The greatest challenge is stating the problem in a way that will allow a solution.

Wednesday, October 17, 2012

The Science of "Chunking"

“Chunking,” Working Memory, and How Pattern Recognition Fuels Creativity

by Maria Popova

“Generating interesting connections between disparate subjects is what makes art so fascinating to create and to view . . . we are forced to contemplate a new, higher pattern that binds lower ones together.”

It seems to be the season for fascinating meditations on consciousness, exploring such questions as what happens while we sleep, how complex cognition evolved, and why the world exists. Joining them and prior explorations of what it means to be human is The Ravenous Brain: How the New Science of Consciousness Explains Our Insatiable Search for Meaning (public library) by Cambridge neuroscientist Daniel Bor in which, among other things, he sheds light on how our species’ penchant for pattern-recognition is essential to consciousness and our entire experience of life.

The process of combining more primitive pieces of information to create something more meaningful is a crucial aspect both of learning and of consciousness and is one of the defining features of human experience.

Once we have reached adulthood, we have decades of intensive learning behind us, where the discovery of thousands of useful combinations of features, as well as combinations of combinations and so on, has collectively generated an amazingly rich, hierarchical model of the world.

Inside us is also written a multitude of mini strategies about how to direct our attention in order to maximize further learning.

We can allow our attention to roam anywhere around us and glean interesting new clues about any facet of our local environment, to compare and potentially add to our extensive internal model.

Much of this capacity relies on our working memory — the temporary storage that holds these primitive pieces of information in order to make them available for further processing — and yet what’s most striking about our ability to build such an “amazingly rich” model of the world is that the limit of our working memory is hardly different from that of a monkey, even though the monkey’s brain is roughly one-fifteenth the size of ours: Experiment after experiment has shown that, on average, the human brain can hold 4 different items in its working memory, compared to 3 or 4 for the monkey.

What makes the difference, Bor argues, is a concept called chunking, which allows us to hack the limits of our working memory — a kind of cognitive compression mechanism wherein we parse information into chunks that are more memorable and easier to process than the seemingly random bits of which they’re composed.

Bor explains:  In terms of grand purpose:

-chunking can be seen as a similar mechanism to attention

Both processes are concerned with compressing an unwieldy dataset into those small nuggets of meaning that are particularly salient. 

But while chunking is a marvelous complement to attention, chunking diverges from its counterpart in focusing on the compression of conscious data according to its inherent structure or the way it relates to our preexisting memories.

To illustrate the power of chunking, Bor gives an astounding example of how one man was able to use this mental mechanism in greatly expanding the capacity of his working memory. 

The man, an undergraduate volunteer in a psychology experiment with an average IQ and memory capacity, took part in a simple experiment, in which the researchers read to him a sequence of random digits and asked him to say the digits back in the order he’d heard them.

If he was correct, the next trial sequence would be one digit longer; if incorrect, one digit shorter. This standard test for verbal working memory had one twist — it took place over two years, where the young man did this task for an hour a day four days a week.

Initially, he was able to remember roughly 7 numbers in the sequence — an average improvement over the 4-item limit that most people arrive at with a few simple and intuitive rehearsal strategies. But the young man was so bored with the experiment he decided to make it interesting for himself by doing his best to greatly improve his limit — which he did.

By the end, some 20 months later, he was able to say back a sequence that was 80 digits long — or, as Bor puts it, “if 7 friends in turn rapidly told him their phone numbers, he could calmly wait until the last digit was spoken and then, from memory, key all 7 friends’ numbers into his phone’s contact list without error,” an achievement that would make Joshua Foer proud.

But how, exactly, was an average person capable of such a superhuman feat? 

Bor sheds light:

This volunteer happened to be a keen track runner, and so his first thought was to see certain number groups as running times, for instance, 3492 would be transformed into 3 minutes and 49.2 seconds, around the world-record time for running the mile.

In other words, he was using his memory for well-known number sequences in athletics to prop up his working memory.

This strategy worked very well, and he rapidly more than doubled his working memory capacity to nearly 20 digits.

The next breakthrough some months later occurred when he realized he could combine each running time into a superstructure of 3 or 4 running times — and then group these superstructures together again.

