I came across the game of life in Darwin’s Dangerous Idea and decided to research it a little further. Designed by John Horton Conway, the game of life is interesting because it is a universe based on two very simple rules that generates surprising organized complexity. Here’s a short video illustrating these properties:
All of this reminds me of a discussion Dominic and I had a year or so ago in another context. In that context, Dominic was objecting to my use of the term emergence. My memory of the actual discussion is a bit fuzzy now, but if I recall correctly Dominic’s concern was with those ideas of emergence that posit some sort of gap between simple rules or principles governing a system and properties of the emergent system. This sort of emergence is akin to magic, in the sense that it asserts the irreducibility of the system to its rules or principles. Emergent properties thus become something “ontologically spooky”.
The game of life perhaps provides the possibility of a more rigorous conception of emergence. The properties that emerge within the system are all made possible based on the rules, constraints, or principles governing the system but are not strictly predictable within these constraints. As Dennett articulates this sort of emergence, “…the properties [are]… unpredictable in principle from the mere analysis of the micro-properties of the [system]” (415). There is nothing magical here, as these properties are products of the constraints underlying the system. Perhaps a clearer way of putting the issue is that the results of the system are not themselves programmed, but rather are rendered possible by the constraints on the system. This seems to be similar to what biologist Stuart Kauffman is getting at in books like The Origins of Order: Self-Organization and Selection in Evolution. What he draws attention to are certain physical constraints in matter that give rise to persistent organization under specific conditions. It is unfortunate that he’s chosen to dress up this thesis in the trappings of a sort of spiritualism in books like At Home in the Universe. Ah well, no accounting for taste.
Larval Subjects . 
August 3, 2009 at 6:06 am
In the past, while practicing math, I used to play with such comps http://video.google.com/videoplay?docid=8355191184253806687 but now being at home (philosophy) I seem to have moved to other attractors.
August 3, 2009 at 2:11 pm
It’s much better to try out Conway’s game of life oneself! Here’s an online version. Also, it’s worth noting that the original idea was implemented on a Go board, manually. An arduous process indeed.
There’s lots to say about emergence (I talk about it at length in Unit Operations), but the one I’ll throw out now: in most discussions of emergence, the typical comparison is not with magic, but with the sublime.
August 3, 2009 at 4:16 pm
I think emergence is a little spooky to us humans, ontologically. What we usually want when describing a system is some sort of short-hand – either a mathematical equation or a set of formal rules. For emergent systems it’s nearly impossible to predict the rules from the system or the system from the rules, even when the rules are dead simple.
The really spooky part, though, is that emergent behavior/properties appear at one level of observation and are seemingly absent at another.
August 3, 2009 at 8:20 pm
[…] a recent post Levi mentioned the Game of Life, which I’d not seen before. I found an online version of it. […]
August 3, 2009 at 10:07 pm
[…] Fiction, Ktismata, Psychology — john doyle @ 3:07 pm This morning Larval Subjects put up a post about emergence, using for illustrative purposes a deceptively simple video game that Dennett discusses in his book […]
August 3, 2009 at 10:21 pm
Levi, in your object-oriented ontology how would you categorize the emergent gliders, guns, rakes, ships and so on? Are they properties of the game? Separate objects spawned by the game? Objects that exist only inside the ontology of the game itself? Artifacts of physical properties of the game interacting with human perceptual/cognitive systems?
August 4, 2009 at 12:15 am
Hi John,
This is a good question because it meshes well with what I mean by a “flat ontology”. Within the framework of OOO if it makes a difference then it is. As such, the various critters that emerge within the game are real objects. Clearly the objects that exist in the game– gliders, guns, rakes, etc., –are dependent on the “physics” of the game (the two basic rules that govern the game), but this dependence doesn’t undermine their status as objects. As a human being I am dependent on all sorts of constraints as well, but am no less an object for all that. Back in December when I first began to develop my particular ontology, I argued that every object is attached to a field (https://larvalsubjects.wordpress.com/2008/12/21/objectile-and-agere/ ). The “physics” of the game of life would be the field in which these objects are produced and organized. I want to emphasize, however, in light of your post that I am not making the claim that systems are strictly deterministic, only that we can get unexpected organized patterns out of simple constraints.
Returning to the discussion of memes that you analyzed so closely over at Dead Voles, one of the reasons I find the ideas so attractive is precisely that meme theory treats signs as objects. Rather than treating signs as mere representations of something else, meme theory treats signs themselves as objective reality. So unlike common views of language where you have one thing, the world of objects, and another things, the world of signs representing objects, in meme theory you have one flat plane where there are physical objects and signs as well. For the memeticist, signs are objects in two respects: first, they must be embodied in some medium whether it be brains, sound-waves, paper, arches in a building, strings of zeros and ones, etc. Second, signs must replicate themselves in order to carry on. Just as you can’t run your television in a region of the world where no electricity is available, certain forms of social organization are not possible without memes having been replicated in that region of the world. The manner and medium through which memes are replicated will make a big difference on the sorts of replication that are possible. Thus in an oral culture preceding the invention of writing, certain meme-patterns will be more suited to replication than others becomes memes must contend with the limitations of human memory in order to replicate themselves orally. Thus, for example, if we find a predominance of rhythm and narrative among the epics and myths of oral culture, then this is because human minds have an easier time memorizing and repeating rhythmic narratives. By contrast, it would be very difficult to memorize and orally recite Hegel’s Science of Logic or Deleuze’s Difference and Repetition. For this signs must be embodied in a different type of medium: writing. What we get from a memes-eye perspective, then, are different fitness-landscapes that play a structuring role in the sorts of signs that get replicated. This, I believe, is what McLuhan was investigating with his notion of “media”.
When signs are themselves treated as objects, the nature of our questions about signs change somewhat. Rather than asking how a sign represents some state-of-affairs, we instead investigate the manner in which one type of object gets related to or linked to other objects in an assemblage. Perhaps more on this later. None of this is to say that I accept Dennett’s or Dawkin’s account of memes. I just think it’s a very promising idea in need of a lot of theoretical development.
August 4, 2009 at 1:44 am
[…] Ontic, Ontology, Organization, Possibility, Realism Leave a Comment Responding to my post on the Game of Life and Emergence, John Doyle, over at Ktismatics, speculates about the ontological […]
August 4, 2009 at 2:45 am
Thanks for the clarification: it’s consistent with the way you’ve presented these ideas heretofore. It’s interesting to think of an object’s properties, emergent or otherwise, cutting themselves loose from their moorings and launching out as objects in their own right. So, e.g., the timbre of Edith Piaf’s voice can be decoupled from the person and the voice to become its own object. Also presumably the reaction these objects generate in other objects are themselves objects; e.g., my cat’s reaction to the Piaf vocal timbre, even if I don’t know what that reaction is, counts as an object separate from Piaf, her voice, and the cat. It seems that in OOO every difference that makes a difference isn’t just real, it’s also an object.
A sign as an object makes sense: it has identifiable content and structure in its own right. The sign’s function as a signifier, symbolically pointing to some other object, is a packet of information encoded in the sign, available only to other objects that can decode the information; e.g., people and machines that understand the language. I guess in your OOO the signifier isn’t just a property of the sign; it’s an object separable from the sign. So too with the information encoded in the Game of Life: to beings like us who are able to attribute not just movement but intentional movement (eating, shooting, flying) to the higher-order patterns of iterative game states, the patterns represent eaters, guns, and spaceships. For OOO the patterns and the critters that the patterns represent are separable objects. Am I following you here?
“that you analyzed so closely over at Dead Voles”
lol.
August 4, 2009 at 7:26 pm
Levi, your answer @7 crystalizes my fascination with your project very nicely. I agree about the value of the meme concept, although I think it’s worth looking more deeply into why the meme meme failed in Anthropology before getting too excited about it, since those guys are tasked with figuring out how cultural systems work.
Anyway I think I’ve got my themes for intro world history now. I’ve been working with autonomy/heteronomy, individual/community, then explicating with ideas of freedom and constraint, but I think I can get a wider throw with constraint/enablement (or affordance).
The key in teaching this stuff I find is to convey that emergence is just as much the product of constraint as enablement, since it’s the dialectic of the two that create the possibility spaces (or fitness landscapes) within which real outcomes occur. As Marx asked in the Grundrisse, is Achilles possible with powder and ball? More topically, would Michael Jordan be His Airness without gravity? It’s Kant’s swallow again.
John, I’d be sorry to lose the whimsical conceptual kludge known as the ‘floating signifier’, but otherwise yes, I gather the sign and the signifier must be their own objects. Or nodes of relation, depending on whether we’re doing Bryant or Latour.
However, I’m not sure I’m on board with this:
Well, other than getting to call things ‘objects’ rather than calling things things, what’s the advantage here? I see that we clean out the mediating discourse of ‘representation’, but if the ‘signifier’ kind of object doesn’t occur without the ‘sign’ kind, and neither occurs without the ‘signified’ kind, isn’t there an important and realistic claim about the nature of those objectivities embedded in the idea of representation that is simply obscured by flattening the ontology?
August 4, 2009 at 7:28 pm
Rats.
