This site is archived. For my brand new live site go to Struggle Forever!

15 June 2008

Gregory Bateson

Even for an anthropology student, at least at the undergraduate level, I'm pretty familiar with most anthropological theory and the major players in the field. I'm not bragging about how smart or dedicated I am, just pointing out how pathetic my life is. There is no-one, though, who impresses me more than Gregory Bateson. His ideas and the way he viewed the world were truly unique, and, as a result of his idiosyncratic vision he was able to lay the foundation for a vast theoretical framework which has gone largely unappreciated so far. Here I would like to introduce you all to some of his ideas in a very basic way and direct you to further readings to find out more. I really believe that, if we were to use some of the analytical tools that he developed, and apply them to our own society, then we would have a better understanding of how to proceed toward a sustainable future.
I first encountered Bateson when I was a teenager. One of my high school teachers recommended his books to me on a regular basis, particularly Steps to an Ecology of Mind and Mind and Nature: A Necessary Unity. I picked up copies of both at one time and tried to read them, but back then it just seemed like overly scientific chatter - I had no idea what he was saying, and the little I did understand seemed uninteresting to me. Things have changed since that time, though. Just as a side thought, it seems to me that I am a completely different person than I was just a short while ago - so much so that I really think that I've been replaced by somebody else, and that my memories were engineered to give me the illusion of continuity. That might make a good sci-fi story at some point, but for now, on with Bateson. Recently, when I picked up Bateson again (because I had been coming across references to him fairly frequently) I found him far more readable and much more interesting than I had before. It's true that he does use a lot of scientific or simply obscure terminology, but if you're willing to do some work push through all of that then there are some significant rewards. Here are just a few of his ideas, as I interpret them, which are revolutionary to say the least.

Logical Types
Bateson didn't invent the idea of logical types - it's actually been around in western philosophy for a long time. The most prominent explication of the concept was written in the early twentieth century by Bertrand Russell and Albert North Whitehead - I forget the name of the book at the moment. Bateson borrowed the concept from them and began to apply it to all kinds of different themes.
Fundamentally speaking, the theory of logical types explains the relationship between classes and their components. For example, "apples" is a class of objects with certain general characteristics, "granny smith apples" is a narrower class of objects that falls into the category of apples, but doesn't include all kinds of apples. These classes can be further narrowed until you get to a particular apple on a particular day at a particular time and place (and you could probably go even further if you liked). Thus, we can create a hierarchy of orders of abstraction. On the first order is the particular apple with all of its particular characteristics, on the next order are "granny smiths" with various characteristics that define those (green, bitter, hard, etc.) and on the highest order of abstraction are "apples" with a different set of characteristics (roundish fruit, thin skinned, grows on trees, etc.). Each lower class includes the characteristics of the one above it, but adds a few more to narrow the scope a little more.
There a lot of applications of this concept, and Bateson used it in almost everything he did, but I'll only explain one here, for the sake of brevity - the concept of deutero-learning. Deutero-learning is essentially learning to learn or learning to deal with classes of circumstances rather than particular circumstances. For example, in one particular case we may be asked to learn the fact "the Declaration of Independence was signed on July 4th 1776," so we write it down, study it, look up more information about it, etc. and get it memorized. This set of actions can be applied to the memorization of any kind of fact not just this one, so making this a habit would be deutero-learning. This type of learning is, often by necessity, unconscious. We aren't aware of how we do it, or what exactly it is that we do, but it's there anyway. Deutero-learning is learning on a higher order of abstraction from first order learning. It is more generalizable, and so more likely to become habitual. There are even higher orders of learning, natural selection, for example, but the higher the order of learning, the more abstract it becomes and the more rigid it becomes. So it's best to keep the lower orders more flexible and the higher orders more generalized. Bateson went on to use this idea and a more complicated extension of logical typing in communication to explain the origins of schizophrenia in what became his most well known theory - the double-bind theory of schizophrenia.

