The Immortal Game (13 page)

A
S MUCH AS FOR HIS
playing skills, Samuel Rosenthal was admired for his teaching, his writing, and his showmanship. “Sitting on his chair, blindfolded and motionless,” recalled an adoring obituary, “he appeared petrified in his extraordinary thinness. Only his stirring lips would indicate his next move.” Even without the blindfold, Rosenthal put on phenomenal public displays, as captured here in an 1891 Paris drawing by the French artist Louis Tinayre.

Samuel Rosenthal, the author’s great-great-grandfather, in a simultaneous display in Paris, 1891

Such demonstrations electrified the public, but were no longer the otherworldly oddity that they had been a century earlier in Philidor’s time. A number of chess masters now took part in blindfold games—so many, in fact, that as a group they attracted the attention of a young French psychologist named Alfred Binet, who was curious to understand the cognition behind them. How on earth did these players juggle memory and analysis so well? In the 1890s, as part of what would ultimately emerge as a career-long dedication to the definition and measurement of human intelligence, Binet was trying to understand the dynamics of memory. He became fascinated by blindfold chess players and their awesome displays of visual memory. Exactly how did they do it?

The conventional wisdom at the time, endorsed by Binet, was that strong visual memory was based in photographic-type recall. It appeared that great chess players somehow had a highly advanced ability to form mental pictures of chess pieces and boards and to preserve those pictures in their minds. They had, Binet theorized, an extraordinary “inner mirror,” which would forever reflect back to them, move by move, every successive configuration of the board. This notion was supported by more than two thousand years of memory literature and science that depicted memory as being visually based. The ancient Greeks, with no printing press and no pen and ink, had developed the art of mnemonics—mental tricks that relied on visualization to remember large amounts of detail. Typically, a mnemonist would “deposit” difficult-to-remember information into imagined compartments, seats, or rooms.

Now Binet wanted to determine how such memory tricks actually worked. Inspired by the work of the British anthropologist Francis Galton, he had developed a passion for exploring the healthy working mind, as opposed to the pathology of mental illness. The blindfold chess study was one of his first as assistant director of the Laboratory of Physiological Psychology at the Sorbonne. His subjects included the accomplished chess masters Stanislaus Sittenfeld, Alphonse Goetz, Siegbert Tarrasch, and the dean of French chess, Samuel Rosenthal. In Binet’s laboratory, they were questioned intensively about what they “saw” when they played chess blindfolded. The results were surprising and instructive. Binet was humbled to find that his “inner mirror” theory did not pan out. Astounding chess memories, he learned, did not resemble a collection of photographic snapshots. They were much more abstract than that, more geometric, and more
meaningful
.

An actual chess position (left) used in the 1890s Binet study, alongside a hand-drawn rendering by the master player Stanislaus Sittenfeld of how, with eyes closed, he pictured the position in his mind.

The intricate chess positions, it turned out, were not stored in chess masters’ brains as distinct photolike snapshots, but as a more abstract set of integrated patterns—like a musician’s chords or a computer programmer’s code. What looked like a chaotic field of data to the nonexpert was to the expert a coherent, meaningful song. “I grasp it as a musician grasps harmony in his orchestra,” offered French master Alphonse Goetz. “I am often carried to sum up the character of a position in a general epithet…it strikes you as simple and familiar, or as original, exciting and suggestive.”

In the mind of Goetz and the other chess masters, each portion of every game triggered impressions, feelings, and observations as meaningful to them as pieces of a car engine are to a mechanic, or as cloud formations are to a meteorologist. In these players’ minds, there were no sterile boards or carved wooden chess figurines—only evocative configurations that were familiar or somehow resonant. Ultimately, it wasn’t even the chess positions themselves that they were warehousing so much as the impressions they sparked. “It is the multitude of suggestions and ideas emanating from a game,” concluded Binet, “which makes it interesting and establishes it in memory.”

This insight was not inconsistent with long-standing visual notions of memory, but it provided a key clarification: visual memory operated not by recording a multitude of snapshots, but by encoding information in a meaningful context. It turned out that mnemonics was not so much visual as it was
meaningful
. Great chess players, then, were not simply finely tuned camera-computers, adept at acquiring and processing visual data with superlative efficiency. Rather, Binet’s study proved their craft to be supremely human—a combination of resonant feelings, meaningful experiences, and rich memories. Studying chess memory proved that abstract thought and memory were fully entangled with human feeling.

