Who has theorized that you can understand a system by understanding just one part of that system?

February 8, 2005 11:40 AM

Intellectual History / History of Science Question: Examples of thinkers, scientists, artists, critics, etc. who theorized that you can understand a system by examining either a single component or a small set of components of that system. [More Inside]

The exact parameters (the system, components, and the rules about each) of an example's specific use of this strategy don't have to be rigidly defined. Right now I want to think broadly about this theme. Ideally, if there were a book or resource written from a comparative perspective about this way of thinking/interpretive strategy that would be fabulous as I'm interested in learning about the disciplinary differences and challenges in the application of this idea. I'm less concerned with whether or not the example made a positive and lasting contribution to the particular field of knowledge as much as I'm interested in how its adherents pursued the idea and to what ends it came. I don't know if my ideal resource exists, so I'm looking to study disparate examples and make comparisons on my own. Any guidance you can give would be appreciated. I have some leads, but would encourage you to presume I know next to nothing about the topic (that as you can see I don't even have a proper name for). FWIW, this is not an assigned task (i.e. school-work).

posted by safetyfork to religion & philosophy (17 comments total)

I don't know if this applies to your area of study, but there is a whole school of thought about understanding one animal's simple behavior and how that can give complex herd behavior when you have many animals. Check out the Boids web page for a better description, and a cool demo.
posted by Triplanetary at 12:05 PM on February 8, 2005

A Pattern Language does just this, but with architecture and how it affects our lives.

Douglas Hofstadter also discusses Holism and Reductionism in Gödel, Escher, Bach.
posted by driveler at 12:10 PM on February 8, 2005

Also, dont know if this helps, but all of modern physics operates on the principle that what is applicable to a subset is applicable to the whole.

At the largest scale, this is applicable to the universe at large. We know what the properties of iron are here and now on Earth. But we dont know how iron behaves on a planet orbiting Tau Ceti. We just assume it behaves the same. Likewise what were the physics of the early universe? Well, we assume they were the same as the physics we have now.
Certainly there have been a few thinkers who have gone *against* this and argued that the laws themselves change wrt to time and/or space but its those guys that are the outliers.
posted by vacapinta at 12:16 PM on February 8, 2005

I haven't read it, but Stephen Wolfram's A New Kind of Science uses simple cellular automata to model complex phenomena.
posted by driveler at 12:19 PM on February 8, 2005

These all look helpful and are things I didn't know about (Gödel, Escher, Bach has been on my must read list for way too long). Thank you. Most of my initial leads (all three of them) are derived from my background in the humanities and social sciences. Also, I'm casting a wide net and including discredited things like Phrenology, which as I understand it would fit into this scheme. Keep 'em coming!
posted by safetyfork at 12:29 PM on February 8, 2005

Intellectuals do this a lot. They focus on one small detail, an ad, a phrase of speech or just about any piece of unseen everyday trivia and proceed to draw conclusions about our society as a whole. I believe Roland Barthes is a good example of this kind of writing – I'm not saying he was the first, but I think he inspired a lot of followers. I like the question and will follow this discussion.
posted by Termite at 12:31 PM on February 8, 2005

Descartes thought you could get [knowledge of] everything by deducing it from a single self-evident premise. Not that most people think of him as succeeding, but he was thought of as a scientist in his day.

Blake -- eternity in a grain of sand.

If phrenology is a good example of what you're looking for, then you might find a treasure trove of information in scatology, which often attempts to make large scale generalizations about animal behavior, even whole ecosystems, by studying excrement.
posted by ontic at 1:42 PM on February 8, 2005

Newton's law of universal gravitation is an example of replacing a very complicated helicentric (Copernican) model, which essentially was a giant kludge to get the model to fit existing data, with a much simpler predictive method.

But physics shouldn't be confused with science in general - see discussions about reductionism. For example, understanding the detailed chemistry of a cell (we don't yet) isn't particularly useful for raising a socially-healthy child, for example, even though the child consists entirely of cells.
posted by WestCoaster at 1:45 PM on February 8, 2005

Thanks, ontic and WestCoaster.
On the phreno-front: I'd say that Phrenology is an example of what I'm willing to look into. I have reservations about calling it a good example though it might be well thought of as an example of the problem of premises (to bring it back to Descartes).
posted by safetyfork at 2:03 PM on February 8, 2005

I suppose the most obvious example of this tendency, at least in the humanities, would be the school of criticism known as 'new historicism'. Famously, new historicists like to start with an anecdote -- the more bizarre and outlandish, the better -- and use this as a point of entry into an entire culture or system of values.

One of the models for this school of criticism is Erich Auerbach's book Mimesis (1946), which takes a series of passages from Western literature and uses them to illustrate general themes in Western culture. Another is Clifford Geertz's book The Interpretation of Cultures (1973), in which he sets out his theory of 'thick description' which aims 'to draw large conclusions from small but very densely textured facts'.

The bastard child of Geertzian 'thick description' is the sort of popular history found in books with titles like The Paperclip: A Short History which try to persuade you that the entire history of modern culture can be deduced from the history of the humble paperclip. That recent book on the cultural history of air-conditioning comes to mind as a good example of the genre.