When most people think of cybernetics they think of robots and computers, at least I do. However, cybernetics mainly deals with recursive or feedback systems. The best example of a cybernetic system is the thermostat. As the room cools down a switch is triggered (by the cold) which turns on a heater. As the heater warms the room the switch is then turned off (by the heat) causing the heater to shut off, and allowing the room to cool down again. The process is then repeated endlessly, keeping the room within a certain range of temperature. Cybernetic systems remain in equilibrium as long as they are not self-reinforcing. If the thermostat, instead of turning off the heater as it got hotter, continually turned it higher then you would get a system where the heater would heat up the room causing the thermostat to turn the heater up causing the room to get even hotter and the thermostat to turn the heater up even more endlessly until something breaks.
This concept of feedback helps to explain a variety of things including biological systems, ecosystems, social interactions and behavior. One way that Bateson used it was to explain how different groups interact with each other and how those relations can escalate beyond control, what he called schismogenesis. He defined two types of schismogenic relationships - complementary and symmetrical. A complementary relationship is where the actions of one person or group bring about a different action in response from a second person or group. For example if the first person exhibits pride then the other may exhibit adoration, this then causes the first person to feel more pride and the second to display more adoration until the system becomes unstable and "something breaks" - this break is the schismogenesis where one party breaks off to form their own independent group. The symmetrical relationship can also lead to schismogenesis. In this case the first person's behavior is responded to by the second person with the same type of behavior. Boasting is a good example here. Person one makes a boast causing person two to make his own boast, forcing person one to make an even grander boast which person two must then top and continuing until the two parties break off. Another good example of this is the nuclear arms race. Schismogenic relationships can relationships can attain a certain level of equilibrium in a number of different ways. One is to inject a small amount of the opposite form of relationship - so two groups engaged in a complementary relationship would inject a small amount of symmetrical interaction, which helps to neutralize the runaway effects of the complementary feedback (and vice versa). An example of this is when two groups are engaged in a dominance-submission relationship (complementary) such as that of an empire to any of its colonies then they may engage in some form of competitive game (a soccer tournament, for example) and this competition (a symmetrical interaction) will take the edge off the dominance submission for a while - kind of a cathartic or orgasmic experience to relieve the tension.
This kind of analysis could be used for relationships between individuals as well as between large groups such as nations. Gaining a better understanding of how these relationships work and how to maintain equilibrium within them could lead to more enlightened foreign policy and better psychological treatments, among other things.

Teleological Versus Stochastic Processes
Again, these concepts are not new to Bateson, they were around for a long time, but Bateson explained them in a more complete way and applied them to areas where they had never been applied before. The best way to explain the two is to look at different theories of evolution. Most people aren't aware of it, but there were actually several different theories of evolution put forward around the same time as Darwin. Darwin's theory of evolution through natural selection, however, proved to be the best explanation for the evidence. One competing view of evolution was put forward by a guy named Lamarck, and so is often referred to as Lamarckian evolution. Lamarckian evolution is teleological because it is purpose and end driven. The most often cited example is that of the Giraffe's neck. Why does the giraffe have such a long neck? Lamarck would say that long ago the giraffe had an ordinary neck, but then this particular giraffe wanted to reach some particularly tasty leaves at the top of a tree. In order to do this it had to stretch its neck, making it slightly longer in the process. This trait was then passed down to the next generation which again had to stretch its neck a little further which was passed to the next generation, continuing until the present day where the giraffe has a very long neck and can reach the leaves on the very tops of the trees. So essentially, the giraffe saw an end, the leaves on the trees and a long neck to reach them, and purposefully adapted itself to those circumstances gradually over its own life as well as over several generations until it achieved that end. There are many flaws with this theory, including the problem of inheritance of acquired characteristics, but the important thing here is that it's an example of a teleological model.
Darwin's theory of evolution through natural selection, however, is stochastic. A stochastic process is one characterized by underlying (lower order) randomness which is filtered through a (higher order) set of simple rules to create a seemingly ordered system. For Darwin, the randomness was provided by variation within a species - no two chimpanzees are exactly alike. The set of rules was natural selection - competition for scarce resources, variable reproductive success and the heritablity of traits. These two combined explain most of the diversity of life on the planet as well as the complex but ordered interaction of organisms in ecosystems.
There are several other examples of stochastic and teleological processes - one that occurs to me is capitalism, which is stochastic versus Marxism which is teleological. It's important to recognize that social and cultural change is stochastic rather than teleological. Revolutionaries would have us believe that it is the latter, that they can set a desired end and direct the culture to that end. What ends up happening, though, is a lot of violence, bloodshed or misdirection. The revolutionary force gets caught up in fighting ideas and things that it views as being counter-revolutionary, forcing people to accept a particular doctrine with which they may or may not agree, or its grand design is revealed as being inconsistent with on-the-ground realities. For this reason, those of us who are seeking a transformation in culture must seek it through stochastic processes by recognizing that we can't direct or predict the ends, we can only preserve and encourage diversity within the system (the problem with capitalism is that it tends to discourage lower order diversity, this diminishing the "gene pool" upon which it relies).

