I glanced at Outliers with a co-passenger in a flight some months ago and noted that I need to look at it. Anand's blog reminded me to follow up on this. This post is mainly about the book. Subsequently I intend to focus more specifically on Anand's comments and questions.
One basic theme of the author is to question the way we do
Grading
I was a lecturer at university for nearly 20 years in which Ive given my share (> 1500) of 'grades' to students — O(outstanding), A... to F(fail). And for all these years it was always an unnerving exercise for me. At the final stage when, after all the papers etc were done, I would sit down to put those fatal letters I would be trembling that I do not do injustice to some young-un by undue harshness nor damage the system (of which I was a part) by undue leniency. I guess I tended to be slightly more lenient than I should have been because there was the lurking fear that somewhere among the almost faceless names there would be an Einstein to whom my careless grading might say: "You will never amount to anything, Albert." These personal memories of mine were rekindled by the Langan-Oppenheimer story.
You see after some years of teaching and (mis)grading one soon comes to realize that the diff between the O/A grader and the D/F grader is not one of degree (of intelligence, hardwork, IQ, study etc) but largely one of
You see after some years of teaching and (mis)grading one soon comes to realize that the diff between the O/A grader and the D/F grader is not one of degree (of intelligence, hardwork, IQ, study etc) but largely one of
Emotional framing
The good students are going for what they enjoy and love while the 'bad' are running away from what they fear and hate. For a neat expo of the difference between these two modes see: PainBrainGainBrain
The same difference in motivation and the corresponding differences are seen in this: In management it has been found that top mgmt is more likely to see mgmt as a game and mid-level mgmt to see it as a struggle - in all such cases, the persons involved are acting from a very different space. (Think of the 'politeness-space' from which the Colombian pilots crash their plane vs the more normal I-gotta-land-buddy that could have saved them)
Brings me to the next resonance I have with Gladwell...
The same difference in motivation and the corresponding differences are seen in this: In management it has been found that top mgmt is more likely to see mgmt as a game and mid-level mgmt to see it as a struggle - in all such cases, the persons involved are acting from a very different space. (Think of the 'politeness-space' from which the Colombian pilots crash their plane vs the more normal I-gotta-land-buddy that could have saved them)
Brings me to the next resonance I have with Gladwell...
Languaging
Why and how are Chinese kids so much better than American ones at maths?
One of my star teachers hvs, would terrify the numerical analysis class by multiplying 5 digit decimal numbers faster in his head than the students could punch into their calculators! Im not sure of the secret of his prowess, but I can tell you this: If you heard him muttering under his breath you'd hear him calculating (and probably thinking) in his mother-tongue Marathi and then translating into English.
Now consider the example of the famous algorithm: quicksort. Hoare — one of the most brilliant CSists — was struggling to 'discover' this without success for a number of weeks. Then he was introduced to Algol which had recursion, a very 'advanced' concept in the early 60s, and he could do the feat in a few days.
Quicksort remains one of the best sorting algorithms today but is still considered advanced to teach in a first programming course – unless one uses a language like Haskell which makes it almost trivial:
But to me this argument misses the point. Hoare could easily write quicksort without recursion once he has used the prop of Algol's recursion to discover it. hvs can think in Marathi and translate to English for the class faster than directly thinking in English.
Problem solving is more about adroitly dancing round obstacles than about using the brain-muscle to do intellectual weight-lifting.
And if body : nimbleness :: mind : ??
Its languaging because as Wittgenstein said:
The limits of my language are the limits of my world
So what lessons do Gladwell's Chinese budding mathematicians have for the IT industry? Lets enumerate his key points:
One of my star teachers hvs, would terrify the numerical analysis class by multiplying 5 digit decimal numbers faster in his head than the students could punch into their calculators! Im not sure of the secret of his prowess, but I can tell you this: If you heard him muttering under his breath you'd hear him calculating (and probably thinking) in his mother-tongue Marathi and then translating into English.
Now consider the example of the famous algorithm: quicksort. Hoare — one of the most brilliant CSists — was struggling to 'discover' this without success for a number of weeks. Then he was introduced to Algol which had recursion, a very 'advanced' concept in the early 60s, and he could do the feat in a few days.
Quicksort remains one of the best sorting algorithms today but is still considered advanced to teach in a first programming course – unless one uses a language like Haskell which makes it almost trivial:
sort [] = [] sort (x:xs) = sort[y| y<-xs, y<x] ++ [x] ++ sort[y| y<-xs, y>=x]Now some may say that this is cheating because quicksort has 2 key ideas: (a) Partitioning xs on x (b) Doing it in-place. The above Haskell captures (a) with striking beauty but misses (b).
But to me this argument misses the point. Hoare could easily write quicksort without recursion once he has used the prop of Algol's recursion to discover it. hvs can think in Marathi and translate to English for the class faster than directly thinking in English.
Problem solving is more about adroitly dancing round obstacles than about using the brain-muscle to do intellectual weight-lifting.
And if body : nimbleness :: mind : ??
Its languaging because as Wittgenstein said:
The limits of my language are the limits of my world
So what lessons do Gladwell's Chinese budding mathematicians have for the IT industry? Lets enumerate his key points:
- Chinese culture encourages kids to work hard and enjoy doing it in general, but specfically in math because of...
- Efficiency of representation: Saying a uni-digit number in Chinese is faster and therefore easier than in English,
- Chinese numerals and number-words have a greater regularity than the corresponding pairs in English:
6 six sixty obviously related 5 five fifty somewhat related 2 two twenty remotely related - Lower translation mismatch from chinese numerals to chinese words necessitates less translation for them than for English speaking children.
- Do our programmers work for pleasure or for money? (Both is ok; only the latter is hmmm...)
- Are the languages we most use (C++, Java, UML etc) geared for work or for play?
- If we can only earn money by delivering code in 'work-languages' (rather than fun/play languages) do we explore the opportunity of playing in a fun language and translating to a work-language?
Assumptions
- Languages dont matter because all languages are Turing equivalent
is one of the general tenets of academic CS. This is related to the underlying still more subconscious and insiduous assumption: - Creativity is not programmable
- Ironical note 1:
- Turings own work supports the first via the universal Turing machine, but negates the second via the Turing test.
- Ironical note 2:
- Turing's life-aim was the second; yet he is remembered for the first!
And if one really wishes to work in this structured self-reprogramming look at the levels in http://www.nlpu.com/Coach2Awakener.htm
Finally, here's myself nearly 20 years before all these guys saying somewhat similar things http://portal.acm.org/citation.cfm?doid=141874.141877 also available here C in Education and Software Engineering
No comments:
Post a Comment