Thursday, November 23, 2006

Faith and evidence-based reasoning


One dimension of the public debate between science and religion is resolved as "faith" versus "evidence."

Yes, I'd agree that where you have tried and true calibrated observational tools, then viewing reality directly should trump legacy impressions or old mental models of what's out there right now. Things change.

The problem with operationalizing that is that we don't generally have calibrated vision. In fact, the observational powers and perception of any one human working alone are generally dismal,
easily fooled, context-sensitive, path-sensitive, and not reproducible.

We are quite biased about seeing personal flaws, so move the discussion outside. Would you trust a person I selected at random from around you to have perfect judgment and perception of the world and to always make good calls about what's going on and what to do Probably not.

Even if you are an academic or scientist or doctor. Maybe, especially if you are a professional, you are able to see flaws easily in others regarding their observational powers, judgment, and freedom from bias favoring mental models clung to despite evidence to the contrary.

And, the field of "science" as a whole, while quick to note reasoning flaws in those outside the field, can be slow to equally weight centuries of evidence that the scientific method is no guarantee of freedom from bias. Stephen Jay Gould documented this very well in The Mismeasure of Man. New data that contradicted existing and beloved mental models have always been rejected more than appropriate, even by scientists, often especially by the best scientists. Thomas Kuhn writes well on the problems of changing the paradigm in The Structure of Scientific Revolutions.

Yes, "science" rejected Bohr's quantum mechanics, Einstein's "theory of relativity", plate tectonics, Barbara McClintock's "jumping genes", etc. Scientists clung to with a religious fervor the idea that DNA and the human genome had 100% of the information in it needed to create an organism, and none of it was "epigenetic", and none of it was, gasp, changeable in real-time by the environment in a heritable way.

And, my favorite - at the current time, "systems thinking" and the need to include closed feedback loops in models of reality is clearly becoming more and more important, but requires letting go of the misperceived "scientific method" restriction that anything important can be isolated from reality and studied under controlled conditions, and described quantitatively. Unfortunately, that removed economics, business, psychology, love, war, sociology, and most of the rest of human activity from the sphere of "science" which, fascinatingly, did not perceive this as a testament to the weakness of its methods, but saw it as a statement of the worthlessness of the problem domains for study. They were, sniff, "soft" sciences, not really the kind one would invite to a garden party.

So, there's the problem. Our individual perceptual system, even among very well trained scientists, remains flawed and biased. And, our perception of the quality of our own personal perception seems curiously exempt from the flaws that everyone else around us seems to have.

So, we don't perceive the world well, and, worse, we don't perceive that we don't perceive well.
The observational equipment humans are born with is blind to its own blind spots.

Thus, grudgingly, we are forced to use statistical methods to try to overcome our own skepticism or gullibility and avoid errors in seeing things that aren't there, or not seeing things that are there. We are forced to submit our beloved ideas to the cold or acidic waters of peer review, where many of them are shown to have flaws we didn't realize.

The blindness problem and lack of humility are true for those in management and leadership positions as well, and the lack of contact with reality are most easily perceived from either below (the staff) or outside (stockholders), but often almost invisible from within, where shared delusions take on the power of reality.

Thus, it becomes critical for safe and sane operations that there be pathways for the staff to point out problems with leadership's mental models and perceptions of the world. These are not always accepted gracefully. Yet, even the US Army Leadership Field Manual, and all the teachings of High Reliability Organizations, stress the need to build a "safety culture" where individuals can question and challenge the prevailing wisdom.

So, finally, swing this back around to questions of faith. If even people who are way above us in IQ and who have ten years more education than us are subject to having their best judgment proven wrong, what level of humility should we have about our own?

If our eyes very often deceive us and our very perceptions betray us, does it make sense to go with personally perceived reality or with a larger, more slowly moving flywheel of socially hard-won and peer-reviewed reality?

So, while, yes, we need to have a culture that allows respectful questions to be raised and taken seriously, and that needs to accept that things change over time, we also need to recognize that any single random observation we have that contradicts that social wisdom is more likely wrong than a brilliant insight.

Regardless of level of experience and education, these factors are at play, and it is a good rule of thumb to have many counselors from many diverse backgrounds vet our observations and point out to us, in as civil and loving a way as possible, where we, uh, are wrong.

In the meantime, whether science or religion, "faith" in the generally accepted corpus of knowledge is the flywheel we have to go on day by day. In either field, that mental model of the world has to be updated if there is consistent, persistent evidence that there is now a mismatch between the model and the reality. The process works better if those with old views and those with new views are respectful of the need for dynamic stability and the ongoing conversation to continue.

What has always proven to be wrong, historically, is to say "Ah, we've removed the last bug, and the system is now perfect." The reality is, very little of what we hold dear seems to survive 100 years intact.

The unbiased forecast, therefore, has to be that most of what we currently believe to be true is equally distorted.

Also, if genes or religious beliefs are preserved over hundreds or thousands of years, they must serve some very important purpose, because preservation is an expensive process. While the value may not be obvious in a quick glance or in a short-run accounting, we should be respectful of the assumption that there is some value, some kernel of truth, something underlying that selection by survival, that would be good to understand, tease out, and hear.

No comments: