I see a parallel between Adorno’s approach and certain methodologies in personal knowledge management (PKM), so I decided to play around with the Obsidian application to represent constellations.
Constellations as Graphs
In graph theory, a graph is a non-hierarchical and interconnected system of nodes, whereas a tree (or taxonomy) is a hierarchical parent-child structure.
Science, and classificatory knowledge in general, uses trees, the prime example being a phylogenetic tree of life:
In contrast, constellations are a lot like graphs. So I created a vault in Obsidian.
Obsidian is a note-taking and knowledge management app, so each node in its graphs is a note. Each is connected with others through links, and what you get is a graph structure representing connected objects (in red) and concepts (in yellow).
Each note opens with a short characterization but doesn’t complete it as you’d expect from a definition. The rest of the note doesn’t contain more specification, rather it lists the things and concepts it stands in relation to, and the note is unusable on its own. To put it like Adorno, what is inside the note is what is outside, i.e., its relations. The note is not defining the thing’s essence “from within”.
Each of the following screenshots shows the note for the selected node, and the local graph view alongside it. As successive nodes are selected—representing the shift of the cognizer’s attention—the graph view changes. To begin with, The Factory is at the centre, but the centre changes.
The screenshots are in order. This doesn’t represent a progression up or down levels, but a walk over stepping-stones with no route mapped out in advance.
1. The Factory (object)
2. The Clock (object)
3. Abstract Labour Time (concept)
4. Exchange Principle (concept)
NOTE: What the graph doesn’t show is that, as discussed before, The Factory, The Clock etc., as well as being objects, are also concepts and can be referred to as such when that’s the focus, as here:
The cognition of the object in its constellation is that of the process, which it has stored up within itself. As a constellation the theoretical thought circles around the concept, which it would like to open, hoping, that it springs ajar like the lock of a heavily guarded safe: only not by means of a single key or a single number, but by a number-combination.
Constellations as Trees
Representing these objects and concepts in a tree would look like something like this:
Notice the directory structure on the left, and that the note now contains a full definition. The object is specified from within, and its hierarchy, i.e., what’s above it, supplies this specification. This is contrary to constellations in that it’s a closed specification: the taxonomic path and the definition agree, and once they do, the object has been fully accounted for. In contrast, a constellation never closes: each concept gets part of the object and misses the rest, which is why you need several.
Also, in the tree, two individual things are related only if some category further up the tree already contains both, and the categories are fixed in advance. The Clock is filed under Technology and Boredom under Psychology, so there is no way to say that measured time and empty time are the same fact (boredom is the form experience takes when time has been made countable and then emptied—free time as merely the negative of work time).
All of this is the subsumption characteristic of identity-thinking, treating a thing as an instance of its kind.
Layers
Maybe what you’re getting at is something like this:
Previously I had the Depth set to 1, so it showed the object’s adjacent neighbours, forming a constellation. Increasing the Depth to 2 causes their neighbours to be added (and so on). That expansion outwards could be seen as going through layers—a temporarily fixed centre wrapped in context ordered by proximity to the centre.
Caveats
The graph structure has limitations as a representation of constellations. Some things a graph structure cannot show:
- The non-identical: It can’t be represented. You can’t represent it by adding more links—this just makes the graph structure bigger. The object is supposed to exceed what you say about it.
- Antagonism: Every link is the same. A graph shows that two things are connected, but not that they are in conflict (or contradict).
- History: Links are all present tense. They record that a connection exists, but not how, when, or through what it came about.
- Construction: The links are already there when you open the vault, so it looks found rather than made. A constellation is both: assembled by someone, for this object. It could have been assembled differently.
It’s also worth noting that this tool (the linked-note vault) ultimately goes back to the Zettelkasten index card note-taking system, significantly used and developed by Niklas Luhmann, and he was a systems theorist, definitely not a dialectician. The notes are meant to form an autonomous system, with structure emerging, producing personal productivity from unforeseen connections. Its notes link to each other, not to an object outside them. In other words, its structure is not meant to model that of reality.
EDIT: But I notice from the Zettelkasten wikipedia page that Walter Benjamin, perhaps Adorno’s greatest influence and the originator of the method of constellation, used Zettelkasten for his Arcades Project.
Anyway, I find it’s a very useful conceptualization tool.