Moonwalking With Einstein: The Art and Science of Remembering Everything

by Joshua Foer

  In Plato’s Phaedrus, Socrates describes how the Egyptian god Theuth, inventor of writing, came to Thamus, the king of Egypt, and offered to bestow his wonderful invention upon the Egyptian people. “Here is a branch of learning that will ... improve their memories,” Theuth said to the Egyptian king. “My discovery provides a recipe for both memory and wisdom.” But Thamus was reluctant to accept the gift. “If men learn this, it will implant forgetfulness in their souls,” he told the god. “They will cease to exercise their memory and become forgetful; they will rely on that which is written, calling things to remembrance no longer from within themselves, but by means of external marks. What you have discovered is a recipe not for memory, but for reminding. And it is no true wisdom that you offer your disciples, but only its semblance, for by telling them of many things without teaching them anything, you will make them seem to know much, while for the most part they will know nothing. And as men filled not with wisdom but with the conceit of wisdom, they will be a burden to their fellow-men.”

  Socrates goes on to disparage the idea of passing on his own knowledge through writing, saying it would be “singularly simple-minded to believe that written words can do anything more than remind one of what one already knows.” Writing, for Socrates, could never be anything more than a cue for memory—a way of calling to mind information already in one’s head. Socrates feared that writing would lead the culture down a treacherous path toward intellectual and moral decay, because even while the quantity of knowledge available to people might increase, they themselves would come to resemble empty vessels. I wonder if Socrates would have appreciated the flagrant irony: It’s only because his pupils Plato and Xenophon put his disdain for the written word into written words that we have any knowledge of it today.

  Socrates lived in the fifth century B.C., at a time when writing was ascendant in Greece, and his own views were already becoming antiquated. Why was he so put off by the idea of putting pen to paper? Securing memories on the page would seem to be an immensely superior way of retaining knowledge compared to trying to hold it in the brain. The brain is always making mistakes, forgetting, misremembering. Writing is how we overcome those essential biological constraints. It allows our memories to be pulled out of the fallible wetware of the brain and secured on the less fallible page, where they can be made permanent and (one sometimes hopes) disseminated far, wide, and across time. Writing allows ideas to be passed across generations, without fear of the kind of natural mutation that is necessarily a part of oral traditions.

  To understand why memory was so important in the world of Socrates, we have to understand something about the evolution of writing, and how different early books were in both form and function. We have to go back to an age before printing, before indexes and tables of contents, before the codex parceled texts into pages and bound them at the edge, before punctuation marks, before lowercase letters, even before there were spaces between words.

  Today we write things down precisely so we don’t have to hold them in our memories. But through at least the late Middle Ages, books served not as replacements for memory, but rather as memory aids. As Thomas Aquinas put it, “Things are written down in material books to help the memory.” One read in order to remember, and books were the best available tools for impressing information into the mind. In fact, manuscripts were often copied for no reason other than to help their copier memorize them.

  In the time of Socrates, Greek texts were written on long, continuous scrolls—some stretching up to sixty feet—pasted together from sheets of pressed papyrus reeds imported from the Nile Delta. These texts were cumbersome to read, and even more cumbersome to write. It would be tough to invent a less user-friendly way of accessing information. In fact, it wasn’t until about 200 B.C. that the most basic punctuation marks were invented by Aristophanes of Byzantium, the director of the Library of Alexandria, and all they consisted of was a single dot at either the bottom, middle, or top of the line letting readers know how long to pause between sentences. Instead, words ran together in an unending stream of capital letters known as scriptio continua , broken up by neither spaces nor punctuation. Words that started on one line would spill over onto the next without even a hyphen.

