..if a writer needs a dictionary he should not write. He should have read the dictionary at least three times from beginning to end and then have loaned it to someone who needs it.
Ernest Hemingway

I am not a literary man.... I am a man of science, and I am interested in that branch of Anthropology which deals with the history of human speech.
J. A. H. Murray (1837-1915), English lexicographer and editor of the Oxford English Dictionary, on which he worked from 1879 until his death; it was not published until 1928.

Entropy Gradient Reversals

Another old one we found mining the hard disk...

Reading the Dictionary

Though I had passed the same buildings nearly every day, walked and driven the same streets, they now appeared alien, even threatening, as if some inimical wind had swept away whatever significance I had once attached to shops and intersections, old meeting places, the houses of friends long gone, or worse, unable to be reached. Whatever love I had woven into these scenes was suddenly lost, the town become a cheap tapestry of mistaken memories, unraveling at the edges. I remember standing on a corner that winter day, whatever naive beliefs I'd cradled undermined at last, whatever vague hopes I'd casually entertained finally and completely shattered. The city looked somehow flat, deflated, like a cardboard stage set after the filming's over. Grief hit me like a blow to the stomach, and an aching sadness too exhausted for tears. Yet I was mesmerized by wonder at where I might have been these past five years. Had I really lived in this picture postcard dream-turned-nightmare? Without words to say why, I was invaded by the memory of a place I'd never been, and of a gate repeatedly slamming and swinging open in an empty yard, no longer separating the space it once enclosed from the now abandoned road.

That road was waiting for me when I left the place, first going into the high range 40 miles to the north. For nine months of healing silence, I lived in a a rough frame cabin tilted out over a rocky slope. The Aztec-psilocybin rocks a mile across the valley formed a backdrop to the Buddhist shrine below my south wall window. While the Milky Way burned itself into the winter night, I would fill the shrine bowls and light incense and candles to forces I tried not to imagine. I sat like a mirror sometimes, no flicker, no breath of wind, holding the crystal reflection of flames in the water. Outside, beyond the falling snow, the forest was haunted with the sounds of other lives which became my familiars, my wise and terrible brothers.

It was in this place, 10,000 feet above the heartland of America, that I began to read the dictionary. I read it as if it were a mystery novel, turning from word to word for clues as to what it might mean when we speak to ourselves together, for a hint as to what had gone wrong somewhere along the way. Slowly, I began to see that the book was a work of archeology, an inventory of artifacts our race has picked up from its own ancestors, often without understanding their use. And I began to sense a deep movement across time and the world time weaves, a wavelength greater than the span of our life on earth. I began to know language as a music so profound it is impenetrable except to those who speak what they hear in it, and whose words are no longer their own.

Six months later, through a tangled concatenation of nearly random choices and chases, a demented search for heaven down innumerable blind alleys, I found myself standing on the platform of the Yamanote-sen, the train line that encircles the city of Tokyo. I remember especially the flashing window-glass reflections of my face, twelve layers deep, as the trains pulled in and out of Shinjuku station. There is no reason for this, I thought, and felt unreasonably alive in those ruins of my own rationality. No reason for me to be here reading signs that make no sense, I thought, reading them nonetheless: a mountain in flames, the book of the sun, a stranger at the gate. This was the beginning of an awareness that we are verbal fish awash in an ocean of language, unable to taste its salt. Except perhaps in rare moments, on the edges of an island understanding where sea meets land.

Well before I reached Japan, I had come to such an island, not so much through longing as through a series of losses that had systematically removed my family, my home, my friends and lovers, my tools, my livelihood, a large measure of my sanity, and very nearly, my life. Arriving in Tokyo, I realized that the last thing I had to lose was my language, and for the first time I saw what language was: a shared dream of the world from which its dreamers seldom ever wake, a virtually seamless emblem representing the accreted memories of a billion years of biological evolution, twenty-five millennia of human culture, and a few dozen centuries of historical wars and wanderings filtered through the personal fantasy of a single lifetime.

In Kawasaki, I taught English for a short time to the giggling receptionists of Fujitsu Electronics. I told them to pick a word, any word, and to repeat it over and over aloud until all sense had vanished, until they felt completely foolish and had to finally wonder at the meaning of the sounds they were making. If you listen closely, I told them, you will understand a great mystery. You will know what it's like for a fish to go to heaven. Management was not pleased with this approach, so we went back to practicing dialogs: "Hello? Mr. Smith? I'm calling about your appointment. Will three this afternoon be all right? Fine." And I circled the island from a great height, imagining I was a flying fish imagining it was a man teaching English. So I spent my youth in Asia.

