So, first I made flippant comments about “Convergence of Computer Science and Critical Theory Cookies”, and cookies that reference other cookies. Then I made some small loaf cakes with brownies for a base. Then Leonard pointed to his 2003 post on meta-desserts — desserts that reference other desserts.
As he points out, desserts can basically be piled on top of each other indefinitely, or at least until you hit the ceiling. This is why I like baking: you can leave out major ingredients, accidentally replace them with something else, freeze or heat up the result or cover it in chocolate sauce, and then when you’re finished you can chop it up, cover it with cream, mix it with fruit — and chances are it will still taste good.
However, there are limits, and also classificatory difficulties. What are the fundamental dessert types, the metaphorical atoms of dessert, or “dessertoms”? A brownie is very “stable”, which is to say it can be combined with many different desserts while still remaining delicious — but surely it isn’t a fundamental dessert type: a brownie is basically just a sulky teenage cake. A crepe, on the other hand, probably is a fundamental dessert type, but it’s a relatively unstable one — it won’t taste good if you put it on a cookie.
Furthermore, desserts can be transformed not just through the application of another sort of dessert, adding dessert type A to dessert type B, but also by the application of a Dessert Function. Dessert Functions are things like “freeze it”, “put nuts on it”, “take out all the flour”, “cover it in alcohol and set it on fire” — stuff you can do to any dessert that has a good chance of leaving it edible, or better still transforming it into an exciting new dessert.
Clearly this is a topic that requires for further discussion:
- a rigorously defined vocabulary;
- extensive research to discover the fundamental dessert types;
- some sort of consistency in what “applying dessert type A to dessert type B” actually entails; and
- Lots of little pictures on graph paper.
Well, if we have a Meta-Dessert Conference and Party, I can bring number 4.
I call it “A Sketch towards a Taxonomy of Desserts and Meta-Desserts”, though I’m thinking of adding a subtitle as well. I’ve listed dessertoms: cookie, cake, sweet bread, pastry, crepe, crumble, fruit, chocolate, cream, custard, egg-white-and-sugar, and ice-cream. (Obviously this is a very broad-grained study, and further research would be well-advised to, eg, clarify that the broad category of “cake” can itself be divided into a number of fundamental types which can have transformations enacted upon them while still remaining cake). These run along the top of the page; following a column down, you can see what might happen to each dessertom when a different dessertom is applied to it (to apply Dessertom A to Dessertom B, you either (a) use Dessertom A as a component ingredient in making Dessertom B; or (b) put Dessertom A inside Dessertom B; or (c) put Dessertom A on top of Dessertom B, in roughly that order of preference).
I’ve also included seven Dessert Functions: shrink, freeze, chill, put in food processor, heat, add leavening, and remove leavening. At this point I ran out of graph paper, and had to leave out “add nuts”, “squash”, “take out flour” etc, but just because they aren’t on the page doesn’t mean they aren’t equally valid.
To reference the entries on the sheet I will refer to the Dessertom in brackets, and prepend the applicable operation: the notation for applying Custard to Crumble is therefore Custard(Crumble); performing Freeze on Chocolate is Freeze(Chocolate). The result of the operation is indicated by an arrow: Freeze(Chocolate) -> Frozen Chocolate.
Many Dessertom-plus-operation combinations provide a dessert that already exists as a major (albeit not fundamental, except in the case of Freeze(Custard)) dessert in its own right, for example:
Chocolate(Cream) -> Ganache
Add Leavening(Crepe) -> American Pancake
I have labelled these in blue.
Many combinations provide something that, though not extant as a major dessert type, is widely eaten or seems like it would be nice:
Put in food processor(Cake) -> A big pile of cake crumbs
Custard(Custard) -> Loads of custard
I have labelled these in green. Finally, dessertoms are not infinitely stable: many, such as “crepe” and “sweet bread”, are actually very unstable, and can easily turn into a dessert that would be horrible. For example:
Freeze(Sweet Bread) -> Frozen Bun
Crepe(Chocolate) -> A crepe hidden in a chocolate bar
I have labelled these failed desserts in red.
