Case Study
On Food Ontologies
This piece came out of one of our “ Walkthroughs,” during which four people to take us through a particular channel, the blocks and ideas held within, and the ways those ideas may have evolved as the channel has grown and accumulated. Our latest one was on October 22 over Zoom, and it featured Leslie Liu, Maya ManMaria Gerdyman, and Agnes Cameron. Here, Agnes shares some of what she talked about while walking us through her “data / food ontologies” channel.
This is a channel I made while thinking about different ways to organize and understand food. It inherits a bit from an earlier channel I made, which looked at different ways to format recipes (a design exercise I wrote for my web programming students), and also from making a recipe website a few years ago which is built around a joke on taxonomies. What goes in here is a mix of things that pretty much call themselves ontologies (e.g. maps, charts) but also increasingly includes things like lists, guides, and narrative/historical accounts that (to me at least) outline a system or set of systems for thinking about food or cooking.
Food ontologies have the nice feature of spanning the subjective (e.g. ranking of different brands of crisps), to scientific analyses of chemical structure, to systems of cooking and eating defined by agriculture, geography, and tradition.
Patrick Gunkel is kind of an amazing character, and this was how I first encountered his work—a chart titled with an assertion that his method of ideonomy (a kind of diagrammatic idea-expansion) could be used to envision more ‘optimal’ foods through technological intervention. Gunkel starts by developing possibilities based on a ‘hint’—in this case, the sausage-form, a filling contained within a taught skin—into a branching diagram of linked, possible forms.
On the face of it, this seems fairly over-the-top, a bit ridiculous even, but there's something about the thoroughness with which he explores the space he sets up for himself that I find very compelling. This diagram is also a few decades old by now, but the ideas he describes around layered/nesting/spherical heterogenous textures, temperatures, and flavors also now characterize a lot of foods that currently have a meme-y popularity—cheese-stuffed things, popping boba, mochi ice cream (perhaps because taste and smell are very hard to communicate online, but texture is not?). 
There’s something very appealing to me about technical manuals, and I think this one in particular has the quality of existing on both the extremely domestic and the heavily industrial simultaneously. I love the tenderness of descriptions of a complex set of chemical processes required to produce a food that creates a particular sensation on the lips, or sits satisfyingly on a spoon, or is suitable for someone elderly or very young to eat. For weeks after reading this book, I started to notice and appreciate the texture of industrially-produced foods like bread and mayonnaise a lot more.
I’ve also always had a soft spot for materials science, so the use of stress-strain curves and phase diagrams to organize foods was also really interesting. I had a hard time picking which of the two volumes was my favorite, but settled on volume 1 for its commentary on jelly.
This article by Suvajit Halder gives a long and detailed history of Bengali cuisine, as affected by migration, colonization, trade, technology, agriculture, religion and medicine, through texts dating back to the 11th Century. 
Different sets of ingredients and techniques come in and out of discussion, but so do systems of talking and thinking about food, and many coexist simultaneously. There’s a part I really like which talks about how the enthusiasm for food of 13th-century gourmands has left a real wealth of historical texts discussing both prescriptions for the sequences of different dishes, and dishes and recipes themselves. I don’t think of this piece so much as an ‘ontology’ in and of itself, but a history that links together webs of intersecting ideas about food.
This piece is also a reminder that taxonomies of food by ‘national cuisine’ are slippery: not only is Bengali cuisine already split by a national border, but events such as the introduction (and subsequent promotion by the colonial British administration to drive profits) of ingredients like potatoes and chillies, or the slow adoption of Islamic cooking techniques during the 12th-18th centuries are all influences on ‘traditional’ recipes.
I remember having a conversation with my friend Dan about how moving around different flavors is a bit like moving through a 3D space, with different areas that coalesce into a family of tastes, or a particular cuisine or culinary convention, but wide areas in between that feel much less well-defined and experimental. This diagram reminds me of that spatial feeling in foods, and was also how I learned the word ‘caprylic’—the flavor of goats (gratifyingly, in the ‘erogenic’ section of the diagram). 
This is also one of the only blocks in this channel that’s linked from somewhere else on, rather than something I came across in the world—perhaps as it’s also visually quite beautiful. 
I’m interested in the process that produces ‘big picture’ Wikipedia pages—pages that try to address a subject with a vast amount of complexity and divergence in very general terms—like ‘human,’ or ‘happiness.’ Take the caption on the main image, for instance:
‘An example of Central European cuisine, the Wiener Schnitzel. It is prepared with regional ingredients and according to the local cooking style.’
Like, nobody can argue with that, I guess? These kinds of pages seem to exist on a continuum from extremely poor quality (too ill-defined for anyone to say or care very much) to incredibly slowly and self-consciously refined (still not saying very much, but with a lot of discussion about what it is that should be said). This page, ‘cuisine,’ tends toward the latter end, with edits dating back to 2001; clicking through the edit history moves you from slowly growing lists of different sub-cuisines, to a more general narrative form with particular dishes given as examples, perhaps in acceptance of the trouble of trying to give an adequate amount of detail about something so vast. 
Agnes Cameron is a hardware and software developer based in London. She has a particular interest in simulation, distributed knowledge, and infrastructural systems. Currently, she works with the Knowledge Futures Group, building tools for working with datasets. She is a founding member of the collective Foreign Objects, and a resident at Somerset House Studios. Blog
Learn about how people use to do work and pursue personal projects through case studies, interviews, and highlights.
See More