You overhear a lot of strange things in coffee shops, but an order for an “almond-based dairy-alternative cappuccino” is not one of them. Ditto a “soy-beverage macchiato” or an “oat-drink latte.” Vocalizing such a request elicited a confidence-hollowing glare from my barista when I recently attempted this stunt in a New York City café. To most people, plant-based milk is plant-based milk.
But though the American public has embraced this naming convention, the dairy industry has not. For more than a decade, companies have sought to convince the FDA that plant-based products shouldn’t be able to use the M-word. An early skirmish played out in 2008 over the name “soy milk,” which, the FDA acknowledged at the time, wasn’t exactly milk; a decade later, then-FDA Commissioner Scott Gottlieb pointed out that nut milk shouldn’t be called “milk” because “an almond doesn’t lactate.” To be safe, some fake-milk products have stuck to vaguer labels such as “drink,” “beverage,” and “dairy alternative.”
But a few weeks ago, the FDA signaled an end to the debate by proposing long-awaited naming recommendations: Plant-based milk, the agency said, could be called “milk” if its plant origin was clearly identified (for example, “pistachio milk”). In addition, labels could clearly state how the product differs nutritionally from regular milk. A package labeled “rice milk” would be acceptable, but it should note when the product has less calcium or vitamin D than milk.
Rather than prompt a détente, these recommendations are sucking milk into an existential crisis. Differentiating plant-based milk and milk requires defining what milk actually is, but doing so is at odds with the acknowledgement that plant-based milk is milk. It is impossible to compare plant-based and cow’s milk if there isn’t a standard nutrient content for cow’s milk, which comes in a range of formulations. This awkward moment is the culmination of a decades-long shift in the way the FDA—and consumers—have come to think about and define food in general. At this point, it’s unclear what milk is anymore.
Technically, milk has an official definition, together with more than 250 other foods, including ketchup and peanut butter. In 1973, the FDA came up with this: “The lacteal secretion, practically free from colostrum, obtained by the complete milking of one or more healthy cows.” (Yum.) The recent guidance doesn’t override this definition but doesn’t uphold it either, so milk’s status remains vague. The agency doesn’t seem to mind; consumers understand that plant-based milk isn’t dairy milk, a spokesperson told me. But the FDA has long allowed for loose interpretations of this standard, which is why the lacteal secretions of sheep and goats can be called “milk.” As time goes on, what can be called “milk” seems to matter less and less.
At one point, names mattered. In the late 1800s, people began to worry that their food was no longer “normal and natural and pure,” Xaq Frohlich, a food historian at Auburn University who is writing a book on the history of the FDA’s food standards, told me. As food production scaled up in the late 19th century, so did attempts to cut corners with cheap products parading as the real thing, such as margarine made with beef tallow. In 1939, the FDA began establishing so-called standards of identity based on traditional ideas of food.
But the agency’s food definitions were malleable even before oat milk. The agency hasn’t been very strict about standards of identity, because consumers haven’t either. Around the 1960s, as people became aware of the ills of animal fat and cholesterol—and purchased the low-fat and diet foods that proliferated in response—the agency moved away from defining the identity of food toward a policy of “informative labeling” that provided nutritional information directly on the package so consumers knew exactly what they were eating. It became accepted that food was something that could be “tinkered with,” Frohlich said, and what mattered more than whether something was natural was whether it was healthy. In the midst of this change, milk was assigned its official identity, which came with caveats for added vitamins. Loosely interpreted, “milk” soon came to encompass that of other ruminants, as well as chocolate, strawberry, skim, lactose-free, and calcium-fortified stuff.
In this context, the FDA’s recent expansion of this standard to accommodate plant-based milk is to be expected; Frohlich doesn’t think the plant-based or dairy industries “are particularly surprised by this proposal.” Very little will change if the new guidance becomes policy. (The decision has to go through a public-comment period before the FDA issues the final word.) If anything, there may be more plant-based products labeled “milk” at the supermarket, and perhaps the new labels will stave off any potential confusion that occurs. Pointing out nutritional differences between plant-based and dairy milk on packaging, the FDA spokesperson said, is meant to address the “potential public-health concern” that people will mistakenly expect these products to be nutritional substitutes for each other. But the nutritional value of dairy milk varies depending on the type, and in some cases, the nutrients are added in. Milk is just confusing, and perhaps that’s okay. For most consumers, milk will continue to be milk—a white-ish fluid, sourced from a variety of plants and animals, and ever-evolving.
Milk aside, for most modern consumers, what to call a food matters less than other factors, such as what it consists of, where it comes from, how it’s made, and its impact on the planet. “Public understandings of food have really changed since the early 21st century,” Charlotte Biltekoff, a professor of food science and technology at UC Davis, told me. In some cases, people don’t define food by what it is so much as what it does. Many plant-based milks, Biltekoff said, don’t look or taste much like dairy milk but are accepted as milk because they’re used in the same way: splashed in coffee, poured into cereal, or as an ingredient in baked goods. In short, trying to define food with a standard identity can’t capture “the full scope of how most people interact with food and health right now,” she said. A name—or, indeed, a label pointing out nutritional differences between dairy and plant-based milk—can encompass only a fraction of what people want to know about milk, all of which is beyond what the FDA can regulate, Biltekoff added. No wonder its name doesn’t seem to matter much anymore.
That’s not to say that all food names will eventually become diffuse to the point of meaninglessness. It’s hard to imagine peanut referring to anything but the legume, but then again, a debate over what counted as “peanut butter” lasted for a decade in the ’60s and ’70s. Naming clashes, in all likelihood, will occur over staple foods that already attract a lot of scrutiny and are produced by powerful industries, such as eggs or meat. For example, Americans use the term meat flexibly: In addition to animal flesh, it can also refer to products made from plants, fungi, or even mammal cells grown in a lab. Just as the dairy and plant-based industries fueled the naming debate over milk, there will undoubtedly be pushback from those holding on to and breaking meat conventions: “You will see the meat industry make similar arguments” about what constitutes a hamburger or what lab-grown chicken can be named, Frohlich said.
So long as technology keeps pushing the boundaries of what food can be, food names will continue to shift, and the results won’t always be neat. Someone can value natural foods plucked from farmers’ markets and served to them at farm-to-table restaurants but at the same time champion technological advances that make different versions of our foods possible. Such a person might exclusively eat free-range organic bacon but demand highly processed oat milk for their cortado. These inner conflicts are inevitable as we undergo what Biltekoff calls “a kind of evolution in our understanding of what good food is.” Milk, for now, remains fluid—simultaneously many things and nothing at all.