Saturation, Distillation, Improvisation: A Story About Procedural Knowledge And Cookies

Most propo­si­tional knowl­edge (knowl­edge of facts) is pretty easy to come by (at least in prin­ci­ple). There is only one cap­i­tal of Venezuela, and if you wish to learn the cap­i­tal of Venezuela, Wikipe­dia will co­op­er­a­tively in­form you that it is Cara­cas. For propo­si­tional knowl­edge that Wikipe­dia knoweth not, there is the sci­en­tific method. Pro­ce­du­ral knowl­edge—the knowl­edge of how to do some­thing—is a differ­ent an­i­mal en­tirely. This is true not only with re­gard to the ques­tion of whether Wikipe­dia will be helpful, but also in the brain ar­chi­tec­ture at work: an­tero­grade am­ne­si­acs can of­ten pick up new pro­ce­du­ral skills while re­main­ing un­able to learn new propo­si­tional in­for­ma­tion.

One com­pli­ca­tion in learn­ing new pro­ce­dures is that there are usu­ally dozens, if not hun­dreds, of ways to do some­thing. Lit­tle de­tails—the sorts of things that sink into the sub­con­scious with prac­tice but are cru­cial to know for a be­gin­ner—are fre­quently omit­ted in ca­sual de­scrip­tions. Often, it can be very difficult to break into a new pro­ce­du­rally-ori­ented field of knowl­edge be­cause so much back­ground in­for­ma­tion is re­quired. While there may be ac­knowl­edged mas­ters of the pro­ce­dure, it is rarely the case that their meth­ods are ideal for ev­ery situ­a­tion and po­ten­tial user, be­cause the suc­cess of a pro­ce­dure de­pends on a vast ar­ray of cir­cum­stan­tial fac­tors.

I pro­pose be­low a gen­eral strat­egy for ac­quiring new pro­ce­du­ral knowl­edge. First, sat­u­rate by get­ting a di­verse set of in­struc­tions from differ­ent sources. Then, dis­till by iden­ti­fy­ing what all or most of them have in com­mon. Fi­nally, im­pro­vise within the re­main­ing search space to find some­thing that works re­li­ably for you and your cir­cum­stances.

The strat­egy is not fully gen­eral: I ex­pect it would only work prop­erly for pro­ce­dures that are widely at­tempted and shared; that you can af­ford to try mul­ti­ple times; that have at least par­tially in­de­pen­dent steps so you can mix and match; and that are in fields you have at least a pass­ing fa­mil­iar­ity with. The sort of pro­ce­du­ral knowl­edge that I seek with the most reg­u­lar­ity is how to make new kinds of food, so I will illus­trate my strat­egy with a de­scrip­tion of how I used it to learn to make mer­ingues. If you find cook­ies a dread­fully bor­ing sub­ject of dis­course, you may not wish to read the rest of this post.

I. Saturation

The first step is to col­lect pro­ce­du­ral in­struc­tions for the ob­ject of your search from many differ­ent peo­ple, sat­u­rat­ing your field of search with a va­ri­ety of recom­men­da­tions. A Google search did it in my case; for more es­o­teric knowl­edge, it might be nec­es­sary to look harder. Half a dozen of the more pop­u­lar sets of in­struc­tions tends to be plenty for recipes, but the ideal num­ber could eas­ily be higher for pro­ce­dures with a wider var­i­ance of de­tail or an un­usu­ally high num­ber of peo­ple who have no clue what they are talk­ing about. Here are four recipes for mer­ingues that I referred to and one recipe for mer­ingue pie top­ping that also in­formed my learn­ing. I also got one recipe from a friend.

All of the recipes pur­ported to teach me to do the same thing: turn some egg­whites and sugar (and vary­ing other in­gre­di­ents) into puffy lit­tle cook­ies. They varied in such de­tails as: in­gre­di­ent ra­tios, type of sugar, other in­gre­di­ents called for be­sides egg­whites and sugar, oven tem­per­a­ture, what to line the cookie sheet with, and men­tions of other fac­tors such as hav­ing a clean mix­ing bowl or hu­mid weather.

