Friday, June 3, 2011

The Human Limits of Reason

This essay is a continuation of a previous one: We Can't (Always) Be Reasonable.


The Human Brain: The most important regions of the human brain. "The key issue as to whether or not a non-biological entity deserves rights really comes down to whether or not it's conscious.... Does it have feelings," said Ray Kurzweil, inventor, futurist and believer in artificial intelligence, in USA Today: Aug 19, 2007.
Credit: National Institute for Aging, a branch of NIH, 2008
Source: Wikipedia

Now, pure reason is a wonderful ability, to think through a problem or idea. It certainly has its place as a means of practicing science, medicine and to a great degree, law. Emotion might just muddle up things, and act as a distraction. But like all intellectual constructs of the human mind, it has its limitations, notably in affairs of the human enterprise.

Arguing for reason is like arguing for goodness or honesty or self-control. Few would argue against such ideals. Yet, arriving at moral decisions, and being fulfilled, for most mortals, requires more than the use of reason and the cold facts of logic. Something else is at play here. The happily fulfilled tend to make decisions and life choices based on emotions, intuitions and other so-called soft decision-making skills. They also have few regrets and face less dissonance in their choices. In fact, successful and fulfilled people do not overly analyse situations and problems as a computer does.

Which might seem obvious to nonscientists such as myself. But not to some cognitive neuroscientists who work hard to compare humans to machines (e.g., the brain-computer argument), even though humans predate machines, and could never be as predictable as machines. Much of the work done to equate human brains to machines is linked to cognitive science, a fairly recent discipline dating to the 1970s, which downplays the importance of emotion, social interactions and the unconscious in human actions and decision-making. Cognitive scientists like to explain human thinking by using models and algorithms. Humans don't always comply.

This is made clear by anyone who observes humans in a nonscientific or literary way. In a recent article (What's the big idea?) in the British Guardian, Stuart Jeffries writes about American social commentator David Brooks' latest book, The Social Animal and his views on the limits of reason:
Brooks says that, overwhelmingly, human decision-making is not rational but unconscious. Much of the book's pleasure consists in reading digests of experiments (such as international differences in the incidence of touching during coffee) that show how non-rational we are and yet how successful the social animal when breaking free of mere rational decision-making. The style and substance will be familiar to readers of pop psychology bestsellers such as Malcolm Gladwell's Blink or Jonah Lehrer's Proust Was a Neuroscientist: for Brooks the unconscious isn't a seething Freudian netherworld of sexual urges, but where we make the key decisions of our lives – whom to date and marry, how to vote.

Most success stories stress academic ability, IQ, hard work, he argues. Brooks rather stresses non-cognitive skills, which, he writes, is "the catch-all category for hidden qualities that can't be easily measured, but which in real life lead to happiness and fulfilment." "By that I mean emotions, intuitions, genetic inheritance. Soft stuff, which is pretty rich given that my wife thinks I'm insufficiently touchy feely."

And what are these mysterious non-cognitive skills? Good character (energy, honesty, dependability, recognising your weaknesses and controlling your worst impulses). He also mentions "street smarts", by which he means reading situations and people, often unconsciously, and developing human relationships. He thinks these skills can be honed.
I would agree. What David Brooks is talking about here is a renaming (or repackaging for modern minds) of the old-fashioned virtues, including honesty, dependability, and self-control. Whether they originate from the unconscious or not is a moot point and not worth arguing about. What is important is that such are human virtues that can certainly be taught, learned and honed, passed on from parent to child. It's a lot of hard work with no certainty of success.

Even so, such is not the way cognitive scientists tend to look at things, even though they themselves might be parents who value such virtues. Cognitive scientists would like to explain human thought by using machine language, since that is all they are left with once words like soul, consciousness, unconscious and emotions are stripped from their lexicon. Admittedly, they work hard to make their models fit human experience. Many are enamoured about such ideas.

Perhaps this speaks about a dark pessimistic view of humanity, and their having given up on the human species. It might also explain why some look to non-biological species, namely machines with high intelligence, as our future. It might seem for some that these non-biological species would be far more predictable, reasonable and intelligent (and some would add ethical), notably if they were fitted with all the best that humans can offer.

I am not convinced, however, that a machine, even one endowed with artificial intelligence (AI) — and despite the best efforts of inventors, futurists, science-fiction writers and Hollywood movie scriptwriters— could ever achieve an authentic human response.

Thus, I vote for the human, faults and all.

No comments:

Post a Comment

All comments ought to reflect the post in question. All comments are moderated; and inappropriate comments, including those that attack persons, those that use profanity and those that are hateful, will not be tolerated. So, keep it on target, clean and thoughtful. This is not a forum for personal vendettas or to create a toxic environment. The chief idea is to engage, to discuss and to critique issues. Doing so within acceptable norms will make the process more rewarding and healthy for everyone. Accordingly, anonymous comments will not be posted.