A Grain of Science

More than a matter of taste

The lowest point in North America looks like a beach. At Badwater Basin in Death Valley, a blinding white plain runs to the horizon — not sand, but salt, the residue of ancient water that evaporated and left its ions behind. Salt becomes visible when it concentrates.

Most seasonings have a personality. They bring volatile molecules, recognizable aromas, a place in the spice rack. Salt is different. It has no aroma, no botanical identity, no signature molecule announcing itself as “salt-smelling.” It is two bare ions.

Good cooking usually asks salt to do the opposite of what happened at Badwater. You do not want it sitting apart as a white crust or announcing itself as a separate presence. You want it dissolved, distributed, and working in the background: changing taste-cell signals, suppressing bitterness, strengthening sweetness and umami, altering protein charge, holding water, and moving slowly inward by diffusion.

Salt does this background work even in foods that do not taste salty. Ice cream is a surprising example. Divide some plain vanilla ice cream between two bowls, then mix salt into one at 0.15% of its weight. Taste them blind and see which you prefer.

If you watch cooking videos by professional chefs, they seem to use a scary amount of salt. Thomas Keller even joked in one video that whenever his kitchen staff asks for feedback on a dish, he says it needs more salt. Many savory foods taste fully seasoned at around 1–1.5% salt by weight, although the right level depends on the dish and the diner’s palate.

That practice runs straight into the health warning everyone has heard: too much sodium is linked with high blood pressure, and U.S. dietary guidelines tell adults to stay under about 2,300 milligrams of sodium per day. If good seasoning really takes that much salt, should you deliberately underseason your food in order to have a healthy diet? Let’s do the math. According to an EPA study, the average adult consumes about 1,500 grams of solid food per day. 1% is 15 grams. Salt is sodium chloride, and sodium is about 39% of salt by weight. That is about 6,000 milligrams of sodium, way above the guideline. Even though home cooks typically use only about half as much salt, that is still about 3,000 milligrams. So why aren’t we all getting high blood pressure?

Well, a lot of us do have high blood pressure. But there are problems with the calculation above. The real answer has a twist: the number one source of dietary sodium in the United States is not the salt shaker, and it is not potato chips. It is something much more ordinary. Before naming it, we need to separate three questions: why sodium tastes salty, why sodium can raise blood pressure, and where sodium in the modern diet actually comes from.

Sodium: Taste, Body, and Diet

Why sodium tastes salty

There are five basic tastes: salty, sour, sweet, umami, and bitter. Sweet, umami, and bitter are metabotropic: they are detected when flavor molecules bind to specialized receptor proteins on taste cells. This binding triggers a cascade of intracellular signaling events that generate an influx of borrowed ions along the way. Salt is ionotropic: it brings its own ion flux. The sodium ions flow straight through the epithelial sodium channel (ENaC). No receptor, no intermediate channel, and no cascade. The brain has evolved to recognize the spatiotemporal pattern of nerve signals generated by this sodium ion influx as the salty taste.

Sour taste works by a related direct-ion logic. Acids donate protons, and sour-taste cells respond to those protons rather than to an aroma molecule with a named identity. That is why salt and acid belong in a different category from herbs and spices: they arrive as small ions that change the conditions of perception.

An interesting question is what happens if you flood the taste buds with other ions? Technically, sodium chloride is only one type of salt. Chemically speaking, a salt is an ionic compound made from the cation of a base and the anion of an acid. In the same column in the periodic table, potassium and lithium are the two closest elements, and their salts do taste salty too. Potassium ions are larger than sodium ions, and ENaC’s narrow selectivity filter acts partly as a molecular sieve: it conducts sodium hundreds of times more readily and is virtually impermeable to potassium. Any taste response to potassium must therefore travel through ENaC-independent pathways, whose molecular machinery is still not fully understood. These pathways are less selective. The result is not that potassium chloride lacks saltiness, but that its salty signal often arrives with bitter or metallic notes. Lithium reveals the other side of the size filter: lithium ions are smaller than sodium ions and pass readily through ENaC. Their similarity to sodium lets them follow some of the same routes through the body, but once inside cells they interact differently. This makes lithium pharmacologically active — and potentially toxic — rather than a culinary salt substitute.

You could also just run a weak current across the tongue. It would generate some nerve response, but you would perceive what is usually described as a metallic or sour taste (sour is another “direct” taste: it’s the result of an influx of protons). As of now, no artificial stimulus can replicate the effect of sodium ions entering through the ENaC. There is an “electric salt” device, but it doesn’t create a salty taste from nothing; it concentrates sodium ions toward the taste cells, effectively boosting the perceived saltiness of food. So people who eat low-sodium food can still taste salt.

How sodium affects blood pressure

Sodium is the main positively charged ion in extracellular fluid. When the body retains more sodium, it also retains more water, expanding the extracellular fluid, including blood plasma. All else being equal, an increase in circulating blood volume tends to raise arterial blood pressure by increasing venous return and cardiac output. Healthy kidneys usually excrete excess dietary sodium, but that protection is not unlimited. In salt-sensitive people, maintaining sodium balance may require higher blood pressure.

So why doesn’t sodium flow into cells to pull water back in? It does, except that it is pumped right out. One of the interesting results of evolution is that most human cells have sodium/potassium pumps. These pumps continuously push sodium outside cells and pull potassium inside cells. This creates sodium and potassium gradients across cell membranes. The cell membrane contains potassium leak channels that are open at rest. These channels let potassium ions pass but block sodium ions. Why? It has to do with how potassium ions and sodium ions exist in water. They attract water molecules and become surrounded by them. To pass through a narrow protein channel, the ion must partly shed that water shell. The channel has a narrow region called the selectivity filter, lined with oxygen atoms. These oxygen atoms temporarily replace the water molecules and “hold” the ion as it passes through. The potassium ion is just big enough to touch the oxygen atoms. The sodium ion is too small and holds tightly to its water shell, so the filter cannot dehydrate and stabilize it.

This leaking potassium flow leads to a slightly negative voltage inside cells (about −70 mV). It creates a sort of biological battery. This is useful for nerve signals, muscle contraction, heart rhythm, and many transport processes.

Potassium helps counter the blood-pressure effect of sodium, but not because potassium inside cells “balances” sodium outside cells. It mainly works through the kidneys and blood vessels. Higher potassium intake can make the kidneys dump more sodium into urine. Less retained sodium means less retained water, lower blood volume, and often lower blood pressure. Potassium also affects vascular smooth muscle and electrical signaling in vessel walls. In simplified terms, good potassium levels can make blood vessels less constricted. So if you are watching your blood pressure, you should pay attention not just to the absolute level of sodium, but also to the sodium/potassium ratio. Common potassium-rich foods include beans, potatoes, bananas, and tomatoes.

Why is sodium outside the cells and potassium inside the cells? The leading hypothesis is that early cells evolved in a potassium-rich and sodium-poor environment.

Where dietary sodium actually comes from

Remember the scary 3,000 milligrams of sodium a day? According to the National Health and Nutrition Examination Survey, the average sodium intake in the United States is around 3,400 milligrams per day. So the calculation lands near the real number, but for the wrong reasons.

The calculation assumes that every gram of food is seasoned equally, that you add all the salt yourself, and that all the salt you use stays in the food. None of those assumptions is true.

We do not salt fruit, yogurt, cereal, bread, vegetables, and steak to the same percentage. Even when we cook with heavily salted water, as with pasta or blanched vegetables, most of that salt goes down the drain rather than into the food. A seasoning percentage that makes sense for one savory dish cannot be multiplied by the total mass of everything eaten in a day.

More importantly, most of the sodium in the American diet is already present before the home cook reaches for the salt. More than 70 percent comes from packaged and prepared foods, including restaurant food. Nutrition labels also count sodium from ingredients other than table salt, such as baking soda, MSG, and sodium phosphates. The salt shaker is visible, but much of the sodium intake is not.

Store-bought bread is the surprising example. In one major U.S. dietary survey, breads and rolls were the largest single food-category contributor to sodium intake. Bread does not usually announce itself as salty, but a moderate dose eaten frequently adds up. The same is true of deli meat, cheese, bottled sauces, canned food, frozen meals, and restaurant dishes. A food’s sodium contribution depends not only on how salty it is, but also on how often and how much of it you eat.

