Whether they're skewering satay in Singapore, patting puri in Poona or tossing salad in Wheeling, Ill., cooks of all cultures have one problem in common: turning a recipe for four servings into a formula for 400 -- or 4,000 -- servings.

The conventional wisdom is that expanding a recipe results in only a shadow of the original's delectability. At the very best, an expanded 36-cookie chocolate-chip recipe results in cookies slightly better than available off the shelf. And, at the very worst, the cook who expands the recipe for a luscious, four-portion stew walks into a 1,000-portion watery grave.

Is there, then, any other conclusion to draw -- that recipes cannot be expanded and, if they are, the end product is considerably different from the original?

In stark contrast to this conventional wisdom, passed on from generation to generation of cooks and bakers, the food industry is expanding recipes, and doing it successfully. Food processors that produce canned, frozen, dried or fresh apple pies, potato salad or zwieback develop their product lines by developing formulas in-house on tiny regulation household-size mixers, then multiplying them a thousand, or even a million, times. Or, they take a Mrs. Smith's or a Colonel Sanders' favorite recipe and expand it into a product that apparently satisfies millions of consumers .

What voodoo do they practice? Can the average cook or baker learn a thing or two?

"Give 40 people the same recipe and they'll create 40 different dishes. Then ask them to double the recipe and you'll have 80 different dishes," explains Wayne Gisslen, author of "Professional Cooking" (John Wiley & Sons, 1983), a popular text at universities teaching quantity food production. He has spent much of his life developing small, expandable recipes based on percentages, for large food processors, and measuring cup-based recipes for the average cook.

"There is a misconception that some secret mathematics drives recipe expansion," says Gisslen. "I do not know of a formula which cannot simply be multiplied by four to obtain that number of portions."

He attaches one big if to that statement: "that is, if other variables are not introduced. Changing mixer size or a longer cooling period may result in a different product. In such situations, one must alter the formula, the procedure or both."

The problems encountered in expanding a recipe are the same, whether you want to make enough of your famous peanut butter cookies for the school bazaar, or your rich uncle wants to produce enough to distribute up and down the east coast. They usually fall into these categories: (1) flawed recipe, (2) measurement, (3) surface area-to-volume ratio, (4) time, (5) equipment, (6) texture and (7) substitution.

Understanding these variables makes the difference between the successful expansion of a recipe and failure. Flawed Recipe

It is not uncommon for a favorite, four-serving recipe to taste, look or feel different when expanded. Sometimes the cause is a recipe flaw that is accentuated because of changed conditions of cooking.

Students in Home economics and nutritional sciences learn that lesson early in the educational curriculum.

According to Joan Koch, senior lecturer in nutritional sciences at Cornell University, most institutional recipes are multiples of a much smaller original and just a little extra salt or spice in the original recipe very often shows up in larger batches. Since larger batches must cook longer, there's a greater possibility of evaporation, which causes the flavors to become more concentrated.

To solve this problem, Koch tells her students to add only half the salt and half the spices, then adjust upward according to taste. They then rate the expanded product on a score card for saltiness and flavor. The new, expanded recipe and the altered salt and spice concentrations are recorded.

One should keep in mind, however, the undependability of the human tastebud. It is much affected by how tired or hungry you are, by the time of day, how many cigarettes you've smoked, what you ate beforehand and a myriad of other seemingly inconsequential factors. The original recipe might have been tasted and the amounts recorded during cooking, not during serving. If the cook had been tasting a little of this and a little of that, over and over, the expanded formula's taste might then reflect the recipe author's sensory overload. Measurement

Using the unreliable volume measurements (cups, teaspoons and tablespoons) so ingrained in our daily lives also makes recipe expansion an uncertain process. Gisslen says, "If everything were weighed and every cook had a balance, recipe expansion wouldn't be nearly the problem it is now."

It's hard, for example, to measure 3 tablespoons of chocolate and obtain an accurate measurement. Three tablespoons of shaved chocolate weigh far less than three tablespoons of chopped chocolate, which weigh far less than three tablespoons of melted chocolate. The problem, here, is density. It varies, depending on how coarsely or finely a substance is chopped or minced.

If the original recipe depended on chocolate grated on a small, knuckle-driven grater, it probably would not work well with chocolate chopped with a knife. The expanded recipe would contain more chocolate owing to the greater density of the chopped form.

Density differences are also caused by packing. Wheat flour is made of irregularly shaped starch granules and bits of protein. The irregular shapes, when sifted, stack upon each other much as cards are stacked to form a house. Disturb that stack by stirring or spooning and the wheat-flour card house collapses.

Cake flour weighs 5.2 ounces per cup unsifted and 3.3 ounces per cup sifted. Thus, the same recipe produces different cake textures and volumes depending on whether the flour was sifted before it was measured by the cup. Further variation can be caused by spooning the sifted flour, by double-sifting it into a measuring cup or by sifting it into a gallon measure, the bottom third of which is considerably packed together by the weight of the top two-thirds.

None of that would happen if the flour were weighed. In expanding someone else's recipe, who knows how much the flour was sifted -- if at all -- and, whether it was spooned or poured into the measuring cup?

