APPRECIATION : An All-American : She wrote the book on Pre-Columbian cuisine.

Sophie Coe was a serious food historian, but at heart she probably always remained the playful schoolgirl who kept a tarantula in her dorm room. She certainly kept an eye for the unexpected: The Aztecs relishing frog tamales and tadpoles stewed with corn. The Mayas raising bees in gigantic bee yards with thousands of wooden hives. Inca gourmets and their taste for a particular kind of clay called pasa as a sauce for potatoes.

She died May 15, just three weeks after the publication of her book “America’s First Cuisines” (University of Texas: $14.95 paper/$35 cloth, 288 pp.). Her death is a misfortune not only for those who knew her but for everyone who’s ever been interested in what people were eating in the Americas before Columbus.

In a way, her book’s title is a misfortune too. It calls to mind the media overkill about New World foods at the time of the Columbus quincentenary. But Coe knew vastly more about Aztec, Inca and Maya food than any of us journalists who were writing about the Columbian Exchange two years back.

She was born in Pasadena in 1933, the only child of the famous geneticist Theodosius Dobzhansky, who was teaching at Cal Tech at the time; later the family moved to New York, where her father taught at Columbia University and the Rockefeller Institute. She met her future husband when she was attending Radcliffe. Her roommate was engaged to Michael Coe’s roommate at Harvard, and the roommates were constantly trying to get the two together.

As it happened, though, Michael Coe and Sophie Dobzhansky first met in a physical anthropology lab. They were working side by side measuring the cranial capacity of skulls by filling them with mustard seeds and then counting them when he suddenly realized that this was the girl his roommate had been telling him about, the one who kept a tarantula in her dorm room. (Sophie had picked up the tarantula on one of her father’s trips to Brazil to collect the insects he studied; the Coe family still has it.) The Coes were married in 1955 and raised five children.

Michael Coe, who joined the Yale faculty in 1960, went on to specialize in the Maya and other ancient Mexican and Central-American peoples. The most recent of his 13 books is last year’s “Breaking the Maya Code,” a history of the decades-long struggle to understand the Maya hieroglyphic writing.

Sophie Coe started writing about food history in the early ‘80s, encouraged by Alan Davidson, who published her first food essays in his magazine Petits Propos Culinaires. “America’s First Cuisines” is a vast elaboration of those articles and the papers she presented at Davidson’s annual Oxford Symposium. At her death she had partly completed a book about chocolate, which her husband reportedly plans to finish.

She was a natural to study Central American food history. She knew the area, she’d read obscure 16th and 17th century European writings on New World foods, and as a Harvard anthropology Ph.D. (and the wife of a major student of the Mayas), she was thoroughly up to date on the latest archeological and anthropological studies.

*

Most of all, she was really interested in food. All kinds of food. In Tashkent last year, when I looked up the family of a great scholar of Uzbek and Central Asian food, they took me into the great man’s study and to my surprise showed me his correspondence with Sophie Coe (whose name they pronounced in quasi-Russian form as ko-yeh ).

Here and there in her writings, as she describes the flavor of some very odd thing, you suddenly realize it’s from first-hand experience. For instance, she mentions that the poisonous hydrocyanic acid extracted from manioc root can be boiled down into “a very tasty sweet-sour sauce.”

She once began a scholarly paper with the all-too-true words, “Archaeologists know an immense amount about cooking pots. Unfortunately, most of them know little or nothing about cooking.” With an attitude like this, it’s not surprising that “America’s First Cuisines,” scholarly though it is, never degenerates into a battle of footnotes.

The book concentrates systematically on the Aztecs, the Mayas and the Incas because most of the available historical material is about them. Aztec food is most familiar to us (as Coe points out, the most readily available pre-Columbian dish is guacamole, which is still often made simply of avocado, tomato, onion and Mexican cilantro). Still, as she shows, modern Mexican food and ancient Aztec food are not the same.

*

A sidelight on the Aztec character is their use of food as weapons. When they were upwind of their enemies, they’d set fire to chiles as a sort of tear gas. Even more diabolical, when they were besieging a city they’d roast and stew the fish and ducks of their lakes “so that the smoke will enter their city, and the smell will make the women miscarry, the children waste away, and the old men and old women weaken and die of longing to eat that which is unobtainable.”

Most people have heard about how the potato (and the health-foodie grain called quinoa) were domesticated in the Inca’s territory, and many have heard of various other tubers that are still eaten in Chile and Peru. It may be a surprise to learn that nasturtiums--among the edible flowers that have been showing up on people’s plates in restaurants--were also food to the Incas, but they only ate the tubers.

The Incas’ drink was corn beer ( asua ); as Coe points out, they hated water. In fact, they hated it so much that when the Spaniards wanted to punish an Inca, they’d force him to drink water.

Maya cooking is the least well known of the three, even though the Maya were the source of chocolate (and subsequent European ideas of flavoring chocolate, such as vanilla). At first glance their cookery looks close to Aztec, but there was always some Maya twist to it. For instance, instead of putting a lump of filling in a tamale, as the Aztecs did, they might spread out a thin layer of masa, cover it with a thin layer of filling and then slice it up, making tamales with a spiral pattern. Even today, the Maya make a 13-layered tamale representing the 13 levels they believe to exist in Heaven.

Not everything that was cooked in the Americas in pre-Columbian was equally good, Coe acknowledges, but in her opinion a number of promising food ideas have been forgotten in the intervening centuries. “We would do well to return to the kitchens of the Emperor Motecuhzoma and see what we could learn there,” she wrote. “Hot chocolate with chile is quite delicious.”

*

Before the coming of the Europeans, New World cooks had no cooking fats, not even vegetable oils. As a result, the pre-Columbian tamale was not the soft, lard-rich product made in Mexico today. In the following recipe, Sophie Coe recreated a variety favored by the ancient Maya. It makes a fat-free tamale with a good corn and bean flavor but quite a different texture--smooth, heavy and gummy.

The Mayas used the same dough for small loaves (“as long as the palm of the hand and four fingers thick”) that they cooked by barbecuing, frying on an ungreased griddle or baking in an earth oven. They seem even more archaic than the tamales. In baking (one hour at 450 degreees, if you really want to try it), the surface of the loaf dries, stiffens and cracks grotesquely, making a rather chewy contrast with the interior.

MAYA BREAD OF MAIZE AND BEANS 1 cup dried beans, soaked overnight Water 2 cups masa flour (not masa dough) or corn meal 1 teaspoon salt

In pot cook beans in water several hours until they begin to fall apart. Result should be about 4 cups of beans and liquid.

If using masa flour, let beans cool, then mix well with masa flour and salt. If using cornmeal, pour boiling beans on cornmeal and stir furiously, then add salt.

Form corn-bean dough into balls about 2 inches in diameter, wrap in corn husks or foil and steam 1 hour. Makes about 20 (2-inch) tamales.

Each serving contains about: 72 calories; 120 mg sodium; 0 mg cholesterol; 1 grams fat; 14 grams carbohydrates; 3 grams protein; 0.57 grams fiber.