I always think of this as the Jurassic Park school of cultural history, as it reminds me of the scientists who reconstruct entire dinosaurs from tiny particles of DNA. In the same way, it seems to me, new historicist critics like Stephen Greenblatt, or cultural historians like Robert Darnton, are reconstructing entire cultures from tiny scraps of historical evidence. Sometimes this is done brilliantly. But the 'culture as DNA' metaphor strikes me as a massive unexamined assumption behind a lot of new historicist writing.

My gut feeling is that this synecdochic way of thinking (i.e. using a part to study the whole) probably doesn't go back much further than the nineteenth century. The idea of statistical sampling, as formulated by people like Adolphe Quetelet in the early nineteenth century, must surely have been an essential precondition. Without some awareness of the principle of statistical sampling, it is hard to see how people could have got their minds around the notion of the-part-as-typical-of-the-whole. Certainly it doesn't seem to bear much resemblance to the theories of knowledge I am familiar with in the seventeenth century, which tend to be more systematic in nature -- i.e. concerned with uncovering the underlying structure of a system, rather than focusing on a particular part of the system and considering it in isolation.

I'm afraid these random thoughts probably won't be much help to you. But it is a topic of great importance, and one that has long fascinated me. As far as I am aware, there is no general/interdisciplinary book on this subject, which is a great pity. Who knows, perhaps you will be the one to write it?
posted by verstegan at 3:47 PM on February 8, 2005

Try some searches on "agent-based modelling".
posted by duck at 4:14 PM on February 8, 2005

You may want to look into Francis Galton's early studies of heredity. He studied a group of talanted people and their families (a considerably small sample to make broad claims) to come up with his "Law of Ancestral Inheritance." Most of his heredity studies were flawed (eugenics!) but this is the one of the few times that I can remember him using a small sample.
posted by sophie at 4:53 PM on February 8, 2005

Astrology, or sympathetic magic in general?

You should probably look for connections between alchemy and the history of the scientific method. What you're looking for is pretty similar to the whole Emerald Tablet microcosm/macrocosm thing: "whatever is below is like that which is above, and whatever is above is like that which is below."

Perhaps Haeckel's Lie, that "ontogeny recapitulates phylogeny" ?

Wikipedia has a nice little section on induction, and another on the problem of induction which might lead you to some other useful names.
posted by jbrjake at 7:08 PM on February 8, 2005

Some very simple mathematical models can be pretty powerful. You might want to try the Lotka Volterra predator prey model. There isn't much good material online though.

Hodgkin and Huxley essentially demystified the neuron with their experiments and model.
posted by euphorb at 9:52 PM on February 8, 2005

Goethe. All is leaf. Get his scientific writings.
posted by kenko at 7:55 AM on February 9, 2005

i hope i'm not too late here. i just came across laughlin's nobel prize acceptance speech and he makes some interesting points about physics and reductionism. you might also like this article.
posted by andrew cooke at 6:07 PM on February 9, 2005

Thanks everyone, these answers are truly great.

For anyone who comes across this later and finds the thread of interest, here are the leads that I started with:

This really came about, like others above have mentioned, through seeing the process repeated in my studies, both by professors who would take the whole seminar session on one passage of a text to explain the whole, and by engaging with the works of people like Barthes. Recently, I was simply struck by how well it was done in chapter 2 of Running with the Devil (a musicological/sociological study of Heavy Metal). For some reason this book really threw into relief this act as a rhetorical strategy, but it also seems like that strategy mimics a fundamental and yet often, or at least potentially, flawed way of thinking. In thinking about the thinking, that's when the real curiosity started to kick in. So I started poking around....

First, the Atomists: Leucippus and Democritus. They developed a fully mechanistic view of nature in which every material phenomenon is seen a product of the atom collisions.

Then, a big ol' leap forward to Wittgenstein and Russell's Logical Atomism. I had read some Wittgenstein, but no Russell before -- this I glossed and probably need to go back to.

A left turn, ha!, to Susan Sontag On Style: The text repeats itself. I remembered I got into Sontag in school because of Barthes. But did not think to go back to him, thanks Termite. For some reason, Paul De Man was creeping in here too, but I wasn't turning up anything about this that I thought I remembered being attributed to him (or Derrida by proxy). This phase was really just thinking of examples of this strategy. I didn't find anyone who was outlining a program based on it. Just a tendency to attribute interpretive weight to parts of the whole.

Before I asked Mefi, I wound up somewhere with Eric Lormand's work on meaning: particularly, How to be a Meaning Atomist. Interesting.

It's right about here that I started to think more deeply about this mode of thinking and started wondering about the differences in the rigor of its application and the differences in function between the Humanities and the Sciences. It was also apparent that not all of my examples were of the same thing exactly but there was a general contour of thought which held my interest. That's when I decided I needed more examples, and more examples I got! Thanks again, these will keep me busy for a while.
posted by safetyfork at 8:10 AM on February 11, 2005