For the typical American diet, undersalting fresh food at home is the wrong lever to pull: it gives up a large amount of flavor while addressing a relatively small part of the average sodium intake. A better strategy is to reduce the background sodium in frequently eaten packaged and prepared foods, then use a measured amount of salt where it has the greatest effect. Do not spend sodium where you cannot taste its benefit.

Consider a 200-gram chicken breast. At 1% salt by weight, it needs 2 grams of salt. In my measurements, a generous three-finger pinch of kosher salt weighs about 0.7 grams, while one full turn of a salt mill produces only 0.03–0.10 grams. Reaching 2 grams would therefore require about three generous pinches or 20–67 turns of the mill — far more than most home cooks use. With fresh, whole foods, undersalting is usually the more likely mistake.

Salt as Seasoning

If sodium is a limited budget, the cook’s question is no longer “Should I use salt?” but “Where does salt do the most work?”

Most seasonings announce themselves by molecular identity: basil tastes like basil, garlic like garlic, smoke like smoke. Salt is different. Like acid, it works less as a named aroma and more as a structural seasoning. It changes how the food that is already there is perceived.

In cooking, salt does two broad kinds of work. First, it changes flavor perception. Later, we will see how it changes texture by changing proteins and water.

1. Salt suppresses bitterness

Bitterness is a special taste. Salty, sweet, and umami all signal something the body wants. The body only needs to detect their presence, so a small number of detectors do the job. Many toxic compounds taste bitter, and they share no common structure, so evolution favors a broad spectrum of bitterness detectors that are extremely sensitive. There are 25 different receptors for bitterness, far more than for any other taste. Sodium appears to interfere with how some bitter compounds activate the receptors. But that appears to be a secondary effect. The main interference happens at central processing. When the salty and bitter signals arrive at the brain, it doesn’t simply add them up. It does a kind of mutual suppression. Bitterness has a very low detection threshold, so a little sodium goes a long way.

Coffee is the standard example. A pinch of salt can push the bitterness to the background without tasting salty. You are not changing anything chemically, just the perception of taste.

2. Salt strengthens sweetness and umami

Salt can also affect sweetness directly. Taste mixtures are nonlinear: at low concentrations, sodium can increase perceived sweetness, while at higher concentrations saltiness competes with it and eventually dominates. The effect depends on the food and the concentrations involved, which is why a pinch of salt can round out caramel or watermelon, but more salt does not simply make them sweeter.

Umami is detected when glutamate and related compounds activate umami receptors. Salt supplies a separate salty signal that the brain combines with the umami signal from glutamate. In the right proportion, the combination increases the perceived savoriness and palatability of the food even though sodium does not activate the umami receptor. That is why a broth rich in glutamate but low in salt can taste flat, then “bloom” when properly seasoned.

The umami receptor has its own lock-and-key trick. Glutamate binds at one site, while purine nucleotides such as IMP and GMP bind at a neighboring site. With both sites occupied, the nucleotide stabilizes the receptor’s closed, active form, amplifying the savory signal far beyond a simple additive effect. This is why kombu and bonito flakes work in dashi: kombu supplies glutamate, while bonito supplies IMP.

3. Salt can increase aroma impact

Flavor is not only taste. Smell is a large part of what we casually call flavor. Salt can change that system too.

One mechanism is perceptual. When bitterness falls and the overall taste profile becomes better organized, the brain interprets aromas differently. Another mechanism is physicochemical. Salt ions in solution attract water molecules tightly around them, effectively making the liquid a less hospitable place for nonpolar aroma compounds — the same molecules that give wine its bouquet or garlic its punch. Crowded out of solution, they escape into the air above the food instead, reaching your nose more readily. The Setschenow relation describes how much:

$\log\left(\frac{S_0}{S}\right)=k_s c_s$

where $S_0$ is the solubility of the aroma compound without salt, $S$ is the solubility with salt present, $c_s$ is the salt concentration, and $k_s$ is a compound-specific constant. This equation shows that solubility decays exponentially as a function of salt concentration.

