10 Herbert Simon
The maze, the labyrinth, drew Herbert Simon irresistibly: a metaphor for understanding human choice and the patterns of a human life. His witty autobiography, Models of My Life (1991) was under the unacknowledged influence of an American classic, The Education of Henry Adams. But Henry Adams was a man suspended in puzzlement between two worlds, old and new, whereas Herb Simon was an active begetter of a very new world.
As we think, as we live, we choose step-by-step, turn-by-turn, Simon writes. Our environment offers those turns, those choices. Having chosen, we can’t go back. What lies ahead, we don’t know. We entertain goals, but they’re often vague. “They do not guide the search so much as emerge from it,” he writes (1991). No single turn we make guarantees to bring us closer to those goals. In such a labyrinth, minotaurs may lurk. We hope not but must press on regardless (Simon, 1991).
Simon was born and raised in Milwaukee—he declared a great book could be written about the disproportionate influence Midwesterners had in shaping 20th century America—and was educated at the University of Chicago, where he received his BA in 1936 and his PhD in 1943. He headed a research group at the University of California between 1939 and 1942 and then taught at Illinois Institute of Technology until 1949, when he moved to Carnegie Mellon, where he remained the rest of his life.
From the beginning, Simon’s interests were capacious: municipal administration, political science, mathematical economics, cognitive psychology, computer science. Yet they circled a central theme: how do people make rational choices?
His first book, Administrative Behavior, was based on his doctoral dissertation and examined the prevailing idea in business and economics that any completely rational choice must take into consideration all the alternatives, the possible consequences of each alternative, and compare the accuracy and efficiency of each consequence.
No, he scoffed. For humans, complete rationality is absurd. Human rationality is bound by constraints, such as time, available information, where the decision-maker stands in an organization, and so on. Humans practice only bounded rationality, as he called this principle. His argument was squarely opposed to that of classical economists, who kept pretending that humans always made omniscient rational economic choices. His dogged belief in the principle of bounded rationality would earn him a Nobel Memorial Prize in Economic Sciences nearly forty years later.
Simon saw that human decision-making and problem-solving—rational but fallible—were complex and poorly understood. By 1954, he thought the best way to understand these processes was to simulate them on a computer. In the younger Allen Newell—whom he met at RAND Corporation in Santa Monica and brought back to Carnegie Mellon as his PhD student and later a faculty colleague—he found a research partner who shared those convictions. With the computer, they were confident they could model some small but important parts of human thinking.
In the pivotal year of 1956, even as Simon and Newell were absorbed with the Logic Theorist, Simon published a scholarly paper, “Rational Choice and the Structure of Behavior.” It proposed ideas that impelled him to write his only short story, “The Apple.” A young man called Hugo lives in a castle with many rooms, each room with several doorways to choose from. Though he can move forward through one of those doorways to an adjoining room, he can never go back. He can see into adjoining rooms, but he can never know what he’ll find far ahead. As Hugo (you go, Everyman) develops preferences—for food, for art on the walls—the story slowly reveals the cost of pursuing those preferences.
Simon thought the story was finally about the burden of choice, the search for meaning. It can also stand for the artist finding a way through choices to be made, a world where overpainting, rewriting, and erasing are forbidden. Nor does Hugo encounter a beloved (or despised) Other in any of those rooms. Though it’s only a fable, Hugo’s lifelong isolation is chilling.
In late 1960 or early 1961, when Simon, on a year’s leave from Carnegie Mellon, was at RAND, Ed Feigenbaum brought him a copy of Jorge Luis Borges’ Ficciones. Simon was thrilled to discover “The Library of Babel,” a tale of one of the greatest labyrinths. Ten years later, as a visiting lecturer in Argentina, Simon asked to meet Borges. In his autobiography, he recounts that meeting in some detail, how much they found to talk about—philosophy, mathematics, poetry—yet how he finally understood that, although the labyrinth is fundamental in Borges’ fiction, no great abstract model lies beneath. “He wrote stories; he did not instantiate models. He was a teller of tales,” Simon writes, reconciled to the difference between impulses that drive scientists and those that drive artists (1991).