Further Reading
All of these different concepts intertwine within Bateson's work to form a larger picture of the world. As I mentioned before, understanding and using these tools that he's laid out for us will give us a much better understanding of the world in which we live. Unfortunately, little has been done in the way of applying or testing his theories - the only recognition he got was from the counter-culture movement of the sixties and seventies which saw him as a kind of scientific guru (a role he wasn't very fond of, I suspect). I hope I've sparked some interest in those of you who read this. I know it was a long post, and probably not very exciting, but maybe some part of it caught your attention and you'll want to delve deeper. Well, here is a list of some books that explain this stuff further:
Steps to an Ecology of Mind
by Gregory Bateson
Mind and Nature: A Necessary Unity
by Gregory Bateson
Angels Fear: Toward and Epistemology of the Sacred
by Gregory Bateson
by Gregory Bateson
Ecology, Meaning and Religion
by Roy Rappaport (he applies some of Batesons ideas to ecological anthropology)
A Recursive Vision: Ecological Understanding and Gregory Bateson
by Peter Harries-Jones
Here is an essay by the Norwegian anthropologist Thomas Hylland Eriksen which uses Bateson's concept of flexibility (which is based on all three of the above ideas) to analyze the "new work."
Here is a site dedicated to Gregory Bateson with some interviews, essays and such (by the way, he was married to Margaret Mead and their daughter Mary Catherine Bateson continues to develop some of his ideas).


Jérémy M. said...

Hi. I have found you blog because of (o)ac.

I also really like Bateson's thought.
I think there is one thing that you did not talk about and that is important. But I do not know exactly why it is important and I would like to think and talk about it with you.
That is a part of the theory of logical types. And I think the original idea is from Russell.
I think this idea has a name (maybe it is called a law, but I am not sure).
The idea is linked to the idea of a hierarchy of logical types, and it says that (to take an example) :

The class of chairs, which is not a chair, does not belong to the class of non-chairs.
Because (if I remember well) the class of chairs and the class of non-chairs are of a different logical type than the chairs.
I think that the more general formulation is that a class can not be a member of oneself.

This idea seems to play a great role in Bateson's theory (notably about communication). But I still quite not understand why and how. And I am not quite sure I have it right.
What do you think ?

Jeremy Trombley said...