This effect is real, but it should not be exaggerated into a universal law that “salt always releases aroma.” One obvious caveat, looking at the equation, is that the salting-out effect only holds when $k_s$ is positive. The physical picture is this: ions are charged and grab water molecules tightly into hydration shells. That “uses up” water and also raises the cohesive energy of the liquid, making it harder to open up a cavity to host a nonpolar molecule. A nonpolar aroma compound, which was only marginally tolerated in water to begin with, gets squeezed out — solubility falls, volatility rises. Other compounds interact with sodium ions and water molecules differently, and that turns $k_s$ negative. We will talk about the other half of the relation when we talk about how salt affects texture later.

Seasoning as a control problem

The tongue has thresholds. Below a certain concentration, a seasoning may be functionally invisible. Above another, it dominates. The best seasoning zone is often narrow. In the language of control theory, salt is a high-gain control with asymmetric failure modes. This is why experienced cooks season incrementally: they treat the result as a feedback loop, adding a little and tasting before committing to more.

We can apply the same principle to decide when to add salt to coffee: to the grounds before the pour or to the brewed coffee after the pour? The amount of salt is so small that it doesn’t affect extraction. So the only consideration is how best to control the final output. Put that salt in the grounds before brewing and you commit to a dose before you can taste the result, leaving yourself with an open loop. Add it to the brewed cup instead and you close the feedback loop described above. Same salt, same chemistry; the only thing that changed is where in the process you add it.

It takes real skill and a lot of practice to get seasoning just right. That’s why you don’t see salt shakers on the dining tables of Michelin-starred restaurants. You would ruin the flavor balance the chef has carefully built.

Salt and Texture

One of the most widely misunderstood claims is that salt tenderizes meat. To fully unpack this, we need to start with what meat is.

Meat is mostly water — about three-quarters by weight. After water, protein is the main component, and it comes in three kinds:

1. Water-soluble proteins, like albumin.
2. Salt-soluble proteins, like myosin and actin, which form the myofibrillar network — the filament lattice that holds most of the meat’s water. “Salt-soluble” doesn’t mean they dissolve in salt — it means they won’t dissolve in water without the help of salt.
3. Collagen, which doesn’t dissolve in water. (With long, moist heat, though, it converts to gelatin, which does — that’s what makes tough cuts tender.)

A protein’s surface is studded with ionizable groups: acidic ones (carboxyl groups, from amino acids like glutamate and aspartate) that can shed a proton to become negative, and basic ones (amino groups, from lysine, arginine, and histidine) that can grab a proton to become positive. Whether each group is actually charged depends on the pH of the surrounding liquid: in acidic conditions (lots of H⁺ around), the basic groups grab protons and the protein goes net positive. In alkaline conditions, the acidic groups give up their protons and it goes net negative. Somewhere in between there’s a specific pH where the positives and negatives are perfectly balanced — net zero. That’s the isoelectric point (often written as pI).

The crucial consequence: at the isoelectric point, with no net charge, protein molecules don’t electrostatically repel each other, so nothing keeps them apart — they clump together. Meat’s natural pH is about 5.5, and the myofibrillar proteins’ pI is about 5.0 — so fresh meat sits close to its isoelectric point, which is exactly the zone of minimum water-holding. The filaments have little net charge or repulsion, leaving a collapsed lattice with little room for water.

When you add salt, the chloride binds to the protein and effectively drags the isoelectric point lower. The meat’s pH stays at 5.5, but now that’s further above the pI than it was, so the protein carries more net negative charge and therefore experiences more repulsion. The filaments move away from each other and there is more room to hold water between them.

You might ask why sodium ions won’t neutralize the negative sites on the protein and move the pI back. It’s the same hydration story as the potassium channel: sodium clings to its water shell. To bind closely to a carboxyl group on the protein, sodium would have to strip off some of that water, which costs energy. So it stays wrapped in water and lingers nearby as a loose, diffuse counter-ion — present, but not latched on. Chloride sheds its water shell easily, so it can move right up against the protein’s positive sites and associate closely. It commits where sodium won’t.

We can also move the pH directly instead of the pI. That’s the baking soda trick used in Chinese stir-fries. The alkaline conditions created by the baking soda cause acidic groups on the proteins to release protons, increasing the proteins’ net negative charge. The resulting repulsion opens the myofibrillar lattice and improves water retention.

