Eccentric Flower:201012/Recipe Tolerance

"Recipe tolerance" is what the people who work for large food conglomerates call the ability of recipes to handle variance and ineptitude. When you put together a recipe to go on the side of a box of vanilla wafers or frozen spinach, you want to make it as failure-resistant as possible. We're not just talking foolproof, you understand. It also has to be reasonably able to handle people like my wife, who is genetically incapable of following a recipe exactly as written.

Actually, I don't think my wife, who should be running a restaurant except the stress would kill her if I didn't kill her first, has ever used a recipe off the side of a food package in her life, even as a starting point ... but I sure have, and I've given a lot of thought to recipe tolerance over the years.

See, recipe tolerance is one of the two big reasons (the other being that it's commercially safer to be bland) that recipes on the side of packages are, for want of a better term, banal. To my mind, the more interesting a recipe is, the higher the chances it's going to contain some step that has to be done exactly thus-and-so or which requires a trained eye or some other trickery.

And then we have the truly frustrating situation: A recipe which looks like it won't require much in the way of care or caution, but then turns out to have hidden pitfalls. Usually these pitfalls are because most people and publications, even to this very day, write recipes like they're sending a telegram and every word counts. (Younger folks may need to go look up what a "telegram" was, he says crustily.)

Now, I write verbose recipes, and I'm sure there have been people who have read my recipes and wished I would cut the chatter and just condense. Today I'm going to show you an example of why I don't.

Today was Confection-Making Day, and I had three recipes planned. One I had made before many times, another I had never prepared myself but had consumed on many occasions, and thus knew what they should end up like, and a third was a complete, utter unknown. That last is the topic under consideration.

The recipe was called Mexican Orange Drops - Now, originally, I found this recipe in The Joy of Cooking, but that recipe led me to believe that it was a sort of dropped-with-a-spoon candy, and for various reasons I specifically wanted something that was rolled into balls, or at least spherical. But it also had oranges and could accommodate pecans, two flavors which I felt definitely provided good contrast with the other two recipes I already knew I was making. So I went to the web and looked for close variations.

All the recipes I found were very clearly based on the same recipe as the Joy version, but some of them implied that the finished product could be rolled into balls, so I decided I'd give it a try, even though the recipes, each and every one of them, had some perplexing-sounding bits. (After years in the trenches, I have a sixth sense for when a recipe is Not Quite Telling Me All I Need To Know, and these were setting off that alarm.)

Here, extracted under Fair Use for purposes of criticism and comment, is the recipe exactly as given in The Joy of Cooking. I do not recommend you follow this one; I will give you my own version later.

Before I comment on this version, now we'll put in the most commonly found version on the web. This version appeared in enough places that I can't attribute it. In some places the nuts were specified as pecans, in other places not.

Now, some observations. I have never cared for the Joy method of integrating the ingredients into the recipe, because it makes it harder to put your shopping list together. On the other hand, Joy has the good sense to split ingredients (that is, they list 1 cup of sugar and 2 cups of sugar separately because they will be used separately, instead of springing that on you during the procedural part). They also don't truck with abbreviating "cup" or "teaspoon," which only gets people into trouble. (I don't mind 'c' for "cup" because in the American recipe world there is nothing else it can be. But 't' vs 'T' confusion has broken many a heart. At least write out 'tsp' or 'tblsp'.)

The web recipe doesn't say why the covered cooking period is needed; that was information I felt was pretty important. Others might disagree, under the "I don't need to know why to do it, just what to do" theory. Alas, I am a "why" cook. If I understand what we're trying to do, it improves my judgement about how it's coming along when I actually do it.

Both recipes assume you know why you shouldn't stir it during cooking (to prevent you stirring in crystals from the side of the pan and thus possibly recrystalizing the whole mixture), but don't say when or if it's safe to stir it later. Joy at least attempts to tell you how brown to get the sugar, although they do a rotten job of it.

To these ambiguities I would add my own mistake: You see, I had to copy the recipe by hand because we don't have a working printer right now (and I forgot I could just take the iPad into the kitchen like I did for the buckeyes, later). I made what you might call a transcription error. It was non-fatal but alarming. We'll come to that.

Here is my recipe, complete with my usual warnings, digressions, and operational anecdotes. But in order that you may separate out what I consider absolutely essential instructions from all the side chatter, I will help you out by italicizing the key bits.

Ingredients:

• 2 oranges (for both juice and zest)
• 1 cup evaporated milk
• Orange extract/flavoring
• 3 cups sugar (will be divided)
• 1/4 teaspoon salt
• 1 to 2 cups finely chopped pecans

Zest the oranges, then squeeze them; you will need 1/4 cup of their juice. One large orange, if you've picked a good one, will give you about 1/4 cup of juice, but you'll still want the zest from two, so you'll just have to do something else with the other, denuded orange. Grocery stores have a bad habit this time of year of only selling navel oranges loose - not the best juice OR zest orange. If you do have to buy navels, pick two that actually have a strong "orange" smell when you hold them up to your nose. These are likely to have the most orange oil in their zest.

