kitchen table math, the sequel: Explicit and implicit learning

Friday, June 15, 2012

Explicit and implicit learning

Explicit (Declarative)
Implicit (Procedural)
facts and experiences
skills and habits, priming,
classical conditioning

knowledge can be verbalized
(conscious)

knowledge is inaccessible
(unconscious)
tested by recall, recognition,
cued recall (as in school)

evidenced via altered dispositions,
preferences, judgements, behavior

one trial learning
often acquired gradually
over multiple trials

requires effort and intention.....
acquired incidentally
(without intention, attention,
or awareness)

flexible knowledge,
available to multiple
response systems

inflexible knowledge,
limited to response systems
participating in original learning

not durable

durable

can form conjunctions
between arbitrary stimuli
(e.g., paired associate learning)

cannot learn conjunctions
“specialized to detect variance,
i.e., what is different or unique
about the events of a particular
time and place” [Eichenbaum].....

“specialized for detecting invariance, i.e., for
extracting what is common in stimulus
environment” (i.e., regularities)


medial temporal lobe
(hippocampus)

striatum (basal ganglia)
source:
Learning and Memory

Various:
  • I always remember the meaning of "declarative knowledge" by the sentence: "I declare that Abraham Lincoln was the 16th president of the United States." Or: "I declare that I was born in Springfield, IL." Declarative knowledge is facts and factoids you can (consciously) declare.
  • I think these two systems correspond to Daniel Kahneman's System 1 and System 2. (Take that with a grain of salt.)
  • It's misleading to characterize implicit learning as strictly procedural. Category learning and probabilistic learning are both handled by the basal ganglia. It's the basal ganglia that allow you to learn that where there is smoke, there is fire.
  • It's the hippocampus that allows you to learn the aphorism "Where there's smoke, there's fire."
  • The basal ganglia-frontal circuit understands language. Procedural memory processes grammar; declarative memory learns vocabulary. 
  • "One trial learning" is a bit misleading (in column one). Declarative knowledge can be acquired in one trial in a way that procedural knowledge (e.g.: how to hit a tennis ball) cannot. But retaining declarative knowledge over time requires spaced repetition and practice.
  • Not sure about 'cannot learn conjunctions [8th row].' Offhand, it doesn't jibe with scenarios like the weather prediction task. update 8/1/2012: paired-associate learning at Cambridge Brain Sciences - vocabulary learning is a case of paired-associate learning
  • Must determine the meaning of "limited to response systems participating in original learning."
  • Alzheimer's affects the hippocampus.
And:

Seems to me constructivism mixes these two systems up.

The basal ganglia are built to look for a pattern naturally and unconsciously. You don't have to think about it, and you don't have to go to school to do it. But constructivists want students to expend a great deal of time and conscious effort figuring out the patterns and regularities in school subjects.

Meanwhile declarative knowledge has to be attended to and consciously acquired, but constructivists seem to want students to learn content knowledge more or less by osmosis. Osmosis is the basal ganglia's department. Ditto learning by doing.

Constructivists seem to want to make easy things hard and hard things easy.

5 comments:

SteveH said...

Implicit = the secret handshake

It's don Juan's Yaqui way of knowing in "A Separate Reality". It's the tonal versus the nagual.

Implicit = modern art

Explicit = me saying that if I could do it, it isn't art. At least with music, you can't sit there and get away with not playing music ... or maybe not.


I remember a PBS special on Robert Rauschenberg when he was setting up a show for his white paintings. Rauschenberg claims he was trying to see how far he could push an object and still have it mean something. It scared him, so it was all about courage. (Or a check to see how fame can test gullibility.)

Interestingly, this inspired his friend John Cage to write his "silent composition" called 4'33". Amazingly, you can see it on YouTube.

http://www.youtube.com/watch?v=hUJagb7hL0E


This is really not new, according to Wikipedia:

"Yves Klein's 1949 Monotone-Silence Symphony (informally The Monotone Symphony, conceived 1947–1948), an orchestral forty minute piece whose second and last movement is a twenty minute silence[19] (the first movement being an unvarying twenty minute drone)."


How about Gehry's Stata Center
for Computer, Information, and Intelligence Sciences at MIT, where implicit architecture meets the concrete, sequential world. Gehry is known for not seeing a lumpy curve he didn't like.

If it's different, it must be good. This is based on the idea that if many great ideas are disliked at first, then anything that gets a bad first reaction will be good sometime in the future. You just have to wait long enough. The cognoscenti try to define art, music, and architecture. They want their sensibilities to define their fields. They want to define explicit.

Sounds like K-6 education; pedagogy defines math.

SteveH said...

Oh, and in the Cage video, I like the way they turn the blank pages of the music. The orchestra is a little slow. I guess they didn't practice the piece much.

Cage is also known for "prepared piano" pieces where you have to carefully place screws between the strings. Some of the sounds might be interesting ... for a minute. Perhaps different is an easier path to success than better.

Catherine Johnson said...

Wait!

How is implicit learning the secret handshake?

Catherine Johnson said...

I wonder whether, after you have some expertise (or background knowledge) in a subject, you start to do more implicit learning in that area....

I notice that people who are more fluent in math than I am "read" numbers faster than I do --- (looking at, say, the SPED spending numbers, for instance).

SteveH said...

“knowledge is inaccessible”

“evidenced via altered dispositions, preferences, judgments, behavior”

“often acquired gradually over multiple trials”


Then you said:

“I wonder whether, after you have some expertise (or background knowledge) in a subject, you start to do more implicit learning in that area....”


Yes, but people don’t come to the same conclusions. Some “experts”, however, want to impose their own opinions. It’s kind of like a club with a secret handshake. Why should I like an art show with all white paintings? Why should I like to listen to Schoenberg’s “Verklärte Nacht”? If I don’t come to that same implicit learning, then what? For some experts, the goal is to control this implicit level of sensibility. It's interesting to see how many novices buy into those ideas without going through the process themselves.

There is someone in our area who loves to question this imposed sensibility in the area of modern architecture. He is constantly telling people to open their eyes and not buy into experts' sensibilities. However, many novices want to learn how to join the club and talk the talk.


This reminds me of ed school. Students are not allowed to discover pedagogy using the same techniques they hold dear. They are led to the secret handshake and never allowed to discover it naturally ... because it might not happen and the experts don't want that. Ultimately, implicit judgments end up defining reality. They get it backwards. "Research shows" ... given my assumptions.