Artichokes; butter

Cook artichokes: peel the stem, cut off the top third or so, rub the exposed cuts with a lemon, and steam. Meanwhile, melt butter with thyme, lemon zest, and a dried hot pepper. When you can easily pull a leaf off the bottom, add lemon juice and extra virgin olive oil to the “sauce”. Eat slowly, saving the heart for last.

Artichokes; butter

In other news, the carote in agrodolce were absolutely delicious. The raisins add a really delightful dimension beyond the sweet vinegar. While I enjoyed their flavor, the pine nuts had a mealy texture—perhaps pre-toasting would avoid that.

Posted in Recipes | Leave a comment

Il Cucchiaio Americano: carote in agrodolce

In order to practice my Italian, I bought a copy of Il Cucchiaio d’Argento, a hefty tome with over a thousand pages. I’d hesitate to say it’s definitive for Italian cuisine—they’ve managed to publish other volumes containing regional cuisine—but it’s something very close. I’ve been told it’s considered a little sciccoso: perfect.

Carote in agrodolce

I’ve made a few things out of it. Here’s carote in agrodolce: sweet and sour pickled carrots. Perhaps the most famous agrodolce dish is caponata. Unlike caponata, which cooks for some time and is served immediately, this recipe cooks quickly and then sits for nearly three weeks. I have to be frank: I have no idea how these carrots will come out—it says to wait twenty days. But this quick cooking will, I imagine, leave the carrots delightfully crispy and pickle-like. This recipe is “an optimal contour for mixed boiled meats”, so there’s that, too.

This recipe plays to my tastes, seeing as I’m a complete pickle pig. My New Year’s Resolution, however feeble, was to “pickle more things”. And also to use our champagne saber as often as possible, though that’s a different story. We almost made this as a contorno for dinner the other night, and then realized we’d be awfully hungry by the time it was ready next month. But I made it this morning, and I’m translating the recipe now, while the spirit moves me. If it’s no good, I’ll come back and say so later.

Recipe: Carote in agrodolce

Sweet-and-sour pickled carrots
Translated and adapted from Il Cucchiaio d’Argento, p. 398


  • 300g carrots
  • 300g white sugar
  • 100g raisins
  • 50g Italian pine nuts
  • 1 L white vinegar
  • pinch salt


Thoroughly clean a wide-mouth mason jar and lid. Peel and thickly julienne the carrots. Soak the raisins in water.

Bring the vinegar, sugar, and salt to a boil. Boil the carrots for three minutes. Drain the carrots but strain and save the liquid.

Squeeze the raisins dry and then mix in the pine nuts. Fill the mason jar with alternating layers of carrots and the raisin mixture. Cover the contents of the jar with the cooking liquid, topping up with more vinegar as necessary. Seal the jar “absolutely hermetically” and let sit twenty days.


Preparation time:

The recipe doesn’t in fact specify peeling the carrots. Nor does say to strain the liquid, but I didn’t want the foam in the final product.

It is important to use real Italian pine nuts—the much cheaper Chinese pine nuts from a different species can seriously mess with some people’s palates. Just say “no” to pinemouth.

Also, don’t be misled by my super professional photo: I normally keep my jars on the windowsill, but in this case I’m worried about the fat in the pine nuts going rancid. It’s probably best to keep this jar in a cool, dark corner.

Posted in Il cucchiaio americano, Recipes | Leave a comment

Chocolate Preferences by Gender

Apparently, I’m what a feminist looks like. When considering the chocolate taste test, it didn’t even occur to me to break down tasters by gender—it took my friend Ruthie’s suggestion for me to realize that I hadn’t considered an obvious variable. (Thanks, Ruthie!) The tasting team comprised 6 men (Justin, Andy, John, Bill, Jonah, and I) and 4 women (Sydnor, Lizzie, Adrienne, and Nicole).

Likes by Gender

Likes weren’t particularly gendered except for samples E (Theo Organic), I (Mast Bros. Brooklyn Blend), and O (Pure Icelandic Noi Sirius).  Only men liked sample E (Justin, John, Andy, and Bill) and sample O (Bill and I); only women liked sample I (Sydnor and Lizzie).  You can see the spread in this updated summary chart, which includes lines for the male and female means.

Normalized Summary with Gender


While there are differences, the mean spread between male and female means is 0.46 points, with a standard deviation of 0.38 points. The “Likes by Gender” graph above shouldn’t be misinterpreted as some chocolates being “for men” and others “for women”—the gender effect seems to be a small one.

Update: Here are the full tasting notes for these samples. Not particularly illuminating, I don’t think.

Sample E Sample I Sample O
brittle cherry overripe fruit way overdone good texture furniture polish gritty strong smell overripe blah dirt sock old moldy fertilizer poop dirt a little bit of rottedness i guess grainy wine high booze blackberry smooth finish gritty waxy sooty stronger fruity smell smooth good texture bland plain aromaless crumbly chalky winey fruity grainy as it melts baking chocolate m&ms crumbly texture layered flavor boozy tart at first mid-density smooth texture boozy fermenting fruit tart rotting fruit boozy acerbic withdrawn will stab you in the night bitter fruit with aluminum not very chocolatey neutral smell texture not good chalky grainy cloying a little little aroma berry light fruit edging on sweet coconut original smell sweet milky light and sweet not complex bubble gum blah couldn’t finish lower cocoa sweeter baby powder sweet coconut oil light saccharine wax coconut and weird fats which overwhelms the cocoa subtle smell fruit light not strong chocolate flavor a little too sweet

One (over)interpretation of this is that women don’t want to eat poop dirt and men don’t want chocolate to stab them in the night—though I would have guessed these things were universal.

