The Prison Steganographer
When a civil war erupted and prisons across his country were flooded with insurgents, a book reviewer at a small newspaper was forced by the government to leave his job and instead become a prison steganographer. A steganographer is a person who works with detecting concealed messages, and his new task was to every morning scrutinise any items that were sent to prisoners from outside, in order to detect the presence of some hidden and encrypted communication that could help the prisoners organise an escape.
Asked for the reason behind this strange appointment, the government had replied: “In order to spot a hidden cipher, a person must be hyper-sensitive to evidence of human agency. From decades of analysing literature, this is a skill you by now should have honed to perfection. When you analyse a novel, your aim is to reverse-engineer the author’s numerous decisions back to what he intended to communicate, which he, for the sake of delighting his reader, has obscured through so many layers of detail and incidentals, that the overarching message now registers only subliminally in the uncritical mind. It is in your job description to extract any meaning that may be buried beneath the meaninglessness with which you are confronted.”
He could see where they were going. He was, it was true, extremely receptive to literary themes, and his professional habits did tend to spill over into his life off-duty, where he found great pleasure in discovering patterns of the most abstract order. He sometimes imagined himself to be the protagonist in some literary kind of Truman Show, determined to out-smart his creators by making explicit the meaning lurking behind life’s many coincidences, and in that way decode the communicative scheme he himself was a medium for. Were there dots to connect, then connect them he would, and if no dots were to be found, he took it as a sign to look harder. He was, beyond doubt, the man for the job.
And yet, when on his first day his elderly assistant hurled on his desk a collection of objects ranging from shampoo bottles to shoes and shortbread biscuits, he was understandably at a loss. Which aspects of these items could conceivably embody a message, and which could not? So he began to think long and hard about the nature of meaning and communication.
For something to be meaningful, he figured, it would have to matter, and something only matters if it can guide the recipient to make sensible decisions he otherwise would not be able to make, such as the course of actions in an escape plot. He figured, moreover, that any decision could be represented as a series of “yes/no” answers for all actions the recipient has to choose between. Should they escape through the window? No. Should they use the door? Yes. Similarly, a word could be represented as a series of “yes/no” to all possible letters. Is the first letter an “A”? No. A “B”? No. A “C”? Yes. This way, if sender and receiver share a context of interpretation, any perceivable physical difference in the objects could theoretically encode a decision. Meaning, he reasoned, was nothing but differences that make a difference.
There were, thankfully, additional constraints in place. The meaning-bearing differences would have to be sufficiently permanent and numerous to have capacity for storing a long message. Also, given the complexity of an escape plot, it would be foolish to use a rigid coding scheme with a limited vocabulary. Very likely it was in English, encoded maybe through barely-noticeable dot inscriptions. So he began to scrutinise the objects for any suggestions of dot-patterns, count the frequency of each type of pattern, and then plot their distribution. From examining his printer type tray, he knew that letters in English occur with specific relative frequencies. This allowed him to infer that the most frequent dot-pattern represented “E”, the second most frequent represented “T”, and so on. Suddenly, he could make plain conspiratorial messages encoded in dents on shampoo bottles, fibre patterns in shoelaces, and subtle marks on shortbread.
Amazed by his success, the prison steganographer grew increasingly obsessed with his work. Secluded in his dark office, his only contact with the outside world was now his sweet, avuncular assistant, who morning after morning carried in new sets of prisoner gifts for him to brood over. His obsession escalated when, all of a sudden, every morning the items included a pot plant. He was convinced they contained some message encoded in their leaf veins, but whose frequency distribution obstinately refused to match that of letters in English. It must, he decided, be encrypted with yet another layer of transformations, to mask the presence of some particularly grand escape plan that could threaten national security.
The mysterious pot plants soon consumed him entirely. For weeks on end, all he did was toying with statistical regularities at higher and higher altitudes of abstraction. His health deteriorated, his paranoia intensified, he no longer greeted his assistant or thanked him for the tea he brought him. Until one day, the collection of items his assistant hurled on his desk no longer included any flowers.
“Where is the pot plant?” He asked, irritated, without lifting his eyes from his desk.
“I am sorry, sir, but it is so hard to find them now as winter approaches.”
“As if the devious escape plot conspirators of the world would care!”
“Sir,” the assistant now said in a voice broken with mortification, “sir, I think… I think you have been reading too much into it. These flowers that I every day have carried into your room… I plucked them myself on my way to work to give you some sweet smell and greenery in this squalid office of yours. They carry no more meaning than my plain and simple sympathies.”
The assistant stiffly turned his back and left, as the steganographer felt the cold, lonely squalor of his circumstances impinge on him. He looked up, out through the window, on the civil war unfolding outside, and thought, with the clarity of a book reviewer, that just like how the prison vainly imposed order on chaos, he too had imposed meaning where there was none, fooled by random noise behaving conspicuously, and – in so doing – he had missed the signs that truly mattered.
It from bit, Or from metaphor
It is a fact oft-forgotten that the scientist has a mind that is very much like ours. One flooded by the same ephemeral streams of fancies, constantly struggling to solidify the ineffable into words. But above all: a lazy mind, sometimes so to a fault, supported by a brain that has limited computational capacities, and which is always in search for opportunities to offload the work onto the environment, much like how the poet seeks to discharge his emotions into a stanza.
So what the scientist does is to think patiently about his intuitions, wallow for a while amidst the airy wisps curling through his consciousness, in an attempt to find structure in them, as if the billowing mind-scape had contours in it if you look really, really hard. That way, he can then try to map – to arbitrarily associate – his ideas onto something external and discrete that is easy to quantify. By letting something else, a representation, stand in for his private thought, he sacrifices the rich experience of the real thing, in exchange for something that he and his collaborators can work mindlessly and mechanically with. Maybe they jot down an idea on a cafeteria napkin, apply some theorems to an equation on a blackboard, or let a computer grind through a dataset on the order of millions. The less brain it requires, the better.