Hey Jérémy - I like your name! :)
Glad you found my blog, how do you like the OAC so far?
I do have a section in the above article on Logical Types, though I may not have explained it very well. I agree, though, that the concept is important, even fundamental to much of Bateson's work.
The idea of logical types has had many iterations since Ancient Greek philosophers, however, it was Neil Whitehead and Bertrand Russell who put together the theory of logical types that Bateson uses. I'm not sure if mathematicians still use their theory, but Bateson has show its value outside of math.
The basic formulation is that no class can be a member of itself. So, an apple is a member of the general class "apples" but the general class "apples" is not an apple - if that makes any sense. The problem is that confusing logical types leads to all kinds of logical fallacies. For example (Russell and Whitehead use this example, but Bateson reiterates it in Mind and Nature p. 117), the paradox of Epimenides - "Epimenides wasa a cretan who said, 'Cretans always lie.'" The problem is that the phrase "cretans always lie" is of a higher order of abstraction than the original statement, and therefore it cannot include itself. The paradox, as a result, is a logical fallacy.
As I said, Bateson used logical types in a lot of different ways - in fact, he even states that it is the basis for his entire epistemology. In communication this lead him to the idea of metacommunication. In any act of communication a message is sent, however, in order for that message to be interpreted properly, there must be some context. In general, this is provided by another form of message, which is of a higher order - a metamessage. So, when a dog goes up to another dog and growls, in order for the second dog not to attack the first and get into a real fight, the first dog must communicate that its growl is "play." It can do this using posture or tone, but this message categorizes the growl as "play" not "fight." This, then, lead to Bateson's double bind, in which the two messages are contradictory. That is, the growl is sent as "play," but the dog wants to fight. His theory of schizophrenia says that, if a person is repeatedly exposed to such mixed messages (i.e. by a schizotypic parent), then the individual will be constantly confused about the context of messages from others and will not be able to communicate effectively or tell the difference between "play" and "fighting" or "fantasy" and "reality."
Furthermore, logical types could provide the interface between biological genetics and learning. Genetics, Bateson says, provide a context for learning just like metamessages. The difference is that genetic learning is of a higher logical type than learning and deutero-learning, and is, therefore, more rigid.
It's an interesting topic. I'd be happy to discuss it more with you.

Jérémy M. said...

Your name is not too ugly either. And I am happy you answered that fast.

If you're ok, we'll talk about (not so O)AC a bit later.

I agree with what you said and thank you for refreshing my memory. I'd love to discuss and explore with you the theory(ies) of Bateson more broadly afterward, but for now I'd like to focus on a precise logical point that disturbs me.

That is, you wrote :

"The basic formulation is that no class can be a member of itself. So, an apple is a member of the general class "apples" but the general class "apples" is not an apple - if that makes any sense."

It makes perfect sense to me. But I think there is something else about that. It is not just that the class of apple is not an apple. What seems a bit less intuitive is the other statement that Bateson has made :

--the class of apples is not a member of the class of non-apples--
(and that, despite the fact that it [=class of apples] is obviously not an apple).

(and that is precisely a discovery of Russell I think, of which I may talk later if i can find my notes)

I am sorry to insist on that point, but do you see what I mean ? I perceive that there is something at stake here, involving Hierarchy in a strong sense.
Without an idea of hierarchy, there is a contradiction here.
But I do not remember what precise function that point exactly plays in Bateson' theory. (maybe none, then we'll forget it). I'll try to find my notes.
What do you think ?

Jérémy M. said...

BTW, thanks for the link to gene poll version 6, i like to play the role of mother nature.

Jeremy Trombley said...

Yeah, I think the (O)AC is undergoing some radical changes right now, and I hope we move beyond the limitations that have been imposed on it.
I see what you're saying about the logical types. Clearly, "apples" and "non-apples" are mutually exclusive categories. I haven't read Russell's original formulation, and I'm not sure how it plays into Bateson's work - I'd be interested to hear more.

My understanding was that the hierarchy comes about with different levels of categorization. For example, there are "granny smith apples" and "macoun apples" and "red delicious apples" which are all sub-categories (lower logical types) of the broader category of "apples." Then there is "this particular apple right here" which is a representative of one of those (or some other) sub-category.
The fallacy comes about when a category attempts to classify itself. In the Epimenides paradox, the phrase "all cretans lie" is a meta-message which attempts to categorize the statement "Epimenides was a cretan who said that 'all cretans lie.'" So it attempts to categorize itself, which doesn't work - you end up with the possibility that the phrase "all cretans lie" is both true and false at the same time.

There is one other thing that I'm learning from my research on complexity science, which is emergence. Classes of objects can have qualities that individuals don't. For example, temperature. A single particle has no temperature because temperature is a statistical measurement of a group of particles. As a result, temperature is an emergent property of groups of particles, which doesn't exist in an individual. I think that's interesting, and important as well.

Let me know if you find out more about the property of classes that you mentioned.

Jérémy M. said...

I'll enjoy to talk about all this more at length. But I have to finish some job in this week. For now I'll share my notes on Russell with you.

It is in a book called (I found it in french, hence i'll translate the french title) : "History of my philosophical ideas".

I translate Russell as well as I can. It talks about the origin of his theory of logical types.