By the way, marinating meat with pineapple works for a completely different reason. The enzyme bromelain cuts the protein backbone, chopping the long protein chain into pieces. It goes after not only the salt-soluble proteins but also collagen. Salt’s ions don’t do much to collagen. Collagen fibrils are locked together by covalent crosslinks — actual chemical bonds between collagen molecules. Covalent bonds don’t care about your ionic environment. You can pile on charge and screen all you like; you’re not going to pull apart a covalently bonded network with electrostatics. So collagen mostly just sits there.

To attack covalent bonds, you need hydrolysis, a reaction in which water splits a chemical bond. At cooking temperatures, hydrolysis proceeds slowly, so it needs both moisture and time. That is why tough cuts are braised: hours of gentle cooking in a wet environment gradually convert collagen into gelatin, which then dissolves in the surrounding liquid. Gentle heat matters because muscle proteins contract more as their temperature rises, squeezing out water; a long braise trades intensity for time, allowing collagen to soften without drying the meat as severely.

Given enough concentration and enough time, salt does more than hold water. Myosin molecules normally stay clipped together by electrostatic attractions between charged spots on neighboring molecules. Add salt, and the ions move in and screen those charges. With the electrical attractions shielded, the molecules stop gripping each other, the filaments come apart, and each myosin molecule becomes surrounded by water. The myosin dissolves. The rough threshold is around 0.3–0.6 molar ionic strength, which is part of why sausage recipes call for a real, nontrivial amount of salt — below a certain point, you simply don’t extract enough myosin. This is “salting-in”: salt causes the molecule to be more soluble in water, and the $k_s$ turns negative in the Setschenow relation. Strictly speaking, however, the earlier equation describes small, electrically neutral compounds in relatively simple solutions. Protein salting-in is more complicated because myosin is a large, charged molecule embedded in an organized filament, so its extraction depends on ionic screening, pH, protein structure, and mechanical mixing — not merely a negative Setschenow constant.

A plain unsalted ground beef patty holds together because grinding ruptures cells and smears protein across the cut surfaces, so the particles are tacky and physically interlock — meat, fat, and connective-tissue fragments tangled together. A little myofibrillar protein is exposed, and there’s some natural ionic content in the meat itself, so a small amount of sticky protein is in play. Then, when you cook it, the proteins on all those particle surfaces denature and coagulate, fusing neighboring particles. That’s enough to give you a cohesive burger. The texture is loose, tender, slightly crumbly — which is exactly what you want in a classic burger.

When you add salt and work the meat (mix, knead, smear), you actively extract myosin from the fibers into a sticky, tacky exudate that coats every particle. On cooking, that exudate sets into a continuous protein matrix gluing everything into one mass. The result is springy, bouncy, dense, sliceable — sausage texture, hot-dog snap, the bounce of a well-made meatball.

So salt isn’t the on/off switch for “holds together”; it’s the dial between “loose and tender” and “springy and bound.” This has a direct kitchen consequence: if you salt ground beef and mix it before forming burger patties, you extract myosin and get a dense, rubbery, weirdly sausage-like burger — a mistake in my opinion. The fix is to salt only the outside right before cooking, keeping the interior loose. For sausage, meatballs, meatloaf, or anything you want firm and cohesive, you do the opposite: salt early and mix hard to develop that bind on purpose.

Before we wrap up our discussion of proteins in meat, a final word about albumin, a water-soluble protein. When salmon is heated and its proteins tighten, they squeeze out water along with the albumin. Once that albumin reaches the surface and heats up, it denatures and coagulates into a white solid. If you brine salmon in salt water, the myofibrillar lattice relaxes and holds onto more water, so more albumin stays inside the muscle instead of leaking out. But that only helps at the margin. The dominant factor is still heat: if you see white stuff on your salmon, you probably overcooked it rather than under-brined.

So does salt make meat more tender? The short answer is: sometimes, modestly, and the effect is much less important than the timing and intensity of heat.