When Herb Simon told me some of this in 1974, we’d already become friends. After work each evening, he walked home past our house on Northumberland Street in Pittsburgh. (He had for years and would for years more after Joe and I left that house.) I’d just be putting the cover on my typewriter and might catch the top of his black beret or his wintertime chu’ulla sailing past the front hedge. I’d lean out the door, ask him if he’d like to stop for a sherry. He nearly always would. These encounters took place almost weekly.
Over sherry, we’d talk AI shop but also range broadly. Simon was interested in everything. In our formal interviews for Machines Who Think, he’d say offhandedly, “Oh yes, I used that because I was teaching myself Greek at the time;” or, “that came about because I was teaching myself Hebrew;” or, “by then I’d got hold of de Groot’s book, and was translating it from the Dutch.” (Adriaan de Groot’s study of how chess masters played, as distinct from ordinary players, would help Simon form his theories of how experts became so, how long it took, and how they operated differently from nonexperts, ideas that animated a later best-seller by Malcolm Gladwell.) Simon had such a good feel for languages that he’d simply open a book in the language he wanted to learn and keep reading until he got it. Eventually he could read professional papers in twenty languages and literature for pleasure in half a dozen.
So in our casual conversations, we liked to talk about cognates and how they changed from language to language. We talked about how languages were structured, verb before subject, verb after subject. Once he’d taken up painting so intensively that he had to step back, afraid it would encroach on his research time, but he often carried a sketchbook—so we talked about art. He was a decent amateur pianist, and we talked about music. Late into our sherry, Simon wasn’t above doing a little dishing about his colleagues, perceptive and funny. He didn’t suffer fools gladly, and he taught me how to look out for them. We talked about everything from the then sterile state of the humanities to the backward state of the board of trustees of Carnegie Mellon, of which Simon was a member. After one such three-and-a-half hour chat, I was sorry when he had to leave and regretted I hadn’t asked him one impertinent question: what’s it like to be so much smarter than everyone else you meet? Is it a drag? Fun? Don’t care? But I’d posed enough for one day, including why he stayed at CMU when more glamorous places were always asking for his hand.
Other people must have wondered because he addressed that question in his autobiography. He conceded how competitive he was, but the competition had to be “both stiff and fair.” He’d never believed he had to be at Harvard, Stanford, or MIT to win the academic game. He wanted “to win without conspicuous social support, whether from family or university. Then it would be clear that I had won ‘fairly,’ and not just by using the hidden, or not so hidden, weapon of a superior environment” (Simon, 1991).
As to the first question I didn’t ask, how did it feel to be so much smarter, I can only guess that because he was enthusiastically sociable all his life, he didn’t mind the Mendelian shuffle that had endowed him so generously and the rest of us with less.
For the history of AI I was writing, I was reading the field’s technical papers, forcing myself to understand them. Sometimes the going was so hard I read through tears of bewilderment. But my agnosticism about AI was giving way.
From my journal, November 3, 1974:
Herb declares I’m not a real humanist, since I’m willing to admit human values could be transmitted by extra-human forms, e.g., computers. I’ve come a long way from the time when I took offense at the idea of computers writing novels. Now I think I’d welcome a new form of intelligence to live in parallel with us. To replace us? That’s Herb’s referendum: we vote to see whether computers (“beasties” he calls them) which are less prone to human frailty, and which share human values, and can perpetuate those values better than ordinary flesh and blood humans, should be allowed to replace us. I didn’t reject it out of hand, which astonished him.
Although I didn’t know it then, Simon had tried out this idea before—he’d once nearly been ejected from a Yale dinner party for even proposing the question. He would’ve agreed that it was misleading to ask what might perpetuate human values better than frail humans, without conceding, first, that human values were elastic and, second, that any such beasties might have their own values.
One afternoon we talked about national senses of humor. “You’re English,” he said. “Maybe you can explain the curate’s egg to me. I just don’t get it.” I’d never heard of the curate’s egg, so he told me. It derives from an 1895 Punch cartoon, where a young curate is having breakfast with his bishop. The caption reads: “Bishop: ‘I’m afraid you’ve got a bad egg, Mr Jones.’ Curate: ‘Oh, no, my Lord, I assure you that parts of it are excellent!’”