"[at the beginning] it seemed to me that a class, sometimes is, and sometimes is not, a member of itself (...) for example : the class of things that are not spoons is one of the thing that are not a spoon.(...)

It drove Russell to consider : "the classes that are not a member of themselves".

And then Russell considered that "the classes that are not a member of themselves" must form a class.

And he asked himself "whether this class is a member of itself or not."

So :
1- If this class is a member of itself, she must possess the defining property of the class, which is not to be a member of itself.

2-If she is not a member of itself, she has to not have the defining property of the class, and so, not be a member of itself.

So, each alternative lead to its contrary. There is a contradiction.

Here he adds that FREGE was so troubled by this contradiction that he abandonned his tentative to deduce arithmetic from the logic, to which he add consacred his whole life. (!!!!!!)


Hence Russell had to introduce his theory of logical types to resolve this contradiction.

Then he goes back to the example of the lying Cretan.

Here my notes are not good enough. I'll try do have it straighter next time I go the library.

It is something like this :

According to Cantor, for a class with N members, there is 2^N sub-classes.

Russell then says that if one applies this proposition to all the things that exist in the universe, one arrives to the conclusion that "Classes of things are more numerous than things",
So : classes are not things

..... well.... I ll have to go back to this book again, because I still do not have the solution of the contradiction noted earlier.

BUT, to go back to what you said, some pages further, Russell says :

"It will appear that every logical paradox presents a kind of reflexive reference to itself that must be condemned for the same reason [??] : that is, that she includes, as a member of a totality, something that refers to this totality that can only makes sense if that totality IS ALREADY FIXED."

Then he talks about the difference between a NOUN [or name ?] and a DESCRIPTION :

"A noun can have a sense [=have the property to be right or wrong] only if THERE IS SOMETHING that is named, while a description is not bounded by this limitation".


Sorry for the poor quality of my notes, but I think there is really good food for thought in those passages.


Maybe we should go back to Bateson's theory now. And I'll try to have it more precisely later.



What are the radical changes that are taking place at (o)AC ?


I'd really love to discuss bateson's theory in the coming days/weeks/months. :)

Jérémy M. said...

Sorry, my thought were really vague. And I also apologize for such a sentence as "I have a lot of work, so blablabala" which is very ugly.

All this can be summarized. Russell has proven that no class is a member of itself, and the correctness of the theory of logical types.

And the contradiction about : "the class of apples is not a member of the class of non-apples" can be resolved through the theory.

I was not precise enough.

The class of things that are not apples is not a thing (it is a class), so it is NOT a member of the class of things that are not apples.

Pfffff. Sorry about that.

Then here we go with Bateson.

I really like what he says about "play" and "eating the flag" (or confusing the map and the territory, confusion in which I've just got stuck).

But how can we construct an anthropological inquiry with "his" thought ? What are the social facts than one can search in which Bateson's thought will be particularly relevant ?

How do you understand the notion of "mind" ?

Jeremy Trombley said...

I looked up some about Russell's theory, but it's based in math and math is not my best subject. Needless to say, I had a hard time understanding it all. I did find that it's not all that important in math anymore thanks to Godel and others. However, the way Bateson describes it is more understandable to me, and more relevant, I think, to anthropology and other (less abstract) fields.
What I've found in Bateson is the basis for a new epistemology, which, I think could provide better analysis of social, cultural and biological factors. In his concept of deutero-learning he provides a link between culture and biology. The stochastic processes he describes offers a possible explanation for cultural change. The concept of feedback systems sheds new light on how to study cultures. And the idea of flexibility will be useful for studying relationships between culture and environment.
I'll try to expand those ideas a little further tomorrow. I'd suggest you look into complexity theory - it draws on some of the same ideas that Bateson was using, but goes into a little more depth. I'm going to do a post about it pretty soon (as soon as I finish reading my book on it), so I hope you'll join me for some discussion.
How do you see Bateson's relevance to anthropology?

Jérémy M. said...

All I've written about Russell earlier was quite irrelevant. (though it helped me see the logical type issue more clearly).

I'll definitely join you for discussion.


Related Posts with Thumbnails