Posted in Taste tests | Leave a comment

Taste Test #1: Chocolate

We tasted chocolates. We were ten tasters: Justin, Sydnor, Lizzie, Adrienne, Nicole, Andy, John, Bill, Jonah, and me—Michael. I asked each taster to bring a bar of “fair trade” chocolate; most people brought dark chocolate which at least claimed to be fairly traded. We had 16 chocolates: 13 dark, 3 with dairy (samples A, L, and M). All but 2 of the chocolates were plain bars: one had cocoa nibs (sample G), and another wasn’t bar shaped (sample F).
Chocolate tasting notes
The methodology was slightly convoluted and didn’t completely blind the tasting. Lizzie chopped the chocolates, put them on plates with distinguishing markings facing down, and numbered each plate 1 through 16. (Thanks Lizzie!) Sydnor then randomly switched the numerical labels for alphabetical labels, A through P. (Thanks Sydnor!) Lizzie did a remarkably good job of camouflaging bars, but some bars retained their distinctive texture or markings. That said, the group was willing to avoid open speculation about which bar was which. In addition to making tasting notes, tasters assigned each chocolate a rating between 1 (out of my mouth!) and 5 (platonic ideal of chocolate), with tasters freely using decimal ratings, continued fractions, etc. Tasters split into small groups: Nicole, Adrienne, and Andy; Justin, Bill, John, and Sydnor; Jonah, Lizzie, and I.

Sample key
Alter Eco Dark Velvet*
Madecasse Rich & Fruity Endangered Species Dark Trader Joe’s Dark
Theo Organic Dark Taza Mast Bros. Cocoa Nibs Trader Joe’s Dark Fair Trade
Mast Bros. Brooklyn Blend Alter Eco Dark Blackout Equal Exchange Very Dark Endangered Species Milk*
Trader Joe’s Milk Fair Trade* 365 Dark Pure Icelandic Noi Sirius Cote d’Or Noir Intense
* Contains milk
Round, not bar shaped
Contains cocoa nibs

The raw ratings are available in CSV; the tasting notes are available in TXT (mildly preprocessed: tasters aggregated, one line per sample).  Statistical amateur hour follows.

Tasters did not rate surjectively: many tasters never gave a 5, while Andy never gave anything a 1. Supposing that a taster “likes” something they rate at 3 or above, here is the breakdown (graphs are clickable). The samples are on the x-axis; the y-axis is the number of “likes”.

Aggregate likes
Sorting chocolates by popularity, we find:

  1. TJ’s Dark (D), Mast Bros. Cocoa Nibs (G) (8 likes)
  2. Alter Eco Dark Velvet (A), Equal Exchange Very Dark (K), 365 Dark (N) (7 likes)
  3. TJ’s Dark Fair Trade (H), TJ’s Milk Fair Trade (M) (6 likes)
  4. Theo Organic Dark (E), Taza (F), Cote d’Or Noir Intense (P) (4 likes)
  5. Madecasse Rich & Fruity (B), Endangered Species Milk (L) (3 likes)
  6. Mast Bros. Brooklyn Blend (I), Pure Icelandic Noi Sirius (O) (2 likes)
  7. Endangered Species Dark (C), Alter Eco Dark Blackout (J) (0 likes)

To my mind, the moral of the story is that supermarket brands are inoffensive: TJ’s and the Whole Foods 365 brand both ranked near the top.  The Mast Bros. bar was the highest rated “boutique” bar. I didn’t like it much but 8 out of 10 tasters did—so what do I know.

Linearly normalizing tasters to their own range of scores, the full results are below.  The taster colors in the legend match the cumulative bar chart of “likes” above. Again, samples are on the x-axis, and the y-axis are the ratings.
Normalized Summary

Looking at inter-taster effects, k-means with k=3 isolates two pairs of tasters with fairly divergent palates: Sydnor and John’s tastes are plotted in blue, while Bill and mine are in red; everyone else is plotted in green, and the mean over all tasters is in black. The axes are the same as the previous graph.


The four of us deviated a fair bit from the mean, yet Bill and I rarely agreed with John and Sydnor. The points where we did agree are telling: the universally despised Endangered Species Dark and the not-much-better Madecasse Rich & Fruity; the classic Trader Joe’s Dark and Dark Fair Trade, the caramelly and accessible Endangered Species Milk.  I don’t think the root cause is differences in perception, since our notes often read similarly: John, Bill, and Sydnor all mentioned coconut in their notes on sample O; we all agreed that Taza was “like putting playground sand in [our] mouth[s]”, but some of us enjoyed the “interstitial delights” while others didn’t.  De gustibus non est disputandum.  Or at least I hope not—Sydnor and Bill are married.

I’m not sure what to make of this heatmap, but a statistician friend of mine recommended making it (thanks for the help Justin!), and it certainly looks cool.  Darker reds indicate higher positive correlation; lighter yellows indicate anticorrelation.  Yes, I know this is the wrong way to do it if the diagonal is white.


All of this aside, I suspect that there was a strong “considered taste” effect. Normally, we are happy to say that chocolate is chocolate, and it goes in the mouth.  Considered next to each other, though, each chocolate’s idiosyncrasies and—let’s not be coy—flaws fall into clearer relief.

I didn’t get what I wanted out of this taste test: a clear winner. I found instead that:

  • 16 samples is way too many,
  • chocolate can taste like “poop dirt” (thanks, Nicole!), and
  • disgusting chocolate still makes great devil’s food cake (happy birthday, Kelly!).

In the end, I doubt I’ll return to Chocodome: 16 bars enter, ten tasters leave… unpleasantly full and distressingly caffeinated.  A square or two after dinner will suit me just fine.

Update: I’ve also done some analysis by gender.

Posted in Taste tests | Leave a comment