A RED DESSERT SHOULD HAVE NO FURTHER OPERATIONS ENACTED UPON IT. You will just be throwing good ingredients after bad. Blue and green desserts, however, can be taken back up to the top of the chart, where they can have another dessertom or dessert function applied to them, at some risk of “flaring”, or turning red (particularly if you operate within an unstable row — “heat” or “freeze” for example). Many quite widespread desserts are the result of iterative runs through the chart, hence:
Chill(Add Alcohol(Cake(Custard(Cream(Fruit))))) -> Trifle
This is a sixth-order dessert, five steps removed from a first-order dessert, the “dessertom” we met earlier. My small brownie-based cakes (recipe later in the week) are slightly more complex:
Shrink((Chocolate(Cream))(Remove Flour(Chocolate(Cake)))
(Chocolate(Remove Leavening(Chocolate(Cake))))
-> Brownie-bottomed mini-cakes
Which is to say, a chocolate cake without leavening (also known as a brownie) has chocolate (chips) added to it, and then has flourless chocolate cake applied to it, and then ganache applied to that. And then it’s all shrunk. Terminology: a third-degree fourth-order dessert, perhaps? Three separate desertoms, with the most extensively developed having undergone four operations.
The most urgent question facing meta-dessert studies today, then, is: what would be the highest-order dessert that still tastes good? Note that you can apply the same operation more than once within a dessert’s construction, but doing it twice in a row reduces down to a single operation:
freeze(freeze(cake)) = freeze(cake)
Furthermore, I think a rule reminiscent of Chess’s threefold repetition rule is called for: if the same sequence of two or more operations occurs three times, then the dessert is over. Hence, no seventieth-order desserts that are just fruit and cream on top of fruit and cream on top of fruit and cream on top of etc.
The whole chart is here, though unfortunately non-Flickr members can’t see it at its proper size, and a medium-sized version visible to all is here.
One of the great advantages of this exciting new dessertonomy is that we now have a neat way to describe things like chill(chocolate(heat(cream))). This formerly had to be described as “chocolate stuff like in chocolate pots”, which was at best inelegant and potentially highly confusing.
It should also be possible to define higher-order functions on desserts, for example a function which replaces all instances of fruit() with chocolate() or which remaps add_flour() to add_flour_and_coconut(). Then recipe books could be rewritten as libraries of higher-order recipe functions, with the formula for each actual dish being expressible in a couple of lines…
…and comprehensible to nobody! :-)
Comment by Bateleur — 4 June, 2007 @ 1:24 pm
This is, quite frankly, genius of the first order. *applause*
Comment by Mike — 4 June, 2007 @ 2:32 pm
Wow! I’m stunned, amazed and inspired by your invention of a culinary calculus.
I think a friend of mine made Sorbet by folding a smoothie into egg white mixture then putting it into the freezer till half frozen. Then beating the mixture so it didn’t separate into different layers, then freezing till actually frozen. Which I guess would be (freeze(fruit(egg white mixture))) and is pretty close to what you predicted would be “an odd light ice cream”, which Sorbet sort of is, isn’t it? And isn’t “relatively light ice cream” almost a Baked Alaska?
It looks like your Scientific theory has two confirmed predictions. You should apply for a research grant.
Comment by obandsoller — 4 June, 2007 @ 5:46 pm
hooray! it’s like the meta-product of a cross between the periodic table of dessert (http://www.eblong.com/zarf/periodic/) and the futurologian congress (http://burri.to/~joshua/fut.html)…
Comment by britta — 4 June, 2007 @ 11:25 pm
This is great! The distinction between the number of dessertoms and the number of operations reminds me of the difference between “nth cousins” and “cousins nth removed”.
Comment by Leonard Richardson — 5 June, 2007 @ 3:10 am
Bateleur: excellent, I’ve had a low-grade ignorance of what a chocolate pot is that I’ve never bothered to correct, and now you’ve done it for me. I’m not sure how well the notation method translates into actual recipes - take the brownies-on-cake version, where the last step is “shrink”, when in actual cooking practice you’d probably have wanted to, I don’t know, put everything in small baking tins to begin with. But surely concision is more important than comprehensibility! Think of how easy it would be to write down online recipes and bring them into the kitchen, taking mere seconds instead of minutes!
Mike and Obandsoller: Thanks. And oh, yes, I suppose so, to a Baked Alaska; I’ve never had one of those.
Britta: Ha, yes, I’d forgotten about the Periodic Table of Dessert, but it’s great; thanks.
Leonard: Someone argued that I need to distinguish between complexity of combination as well, between a mere “put on top” and a “fundamentally change the nature of”, but I think that’s more the sort of subject to be treated in an appendix than in the base terminology.
Comment by Holly — 5 June, 2007 @ 8:57 am
Just out of interest :- http://www.rampantscotland.com/recipes/blrecipe_blackbun.htm
Which I think is a cake in pastry. It’s also very nice.