August 4, 2009 at 11:33 pm
Carl,
I’m still working out how far I’m willing to go with the whole treatment of signs as objects move as things get complicated very quickly. Part of this is a self-reflexive demand of my own philosophy. Insofar as I’m trying to break down the whole distinction between nature and mind that’s vexed philosophy, this leads to the conclusion that any philosophy (or other cultural artifact) is itself an assemblage of objects. The question then becomes that of determining what sorts of peculiar objects signs are. I suspect that anthropologists are critical of memes for the same reason that I was critical of memes when I first encountered the theory about five years ago: Here we have these cowboys claiming to have discovered a whole new realm of investigation– memes –when we have had semiotics and linguistics for decades now. When you read Dawkins and Dennett on memes you get the sense that they are reinventing the wheel, and in a number of instances poorly. In a number of respects, I think the meme theorist stands to learn far more from the semiotician (and cultural theorists like the anthropologist) than the semiotician has to learn from the meme theorist.
Nonetheless, I do think that the concept of memes draws our attention to three very important features of signs that are often overlooked. First, I very much appreciate the manner in which meme theorists treat signs as material realities. This is already something that was implied by structural linguistics, where the signifier had a certain materiality to it, but it really gets brought into relief in meme theory. Although the meme theorist is agreed with the semiotician in claiming that memes are ideal entities, memes/signs are ideal entities in the sense that DNA is an ideal entity. DNA is an ideal entity in the sense that it is a pattern that can be copied or replicated so there’s a very real sense in which it persists regardless of what it’s copied into. However, while it is an ideal entity it must nonetheless be embodied in some form of material. Similarly in the case of memes. Memes are ideal patterns but are nonetheless always embodied in some material medium whether this be brains, sound-waves, paper, clay, the wood of my desk, electronic pulses, zeros and ones, etc. Although this point seems slight, I believe it is filled with all sorts of profound consequences.
Second, I think meme theory draws attention to the replication of signs throughout the world. As a historian and a Gramscian to boot, I’m sure you can appreciate that. As we move from the Middle Ages to the Renaissance to the Enlightenment, we have the emergence of certain memes, but this emergence has a frequency that can be described in terms of populations. Just as we can describe a particular species as rare and bordering on extinction, we can talk about how signs pass from the rare to the ubiquitous and examine the manner in which this takes place. Because memes or signs must be replicated and embodied in a physical medium, we get all sorts of questions about how this replication takes place. In many respects, I think this is precisely the sort of issue Badiou is working on with his account of truth-procedures. The subject in the grips of a truth-procedure is a subject that is seeding the social world in which he is embedded with a new set of memes.
Third, I think mimetic theory has made a significant contribution to semiotics by situating signs in terms of questions about fitness-landscapes. I am perpetually fascinated by questions of why certain signs come to resonate or replicate at a particular point in history and why at other times they do not. This, I think, can be tracked according to two axes. On the one hand, coming from the perspective of media and technology studies, we can explore the manner in which media functions to constrain and enable the replication of particular types of signs. When I refer to “media” I am referring to the medium in which signs are embodied. Thus, for example, in oral cultures you get very real constraints as a result of the medium of speech. If, in transmitting signs, oral cultures tend towards narrative or stories and rhythmic, song-like modes of composition, then this is because cognitively these sorts of patterns are more susceptible to being remembered. This can be witnessed among very young children that can memorize an entire Dr. Seuss story precisely because its narrative structure and rhyme scheme. Here the medium defines, in part, a fitness-landscape that is not very congenial to, for example, a thought or meme like Hegel’s Science of Logic. On the other hand, we get fitness landscapes defined by relations among memes themselves. What is it, for example, that makes the current fitness-landscape of memes in the United States particularly inhospitable to memes pertaining to socialism and Marxist thought? Where are the reigning assemblages that make it so difficult for these memes to replicate and spread throughout the social world? What sorts of strategies can be devised to change this? And so on?
At any rate, all of these are points about materiality. Treating signs as objects helps draw attention to the material embodiment of signs and the role this plays in proliferating signs.
August 5, 2009 at 12:23 am
[…] Just a quick note before I get down to grading. In response to my post on the game of life, Carl writes: I’m not sure I’m on board with this: [O]ne of the reasons I find the ideas so […]
August 5, 2009 at 6:26 am
[…] resistant to interventions that disrupt these tendencies. Not unlike the emergent patterns in the Game of Life, you get, as it were, systems that can only evolve diachronically according to the synchronous […]
August 5, 2009 at 4:39 pm
Levi, thanks for cleaning up my sloppy html tag and for this terrific response. Busy with siding repair – termites having left signs of their full objectivity – but I’ll try to pick it up as it deserves at its new location.