Interestingly, the number of holders he used never went above his initial capacity of just a handful of items.

He just learned to cram more and more into each item in a pyramidal way, with digits linked together in 3s or 4s, and then those triplets or quadruplets of digits linked together as well in groups of 3, and so on.

One item-space, one objet in working memory, started holding a single digit, but after 20 months of practice, could contain as much as 24 digits.

This young man had, essentially, mastered exponential chunking. But, Bor points out, chunking isn’t useful only in helping us excel at seemingly meaningless tasks — it is integral to what makes us human:

Although [chunking] can vastly increase the practical limits of working memory, it is not merely a faithful servant of working emory — instead it is the secret master of this online store, and the main purpose of consciousness.


There are three straightforward sides to the chunking process

1. the search for chunks,
2. the noticing and memorizing of those chunks, and
3. the use of the chunks we’ve already built up.

The main purpose of consciousness is to search for and discover these structured chunks of information within working memory, so that they can then be used efficiently and automatically, with minimal further input from consciousness.

Perhaps what most distinguishes us humans from the rest of the animal kingdom is our ravenous desire to find structure in the information we pick up in the world. We cannot help actively searching for patterns — any hook in the data that will aid our performance and understanding.

We constantly look for regularities in every facet of our lives, and there are few limits to what we can learn and improve on as we make these discoveries.

We also develop strategies to further help us — strategies that themselves are forms of patterns that assist us in spotting other patterns, with one example being that amateur track runner developing tactics to link digits with running times in various races.

But, echoing Richard Feynman’s eloquent lament on the subject, Bor points to a dark side of this hunger for patterns:

One problematic corollary of this passion for patterns is that we are the most advanced species in how elaborately and extensively we can get things wrong. 

We often jump to conclusions — for instance, with astrology or religion.

We are so keen to search for patterns, and so satisfied when we’ve found them, that we do not typically perform sufficient checks on our apparent insights.

Still, our capacity for pattern-recognition, Bor argues, is the very source of human creativity.

In fact, chunking and pattern-recognition offer evidence for the combinatorial nature of creativity, affirm:

- Steve Jobs’s famous words that “creativity is just connecting things”,

- Mark Twain’s contention that “all ideas are second-hand”, and

- Nina Paley’s clever demonstration of how everything builds on what came before.

The arts, too, generate their richness and some of their aesthetic appeal from patterns.

Music is the most obvious sphere where structures are appealing — little phrases that are repeated, raised a key, or reversed can sound utterly beguiling.

This musical beauty directly relates to the mathematical relation between notes and the overall logical regularities formed. Some composers, such as Bach, made this connection relatively explicit, at least in certain pieces, which are just as much mathematical and logical puzzles as beautiful musical works.

But certainly patterns are just as important in the visual arts as in music. Generating interesting connections between disparate subjects is what makes art so fascinating to create and to view, precisely because we are forced to contemplate a new, higher pattern that binds lower ones together.

What is true of creative skill, Bor argues, is also true of our highest intellectual contribution:

Some of our greatest insights can be gleaned from moving up another level and noticing that certain patterns relate to others, which on first blush may appear entirely unconnected — spotting patterns of patterns, say (which is what analogies essentially are).

Best of all, this system expands exponentially as it feeds on itself, like a muscle that grows stronger with each use:

Consciousness and chunking allow us to turn the dull sludge of independent episodes in our lives into a shimmering, dense web, interlinked by all the myriad patterns we spot. 

It becomes a positive feedback loop, making the detection of new connections even easier, and creates a domain ripe for understanding how things actually work, of reaching that supremely powerful realm of discerning the mechanism of things.

At the same time, our memory system becomes far more efficient, effective — and intelligent — than it could ever be without such refined methods to extract useful structure from raw data.

Though some parts of The Ravenous Brain fringe on reductionism, Bor offers a stimulating lens on that always fascinating, often uncomfortable, inevitably alluring intersection of science and philosophy where our understanding of who we are resides.

tags: books culture philosophy psychology sc


The Science of "Chunking," Working Memory, and How Pattern Recognition Fuels Creativity | Brain Pickings