  ASYOUCANSEEITSNOTVERYEASYTOREADTE XTWRITTENWITHOUTSPACESORPUNCTUATI ONOFANYKINDOREVENHELPFULLYPOSITIO NEDLINEBREAKSANDYETTHISWASEXACTLY THEFORMOFINSCRIPTIONUSEDINANCIENT GREECE

  Unlike the letters in this book, which form words that carry semantic meaning, letters written in scriptio continua functioned more like musical notes. They signified the sounds that were meant to come out of one’s mouth. Reconstituting those sounds into discrete packets of words that could be understood first required hearing them. And just as it is difficult for all but the most gifted musicians to read musical notes without actually singing them, so too was it difficult to read texts written in scriptio continua without speaking them aloud. In fact, we know that well into the Middle Ages, reading was an activity almost always carried out aloud, a kind of performance, and one most often given before an audience. “Lend ears” is a phrase often repeated in medieval texts. When St. Augustine, in the fourth century A.D., observed his teacher St. Ambrose reading to himself without moving his tongue or murmuring, he thought the unusual behavior so noteworthy as to record it in his Confessions. It was probably not until about the ninth century, around the same time that spacing became common and the catalog of punctuation marks grew richer, that the page provided enough information for silent reading to become common.

  The difficulties associated with reading such texts meant that there was a very different relationship between reading and memory than the one we know today. Since sight-reading scriptio continua was difficult, reciting a text aloud with fluency required a reader to have a degree of familiarity with it. He—and it was mostly he’s—had to prepare with it, punctuate it in his mind, memorize it—in part, if not in full—because turning a string of sounds into meaning was not something you could do easily on the fly. The text had to be learned before it could be performed. After all, the way one punctuated a text written in scriptio continua could make all the difference in the world. As the historian Jocelyn Penny Small pointed out, GODISNOWHERE comes out a lot differently when rendered as GOD IS NOW HERE versus GOD IS NOWHERE.

  What’s more, a scroll written in scriptio continua had to be read top to bottom if anything was to be taken from it. A scroll has just a single point of entry, the first word. Because it has to be unwound to be read, and because there are no punctuation marks or paragraphs to break up the text—to say nothing of page numbers, a table of contents, chapter divisions, and an index—it is impossible to find a specific piece of information without scanning the whole thing, head to toe. It is not a text that can be easily consulted—until it is memorized. This is a key point. Ancient texts couldn’t be readily scanned. You couldn’t pull a scroll off the shelf and quickly find a specific excerpt unless you had some baseline familiarity with the entire text. The scroll existed not to hold its contents externally, but rather to help its reader navigate its contents internally.

  One of the last places where this tradition of recitation still survives is in the reading of the Torah, an ancient handwritten scroll that can take upward of a year to inscribe. The Torah is written without vowels or punctuation (though it does have spaces, an innovation that came to Hebrew before Greek), which means it’s extremely difficult to sight-read. Though Jews are specifically commanded not to recite the Torah from memory, there’s no way to read a section of the Torah without having invested a lot of time familiarizing yourself with it, as any oncebar-mitzvahed boy can tell you. I can personally vouch for this. On the day I became a man, I was really just a parrot in a yarmulke.

  Though years of language use condition us not to notice, scriptio continua has more in common with the way we actually speak than the artificial word divisions on this page. Spoken sentences flow together seamlessly as one long, blurry drawn-out sound. We don’t speak with spaces. Wher
e one word ends and another begins is a relatively arbitrary linguistic convention. If you look at a sonographic chart visualizing the sound waves of someone speaking English, it’s practically impossible to tell where the spaces are, which is one of the reasons why it’s proven so difficult to train computers to recognize speech. Without sophisticated artificial intelligence capable of figuring out context, a computer has no way of telling the difference between “The stuffy nose may dim liquor” and “The stuff he knows made him lick her.”

  For a period, Latin scribes actually did try separating words with dots, but in the second century A.D., there was a reversion—a giant and very curious step backward, it would seem—to the old continuous script used by the Greeks. Spaces weren’t seen again in Western writing for another nine hundred years. From our vantage point today, separating words seems like a no-brainer. But the fact that it was tried and rejected says a lot about how people used to read. So, too, does the fact that the ancient Greek word most commonly used to signify “to read” was ánagignósko, which means to “know again,” or “to recollect.” Reading as an act of remembering: From our modern vantage point, could there be a more unfamiliar relationship between reader and text?