I was convulsed. I couldn’t explain why, but I was. Much later I decided that this was the reduction to the absurd of all my proper English parents’ admonitions: “Don’t make a fuss. Make the best of it. Always be polite. If you can’t say something nice, say nothing. No matter what.” But that was later. Herb shook his head in puzzlement and left no wiser than he’d arrived.
February 3, 1974:
Herb Simon said tonight, quoting St. Augustine, that thought was pure form, an idea I have to mull over, though it sounds plausible. A lively lecture to a packed hall. A delightfully sensual talk, too: “Our intelligence makes sex the set of rich fantasies that it is;” or, “What if you have to pull a little fact out of your head, like the Latin for ‘I love,’ Amo.” “Is he signaling someone in the audience?” Joe murmured, to tease me, because he knows I have a sweet tooth for Herb.
April 13, 1976:
Then home, and over dropped Herb for tea. A lovely talk about a million topics, most particularly, did he feel let down after he’d given a talk? Absolutely, even after all these years. What makes you good is the adrenaline you pump into yourself for the occasion, and it seems your body can’t let you down gradually but slumps instantly. “You’re so grateful for crumbs of praise,” he added, and I agreed, though wondered, even he…?
May 11, 1976:
Reading Swann’s Way again. Like rich food, one must do it in small portions. But oh! How did I read it eighteen years ago, enchanted by it, when I’d never thought of writing myself? It’s such a writer’s book. Yet much beloved by Herb Simon, who’s read all volumes through twice in the original French.
For Simon, Proust was the exquisite artistic representation of memory, which had preoccupied him for a long time. But reading, he said in his autobiography, was more than a mere hobby; it was one of life’s main occupations. “As with eating, so with reading. I am nearly omnivorous. But my stomach for words is hardier than my stomach for rich foods, so I do not ration myself” (1991).
September 22, 1976:
A long talk with a student who reads my published words back to me very flatteringly. But as I said to Herb, who stopped for a brief afternoon sherry, we do it, writing, for love. And Herb said, yes, I do science for love too, and we whooped and giggled about our vulnerability—God bless him for being so honest. And laughed too at how we participate in that stylized game: we say in print “Love me, for here I am,” and then the critics say, “I don’t love you,” also in print for everyone to see. We could laugh and recognize our own idiocy. Herb is thinking over doing his memoirs, and is dubious. “Writers do it so well because they’re writers. Mark’s autobiography was so fine it really discouraged me.”
Our friend, novelist Mark Harris, had just published an autobiography, The Best Father Ever Invented.
Over those sherries, the Squirrel Hill Sages were born. I was complaining to Simon that my students had the really interesting conversations—the meaning of life, and all—whereas in the University of Pittsburgh’s English Department, I was stuck with squabbles over the photocopying budget, whether the course in Romantic poetry should be one semester or two, and other chutes to tedium. Simon said that he and Dorothea, his wife, had run a little salon when they were at the University of Chicago. People met every Sunday night with the understanding that the gathering wasn’t for small talk, but for tackling the important themes.
“You could do that here,” he said. “But it would be wise to choose a topic ahead of time.” Over the summer of 1976, Simon and I corresponded (by U.S. mail in those days) refining the idea: there would be eight of us only; we’d choose a topic ahead of time; we’d meet after dinner, so nobody had to rush around hosting; and if it worked out the first time, once a month seemed fine.
For the first few months, I privately called the group the Squirrel Hill Sages, after Squirrel Hill, the Pittsburgh neighborhood where we all lived. I soon confessed to my fellow Sages, and they laughed and adopted the name. We were Herb Simon and his wife, Dorothea; Allen Newell, and his wife Noël; Mark Harris, the novelist, and his journalist wife, Josephine; Joe and I. None of us was exactly shy, and the conversations were deliciously lively.
Two topics kept us going long after our usual 10 p.m. ending time. The first was arete, the Greek notion of excellence.
