In 2007, the Campbell Soup Company started slashing the salt content in its canned soups in an effort to be more health-conscious. But the move, while it may have been well-intended, wasn’t well received: By July 2011, soup sales had dropped, and stock-market shares were down 5 percent. Faced with disappointing numbers, Campbell’s CEO Denise Morrison announced that month that the company was bringing back the salt, telling investors that the company would increase its sodium levels from 400 milligrams per serving to 650 milligrams.

The food industry is full of cautionary tales like this one. In 2004, General Mills released reduced-sugar versions of its cereals to lackluster sales, only to pull the cereals three years later. The British publication The Grocer reported in 2014 that sales of Sprite took a dive after Coca-Cola cut the drink’s sugar by one-third. In 2012, Mintel, a consumer-research firm, released a report that found that products with “low-sodium” labels had declined by 5 percent in the previous year due to low consumer demand.

“Existing salt replacements have not caught the imagination of consumers,” the report said. “Consumers are concerned about salt intake, but are not willing to compromise on taste.”

In the past few decades, salt and sugar consumption in the U.S. have both increased dramatically. According to the U.S. Department of Agriculture, in 2005 and 2006, the average American devoured around 3,400 mg of salt per day—almost 50 percent more than the recommended daily intake of 2,300 mg. The American Heart Association reports that the typical American man consumes 335 calories’ worth of added sugar each day, and the average woman 230 calories’ worth; nutritional recommendations, meanwhile, advise men to limit their added sugar to 150 calories and women to 100.

There are many reasons for the stubborn presence of salt and sugar in our diets. Both can act as preservatives or rising agents in processed food, and thus far food companies haven’t been able to find good substitutes for those functions. Sugar can lower a food’s freezing point—important when creating ice creams, frozen desserts, and freeze-dried foods—and can stave off staleness in baked goods. Salt has similar preservative properties and also contributes to texture of processed foods, like the stickiness of some doughs.

But amid mounting pressure from the USDA and other government agencies to make healthier processed food, the food industry’s biggest obstacle is still the simple fact that people won’t buy what doesn’t taste good. As an article in the 2013 issue of Food Business Industry News cautioned, failed early efforts to reduce salt and sugar made consumers more averse to labels like “low-salt,” “low-sugar,” and “low-fat.”

The solution, some food scientists believe, may not involve the taste buds at all.

Robert Sobel is the vice president of research and innovation at the flavor company FONA International. In the last few years, he’s been researching ways to use smells to trick our brains into thinking food contains high levels of sugar and salt, even when it doesn’t. Sobel first came across this concept, called “phantom aroma,” in a 2009 article called “Taste, Aroma, and the Brain” in the magazine Perfumer and Flavorist. The term, inspired by the neurophysiological phenomenon of phantom limbs, is the process by which the brain fills in the perception of a certain taste perceptions even when the ingredient may not exist.

We perceive a food’s flavor through a variety of stimulants—taste, of course, but also texture and smell. There’s a lot that’s still unknown about the neuroscience of taste, but the current prevailing theory is that we take in taste through the gustatory nerve and smells through the olfactory nerve, and information from both receptors combines in the orbital frontal cortex. In addition, past research has shown that we learn to associate certain smells with certain tastes—we associate the strawberry aroma with sweetness, for example, and lemon aroma with sourness.

So what happens when we smell something but don’t taste it? In one 2009 study, French researchers found that when participants ate something treated with a salt-associated aroma (like the smell of ham), they perceived that the food had more salt in it. The researchers theorized that when the brain has a learned perception of a certain type of food, it fills in for the missing taste on its own. “When you are tasting food you are perceiving several sensory dimensions—smell, taste, texture—and the brain is making a synesthetic perception,” said Thierry Thomas-Danguin, one of the study authors and a flavor scientist at the French National Institute for Agricultural Research. “When you are exposed to one dimension, your brain is reconstructing all the flavors and all the sensory dimensions, even if they aren’t there.”