  Today, when we live amid a deluge of printed words—would you believe that ten billion volumes were printed last year?—it’s hard to imagine what it must have been like to read in the age before Gutenberg, when a book was a rare and costly handwritten object that could take a scribe months of labor to produce. Even as late as the fifteenth century, there might be just several dozen copies of any given text in existence, and those copies were quite probably chained to a desk or lectern in some university library, which, if it contained a hundred other books, would have been considered particularly well stocked. If you were a medieval scholar reading a book, you knew that there was a reasonable likelihood you’d never see that particular text again, and so a high premium was placed on remembering what you read. You couldn’t just pull a book off the shelf to consult it for a quote or an idea. For one thing, modern bookshelves with their rows of outwardfacing spines hadn’t even been invented yet. That didn’t happen until sometime around the sixteenth century. For another thing, books still tended to be heavy, hardly portable objects. It was only in the thirteenth century that bookmaking technology advanced to the point that the Bible could be compiled in a single volume rather than a collection of independent books, and yet it still weighed more than ten pounds. And even if you did happen to have a text you needed close at hand, the chances of finding what you were looking for without reading the whole thing start to finish were slim. Indexes were not yet common, nor were page numbers or tables of contents.

  But these gaps were gradually filled. And as the book itself changed, so too did the crucial role of memory in reading. By about the year 400, the parchment codex, with its leaves of pages bound at the spine like a modern hardcover, had all but completely replaced scrolls as the preferred way to read. No longer did a reader have to unfurl a long document to find a passage. A reader could just turn to the appropriate page.

  The first concordance of the Bible, a grand index that consumed the labors of five hundred Parisian monks, was compiled in the thirteenth century, around the same time that chapter divisions were introduced. For the first time, a reader could refer to the Bible without having previously memorized it. One could find a passage without knowing it by heart or reading the text all the way through. Soon after the concordance, other books with alphabetical indexes, page numbers, and tables of contents began to appear, and as they did, they again helped change the essence of what a book was.

  The problem of the book before the index and table of contents is that for all the material contained in a scroll or between the covers of a book, it was impossible to navigate. What makes the brain such an incredible tool is not just the sheer volume of information it contains but the ease and efficiency with which it can find that information. It uses the greatest random-access indexing system ever invented—one that computer scientists haven’t come even close to replicating. Whereas an index in the back of a book provides a single address—a page number—for each important subject, each subject in the brain has hundreds if not thousands of addresses. Our internal memories are associational, nonlinear. You don’t need to know where a particular memory is stored in order to find it. It simply turns up—or doesn’t—when you need it. Because of the dense network that interconnects our memories, we can skip around from memory to memory and idea to idea very rapidly. From Barry White to the color white to milk to the Milky Way is a long voyage conceptually, but a short jaunt neurologically.

  Indexes were a major advance because they allowed books to be accessed in the nonlinear way we access our internal memories. They helped turn the book into something like a modern CD, where you can skip directly to the track you want, as compared to unindexed books, which, like cassette tapes, force you to troll laboriously through large swaths of material in order to find the bit you’re looking for. Along with page numbers and tables of contents, the index changed what a book was, and what it could do for scholars. The historian Ivan Illich has argued that this represented an invention of such magnitude that “it seems reasonable to speak of the pre- and post-index Middle Ages.” As books became easier and easier to consult, the imperative to hold their contents in memory became less and less relevant, and the very notion of what it meant to be erudite began to evolve from possessing information internally to knowing where to find information in the labyrinthine world of external memory.