November 14, 1976:
Last night was the second meeting of the Squirrel Hill Sages. Our topic was arête, Excellence, and it was a big hit—we yakked for three and a half hours without a break. We decided that arête was both private and public, that you must know yourself to have it (hence no false modesty) and you must demand an acknowledgment from others. You must be excellent in comparison to others of the species (thus relative rankings of arête). We decided it implied for us specialization, though that might not be what the Greeks intended. A funny insight: Allen Newell isn’t at all turned on by the winner of the decathlon, but loves the idea of the other winners of single events—save the pole vaulters, “because that’s only a matter of better technology!” Joe and Herb and I howled; was this Allen Newell, a man whose whole career is based on a better technology? As Joe said, he didn’t even look contrite. The only people, we decided, who are worse than those who have arête and don’t know it are those who don’t have it and think they do.
The talkers were mostly the men (Noël had already pointed out what a competitive, masculine notion arête was). The scholars among us were dazzling—Dorothea noting that St. Paul had used the word three times in one of his letters, variously translated as triumph; virtue or excellence; and knowledge. Allen noted its root suggested warlike valor, and also quoted the Meno, which Herb then objected was middle Greek sort of thought, unlike the early phase of arête, which intended war-like virtues. Josephine compared arête to the Japanese concept of shibui, saying that the economy and restraint inherent in or implied by shibui is different from the ostentation—or at least lack of humility—arête implies. And so it went. Me, I was astonished that everyone had taken it all so seriously that they’d do homework. But pleased too, of course. Much fun, much fun. But how exhausted I was at the end of it; what concentration it takes to toe dance with the Sages.
The second topic to keep us late was Criticism, a discussion that took place in Mark and Josephine Harris’s living room. Nearly everyone in the room had put something out in the world and watched the critics gorge themselves.
We spoke of doing our work for love, for future love, for present gratification; of the differences between art and science (although art is for laypersons and science for specialists, in places the two overlap). Of the difference in caliber of critics—critics were often nonpractitioners in the arts, quasipractitioners in the sciences; of different sets of standards, slightly more uniform in the sciences than the arts.
The discussion got personal. I watched as Simon’s face, usually guarded, displayed a spectrum of expressions. Finally he spoke. “There are two definitions of criticism. One is where the critic looks at your work, explains it or describes it to people who wouldn’t otherwise know about it, and if judgment is made, it’s tentative and with some respect for the effort that has gone into the work under discussion.” (That’s how he spoke, in language precise and informed.) “Then there’s the other kind.” He looked around the room, a grin that might be a snarl at any moment. “This is when the critic is only interested in advancing an agenda, and your work is the innocent victim.” Ah yes, we all knew.
After a while, I said to Simon: “Who do you write for?”
“Come on,” I teased.
“They’re not all alive.”
“Come on,” I coaxed.
He gave me a look that was jokey, sheepish, and maybe a little proud. “Oh, Aristotle, for one.”
The Squirrel Hill Sages met regularly for perhaps two years, until Joe and I left Pittsburgh. In one sense, we were the living embodiment of cross-disciplines, feeding each other gladly, a negation of the premise of the Two Cultures. Recall my description of how earnest the life of the mind is outside the cultural capitals. But it was also me, my own yearning to bridge the Two Cultures. I started the Sages. At the turn of the millennium, Joe and I started a salon in Santa Fe, New Mexico, where we then had a second home, and the gathering extended for a week between Christmas and New Year’s to tackle meaty issues. The salon met annually for more than fifteen years. I still belong to informal discussion groups with serious purposes in New York City. Mea culpa. But such groups coalesce because so many of us yearn for those bridges.
“Terminally earnest,” a friend once teased me. At least I think he was teasing.
Herb Simon was finally a paradox. He was brilliant, generous, visionary, and deeply engaged with the world, at the same time he was shy and despised smalltalk, which made him seem remote. He loved the arts—that obsession with painting, that omnivorous polyglot reading—as much as he loved the sciences. He had not a shred of false modesty, and to paraphrase Churchill, he had little to be modest about. He was intellectually self-confident enough to admit that the Simons subscribed to the daily newspaper so Dorothea could read the news while he read only the comics (a taste he and I shared, though I read the news too). He dressed like the color-blind man he was, blacks and browns, sometimes together, and aside from his few miles walk each day to and from work, I don’t believe he ever indulged in other exercise. (He claimed his ambivalence about owning a car had only been resolved when he and Dorothea became too embarrassed to ask their friends to drive babysitters home.)
