Sometimes a new dish lays at the end of a single line of thought and other times it’s at the intersection of many. In the case of tarte chocolat-noix de pécan, the lines of thought converged without quite crossing.

One of the standard items on the mignardises plate at La Folie in San Francisco is a sort of soft, crunchy, chocolate-toffee base with an equally thick layer of ganache on the top. When I was a stagiaire there for a couple of weeks in 2009, I never got around to asking Mie Uchida, the pastry chef, how it was made. I’m not sure I even saw anyone constructing it, although I did help make some of the other mignardises that were being served. Because this mignardise is made in a large baking sheet, hundreds of pieces can be prepared at once. It lasts well in a refrigerator, and so it is not something made fresh each day. When I started making mignardises for my guests, this particular one would pop up in my mind, but it wasn’t until one particular meal was being planned that I decided to create my own version of it.

After many years had passed, I decided to reconnect will some old friends. It had been almost ten years since we had any contact other than a Christmas card or two, and even those had ceased. When I tried calling the only telephone number I had for them, it no longer was connected. So I tried the email address I had in my address book. A few days later, I received an answer. It turned out that they had moved about an hour south of where we live, and that most of the reasons why they were too busy to get together in the past had ceased to be a problem.

We arranged a time to get together for lunch. My wife and I drove over-the-hill to their new house near the ocean. They had purchased a small fixer-upper and remodeled it into a dream house. We drove a short way to a restaurant and had a very pleasant French-style, bistro meal. The food wasn’t anything to write home about, but the company more than made up for it. Before we left for our next engagement, we agreed upon a date for them to come to our home for a Sunday lunch. We also decided to invite another couple that was part of our original circle. Our contact with this couple had also faltered over the years, especially after they had moved to a location almost three hours east of us, although it was a bit more frequent than with the first couple.

Our group relationship started in 1974 when I landed my first job after college. At the time, I was still living in Rochester, New York. In August of that year, I came home to visit my parents and do some job hunting. The main requirement for any new job was that the company move me back to Palo Alto where I owned a house.

I found a job at ILC Technology. They had just accepted a contract to design and build the interior lights for the Space Shuttle. My title was Optical-Mechanical Engineer, and my responsibility was the optics and housings for the lights. I later learned that there had been three applicants for the job and I was the third choice. The other two applicants had said the annual salary, about $16,000, was too low. John, half of the couple an hour to the south, was my immediate supervisor and Project Engineer. Hal, half of the couple to the east, was his boss and the Project Manager. Linda, John’s wife, also worked at the same company but in a different division. Barbara, Hal’s wife, wasn’t quite on the scene yet. This was also seven years before I met my current wife.

Over the years, the six of us have done many things together and had many enjoyable expanses of time filled with much food, drink, and conviviality. There’s also been numerous anniversaries, celebrations, deaths, projects, dance lessons, and obligations. Some of these brought us together, others prevented us from finding time to meet. This newly planned Sunday lunch was probably the first time the six of us would be together in almost ten years. Bringing the lunch together, food wise, wasn’t going to be a problem. I just planned a dinner that would start at one in the afternoon rather than six in the evening. There would be my usual three amuse-bouches, a first course, a main course, a palette cleanser, a cheese plate, a small salad, dessert, and a selection of mignardises. The only wrinkle in this meal is what caused it to converge with La Folie: John had developed an aversion to glutton. No wheat flour could be used in the meal. I thought now was the time to reproduce that one mignardise from my many memories.

When I think of toffee, I think of what the French call pralines, caramel-coated almonds. This in turn is often ground into a powder called praliné. Its relationship to toffee is only the caramelized sugar, but one still reminds me of the other. And in this instance, for reasons unknown, my mind went to pecans. So I decided to make a pecan crust for the these little treats. The top coat would be a simple, not too hard ganache. The end result would be cut into small squares.

This formulation is really two distinct preparations, prepared and applied separately. The pecan crust is based on the classic Graham-cracker crumb crust used for many cheesecakes. After sweetening the ground pecans, I simply add enough melted butter to hold the two ingredients together when firmly pressed into the bottom of the cake pan. The ganache is a standard formulation with enough fat to prevent the end result from being overly hard.

I use my 11-cm by 14-cm (4-1⁄3-in by 5-1⁄2-in) removable bottom, rectangular, Japanese cake pan. The large amount of butter in the crust means that no special preparation of the cake pan is required. This size pan produces 20 pieces and enough scrap for ample testing.