A further surprising revelation of the Binet chess study was the degree to which photographiclike recall of visuals could actually hamper visual memory. “Some part of every chess game is played blindfold,” explained leading German player Siegbert Tarrasch in a letter to Binet explaining his thought process. “The sight of the chessman frequently upsets one’s calculations.” This comment echoed the sentiment of other top players. What they remembered was not a tactile reproduction of the pieces on the board, but rather an abstract sense of each piece’s properties and movement. In fact, it was the mediocre players striving to recall actual pieces on a board who inevitably fell short. Binet’s photographic theory had not only been wrong; the truth was quite the opposite.

Binet’s observations marked the first stage in a century-long effort by scores of psychologists and cognitive scientists to understand how great chess players think—and to incorporate those lessons into other areas of cognition research. Justifiably proud of his pioneering discovery, Binet was also humbled at the study’s conclusion by how much more there was to understand about memory and thought. “Though we search and examine in the most minute details,” he wrote, “we cannot comprehend with precision the complexity of intellectual activity.”

He was impressed, too, by the degree to which chess turned out to be a model for the mind’s intricacy. “The blindfold [chess] game,” Binet observed, “contains everything: power of concentration, scholarship, visual memory, not to mention strategic talent, patience, courage, and many other faculties. If one could see what goes on in a chess player’s head, one would find a stirring world of sensations, images, movements, passions and an ever changing panorama of states of consciousness. By comparison with these our most attentive descriptions are but grossly simplified schemata.”

Binet’s original hypothesis might have been wrong, but his insight of chess as a powerful lens into the workings of the mind was astoundingly prescient. In fact, it gave birth to a century of chess investigation that would substantially help rewrite our understanding of the human mind. In the ensuing few decades, a few researchers followed up on Binet’s important work. But it wasn’t until 1946 that Dutch psychologist (and master chess player) Adriaan de Groot picked up where Alfred Binet had left off fifty years earlier. De Groot published a study called
Thought and Choice in Chess
, which investigated the skills, speed, style, and articulation of four separate skill levels of the chess player—from grandmasters to ordinary club players. Among his conclusions, de Groot startled the cognitive world with the observation that great players did
not
actually calculate significantly more or faster than lesser players. They also did not have better memories for raw data. Instead, they recognized more patterns more quickly, so as to make more relevant calculations and therefore better decisions.

With his work, de Groot helped to invent a new field of study—cognitive science—that aimed to systematize and deconstruct the thought process. Cognitive science was created by members of older, more established disciplines—psychology, neurology, linguistics, sociology, and anthropology. It was inherently interdisciplinary, a recognition that better understanding of the mind could be gained only through a steady dialogue among experts from these disparate fields. Chess was considered an essential tool of the new field, allowing researchers to study how the mind works as a machine, combining memory, logic, calculation, and creativity.

In 1973 Carnegie Mellon psychologists William Chase and Herbert Simon published two landmark works that built on de Groot’s chess work and that introduced one of the most important cognitive concepts in the twentieth century: a new understanding of memory called “chunking.”

Chunking is a memory technique used by all human beings to convert a collection of details into a single memory. Phone numbers, for example, are stored not as ten separate numbers but in three easy chunks: 513-555-9144. Remembering ten unrelated items in the right order is difficult; remembering three is no problem. The same technique applies to reading words, music, or any other complex array of symbols—including chess positions.

Chess, in fact, helped Chase and Simon formulate their theory in the first place. In their experiment they assembled three groups of chess players:
masters
(among the top twenty-five players in the nation);
experienced players
(ranked in the eighty-fifth percentile); and
novices
(who had spent little or no time studying the game). Each group was asked to:

1. Reproduce a particular board position after viewing it for five seconds.

2. Study an entire twenty-five-move chess game and recall a series of different positions from the game.

Based on their reading of de Groot’s work, Chase and Simon hypothesized that chess skill depended largely on what players already knew—as opposed to how much new data they could remember. The data fit the hypothesis. Superior players did not have intrinsically faster or better memory skills, but their vastly deeper, broader, and better-organized store of chess knowledge allowed them to recognize patterns faster and to form chunks quicker and more reliably. Their brains were not necessarily any faster than other brains; through much work, they had tuned them to be more
efficient
.

Chunking was a landmark discovery, one of those ideas so brilliant it immediately seemed obvious. But for cognitive scientists working with chess, it was only just the beginning. Chase and Simon declared chess to be the drosophila, or fruit fly, of cognitive psychology. Just as the fruit fly was the ideal laboratory model for heredity—the right genetic complexity, quick to reproduce, physical traits easily manipulated by genetic tinkering—so chess was for the study of the human mind. Its attributes were particularly suitable for scientists seeking to unlock questions about decision making, attention, and consciousness.