Just as the dictionary had previously appeared as a fragile map of the territory I was about to venture into, on the eve of my departure to Japan I discovered a guide from another unsuspected quarter: artificial intelligence. The Japanese had just launched an ambitious research project in this relatively new and exciting area of computer science. I read a book about AI before leaving the U.S. and was immediately captured by some of the ideas it described, especially "knowledge engineering" and "natural language processing." Much later I learned that the philosopher Ludwig Wittgenstein had written "Die Welt is Alles was der Fall ist," which is to say: the world is everything that is the case. That covers a fair amount of ground, but given the rapid decomposition of everything I had assumed to be the case, perhaps it will be immediately obvious what so attracted me about artificial intelligence.

Here's the situation. Suppose you want a computer to solve a problem that human beings typically use their intelligence to understand. First you have to describe what's going on. Remember those story problems everyone always liked so much in algebra? "Johnny lives in Trenton, New Jersey and Pauline lives in Paris. If the Eiffel Tower's shadow measures 987 feet at 3 PM Eastern Standard Time, when will Johnny meet Pauline in Paris if he leaves New York at 6 AM and travels at a speed of 750 miles per hour?" Give up? Well, the computer gives up at "Johnny." It has literally no idea what a Johnny is, and no way at all of interpreting this tangled query. The computer's memory is a blank slate of millions of logic gates which can flip to 1 or flop to 0. That's about it. Try building a representation of everything that is the case from that.

Teach a computer to speak English. Now there was a challenge, especially for a sky-bound fish who no longer knew up from down, who had lost practically all sense of what people meant when they spoke to each other, could no longer tell how they knew when they knew -- if they knew -- what was going on in the words they used to send and receive signals representing internal states and the desire to modify same. We're talking an information-processing model of stone solipsism here, if you catch my drift. Immense storage, exquisitely sensitive receptors, sophisticated parallel architectures, representation schemata that loop back on themselves like infinitely nested Russian dolls: people. All dressed up and nowhere to go. "Good morning.... Fine, and you?"

The idea that captured me was not that this confusion of transmitted signals and their generators could be reduced to some sort of algorithmic explanation and the problem of language finally put to rest. Quite the contrary. I began to see the AI approach to modelling mind as allegorical proof of the power of ambiguity. The attempt to represent the world encounters deep paradox; the representation ultimately points back to its own symbols in a great circle. The language cannot escape itself. Likewise with human beings. Something happens in the heart and we don't know what. We see the moon and it tantalizes us, reminds us of something we once recognized somewhere else. But where? Is it a million-year old memory or the momentary resonance of deja vu? We assign symbols to our longing and build entire worlds from the names of our desire. There is a period of childhood wonder, for the individual as for the race, then a kind of blackout and we awaken in the ocean of language. "Good morning. Good evening. I love you too."

The solution is poetry not expert systems. AI and computational linguistics have tried to reduce language to formal logics, to a tractable system of rules and conditions, but it hasn't ever worked. We have been told for years to expect highly capable machine translators, natural language understanding systems, intelligent computers that can correctly interpret the full meaning of human speech. But they haven't come. The systems that have been created have turned out like brain-damaged genetic engineering experiments, lisping a few words that seem to make sense, then launching into demented gibberish. A few more rules here, a couple more conditions there, perhaps a brand new model, and maybe it'll work. What drives this search for intelligent computers? Is it simply the enormous profits they could generate by cloning the experientially acquired knowledge of expert human problem solvers? Or are the researchers looking for something else in there, some one who could answer all our questions, some reflection of our own original face? Does the fear ever arise that the ghost we find in the machine might be insane? Or that we may create a reflection of our own insanity, a la the Star Wars project? "Good morning. I love you too. Launch all delivery vehicles. Have a nice day."

The solution is poetry.

Entropy Gradient Reversals
All Noise - All the Time


Nothing to disclaim at this time.


This is the greatest electronic newsletter ever created. If you think so too, it's free. If you don't think so, the annual subscription rate is $1000. Either way, to subscribe send email to egr-list-request@rageboy.com saying simply "subscribe" on a single line in the BODY of the message. Or, just go to:


No Animals Will Be Harmed in the Making of This Subscription.

   Entropy Gradient Reversals
   CopyLeft Christopher Locke


"reality leaves a lot to the imagination..." John Lennon
Back to EGR HomePage

FastCounter by LinkExchange