II. Distillation

The sec­ond step is to ex­tract what all of the pro­ce­dures have in com­mon, and de­cide which non-ubiquitous steps to in­clude. In this case, I first had to mul­ti­ply all the recipes to make them call for the same num­ber of egg­whites (since those are very difficult to halve or oth­er­wise ad­just, I chose them in­stead of sugar as my start­ing point). All five of the recipes (af­ter this re­vi­sion) called for four egg­whites; all of the recipes call for ei­ther caster/​su­perfine sugar or un­speci­fied sugar1; all of them call for vanilla; all of them in­struct me to beat the egg­whites to peaks first and then add the sugar and beat it in. Four of them call for salt. Four of them call for cream of tar­tar. Most of them call for com­po­nents like candy and nuts, but since I know that mer­ingues come in a wide va­ri­ety of fla­vors (by, for ex­am­ple, read­ing these recipes) I treat these all as op­tional. Pro­posed oven tem­per­a­tures/​bak­ing times are (200/​1.5 hours), (250/​30 min­utes), (300/​25 min­utes), and (325/​15 min­utes). They vary in whether the cook­ies are to be baked on a greased cookie sheet, a greased and floured cookie sheet, on bak­ing parch­ment, on pa­per tow­els, or on tin­foil.

A good place to start is to go with the ma­jor­ity: I de­cided to in­clude both salt and cream of tar­tar. Next, I elimi­nated the im­prac­ti­cal: I could not find su­perfine sugar at the store and I don’t own a food pro­ces­sor, so I went with gran­u­lated sugar. As for the rest of the in­struc­tions, it was a free-for-all. No two recipes agreed about the cookie sheet ar­range­ment; the two of them that men­tioned “crack­ing” dis­agreed on whether it was a de­sire­able out­come; and worst, none of them ex­plained why ev­ery time I tried to make these cook­ies, they re­fused to foam up and form peaks2. Time for the last step.

III. Improvisation

A close read­ing of the more ver­bose recipes turns up ur­gent cau­tions about not let­ting any grease into the bat­ter, be it a smear from a prior cook­ing ad­ven­ture left on the mix­ing bowl, a bit of yolk, or—in one recipe—the oils that are nat­u­rally on skin. This last, it turned out, was the key: I was sep­a­rat­ing eggs by hand, and was not very neat about it. I switched to a tech­nique recom­mended by a friend in­volv­ing spoons, and presto, I could get the mer­ingue bat­ter to hold peaks...

But how long to cook them, at what tem­per­a­ture, and sit­ting on what? There, it was nec­es­sary to ex­per­i­ment (for­tu­nately, af­ter hav­ing nar­rowed the search space some­what). This stage de­pended as much on my per­sonal taste, the lo­cal weather, and the be­hav­ior of my oven as on the ac­cu­racy of the origi­nal recipes; it seems that my oven runs hot, so I need to bake them at 250 de­grees or cooler and babysit them af­ter the first ten min­utes, or they will burn. Ad­di­tion­ally, parch­ment pa­per and tin­foil3 wound up burn­ing the bot­toms of the cook­ies be­fore the tops were even dry; pa­per tow­els worked.

1Caster and su­perfine sugar are the same thing, and you can make a rea­son­able fac­simile us­ing a food pro­ces­sor. When the type of sugar is not speci­fied in a recipe, it means to use gran­u­lated white sugar; other kinds are named (e.g. light or dark brown sugar, turbinado sugar, con­fec­tioner’s sugar, etc.). This is one of the ex­am­ples of a situ­a­tion where back­ground knowl­edge of the field comes in handy.

2Not that this stopped me from bak­ing the bat­ter any­way. It just turned into round, flat cook­ies in­stead of puffy, light ones.

3I didn’t get around to try­ing greased nor greased and floured bare cookie sheets—I pri­ori­tized these tests last be­cause they in­volve more dishes to wash.