Zesting and squeezing the oranges first is an example of a mise-en-place deficiency in both of the other two recipes. Clearly you are not going to be able to give your attention to zesting oranges while you are attending a molten caramel mixture. Always, always break out a recipe into preparatory steps which can be done first (mise-en-place) vs. live procedure, especially where temperature-sensitive or timing-sensitive steps are involved. This is why my gumbo recipe makes you chop and prep EVERYTHING before the roux begins - because I want you to give the roux your undivided attention. Hot sugar is like that too.

Heat the milk in a double boiler; it should be hot but not even showing signs of a boil. The tricky bit of this recipe is the fact that you have to have two hot liquids ready and hot while you're timing a dangerous sugar. Since I couldn't find any reason why the orange juice needed to REMAIN boiling - hell, I'm not even sure why it had to come to a boil in the first place, I heated the orange juice to a boil in the microwave and then let it sit in its cup on the stove while the sugar part was happening. But you could also keep it at a gentle boil in a pot, I suppose, if you don't mind watching three pots at once.

If you like, add 2-3 drops of orange extract to the orange juice after it is heated. I did this because I didn't trust the flavor of the juice from navel oranges, but it is optional. If you do it, do it now, you won't have time or attention for it later.

Now the sugar. In a heavy pot - you want one that seems at least a size too large for the amount of sugar you're heating - heat 1 cup of the sugar over medium heat until it is a brown syrup. Stir constantly while heating. Be careful; sugar passes brown and goes to burnt very quickly.

I had never heated sugar dry before. I'd always made recipes which called for at least a little liquid added to the sugar. Dry sugar does not color like dry flour. It will appear to do absolutely nothing for a while. Then clear liquid will appear to be seeping out from under the sugar. That's the sugar melting. When you stir, the liquid will wet the rest of the sugar and it will clump into rocks. Keep stirring. The rocks will slowly dissolve as the liquid first turns gold and then begins to brown. Try to break them up with the spoon. If the syrup gets too gold before you have dissolved all the lumps, take it off heat for a bit until you've gotten them handled. Don't be scared to remove the pot from the heat or turn the heat down if you have to.

Once you have gotten this to brown, turn the heat down or do this next step off the heat: Stir in the orange juice slowly, then stir in the milk. If you miscopy the recipe and you add the milk first, and you add it too fast, it will foam TREMENDOUSLY - gigantic bubbles of froth filling the pot - and you will think you just ruined your recipe, and you'll have a mess if your pot wasn't big enough. (Mine was, thank heavens). The main reason to do it slowly, though, is not to protect the delicate sensibilities of the milk, but to keep the mess from cooling down too fast. I added mine too fast, and even though both liquids were hot, the near-caramel I'd been building in the pot suddenly congealed into this nasty, sticky, thready stuff that wanted to clump on my spoon and adhere to the surface of the pot. Don't worry, though; it is possible to work it all back in, it just takes time and heat.

Add the remaining 2 cups of sugar and the salt. Stir well until fully combined. Stir on heat, but don't let it come to a boil just yet if you can avoid it. We want to make sure everything is combined, including all that caramel that you had to scrape in from your spoon and the bottom of the pot, before boiling.

Bring to a boil, then cover and let boil for 3 minutes. No peeking. As Joy says, the idea is to let steam dissolve sugar crystals back into the pan.

Cook uncovered without stirring until the mixture reaches 238°. Good luck with your candy thermometer. It can't touch the bottom or sides of the pot (gives false readings if you do) and this stuff is sticky and can't be agitated too much. I had to hold my thermometer in the pot at exactly the right depth without bumping the bottom, for several minutes at a time so I could get a reading. It was like playing a particularly unrewarding game of "Operation."

How long will this take? I have no idea. If you don't have a candy thermometer this is probably not a good recipe to make. I didn't look at the clock.

Remove from heat and carefully stir in the orange zest. I was like, "Oh lord, is it safe to stir it now?" I didn't get any crystals, but I was careful to stir without touching or scraping the sides. If you don't stir at this point you get a large lump of caramelized orange zest that will never come apart later.

Let mixture cool. How cool? Cool enough that you can dip your finger into it to get a taste without suffering mortal harm, at the very least. What I did, I took it off the heat, I went to go chop the nuts, I prepared the dry-mixture part of my next recipe, and then I came back to the mixture. That would have been ten to fifteen minutes, I suppose. I probably could have waited even longer, but if you get it too cool, the nuts will be hard to stir in, because this stuff is thickening as it cools - at least to a point. We'll come to that.

Beat until creamy. If you don't mind cleaning this glop out of a wire whisk, that's the tool to use. But you'll need a sturdy one. By this time, stirring some air into this mixture is going to take considerable strength of wrist. Stir in the nuts. I used 2 cups chopped, twice the amount the original recipe called for, both because I like pecans and also because, by this point, I was convinced the nuts were going to be the only thing that gave this mess structural integrity.

