Sunday, March 25, 2012

The Ineffable Intuition of Chicken Sexers


In the book Incognito by David Eagelman, the author discusses the strange nature of chicken sexing. This is the valuable process of separating female and male chicks as soon as possible, because each sex has different diets and endgames (most males are just destroyed). The mystery is that when you look at the vent in the chick’s rear, some people just know which are female. It is impossible to articulate, so the Japanese figured out how to teach this inarticulable knowledge. The student would pick up a chick, examine its rear, and toss it into a bin. The master would then say ‘yes’ or ‘no’ based on his generally correct observation. After a few weeks, the student’s brain was trained to masterful levels.

This is rather fascinating. I wonder how many things are taught this way, things related to intuitions, usually focused on relationships. I don’t think this would work in stochastic cases like stock picking because even a good stock might have succeeded in spite of itself due to some random luck, making it no more useful to generalize upon than a winning lottery number. A good mentor could really generate a first-order advantage, and perhaps that is relevant to mystery of what generates most of the intergenerational income correlation, because IQ seems only to explain 10 to 25% of it.

We all know a lot more than we can articulate, stuff that can't be translated by our conscious self, which is why we still can’t generate a robust program to distinguish between a dog and a cat, even though a 4-year old has no problem. Nonetheless, we can train our unconscious thoughts via methods like the chicken sexer, primarily by emulating others who are good.

As Oscar Wilde said, the things worth learning can't be taught, by which I think he meant taught via some schoolbook. Watching a father deal tactfully with anger, frustration, and persuasion is surely as valuable as learning the times tables. Many people have this emotional intelligence, and it is very beneficial, probably something that isn't taught via reasoning, rather simply showing.

7 comments:

B. A. said...

isn't it a bit strange that they had to pick an example like the "chicken sexers", so remote from every day life, to illustrate such a general and fundamental claim?

to some people, another example would be grammar. depending on language, it can be very complex. we pick up as children without conscious effort. most speakers of a language can say whether a sentence violates grammar rules or not and where, but without some formal study of grammar, it is very hard to articulate the rules that apply to that example. when asked why is it wrong to say so, there will be people making their own funny rules, and others claiming that the rules are actually impossible to articulate, and come from some higher sense of the language that only native speakers possess (particularly among those who never read a grammar book about their own language).

I guess it's tempting to excuse our inability to articulate part of what we know by claiming it's actually impossible to do so, as it represents some higher form of knowledge that only a restricted group of people possess. a bit like what Derman says of EMH "converting […]]failed attempts at systematic stock price prediction into a fundamental postulate of their field was a fiendishly clever jujitsu response on the part of economists”.

Aretae said...

I've taught dozens of subjects (Math, Programming, Soccer, Latin, English, etc...). This is the fundamental insight of learning. Feedback wins. Unfortunately, in many cases, the professors aren't smart enough to build good feedback systems...and so you're stuck with other methods that work much less well.

Mercury said...

I hate to bring this up but I think Taleb discusses this at some length in (I think) his ‘Black Swan’ book. One example he uses is that of Henry Bessemer and his patented process for mass steel production. Apparently, early patent purchasers had trouble developing the knack for when to proceed from one critical step to the next and Bessemer had to personally mentor some of them in much the same way as described here with the chick-sexers.

Once when I was much younger, my parents hired an old Italian guy to “re-point” (fill in eroded mortar) a brick wall. It was one of those deals where the guy just happened along and said: “gee, I couldn’t help but noticing your wall is in rough shape…” and the job/price is negotiated on the spot. Anyway, the guy does a nice enough job on one side of the wall but whips through it in an incredibly short amount of time. To say he makes it look easy is an understatement. My parents put up a bit of a fight about their having been misled into hiring the guy for an astronomical $per hour rate but a deal is a deal and it is decided that I will do the other side of the wall. Four hours later, despite having followed what I thought were obvious and simple steps, I have glops of mortar mix all over the place and only 1/3 of the job done. Lesson learned!

Anonymous said...

I had much the same experience when I tried bass fishing on Saranac Lake. My friend, a veteran fisherman, was in a small boat identical to mine, and on the face of it was visiting the same spots and casting the same way, but I caught nothing all day, while he pulled out lunker after lunker.

B. A. said...

the point of eric's post is that people can acquire and consciously put use knowledge that impossible (for anyone) to articulate. some people "just know", and have picked this skill up from others that "just know". is bass fishing and fixing mortar on brick walls impossible to articulate? I doubt it. a friend of mine became very good at fishing by the age of 14, and he would routinely embarrass much older and experienced fishermen around him by catching more fish than them. he learned this skill by reading books, and by getting veterans to teach him their tricks. it's all very articulate knowledge and practice.

Mercury said...

B.A. - I don’t know that any knowledge is *impossible* to articulate but I think the point is instead that some skills require more learning-by-doing than others. It’s not impossible to articulate how to ski but 99.9% of your well-articulated wisdom is going to go right over the head of a five year old. Eventually the kid just intuits the physics involved and in a few years he’s better than you.

Teaching an adult female to ski however a completely different proposition and yes I will take the other side of any bet you make involving your knowledge articulation prowess in this area.

B. A. said...

@ mercury

I don't dispute that learning by experience is very important. was just puzzled by the claim that the task requires knowledge that is mysterious and inarticulable. turns out that a better way to describe it is "inaccessible to introspection". in that respect, it is no different from our ability to recognize words or faces, or tell a cat from a dog. it may well feel to us that we "just see" things, however it can be proven that we do it by unconsciously processing individual features and cues. such knowledge can be articulated, but through repeated tests that slightly alter individual features one at a time. it's definitely not possible by asking the guys how they do it, because they don't know. very interestig article on this here, if anybody is interested
http://cogprints.org/3255/1/chicken.pdf