't address the assumption because he didn't want to deal with whinny morons. Maybe Einstein ran the numbers with some various ellipsoids and while he couldn't define the ellipsoid he knew that identifying the sun as an unidentified ellipsoid would confuse the issue and create wasted argument. Maybe.

In reality a lot of scientists make assumptions where they know their data is not "exactly" correct, like the spring function,and they publish anyway. A shallow curve can be defined as a line segment and within a specific range of the curve a linear equation will closely approximate the curve. Yeah, tech speak for estimate.

The thing is, the author has to run the numbers enough to understand the curve. Was that done? Did Einstein do it? I know authors I have worked with have not done the homework behind the estimates and have made invalid assumptions.

When I read a paper I identify the assumptions, often writing them out as I read. Then I figure out where the assumptions came from.

The shear and tensile strength of a material is given as a number. This is an assumption or estimate based on statistical data from testing of materials within the standards. In reality the material breaking strength is a curve with a mean, a modality, a median and a standard deviation. If all four of these numbers are not available the data is an estimate.

People do not publish these four numbers in tensile strength testing data so all tensile strength numbers are estimates or assumptions.

Pretty crazy huh? Communication is beginning to seem like a joke even when dealing with "factual scientific papers" or "factual scientific data".

Of course what is happening is people are making simplifying assumptions about the data so that not every calculation is a probability equation based on other probabilities.

So what happens? The simplified assumption becomes invalid data. I wrote a recent blog about how Lawrence Block's character was whining about Batman killing people and how the original Batman had killed people. Ignorant assumptions based on insufficient data always make me laugh.

So some people take the em-PHA-sis (emphasis or Em-pha-sis) off the data and place it on the spelling, grammar and other assorted crap.

When people complain about spelling and grammar they are changing the subject from the data to the methodology or form of communication.

I like talking to and reading papers by people whose first language is not English. Typically people whose first language is not English are more interested in the content of the communication. People whose first language is not English often work very hard to understand regardless of the form of communication.

That is why people of other nations will surpass the United States in technology as long as the primary language of science is not their first language, they strive for understanding of the content while morons in the States focus on form.

You see this everywhere in the culture of the US, "You can't just do the job, you have to do it with style" and so people in the States ignore the work if it isn't stylish enough, pretty enough, "kewl" enough.

Just like Lawrence Block communicated his ignorant perception of Batman as being a mistake by movie producers more and more scientists in the States perceive things incorrectly because they focus on the form instead of the content.

By focusing on the form, the style, the "wow factor" the people in the US flawlessly present incorrect assumption built on incorrect assumption until they have developed a service based economy asking an uninterested world "would you like fries with that?" because some idiot has assumed that people want services from a nation who cares more about form than content. People who care about the Styrofoam packaging than the McBurger inside.

Imagine for a moment that you taste a standard cardboard burger for the first time and you hear the people in the nation which produces these things is complaining about the wrapper instead of the burger. Would you trust these people to do anything right?

No comments: