I've written it dozens of times: Processing olive oil in a blender can make it unpleasantly bitter. I've crafted recipes to ensure it doesn't happen, instructing readers to make mayo by blending in a neutral oil first and then working in some olive oil by hand afterward. Then, several weeks ago, I was working on a little story about just how different the results are between hand-whisked and machine-blended mayo. The differences are mostly textural, but I noted a flavor divide as well—the blended mayo was more bitter than the hand-whisked one. Exactly as expected.
I wrote my article, and I explained the bitterness as the consequence of blending olive oil at high speed. But I also casually floated a second explanation: that the bitterness was caused by the more aggressive puréeing of garlic in the blender, while the hand-whisked one contained garlic that I'd minced more coarsely with a knife.
The more I thought about it, the more those theories weighed on me. I know for a fact that how thoroughly you mince garlic can have a dramatic effect on its flavor, with a more intense bitterness emerging the more finely the garlic is broken down. And I also know that sometimes at home I break my own rule and blend olive oil at high speed...and I can't remember ever having been bothered by the results. My brain told me this was a settled issue, but my gut told me it deserved more methodical attention.
What do we know about olive oil and bitterness? Well, we know that its bitterness, which is not at all considered a flaw (good olive oil often will be bitter), is caused by phenols, a large family of antioxidant molecules found in both plants and animals. It's thought that secoiridoids, a class of phenols, are most directly responsible for any perceived bitterness and pungency in olive oil.
The idea that the bitterness of olive oil is exacerbated by high-speed blending or food-processing has been around for a while. The food writer James Peterson mentioned this phenomenon in his book Cooking, and the topic has popped up on food message boards for years. Peterson gives no explanation for the underlying mechanism that causes this bitterness, but almost all of the internet chatter I've seen points back to this 2009 Cook's Illustrated article on the topic. According to that article, the bitterness is increased when the spinning blades of a food processor or blender disperse the oil and its fatty acid–coated polyphenols into tiny droplets in an emulsion, making them more easily detected by the taster. The problem, they say, is worse with higher-quality extra-virgin olive oil and less of an issue with pure olive oil, due to a lower phenol content in the latter.
That sounds really convincing. Indeed, it might be true. But I can't tell, because I've been unable to turn up any scholarly articles backing up this explanation. (If you know of any, please send them my way!) Every source I've found that has given this explanation points back to Cook's Illustrated, and the trail goes cold there.
Reading the Cook's Illustrated story, though, got me wondering: How do you even test this? According to Cook's Illustrated, the phenomenon is specific to emulsions, like mayonnaise, that contain olive oil. To test it, you'd have to make two identical emulsions, one by hand and the other with a blender or food processor, and then have blind-tasters evaluate the results. But, as my article on whisking mayo by hand shows, the results are so markedly different that there'd be no way to blind-taste the two side by side without knowing which was which.
Putting that blind-tasting challenge aside, there's an even deeper potential problem in how Cook's Illustrated conducted its tests. According to the article, they made two batches of mayo, one blended and the other hand-whisked, and compared the results. But both those batches also contained minced garlic. That's significant.
Garlic, as I know from my mincing tests, can have a wide range of flavors depending on how (and how thoroughly) you mince it. That's because, as I wrote, "When [garlic's] cells are damaged, say, by a pest, two molecules, one called alliin and an enzyme named alliinase, come into contact with each other, and together produce a new compound called allicin, which is responsible for the pungent, vampire-repelling smell we associate with garlic. The more cell damage that occurs, the more allicin is produced, and the stinkier the garlic becomes." In the kitchen, this means that the more thoroughly we pulverize garlic, the more bitter it will become.
Keeping that in mind, let's think about what happens when you process minced garlic in a food processor or blender while drizzling in oil to make mayo: You're mincing the garlic even more finely. Compare that with mincing garlic and then whisking in the oil by hand—the whisk isn't doing much more to the garlic than has already been done to it with a knife.
Given this, it's hard to imagine that the testers for the Cook's Illustrated article were able to differentiate between bitterness caused by the olive oil and bitterness caused by the garlic, since either one (or both) could have been the culprit. This doesn't prove Cook's Illustrated wrong, but, given the seeming lack of clear evidence on the matter, it at least indicates that more testing is in order.
To see if I could get more clarity on just what was responsible for the perceived bitterness—the olive oil or the garlic—I ran a handful of tests.
Blended Versus Unblended Olive Oil Taste Test
I started with the most basic test: I gathered up five different bottles of extra-virgin olive oil, from four different sources (Australian, Italian, Californian, and Spanish), and blended each one with an immersion blender for one full minute. I let all the samples sit until any cloudiness from air bubbles introduced during the blending had dissipated, so that there was no visual indication of the blending. I then asked blind-tasters, who didn't know what I was testing in even the most general sense, to taste each blended sample against its unblended counterpart (varying the order of which sample came first), and to give feedback on the differences, if any, they noted between the two.