In 2011, researchers from the Netherlands added support to these results in a study published in the Journal of Food Science, where the scent of beef boosted participants’ perception of salt in a low-sodium broth by 15 percent. (You can also see examples of the phantom aroma at work if you look at the ingredient label of a carton of low-sodium beef broth and see the words “natural beef flavoring.”)

It’s difficult to say how much phantom aromas are being used in the food industry—because flavors can’t be patented, most flavor companies are extremely secretive about their work—but judging from its frequent appearances in industry journals, presentations, and white papers, the concept seems to be gathering steam.

Over the last six years, Sobel has taken different aromas like vanillin—the compound that gives vanilla its distinctive smell—and tested them with people who rate how salty or sweet they think the food would taste. Predictably, people who smelled odors most commonly associated with sweet tastes (like vanilla), said they expected the food item to taste sweet. After each consumer panel, Sobel returns to his lab and tinkers with the concentration of aroma that he adds to the test food products. The trick, he says, is to get the aroma to barely detectable levels so that we don’t actually know that we’re smelling it, but we can perceive it when we’re eating the food. While we don’t particularly want a ham-smelling bread, for example, it may be possible to add just enough ham flavor to bread that we still associate it with a salty flavor without realizing the salt isn’t actually there.

The challenge for the food industry is that there is no-one size-fits-all rule for all aromas, says Pam Dalton, a researcher at the Monell Chemical Senses Center—each smell will require a different concentration, meaning that each food will require flavor scientists to start a new experiment. Sobel says that vanillin, for example, is less volatile, making itself known only in relatively high concentrations. Butyric acid, the aroma that gives off the cheesy smell in products like Cheetos, is much more volatile, meaning a smaller amount packs more of a flavor punch.

Because flavor is so associative, people tend to perceive only the aromas that fit with their ideas of how a food should taste, meaning much of the science of phantom aromas is culturally based. For example, Dalton’s team once experimented with benzaldehyde, which has a cherry/almond scent. They found that the benzaldehyde aroma boosted the perceived sweetness of a food product if the food also contained a sugar, like saccharine. But the aroma was undetectable if it was paired with a flavor we don’t normally associate with a cherry/almond scent, like monosodium glutamate (MSG). As Dalton recalled, only one researcher was able to detect the flavor in that case—he was native Japanese, Dalton said, and was used to associating a fruity flavor with MSG, which has an umami flavor, because it’s a common pairing in Japanese cuisine.

While phantom aroma can’t entirely replace a food’s salt or sugar content, it’s a good place to start, says Sobel. Through his experimentation, he’s been able to achieve a 10 percent reduction of salt in food products like chips, sauces, and soups. Based on past studies, researchers like Thomas-Danguin believe that phantom aroma, when combined with other salt-reduction techniques, may have the potential to reduce salt by more than 35 percent.

Many of the aromas the food industry works with can come from natural sources, but due to rising demand, a lot of them are synthetically produced. Vanillin can be extracted from the vanilla bean, but today, less than 1 percent of vanillin sold annually originates from vanilla-seed pods. Most is chemically synthesized from lignin or fossil hydrocarbons.

For this reason, not everyone in the food industry is impressed with the concept of phantom aromas. “I don't want to sound like a luddite, but the vast bulk of our salt intake comes from salt added by food processors who have so corrupted America's tastebuds,” said Michael Whiteman, a food and restaurant consultant and the founder of Baum and Whiteman, a restaurant-consulting company. Treating a food with more artificial flavors, he argued, isn’t the way to make it healthier.

On the other side, phantom-aroma proponents argue that it doesn’t make a difference whether the compounds are synthetic or natural. Karmella Haynes, an assistant professor at Arizona State University, argued in Slate in 2014 that synthetic vanillin has exactly the same chemical structure as natural vanillin. “Chemical analysis of vanillin from yeast reveals no additional atoms [and] no alien side groups attached,” she wrote. “And no tiny molecular boogie men that are prepared to pounce and kill the unsuspecting consumer.”

Still, the research into phantom aromas is so new that the verdict is still out. “We’re still looking into the implications of this research,” said Thomas-Danguin. “But it could help reduce a lot of salt and sugar in our diets, and that would be a good thing.”