See, around this point, it became clear that while I had not ruined the recipe and wasted time and ingredients, it was also never, ever going to be rolled into balls. It was a soft, extremely sticky mass, even with the pecans firming it up. You could still have spread it on a cracker without much difficulty. The question now was how to rescue it. I decided to butter a cookie sheet and spread it out thinly onto that sheet, then refrigerate the whole sheet - the idea being that once it was stone cold, it might have solidified enough to be cut into squares or rescued some other way.

I then went about my business, making the bourbon balls, an absolutely foolproof recipe, and the buckeyes, whose cores were foolproof but which needed to be partially dipped in chocolate, which is always fussy. Once that was all done, I revisited the now quite-cold toffee/caramel stuff. It was reasonably firm to the touch, but not stiff - I mean we are not talking brittle here, if you pulled a corner up from the pan it acted like a very soft taffy, bending and stretching.

I cut it into rough rectangles with an oiled pizza cutter, and pulled one of them up. No, it was not nearly firm enough to stay as a rectangle. But - hey! - it was now firm enough to roll into balls. Barely. The heat of rolling one into a ball softened it up enough that the four experimental models I rolled immediately began to sag and ooze back into one another, trying to recombine into an amorphous mass.

I yelled for my wife and by working tag-team, me pulling up rectangles and making them into balls, her immediately rolling them in cornstarch, we believe we made a batch that will ideally stay separated - although we don't expect them to stay spherical; in fact we had to store them in layers separated by wax paper because the weight of piling them caused the bottom ones to smush. And then they went right back into the fridge. This stuff is firm enough to handle, but only if you keep it cold. If you had chilled the entire mass as a blob, instead of spreading it out into a sheet first, you could probably have pinched off bits and rolled them into balls - if you could work fast enough.

So: Chill mixture thoroughly, divide into pieces, roll pieces into balls and then roll in cornstarch to coat. Keep candy refrigerated.

Now the punch line: As much as a pain in the ass as this was, this candy tastes fabulous. Absolutely amazing; like a pecan praline only with caramel and orange undertones. So I regret nothing. And if you make them, you will have no regrets either - but I still felt you should be forewarned. Some recipes don't tell you nearly enough about what's actually going to happen when you try to make them.

Here is an example of an extremely tolerant recipe. I've been making them since I was very small and have been winning friends and influencing people with them for years. I've come to the conclusion that, as long as you have a food processor with a big enough bowl and the patience for the rolling process, it is impossible to screw up this recipe.

Ingredients:

• 1 to 2 cups finely chopped walnuts (see recipe)
• 60 vanilla wafers
• 2 tablespoons cocoa powder
• 1 cup confectioner's sugar, plus more for rolling
• 3 tablespoons white corn syrup
• 1/4 cup good bourbon
• 2 to 3 drops vanilla extract (optional)

You will want: A food processor with a big bowl, a plate with a rim, and something with a lid to store the balls in as you roll them.

I am ambiguous about the amount of nuts because I start with intact walnut halves and chop them first in the food processor. If you end up anywhere from one to two cups of finely chopped nuts (small chunks okay, they will keep getting chopped later), you're fine.

Add the vanilla wafers to the chopped nuts in the processor bowl. Yes, I do mean for you to count them out. Sixty is the number and sixty it shall always be. It will use almost all of a standard box. Give the processor a few good pulses to pulverize the cookies.

Add the cocoa and sugar and pulse a couple of times until thoroughly combined.

Add the bourbon, syrup, and vanilla. (The vanilla is my personal addition to the recipe because vanilla wafers just aren't very vanilla-y these days. It's probably purely psychological, because I'm not sure you can taste a couple of drops of it in all this. Omit it if you like.) Hint: If you measure 1/4 cup bourbon into a measuring cup, measure the syrup into that, and give it a stir, the booze dilutes the syrup so that less of it sticks to your measuring spoon and cup and it pours into the food processor more easily.

Run the food processor until the mixture forms itself into a single big ball of dough.

Rolling time! You want to do this with either two people or you want to keep your hands separate. If you are alone: Pour a decent pile of powdered sugar onto the rimmed plate (about half a cup, but I never measure it; you may need more). With one hand, pull off chunks of the dough and form them into balls using only that hand. This is possible. Your dough hand then drops that ball onto the powdered sugar plate; your other hand rolls that ball in powdered sugar to coat and drops the ball into the storage container.

This means you do not get powdered sugar in the dough (which weakens its spherical integrity) and that you don't get loose scraps of dough in the powdered sugar, but most importantly, it means you don't get the buildup of successive strata of dough and sugar on your hands which leads to what Alton Brown calls "club hand."

Of course if you have a second person, lucky you, then one person handles the dough and the other handles the sugar, and you get it done in much less time.

Makes, I dunno, how big did you make the balls (smaller is actually better), and how many walnuts did you put in? At least three dozen would be my guess.

Store in the refrigerator. They will taste better after they've had a day to cure, and even better the third day, and they don't usually make it much longer than that. Warning: These do contain a quarter cup of absolutely uncooked bourbon, so teetotalers beware! Although if you ate enough of them to get a buzz on, you'd go into sugar shock first.