Comment by Zobert — 5 June, 2007 @ 9:54 am
“Three separate desertoms” would be a terrible faux pas and take the resulting dessert, if it could still be described as one, straight through red into the invisible end of the spectrum.
An outstanding contribution to science and dessert, someone get Nobel on the phone (and keep Heston Blumenthal away from the internet while you’re at it, he’ll have this away as his own in no time flat).
Comment by Nik — 6 June, 2007 @ 4:28 pm
I am so impressed. I can’t wait to delve into this big chart more!
Comment by Kristen — 18 June, 2007 @ 1:03 pm
You are a genius! This combination of the periodic table and the dessert chapter of a recipe book is the best thing I have read on a blog all year. Thanks for making my day!
Comment by Jeanne — 21 June, 2007 @ 12:22 pm
You missed one. Your cross of “Cream” and “Heat” is desribed as “warm cream, I suppose?”. No! That’s Creme Brulee!
The cross of “cream” and “pastry” should be “cream puffs”.
Comment by Anon — 5 July, 2007 @ 5:53 pm
Nonono, Anon! Creme Brulee, despite the name, is heat applied to what is basically custard, and it’s on the chart as such.
Fair call on the cream-and-pastry, though; I’ve put applying cream to pastry as “eclairs”, which I see as pretty much the same thing as cream puffs, but I agree that there’s room for a difference of opinion.
Comment by Holly — 5 July, 2007 @ 9:31 pm
Do you also have your taxonomy in Dutch? It will make my shopping much lighter!
Comment by flox — 6 July, 2007 @ 7:53 am
[...] A Sketch Towards a Taxonomy of Meta-Desserts — in a word, amazing. [via waxy.org] [...]
Pingback by i prefer pi » Cooking for Formal Linguists — 6 July, 2007 @ 8:38 pm
[...] Raspberry Debacle » Blog Archive » A Sketch Towards a Taxonomy of Meta-Desserts [...]
Pingback by Raspberry Debacle » Blog Archive » A Sketch Towards a Taxonomy of Meta-Desserts » Wayne and Layne — 6 July, 2007 @ 11:06 pm
[...] taxonomy of meta-desserts wow (tags: chart delicious food funny humor research dessert desserts taxonomy geekiness) [...]
Pingback by kenspeckle » 67% fun, 33% scary, 100% not for those prone to seasickness — 7 July, 2007 @ 12:27 am
[...] taxonomy of meta-deserts wow. (tags: chart delicious food funny humor information math research science dessert desserts taxonomy geekiness) [...]
Pingback by kenspeckle » links for 2007-07-07 — 7 July, 2007 @ 11:19 pm
This chart /so/ must be written on a cake.
Comment by Andreas Fuchs — 8 July, 2007 @ 8:57 pm
I love this.
Frozen Custard isn’t actually ice cream, though. It’s a dessert all of its own called (Surprise!) Frozen Custard
Comment by Shannen — 8 July, 2007 @ 10:20 pm
Andreas Fuchs: If anyone wants to write it on a cake they should go ahead, but I think it’s beyond my decorating talents.
Shannen: Ah, interesting, I haven’t encountered frozen custard. I’ve made something that tasted like ice-cream by freezing custard, though; fun to find out I accidentally made something else entirely.
Comment by Holly — 8 July, 2007 @ 10:28 pm
Pastry + chocolate == ‘pain au chocolat’
Comment by Rik Hemsley — 9 July, 2007 @ 7:26 am
extremely cool! I love it! I want to frame it and put it in my kitchen!
Comment by michael — 9 July, 2007 @ 2:07 pm
[...] July 31, 2007 ~ A Sketch Towards a Taxonomy of Meta-Desserts, is an excellent example of “people with too much time on their hands”. Remember, click [...]
Pingback by Informationally Overloaded » Blog Archive » Taxonomy of Meta-Desserts — 31 July, 2007 @ 8:04 am
[...] reader, thought, and my appetite came back. Because, courtesy of Informationally Overloaded, I read A Sketch Towards a Taxonomy of Meta-Desserts.* Hurrah! At last, someone has invented a systematic way to describe pudding. My appetite came back [...]
Pingback by Symbolic Forest » Appetite — 2 August, 2007 @ 5:45 pm
Hi there,
This is absolutely brilliant. Never thought of such a good idea. This makes the field of taxonomy accessible for everybody. If you see all these nice desserts you realize that taxonomy is not a boring science at all. In the birth year of Linnaeus this is far most the most creative thought about taxonomies.
Greetz
Evert
Comment by Evert Jagerman — 12 September, 2007 @ 1:54 am