It is this laziness that language is for. No two persons conjure up the exact same brain state upon reading the word “cat”, but the brain states are similar enough to produce roughly the same behaviour-guiding inferences. We may say that the brain states are “fungible”, and by mapping this overlap onto an external symbol – in this case a word – we get something that is sensorially impoverished but capable of cuing retrieval of information from the vastly more capacious long-term memory.
This way, our language can be seen as a culturally achieved notebook for storing intellectual resting-points that we now can use our computational resources to build upon instead of reinventing the wheel. The idea of language as an ever-shifting springboard for intellectual progress – something that transforms our thinking as much as it mirrors it – has by philosopher Andy Clark been compared to Mangrove trees. Mangrove trees grow in water, and quite like how islands emerge under Mangrove trees from soil accumulating between their roots, the coinage of a word becomes an island for further exploration in the balmy sea of thoughts our brains are bathing in.
Laziness is also what mathematics is about. To say that a fruit bowl contains three fruits is to disregard the fact that one may be a banana and the others two apples, and the number preserves no information about how the fruits are arranged. In some contexts, to represent the bowl by a number would be terribly insufficient, but across contexts, most people would at least agree that it contains three discrete objects. Like how all people generate fungible brain states upon reading a word, all people can easily extract numeracy – quantify – a fruit bowl. By doing so, people can reason about it using mathematical theorems inherited from other people living centuries ago, who themselves mapped their vague intuitions onto formal notations to play around with absent-mindedly. And because numbers are recursively defined (2 is 1+1, 3 is 2+1, etc.), we can offload their transformations to machines.
Perhaps it is due to the very same laziness that we also tend to forget this artificial, brain-serving, and brain-dependent nature of formalisms, and give them more metaphysical significance than they deserve, mistaking simulation for reality. Like how wars are fought over territory lines – lines that exist nowhere else than on our maps – we start unwinnable debates about abstractions as “true” or “untrue”. But when we take formalisms for what they are – external structures to bootstrap ourselves beyond our computational limitations – and give due recognition to the poetic impulse that led to their construction, a more fulfilling kind of insight becomes obtainable.
We see, for example, how deceptively simple concepts evolved gradually, carried and modified from mind to mind over the course of millennia, only recently mapped onto mathematical structures. The word “matter” was first coined by Aristotle, who borrowed the Greek word for “timber” to represent the stuff everything is made of, and acquired, during medieval times, the connotations of something solid and massy, proportional to our intuition of “resistance”. Later it was thought of as invisible corpuscles in an empty space. “Space” was conceived by Newton as an “immaterial medium”, and “gravity” was by him thought of as “Earth like a sponge, drinking up the constant stream of fine ethereal matter falling from the heavens, this stream by its impact on bodies above the Earth causing them to depend”. “Energy”, deriving from our intuition of “potential”, was in the early 1800s thought of as a ghostly substance like phlogiston and élan vital. When eventually we formulated it mathematically, enabling external manipulation, we were no longer required to consciously maintain such expensive imagery in our mind.
So when a scientist waxes lyrical, he does so a little bit less alliteratively than the poet, but his chalk-dry equations spring from sensibilities equally romantic. By encapsulating his experiences into formalisms, he lets formalisms flow through his mind in their stead, as building blocks on their hierarchical course towards self-complication. And so, picking up where hierarchy theory left off, the following sections curate a collection of powerful concepts to help explain these mechanisms of human cognition. Together, they represent probably the closest science has come to a “Grand Unified Theory of Thought”.
Part 2 lays the metaphysical groundwork by explaining how statistical discontinuities in the fabric of reality are the source of meaningful patterns, by reference to concepts like “dynamic systems theory”, “randomness”, “entropy”, “Maxwell’s demon”, “information”, “affordance, and “umwelt”.
Part 3 introduces ideas on how these patterns are “inferred” bottom-up by gradually refining the mind-as-a-computer metaphor, from the early mechanistic conceptions of “cybernetics”, via the serial symbol-crunching computer of “computationalism”, to the parallel networks of modern “connectionism”, and recent ideas within “radical embodied cognitive science” that dispense with the computer metaphor altogether.
Part 4 elaborates on this pattern-completion process by casting the brain as an active predictor of its environment, implementing so-called “Bayesian reasoning”, as posited by the “predictive coding approach” within computational neuroscience. It also introduces a close relative, the “embodied simulation hypothesis”, which theorises that metaphor – the reuse of circuitry used for perception and motor action for abstract thinking – is the central mechanism in human cognition.
Part 5 places low-level categorisation on a continuum with poetic metaphor by presenting theories on “analogical reasoning”, and how the brain maps inference structures onto each other in order to make sense of new information, as well as “blending” them to engage in imaginative, counter-factual reasoning.
Part 6 looks at how the idea of abstract thinking as essentially metaphor forces us to re-evaluate many of our metaphysical commitments, such as the nature of time, causation, and morality, to investigate how such concepts arise from the biological constraints of our brain.
Finally, Part 7 steps down from the lofty heights of cognitive scientific theory to engage in more down-to-earth matters, discussing how to apply this knowledge to improve instructional practice. It presents the idea of “metaphor-driven learning”, of collecting the implicit metaphors students use to understand subjects at different levels of expertise, in order to advance between-stage transitions. It is a project I have chosen to call “menimagerie” – one of capturing mental imagery from the wild, and put it on public display, like a glass menagerie of insight.