  To our memory-bound predecessors, the goal of training one’s memory was not to become a “living book,” but rather a “living concordance,” a walking index of everything one had read, and all the information one had acquired. It was about more than merely possessing an internal library of facts, quotes, and ideas; it was about building an organizational scheme for accessing them. Consider, for example, Peter of Ravenna, a leading fifteenth-century Italian jurist (also, one gets the impression, one of the fifteenth century’s leading self-promoters) who authored one of the era’s most successful books on memory training. Titled Phoenix, it was translated into several languages and reprinted all across Europe. It was just the most famous of a handful of memory treatises created from the thirteenth century onward that helped make memory techniques that had long been the exclusive purview of scholars and monks available to a wider audience of doctors, lawyers, tradesmen, and everyday folks who just wanted to remember stuff. One finds books from the period on every variety of mnemonic subject, including how to use the art of memory in gambling, how to use it to keep track of debts, how to memorize the contents of ships, how to remember the names of acquaintances, and how to memorize playing cards. Peter, for his part, bragged of having memorized twenty thousand legal points, a thousand texts by Ovid, two hundred of Cicero’s speeches and sayings, three hundred sayings of philosophers, seven thousand texts from Scripture, as well as a host of other classical works.

  For leisure, he would reread books cached away in his many memory palaces. “When I left my country to visit as a pilgrim the cities of Italy, I can truly say I carried everything I owned with me,” he wrote. To store all those images, Peter started with a hundred thousand loci, but he was always picking up new memory palaces on his travels across Europe. He constructed a mental library of sources and quotations on every important subject, classified alphabetically. He boasts, for example, that filed away in his brain under the letter A were sources on the subjects “de alimentis, de alienatione, de absentia, de arbitris, de appellationibus, et de similibus quae jure nostro habentur incipientibus in dicta littera A”—“about provisions, about foreign property, about absence, about judges, about appeals, and about similar matters in our law which begin with the letter A.” Each piece of knowledge was assigned a specific address. When he wished to expound on a given topic, he simply reached into the proper chamber of the proper memory palace and pulled out the proper source.

  When the point of re
ading is, as it was for Peter of Ravenna, remembering, you approach a text very differently than most of us do today. Now we put a premium on reading quickly and widely, and that breeds a kind of superficiality in our reading, and in what we seek to get out of books. You can’t read a page a minute, the rate at which you’re probably reading this book, and expect to remember what you’ve read for any considerable length of time. If something is going to be made memorable, it has to be dwelled upon, repeated.

  In his essay “The First Steps Toward a History of Reading,” Robert Darnton describes a switch from “intensive” to “extensive” reading that occurred as books began to proliferate. Until relatively recently, people read “intensively,” says Darnton. “They had only a few books—the Bible, an almanac, a devotional work or two—and they read them over and over again, usually aloud and in groups, so that a narrow range of traditional literature became deeply impressed on their consciousness.”

  But after the printing press appeared around 1440, things began gradually to change. In the first century after Gutenberg, the number of books in existence increased fourteenfold. It became possible, for the first time, for people without great wealth to have a small library in their own homes, and a trove of easily consulted external memories close at hand.

  Today, we read books “extensively,” without much in the way of sustained focus, and, with rare exceptions, we read each book only once. We value quantity of reading over quality of reading. We have no choice, if we want to keep up with the broader culture. Even in the most highly specialized fields, it can be a Sisyphean task to try to stay on top of the ever-growing mountain of words loosed upon the world each day.

  Few of us make any serious effort to remember what we read. When I read a book, what do I hope will stay with me a year later? If it’s a work of nonfiction, the thesis, maybe, if the book has one. A few savory details, perhaps. If it’s fiction, the broadest outline of the plot, something about the main characters (at least their names), and an overall critical judgment about the book. Even these are likely to fade. Looking up at my shelves, at the books that have drained so many of my waking hours, is always a dispiriting experience. One Hundred Years of Solitude: I remember magical realism and that I enjoyed it. But that’s about it. I don’t even recall when I read it. About Wuthering Heights I remember exactly two things: that I read it in a high school English class and that there was a character named Heathcliff. I couldn’t say whether I liked the book or not.