The secret to repeatable results, then, is to record a formula by weight, not volume. Andy Wolf, vice president of research and development for Sara Lee, Inc. of Chicago, says that a recipe recorded as weights is in most cases expandable without any further adjustments.

"This is done by preparing the original recipe using the customary measuring methods," says Wolf. "Then weigh each ingredient and record." Surface Area-to-Volume Ratio

This fancy-sounding term causes trouble everywhere and for everyone. Each object has a surface and occupies space. How much surface and how much space makes the difference. Example of this ratio's importance are:

A restaurant cook expands a four-serving poulet chasseur (chicken in a tomato-onion-mushroom sauce) recipe to feed 100. She prepares it in a tilt-skillet (a large, shallow, stainless steel, steam-heated pan). She adds just the right amount of salt -- that is, according to the standard recipe which has always tested "just-right" in the past. The resulting sauce is a greasy brine.

What happened? The large surface area of this utensil allowed exessive evaporation of the sauce. This concentrated the salt and broke the fat-in-water emulsion normally stabilized by the sauce's flour. An emulsion is a suspension of minute fat droplets in water or stock; concentrating the water or stock leads to coalescence of those droplets into bigger drops that rise to the surface of the sauce.

Such a problem could be avoided by reducing cooking time.

Just the opposite problem may also occur in expanding a stew, soup or sauce recipe. Unlike the tilt-skillet batch, a poulet chasseur prepared in a large, deep, aluminum pot would have less evaporative surface exposed. Hence, the cooked stew would be more watery than the original, four-portion batch simply because less evaporation occurs. Wolf says that "in industry, we cut down on the moisture content by 2 percent when going from smaller to larger batches."

Surface area-to-volume ratio is important in bakeries, too, according to Ray Thelen, who writes "Better Baking," a troubleshooting and recipe column for Bakery Magazine, a production and marketing magazine published by Gorman Publishing Company.

"There is considerable difference in behavior between a large and a small dough," says Thelen. "In the household batch, the dough acquires more frictional heat owing to agitation by the mixer. That heat is quickly lost through radiation, as a small dough quickly adjusts its temperature to that of the surrounding air."

Large batches have a much greater volume in relation to their surface and "are therefore relatively insensitive to the temperature of their environment," according to Thelen, "and, instead of cooling rapidly after mixing, they actually gain heat and rise more rapidly toward the end."

Thus, the handling procedures for small bread batches do not apply for the larger batches. A large batch would take over the bakery if proofed for the same time as a domestic recipe would stipulate.

How does one adjust to this apparent dough schizophrenia? First, add ice water to larger batches. This does not hurt the yeast; it only slows it down. Second, keep large doughs well covered to prevent scaling, but do not put them in a warm spot to rise (unless you are baking bread in a Quonset hut along the DEW-line). Third, raise the salt content of the recipe. The normal salt threshold (before it tastes too salty) for bread dough is 1 1/2 teaspoons per 4 cups of flour. And fourth, lower the water content of the bread dough -- not too much, though, as an overly stiff dough gives bread the texture of a soft pretzel. Time

The time spent preparing a recipe is a variable not to be ignored. Robert Baker, a professor of food science at Cornell University, specializes in developing formulas for new poultry products, particularly sausages such as chicken or turkey hot dogs. He has been an amateur barbecuer for years, preparing feasts ranging in size from 10 to 500 chicken halves.

"The larger a barbecue's number of birds, the less barbecue sauce needed per bird. After years of discarding more than a gallon of extra sauce after each of the larger barbecues, I finally made this conclusion," he said. He says less time is spent basting, which in turn means less sauce drips into and is consumed by the flames.

Time is particularly crucial to doughs and batters containing butter and/or baking powder.

"The main thing that people forget when they go about converting and multiplying recipes, besides measuring problems, is time," says Bob Rodriguez, an instructor at the American Institute of Baking in Manhattan, Kans. The institute has trained a many bakers involved in large-scale production.

"They whip up a batch of 24 cookies quickly enough, think they're great, and then expand the batch only to find them greasy and crumbly," he says. The reason? In larger batches, the batter must wait while the oven of limited capacity bakes them in small increments. Someone accustomed only to the smaller formula might not stop to consider that the butter in the waiting batter will have melted before the batch goes into the oven.

Timing is crucial to seasoning, also. This is particularly the case with herbs.

"There are few volatile oils in herbs. These few are easily lost if cooked too long with a stew or soup," says Tom Burns, vice president of the American Spice Trade Assn., which supplies spice use and availability information to spice-import houses and food companies. "If you want to preserve an herb's full complement of flavors and aromas, you must add them only 15 minutes before the food is removed from the flame."

This is significant because if a four-portion chicken tarragon recipe requires five minutes to come to a boil, half an hour to cook and is served immediately, the tarragon retains most of its flavor and aroma. If expanded 100-fold and cooked in a large pot on top of a stove, it will take 30 minutes to come to a boil, 15 minutes to cook, and then most likely it will be held warm until served or allowed to cool.

