Ounces and Grams: Why Mass Is Not the Best Way to List Ingredients

20151111-measuring-rant-nila-jones-01.jpg

[Photograph: Nila Jones]

Since releasing my first book two months ago, I've been getting far more feedback—both positive and negative—than I'm used to. One big category: international readers upset at my choice to use imperial units of volume (tablespoons, teaspoons, cups) over metric units of mass (grams, kilograms). One well-known British television show producer even went so far as to publicly announce that she was "furious" over my decision. Furious!

At first glance, the complaint seems like a perfectly valid one for a book like mine. Units of mass are inherently more accurate than units of volume because they mostly ignore confounding factors, like how tightly an ingredient is packed into a cup or what shape you've chopped it into. As a science-based book (heck, "science" is even in the title), it ought to use the system of measurement that is most precise, right?

Not only do I disagree with this perspective, but I believe that more often than not, in cooking, using mass as a unit of measure can actually lead to less consistent results. The reason I use a mix of cups, teaspoons, ounces, and pound measures in most of my recipes is not because I don't know how to use a scale or because I'm trying to hang on to some antiquated form of measure. In fact, I've written at length about using both mass measures and the metric system for times when precision is necessary.

However, I use a mix of volume and mass because, for the vast majority of non-charcuterie and non-baking recipes, the level of precision of mass measurements isn't just overkill but can actually do more harm than good. I'm going to argue that the best, most repeatable, most user-friendly system of measure for home cooks is actually one that includes a mix of both mass and volume measures.

Before we jump in, let's make sure that we're all addressing the same issue here. To be clear: I am not arguing that ounces and pounds are better than grams or kilograms (in fact, on the contrary, I do think that Americans should figure out a way to make the switch to the metric system). I am arguing that cups, teaspoons, and tablespoons are as accurate as you need to be for nearly all cooking applications outside of baking and charcuterie.

I'll start by demonstrating that using mass (weight) to measure ingredients for cooking is not always the most accurate or precise method. Then I'll argue that precision is not actually necessary or beneficial in most cooking applications (and that pretending otherwise can be detrimental to your food).

The Precision We Want Versus the Precision We Get

There are two main advantages that proponents of the 100%-by-mass camp will espouse: It's easier to measure, and it's more precise. Let's look at these factors.

When we're talking baking, working by mass does make it easier to measure. With volume measure, you need to pull out multiple cups and spoons and wash each one of them at the end. With mass measure, you need only a single bowl and a scale that you tare after each ingredient is added (we've written extensively about this in the past). I strongly recommend owning a scale and using it any time you bake.

Using mass for baking is also more precise. Depending on how it's packed, a cup of flour can weigh anywhere from around four ounces (113 grams) to six ounces (170 grams), a difference of 50%! This can have a real effect on the food. Weighing your flour and other dry ingredients eliminates this variation.

With cooking, on the other hand, the advantages are less clear.

How Much Does Precision Matter?

20151111-measuring-rant-kenji-01.jpg

[Photograph: J. Kenji López-Alt]

Let's look at a few real-world examples—take onions, to start. I'm currently working on an onion soup recipe. In my final recipe, I'm going to be calling for "4 large onions, finely diced (about 6 cups)"—a relatively imprecise measure. A quick search for onion soup recipes in the UK reveals this one from the BBC, which calls for "1kg onions," a far more precise measure. So precise, in fact, that unless you're really lucky, you're going to have to use some fraction of a whole onion in order to land right at one kilogram. What does this level of precision get you?

I'd argue that it doesn't get you better food, and in fact gives users a false sense of precision, which, in reality, doesn't even exist. Calling for "100 grams of diced onion" or "5 grams of black pepper" gives you the mistaken impression that diced onions and black pepper are completely uniform products that don't vary from batch to batch, and that using the same weight of these ingredients will give you the exact same results every time. In fact, flavor, moisture content, sugar content, etc. can differ greatly between onions, which means that you're both wasting your time by worrying about that level of precision and also creating a less consistent product by relying on your scale instead of your tongue or your other senses.

Let's first establish how much variation there can be in six cups of onions. Starting with a couple of bags of onions, I cut them into dice of various sizes and transferred them to a two-quart measuring cup until they came up to the six-cup line. Onions that are diced more finely will pack into the container more tightly than those that are very roughly diced, so six cups of finely diced onion will weigh more than six cups of roughly chopped onion. At its realistic extreme, this difference comes down to around 20%. How much of a difference does 20% more or less onion make in a given recipe?

20151111-measuring-rant-kenji-03.jpg

[Photograph: J. Kenji López-Alt]

I made two batches of my Slow-Cooked Bolognese Sauce last week, one batch made with eight ounces of onion, the other with 10, a difference of 25%. Tasted side by side, the two are indistinguishable from each other. In fact, you can get far more flavor variation simply by giving the same amount of onion 25% more or less browning time.

You want an even starker demonstration? I did another quick experiment. I picked up a few pounds of onions from my local Whole Foods, then drove to the Safeway supermarket a few miles down the road and bought some of their onions. Back at home, I made four quick batches of onion "tea" by steeping sliced onions in boiling water for 10 minutes before tasting it.

  • Batch 1: 200 grams* of water with 100 grams of Whole Foods onions, sliced paper thin. (This is the control batch to which I'll compare other batches.)
  • Batch 2: 200 grams of water with 100 grams of Safeway onions, sliced paper thin.
  • Batch 3: 200 grams of water with 125 grams of Whole Foods onions, sliced paper thin.
  • Batch 4: 200 grams of water with 100 grams of Whole Foods onions, sliced a quarter inch thick.

* Notice that when I'm doing an actual experiment, where precision is important, I use precise metric units of mass.

Try it out for yourself, and what you'll find is that the batch that varies the least in flavor and aroma from the control batch is in fact the one that uses 125 grams of onions instead of 100 grams. In other words, where and when you buy your onions and how you chop them has far more effect on their flavor than how much you use. This is one of the big reasons good chefs place so much emphasis on sourcing and knife skills. I want people to be imprecise in their measure of ingredients, because I want to actively encourage people to pay attention to their food and taste as they go, relying on their senses to deal with the dynamics of locational and seasonal variations. Humans have the potential to cook better than robots because despite Big Ag's best efforts, our meat, grains, and produce are still not perfectly uniform across the globe and throughout the year.

I'd encourage you to repeat this experiment at home. Better yet, get some friends over to evaluate this onion water in a blind tasting. (Admittedly, this is a good idea only if your friends value science over flavor, or if you value science more than friendship.)

When Is Volume More Precise?

20151111-measuring-rant-nila-jones-02.jpg

[Photograph: Nila Jones]

Of course, with rare exceptions, when recipe writers from mass-friendly countries call for "1 kilogram of onions," it's understood that it's an approximate value and that nobody is expected to whip out the scale. For things like onions, carrots, celery, and other staples that we use all the time, it's very easy for most home cooks to perform a quick conversion in their heads: one kilogram is about three large onions, or one large onion weighs about 12 ounces, or one medium diced onion is about a cup.

I personally believe that humans are inherently visual creatures, and that, all else being equal, approximating measures by volume will give you more consistent and precise results than approximating measures by mass. But all else is not equal, and in cases like these, when we're expected to approximate, mass versus volume largely comes down to our own personal comfort zones and what we're more familiar with. Nobody is going to win that particular argument.

This breaks down as we move toward things on the smaller end of the scale, however: 10 grams of ground cinnamon, five grams of cumin, two grams of ground cloves. Here's where I find that mass measures can not only get you into trouble but are actually nearly impossible to accurately measure. A standard home scale is simply not precise enough. My scale, a mid-range digital version that is popular with home cooks, has a maximum readout resolution of one gram, which means that at any point between 0.51 grams and 1.49 grams, the scale is gonna read the same figure—and, in fact, it gets even worse than that, because its accuracy is only guaranteed to ± 2 grams. You want five grams of cinnamon? Sorry, but unless you're using a jeweler's scale, anywhere from 2.51 to 7.49—a range of 300%—is as close as you're gonna get. Adding three times the cinnamon to a recipe from one batch to the next has a serious impact on flavor.

Volume measurements for small quantities of ingredients are much more precise. The variation in weight in a teaspoon of cinnamon ends up at around 35% or so. (I tested this using a fancier scale, with 0.1 gram resolution and ± 0.1 gram accuracy.) Not perfect, but a heck of a lot better than 300%.

Funny American Things: Ounces and Butter

20151111-measuring-rant-kenji-02.jpg

How many ounces of cream am I adding, and does it really matter? [Photograph: J. Kenji López-Alt]

There are two big elephants in the room when it comes to the American system of measure: fluid ounces and butter that comes in sticks. One of them is defensible; the other is not. Let's start.

Meet the Ounce, the World's Most Confused Unit

If any common unit of measure were to have an identity crisis, it'd be the ounce. The conversion of ounces to pounds, cups, pints, and quarts is confusing enough,** especially when you're trying to scale a recipe, but what makes it worse is that even within the same system of measurement, ounces sometimes denote mass (regular ounces) and other times denote volume (fluid ounces). And, of course, it's not always clear which one we're talking about. This is a problem.

** Quick reference: three teaspoons per tablespoon; two tablespoons per ounce; eight ounces per cup; two cups per pint; two pints per quart; four quarts per gallon.

Even I'm not consistent with my use of fluid versus regular ounces. In the United States system, the general rule of thumb is that if you're writing about things that you measure in liquid measuring cups—water, heavy cream, or honey, for instance—you're using fluid ounces, a unit of volume. If you're writing about solid things—flour, nuts, rice, beef, and so on—you're using regular ounces, a unit of mass.*** Luckily, for liquids that are mostly water, a fluid ounce weighs almost exactly a regular ounce. Close enough that it won't make a difference in all but the most precision-intensive recipes.

*** To make matters yet more confusing, the ounce we use to measure food—the avoirdupois ounce—weighs less, at 28.3 grams, than the 31-gram troy ounce we use to weigh gold and other precious metals. Which raises the question: If we're scaling up a recipe that calls for gold leaf, do we use the avoirdupois ounce or the troy ounce?

But this changes when we're talking about charcuterie or baking, where quantities are almost always expressed in mass, not volume. The 22.5 ounces of flour and the 15 ounces of water called for in my New York–style pizza dough both represent units of mass. Confusing, right? Let's lobby to give fluid ounces a completely different name and end this ridiculous confusion. Or, better yet, let's figure out a way to get Americans to adopt the metric system for weight measures, while still keeping our cups and spoons, the way the Brits do. Even better, let's get all of us to adopt the Australian method, which has cups and spoons that are calibrated to be easily divisible numbers of grams.

Butter Battles

Americans buy their butter in sticks; everyone else buys their butter in tubs or in blocks. Here's where I find the American system to be more user-friendly, more versatile, and faster. Butter generally comes four sticks to a pack, and while the exact dimensions of a stick of butter vary by region, the following is true no matter where you are: Each stick weighs a quarter of a pound (four ounces, or about 113 grams), and its packaging has eight demarcations along its side, representing tablespoons. This makes portioning butter extraordinarily easy, with no need for any measuring apparatus whatsoever, whether you're using mass or volume measurements. You have a recipe that calls for two ounces of butter? Easy, just cut a stick along the middle demarcation. Three tablespoons? That's line three.****

****It's been brought to my attention that in most European coutries, butter blocks are demarcated into 25 gram units. This is good.

You may find it an odd coincidence that 32 tablespoons of butter is exactly equal to a pound. That's because it's not. How can it be, when butter itself is not a uniform product but actually a mixture of water, butterfat, and milk proteins and sugars, and can vary in its composition (and therefore its density) from brand to brand?

I calculated the difference in density between European butter (which averages around 83% butterfat) and American butter (around 80% butterfat). Euro butter clocks in at around 0.909 grams per cubic centimeter, while American is closer to around 0.911 grams. That's a difference of 0.02 grams per cubic centimeter. Is this going to make a difference in your cooking? Probably not. It's a level of variance that a casual drug dealer wouldn't trouble you for, and again—remember that false sense of precision we have about our other ingredients?

If I had my way, the whole world would have sticks of butter with demarcations for ounces, tablespoons, and grams.

The Ideal Form of Measurement Listings

20151111-measuring-rant-vicky-02.jpg

[Photograph: Vicky Wasik]

The final factor to consider in all of this is user-friendliness. Most folks cook either as a means of feeding themselves and their families, or simply for the joy of cooking. The fact is, most home cooks are not scientists, nor do they want to be. Many might understand science and understand the basic scientific principles that can help them cook better, but very few care about the level of precision required for hard science.

For some folks—the engineers, the anal-retentive, and, most of all, those that overlap between the two—precision (even when it is, as we've demonstrated, false precision) is comforting, reassuring. But for most folks, it's intimidating and fussy. If I'm making dinner for my family or friends, I want the version of the recipe that makes it clear when I need to be precise, but also, more importantly, when it's okay for me to stray and relax while still arriving at a good result.

In other words, I could write all of the recipes in my book or on this website in precise mass units, but I would instantly be alienating my core readership: the average home cook. Not only that, but as I've shown, there would be no benefits in terms of the results of those recipes.

So, how should we write recipes? I think the ideal form of measurement listings in recipes is somewhere in the middle of the volume-based system and the mass-based system. Here's my simple plan for non-baking applications, whether metric or imperial units are used, and one which we generally stick to on this site (the main exception being the use of "fluid ounce," which we'll start doing from now on):

  • Use mixed units of volume for vegetables, calling for mass only when using large quantities. For a single onion, "1 medium onion, diced (about 1 cup)" works; for a big ol' pot of stock, I might write "3 pounds onions, peeled and roughly chopped (about 6 medium)".
  • For foods that are typically bought by weight and not by the unit, stick with weight measurements: "1 1/2 pounds ground beef."
  • For dry goods, stick with weight when quantities larger than a few tablespoons are called for; use volume for anything less. When calling for mass, include approximate volumes in parentheses for folks who don't own a scale. E.g.: "5 ounces all-purpose flour (about 1 cup)"; "1 ounce powdered gelatin (about 3 tablespoons)"; "1/2 teaspoon powdered cinnamon."
  • Measure liquid ingredients by volume, whether fluid ounces or milliliters. Always specify "fluid" ounces when you're talking about volume.

And for baking and charcuterie, I, for one, wish that we could all switch to the metric mass system 100% of the time. It simply makes the math easier. I know the salt in my Neapolitan pizza dough is 3% of the weight of the flour, which is simple to calculate using metric units. I have 500 grams of flour? Good: I'll add 15 grams of salt. With ounces and pounds or cups and spoons, that conversion is not so easy.

I can actually only think of three different occasions when measuring everything in grams is the best method: when baking, when making charcuterie (and even with baking or charcuterie, small quantities of spices are still best measured by volume), and when cooking in bulk in an industrial setting where scaling is necessary.

Science Is the Roadmap to Delicious Food

Using mass as the standard may well be how most of the hard sciences work, but this is emphatically not the same thing as saying that all good science requires mass. Science isn't always about using the most precise means possible, any more than it is about using big words to describe things.

Science is a map that can help you to get from point A to point B. The map doesn't care how you get there; your own motives and vehicle drive you. Science doesn't have a final destination or a prescribed path. Science is simply a method of showing you the lay of the land. Science is the "zoom" button on the Google Map of the universe. It'll show you the curves in the road coming up, or the scenic bypasses along the way.

To require that a recipe stick to gram levels of precision is to plan out your exact route before heading out on the road, to draw a red line along the highway and adhere to it, ignoring dynamic real-life factors like the day-to-day traffic or weather, or perhaps that little hot dog stand in Connecticut that you forgot about, but would sure love to pull over for. So long as you get to your final destination and you enjoy the process that gets you there, you are following exactly as precise a path as you need to follow. Having a good, well-drawn road map—i.e., understanding the science and techniques behind the cooking—can help ensure that you get to your destination no matter how many bypasses you take or hot dogs you eat along the way.

So the next time someone tries to come at me with a mass superiority complex, I'll just try to take their comment with .00048 grams**** of salt. Measured on a scale, of course.

**** According to Cooking Light, the average pinch of salt weighs 0.48 grams, and according to Vendian.org, there are approximately 1,000 grains of salt in a pinch. See how useful and precise mass measure can be?