But contradictory aspects lurked and drove him. For a man who studied and honored rationality, however bounded, he was deeply passionate (his reputation for having a significant temper; those keen flirtations he confessed to in his autobiography that remained chaste, because he was too fearful of being rejected, and because of his deep loyalty to Dorothea). When critics claimed that AI was failed promises, they might have meant some of the claims Simon made: the four predictions he made in 1957, which weren’t soon to be fulfilled (he explained why, not entirely convincingly, for Machines Who Think); his flat 1965 assertion that any job a human could do would, in twenty years, be done by a machine (not in twenty years; not in fifty). What impelled him to such farfetched stuff?
He was sharply competitive, with a perplexing belief that, for winning to count, he must come from behind. His psychological needs must have been demanding, needs that left little room for his children, two of whom eventually went off to boarding school. (I speculate. The issue is opaque, and might have been Dorothea’s; might have been the children’s.) He had a well-honed debating style. I’d watch him tangle with anti-AI people in formal debates, and though I agreed with him, wouldn’t willingly have exposed myself to that weed-whacker rhetoric.
People like me, like most of his students, like many of his colleagues, loved and admired him. Manuela Veloso, for many years the Herbert A. Simon Professor of Computer Science at Carnegie Mellon, and a world renowned roboticist, remembers that when she was a young faculty member, struggling for publications, she felt particularly down when a paper she thought was excellent was rejected by a journal. Over lunch with Herb, she confided in him. “Of course that kind of thing never happens to you. I mean, you’ve got a Nobel Prize, and all.” He shook his head. “Of course it happens to me. But Manuela, you’ve got to be your own evaluator. It’s got to come from inside. You know when you’re doing something good. Don’t be too dejected by lack of recognition, or even elated by the opposite. That’s a roller coaster. Keep faith in yourself.” Veloso never forgot that crucial advice.
People exposed to his darker sides were less warm. As well as affectionate friends, he had an abundance of bitter lifelong enemies and, I now see, he generated considerable professional jealousy. Nobody needed to be that good at so many things.
With Dorothea Simon, I never got past a certain cool friendliness. In retrospect, it couldn’t have been pleasant for her to think of her husband stopping weekly for sherry and talk with a much younger woman, but those conversations were so cerebral that, at the time, it never occurred to me. She’d been a gifted and, from photos, a startlingly beautiful young woman, already doing graduate work at the University of Chicago, when she met and married Herb. She was expected then to give up her own intellectual and professional life to serve his. She mostly did, until their children were grown, and she could go back to school for research in education and learning. The Simons even published a few papers together on those topics. When the Squirrel Hill Sages met, she was penetrating, well-informed, and articulate.
She was noticeably cool to my 1970s feminism, perhaps because it seemed to rebuke everything she’d sacrificed herself for. Yet one night after one of Herb’s public lectures, I saw her receive gushing praise from a member of the audience for her husband’s talk. She responded graciously. But in her eyes, in her stance, was pain. I wondered if she was regretting what it might have been like to receive that praise on her own behalf.
- In subsequent years, two more Nobel Memorial Prizes in Economic Sciences were awarded to researchers who pushed the idea of behavioral economics further: Daniel Kahneman (2002) and Richard Thaler (2017). Arguably, the 2018 laureates, Paul Romer and William Nordhaus, also represent behavioral economics, though indirectly. ↵
- “The Apple,” whose structure would be familiar to any videogame player, also anticipates the idea of “expanding into the adjacent possible,” the mechanism of biological and social evolution proposed by theoretical biologist Stuart Kauffman in Investigations: The Nature of Autonomous Agents and the Worlds They Mutually Create. Kauffman, however, despises the concept of artificial intelligence, no matter how hard I try to educate him. ↵
- At the time, French postmodernism and other word-salad nonsense began to capture an exhausted field. The physicist Murray Gell-Mann swore to me that in his archives is a note Simon passed to him during a meeting of the President’s Science Advisory Committee (an body of eminent scientists that would be swiftly disbanded when it failed to support President Richard Nixon’s pet project, the supersonic transport plane). According to Gell-Mann, Simon’s note says: “Help me stamp out the humanities.” ↵