To retain the more volatile tarragon aromas and flavors in the large batch, one would be wise to add the tarragon while the chicken is cooking, not at the beginning. One would also be wise to cool the batch quickly or serve it immediately. Equipment

As a batch increases in size, so must the equipment that chills, heats, stores and mixes it. Changing from the 1-quart to the 3-quart bowl on a domestic mixer with an increase in batch-size illustrates the importance of equipment differences.

In the 1-quart bowl, producing 2-cup batches of mayonnaise or whipped cream is especially easy. The beaters are sufficiently submerged to beat in the oil for mayonnaise and supply sufficient air to thicken the cream.

The story changes when either recipe is doubled. The 3-quart bowl, which has a flat bottom, allows only the tip of the beaters to be submerged. Unless you are prepared and know what to expect, 10 years of experience with the smaller bowl are of no use. The mayonnaise curdles every time. The reason? Little agitation occurs and an oil-in-water emulsification does not develop.

The same is true of the whipped cream, which is firm and smooth when whipped in the smaller bowl. Before whipping in the larger bowl, the cream only covers one-fourth the length of the mixer blades. That means that they are turning furiously, agitating the cream while incorporating less air. The result is at best a thin, lumpy foam and at worst, sweet butter.

A similar situation occurs with bread and pastry doughs.

"When a dough or batter mixes, just as much of the agitation comes from the surrounding dough as from the mixing or blending attachments, says Ron Zelch, who teaches bread production at the American Institute for Baking. "When a batch does not fill the bowl, the dough must be kneaded longer to develop the smooth elasticity characteristic of a well developed dough."

When expanding a pastry or bread dough recipe, then, one should take into account both the size of the dough and of the mixing bowl. The original recipe, which may fill the 1-quart bowl and take only 8 minutes to form a smooth, velvet dough, may if doubled only take up one-third of the 3-quart bowl. Because there isn't enough bulk in relation to mixing bowl and dough hook, 12 to 14 minutes mixing time may be necessary. Texture

The texture of a product is sometimes improved, other times damaged by recipe expansion. Croissants, danish, puff pastry dough and many breads are much superior in lightness and crispness prepared in large batches. That is, if they're prepared in a well-equipped bakery by a well-trained baker. If they're kneaded and rolled out by hand, however, large batches may be coarse texture and low in volume.

Other products -- stews, soups and sauces -- fare poorly when the recipe is expanded. Most home sauce and stew recipes are thickened with cornstarch or with a roux (flour and butter cooked together). When prepared skillfully, the sauces, stews and soups have a melt-in-the-mouth texture and pleasing appearance.

This cannot happen with large batches. "In the food industry, we must use starches that withstand the shearing action of pumping or the additional heat during canning," says Wolf of Sara Lee. "Otherwise, they would thin," he says.

This has little bearing on the amateur cook who quardruples her favorite cream soup for the church social. It does bear on freezing the leftovers or canning the soup so it can be enjoyed all winter.

In such situations, a roux -- the customary thickening for a cream soup -- could not be used. The canned soup would be thin even if the roux were doubled. When thawed, the frozen soup would turn to water with rubbery lumps floating throughout.

Even if a starch-thickened stew, soup or sauce is not to be canned or frozen, it can present problems when the recipe is expanded. For example, Aunt May's beef bourguignon for four thickens without fail. Normally, she stews the marinated beef and vegetables in wine and beef stock, then thickens it 15 minutes prior to serving. If she expands the batch size for the family reunion, two things may occur: One, the sauce never thickens. And two, it may thicken but then it thins when reheated.

If it never thickens, it's because she follows time instead of temperature. That is, she adds the roux and heats the sauce for 10 minutes on medium heat. That works for a small batch, bringing it to a boil so the starch granules gelatinize (absorb water and swell). It does not work for the large batch, which requires considerably more heat to reach the boiling point.

Thinning of the sauce on reheating is caused by the roux's sensitivity to acid in the presence of heat. A beef bourguignon contains tartaric and malic acids, put there by the wine. If the expanded stew takes half an hour to return to the boil and then an hour to cool, the starch might well lose much of its thickening capacity. The cause is dextrinization, an acid-induced breakdown of starch granules.

The secret to preventing both of these big-batch-related changes is to control time and temperature. To induce thickening, cook the stew in a large pot on a commercial stove. Better yet, find a steam-heated kettle, which can bring 20 gallons of water to a boil faster than a teakettle can muster a whistle. To prevent thinning, transfer the stew to smaller containers and plunge them into a running, cold-water bath. This will lessen dextrinization. Substitution

Expanding a recipe is hard enough. Substituting another ingredient or two at the same time will surely get you in trouble.

"People forget the differences between ingredients," says Rodriguez. "To prepare a cake, one should cream butter more than one would vegetable shortening." In the process of expanding a small butter cake recipe and substituting shortening for economic reasons, a shorter creaming period would be required. Otherwise, the batter would be too light, giving the cake an undesirable crumbliness.

It's not just a matter of changing procedure, either. If you substitute one ingredient for another -- say milk and vinegar for buttermilk or all-purpose flour and cornstarch for cake flour (two common substitutions) -- you will obtain a different product.

In so doing, you won't know whether it was the expansion or the substitution that caused the 100 pancakes more suitable for kneepads.