Homiquest Creme Anglaise

This adventure has its roots in Christmas 2004 when I made a chocolate mousse pie with toasted pecan crunch. The recipe called for mounds of whipped cream to top the concoction off; however, my mother and I, prefering creme anglaise to whipped cream, decided to venture into the world of custards instead. It was simple enough when I made it the first time, so I have since considered creme anglaise part of my cooking repetoire. I had been craving this pie for about four months, so I made it again for Christmas 2006. However, it was an abject failure. The mousse was gritty, not sweet and didn't set well. The creme anglaise was an adventure of its own.

First, I started with Julia Child. Because where better to obtain a french sauce recipe than Mastering the Art of French Cooking? Nowhere. As I get into this story, I must first say that I am an engineer through and through. I like specific instructions. I think one of the best demonstrations of this was one time when I started screaming at a lemon curd recipe (after stirring patiently for 30 minutes): "What the fuck does stir until thick mean????? Seriously! What are you doing to me? Huh?" It's even better is when a recipe gives me a "very thick" instruction without any reference section in which the differences between thick and very thick are outlined (an engineer's cookbook would refer to viscosity). So Julia's instructions to stir my creme anglaise until "mixture coats the back of a spoon" was none too helpful. I know this coating of back of spoon probably means something significant. But it's probably like the difference between a simmer and a boil: you just don't know exactly what it is until someone explains it to you, but every cookbook assumes that you already know it.

So, I made Julia's recipe. When it looked to me like my mixture was coating the back of a spoon, took it off the stove, dutifully covered mixture with plastic wrap, and put it in refrigerator. An hour later, I checked on my mixture and it was not a thick creme, it was a liquid milk. My mother and I had a conference. This would not do. We decided to remake the recipe. Take two.

Some background. Custards are made like this: you heat up some milk, (if you are really gourmet) you steep your vanilla bean for 20 minutes or so, reheat the milk, pour the milk in to a mixture of egg yolks and sugar at a slow drizzle so as not to curdle the egg yolks, pour everything back into the pot and cook the mixture on the stove until it thickens. Or reaches a temperature between 155 and 170 degrees (the range I saw in the multitudes of recipes I have read). Or coats the back of a spoon. Then you strain the mixture, cover it with plastic wrap (placing the wrap right against the mixture so as not to form a nasty skin), and cool in fridge.

Now, the three options for knowing it's time to take your custard off the stove can be broken down as follows:

1. Until mixture thickens: useless for someone like me whose thoughts while cooking go like this: "Is it thick? Maybe I just think it's thicker? Maybe it is thicker? Is it thick? It looks thicker. I think the relative effort to pass spoon through mixture is larger, but is it?" Etc. Unless provided with a rheometer and a target viscosity, this is useless to a novice.
2. Until reaching a given temperature: This is the ideal solution; however, the accuracy of candy thermometers is very poor. Very, very poor indeed. They sell candy thermometers for $95 at some stores. Those are probably the accurate ones.
3. Coats the back of the spoon: Someone has to explain this to you. It's really meaningless until someone does. Really, unless you have magical custard knowledge, you have no idea what this really means.

Take two of our creme anglaise was better. My mother showed me exactly how to temper egg yolks. As I was stirring my mixture on the stove, I noticed it getting thicker. COOL. Now, the recipe does not give specific thickness instructions, but it does give an approximate time frame and temperature range. I was using my mother's candy thermometer which came to me with the caveat: "It might read 10 degrees too hot. Or too cold. I'm not sure, but it's not entirely accurate." My mixture was thick before I had reached the low end of my time or temp range, so I dutifully continued to stir. Then I noticed chunks. And more chunks. My mixture curdled. Gross. We tossed it out.

Take three. I was a bit gun shy, so I kept the stove low. I stirred for something like 40 minutes. The temperature never got very high on the broken candy thermometer, but I was under the illusion that it had thickened. But it hadn't, and we had anglaise soupe.

So yes, on Christmas Eve, I made creme anglaise three times and ended up with no creme anglaise. Which is ok, cause we ended up with crappy pie, too.

On Friday, I had a party. My craving for this pie had not gone away; I was going to make it. I was going to succeed. I researched chocolate mousse and creme anglaise recipes. I settled on the Cooks Illustrated Chocolate Mousse recipe from 1996, which makes a fabulous mousse that is creamy, thick, chocolaty though very complicated. I also settled on the Cooks Illustrated Creme Anglaise recipe. I bought a $35 candy thermometer from Williams-Sonoma to assist with my efforts. I was ready. But anyway, I'll get to my point and say, that my mixture started to curdle because my damned thermometer was off... way off. It curdled at 120 degrees when every recipe I've read says that the low end of curdling range is 155 to 165 (later testing revealed water boiling at 190ish degrees per this thermometer). But I strained it multiple times to remove all curdles and to cool mixture down and was successful.

Lesson of all this: Unless you are a stubborn jackass, use whipped cream.