The tasters were all over the place: Several said the unblended samples were more bitter or pungent or astringent or spicy, a couple said the blended samples were more harsh, and one couldn't detect much of a difference at all.
Since I was conducting the test, I wasn't able to do the tastings blind, but I tasted all the samples as well and found no obvious differences between the blended and unblended oils. The biggest factor, I found, was the order in which the oils were tasted, since the intensity of olive oil can be cumulative—the second spoonful can hit you harder than the first one.
Mayo Made From Blended and Unblended Oil
I wanted to run one test that compared two batches of mayo, while making sure the two were indistinguishable from each other in terms of consistency and texture; that way, blind-tasters wouldn't be swayed by differences in those qualities. This meant I couldn't blend one and whisk the other, since those two methods produce such different texture and consistency results. Instead, I hand-whisked two bowls of mayonnaise, one with oil that had previously been processed in a blender for one minute and the other with oil straight from the bottle. To avoid any confusion possibly caused by bitterness from minced garlic, I simply omitted the garlic from both batches. I then asked blind-tasters to try both samples and tell me what they observed. (Once again, they had no idea what I was testing.)
My tasters were evenly split, with two identifying the mayo made with unblended oil as more bitter and two identifying the mayo made with blended oil as more bitter. Several other tasters noted details apart from bitterness, saying one sample was "more buttery" or another "more olive-y," indicating that as far as bitterness went, they didn't notice anything in particular.
Once again, my own tastings of the samples left me feeling like there wasn't much of a difference between the two. Certainly, neither was obviously more bitter.
To try to determine whether the garlic made a clear difference, I next made two batches of mayo using an immersion blender. I added a clove of garlic to one batch at the beginning, allowing it to be fully processed during the making of the mayo. With the other batch, I minced the garlic by hand, but withheld it until after the mayo was blended together, after which I stirred in the garlic by hand.
In this test, four of my tasters identified the blended-garlic mayo as being more bitter and intense, while two thought the batch with minced garlic was more bitter. "Bitterness" was the word of the day in this test, with only one taster not remarking on it, and most people thought the mayo with garlic that had been blended was more bitter.
While not perfectly decisive, this test does add weight to the idea that the garlic is a more significant factor in the perception of bitterness than the oil itself is.
If you read the Cook's Illustrated article, it seems to suggest that the phenomenon is particular to olive oil emulsions specifically. Thus far, in the interest of making my samples as indistinguishable as possible, I had blended the olive oil first, then made the mayo by hand (except for in the garlic test, for which both samples were blended). Perhaps they were right—perhaps this spike in bitterness happens only when olive oil is emulsified into a liquid.
Since there's no way to make two emulsions, one with a blender and one by hand, that look identical, I decided to try an even simpler test: I would make a mayonnaise emulsion of olive oil worked into a whole egg using an immersion blender, with no other ingredients at all—no vinegar, no salt, no garlic, no nothing—and then I would compare it with the same oil in its liquid state. Either the mayo would be dramatically more bitter than the oil in its original condition, or it wouldn't. To avoid the possibility that I just happened to be using an olive oil that wasn't very susceptible to growing more bitter, I decided to make a blend of four different extra-virgin oils and use that. Surely one or two of them would have the characteristics necessary to become more bitter, if the phenomenon were real.
I blended the mayo. I tasted the liquid oil I'd used to make it. Then I tasted the mayo, then the oil again. Back and forth I went, tasting the two, waiting for an obvious bitter note to emerge from the emulsified one. I noticed absolutely nothing. The mayo was creamy and light, even slightly sweet, and hardly bitter at all. There was a light spiciness that both the mayo and the oil blend shared, though the liquid oil possibly had even more of it.
Through multiple rounds of testing, the phenomenon of bitterness in blended olive oil failed to make itself obvious. I never noticed it, and my blind-tasters failed to describe any clear correlation between blending and increased bitterness. The garlic, meanwhile, did seem to have an effect.
In my research, I've found another expert who has also not noticed any problems when blending olive oil. Nancy Harmon Jenkins, in her book Virgin Territory: Exploring the World of Olive Oil, writes, "I've also seen mutterings on the Internet that...olive oil is made bitter by vigorous beating.... I must confess, honestly, that I've not had that experience, nor have any of several other cooks or chefs I've queried."
Does this mean blending doesn't change olive oil at all? No, certainly not. My tests failed to find a connection, but that's not enough to conclude that one doesn't exist. But it is enough to cast quite a bit of doubt on the degree of the phenomenon, if it exists at all. I fear, based on these tests, that we've been overstating the problem for some time while missing the more significant culprit: blended garlic.