Getting Salt into Food

Salt has to penetrate meat before it can do anything to the proteins. When we salt meat, we normally salt the surface. (We could, and probably should, inject salt water deep into a large turkey with a syringe if we want salt on the inside — although this is arguably just a way of salting surfaces that happen to be internal.) Wet and dry brining work differently. Wet brining is straightforward: salt diffuses from the area of high concentration to the area of low concentration. The speed of that diffusion is governed by Fick’s second law:

$\frac{\partial C}{\partial t}=D\frac{\partial^2 C}{\partial x^2}$

where $C$ is salt concentration, $t$ is time, $x$ is distance into the food, and $D$ is the diffusion coefficient. In ordinary language, the equation says that whenever salt is unevenly distributed, nature smooths the profile out over time. Where the concentration curve is most bent — most uneven — change happens fastest. A steep pileup of salt concentration near the surface gradually relaxes toward a flatter, more even distribution.

The form of this equation is the same as the heat equation discussed in the heat transfer chapter. This is not a coincidence. Diffusion and heat conduction are both transport phenomena driven by gradients.

Looking at the dimensions in the equation, the left side scales like $C/t$, and the right side scales like $DC/x^2$. For the two to balance:

$\frac{C}{t}\sim \frac{D\,C}{x^2}\quad\Longrightarrow\quad x^2\sim D\,t \quad\Longrightarrow\quad x\sim\sqrt{D t}.$

What this means is the penetration depth of salt is proportional to the square root of time. To push salt twice as deep takes four times as long. To anchor it numerically:

Salt’s diffusion coefficient in meat is somewhere around $D\approx 3\text{–}5\times10^{-10}\,\text{m}^2/\text{s}$. Taking $D\approx4\times10^{-10}$ and $t=1$ day:

$\sqrt{Dt}=\sqrt{4\times10^{-10}\times 86400}\approx 6\ \text{mm}.$

So the characteristic depth is roughly half a centimeter to a centimeter per day. This aligns with the empirical rule of thumb used by experienced cooks and helps explain the value of injecting brine into a turkey.

In reality, $D$ depends on many things: salt diffuses more slowly at lower temperatures; salt diffuses mainly through the water phase, not through dry protein or fat, so more available water usually means faster diffusion. Muscle is directional. Diffusion along fiber bundles differs from diffusion across them. Salt diffuses more slowly in fat and connective tissue than in lean muscle. Finally, grinding, injecting, tumbling, cutting, and even freeze-thaw damage can shorten diffusion distances or open transport pathways. This is why sausage, ground meat, injected poultry, and industrial cured meats can become seasoned and salt-bound much faster than an intact roast. As order-of-magnitude examples, a thin fish fillet can be seasoned meaningfully in an hour or two. A chicken breast benefits from overnight salting.

Dry brining adds two overlapping processes to ordinary diffusion. As soon as salt dissolves in the meat’s existing surface moisture, it creates a concentrated salty film. Water begins moving outward while sodium and chloride ions begin diffusing inward. For a steak or chop, the dry-brine dead zone runs from roughly 3 to 40 minutes after salting: water accumulates on the surface faster than the salt moves inward or the moisture evaporates. Some salt is already entering the meat, but not enough to season it deeply, while the wet surface interferes with browning. The exact window varies with thickness, temperature, humidity, and airflow, so a visibly wet surface is the more reliable cue.

With more time, the exuded water dissolves the remaining salt and the ions diffuse farther into the meat, gradually smoothing the concentration gradient. Salt also begins altering the myofibrillar proteins near the surface. If the meat remains uncovered in the refrigerator, evaporation eventually dries the surface again while the interior becomes more evenly seasoned. This improves browning. As discussed in the steak chapter, thorough seasoning also improves the perception of juiciness.

So the worst thing you can do is to season the meat a few minutes before cooking. For steak-sized cuts, either season immediately before cooking or at least 40 minutes ahead; avoid the window in between.

Salt is the plainest ingredient in the kitchen and one of the least replaceable. It has no aroma, no color, no botanical heritage — just two bare ions, pulled from the earth or evaporated from ancient seas. And yet it is the single most powerful tool a cook has.

Used badly, it sits on the surface and tastes like salt—coarse, obvious, a failure of integration. Used well, it disappears into the food and makes everything taste more like itself.