模倣子 Helper Memes & The Skin

 Previous Essay - Memetic Index - Glossary - Next Essay 

Introduction

I've been writing a bit about the idea of submemeplectic skins (1). These are interesting since this concept must be understood and well-defined in order to identify "a group" of agents and a memetic inventory as distinct from other groups and from the larger society as a whole. This has memetic engineering implications to do with improving the "effectiveness" of an existing group (such as a corporation, government agency, etc.) or dismantling "harmful" groups (cults, terrorist movements, etc.) without the need for resort to mass arrests or other draconian measures (which are generally not very effective anyway).

The value of skin theory to identifying groups in a rigorous way not to be understated, but a greater value may be found in understanding how the group under study protects itself from outside influences (2) and also projects its power into the outside world. Furthermore, identifying the skin of a group should allow elaboration of the memetic environment inside the group, which may inform memetic engineering efforts to create a suitable environment "outside" the submemeplex, which could enable effects such as memetic tunneling (3) to "evaporate away" member agents into the outer society, or supermemeplex.

Helper Memes (4)

In my previous essay, I looked at the probabilities of certain pathways between a simple set of states with a simple set of agents, however, I left out the effect of immunomemes and alliance memes. It is probably axiomatic that any action taken by any member of a cohort gives other, perhaps all, members of that cohort the ability to enact a response, perhaps in an infinite cascade (6).

In my previous essay, I looked at a simple memetic system with two agents and three states. An even simpler system would have only two states:

State A	agent-a	agent-b		State B	agent-a	agent-b
meme-a!	0%	0%		meme-a!	40%:StateA	60%:StateA
meme-b!	10%:StateB	90%:StateB		meme-b!	0%	0%

fig. 1. Transition Matrices

This seems too simple to be interesting, but one could perhaps imagine a conversation between two people. The State A could be the one in which agent-a is speaking, State B while agent-b is speaking. The transition between states is the passing of the "talking stick" from one speaker to the other.

Unless the two speakers are just talking over one another, there must be some kind of orderly exchange, in other words, one speaker must "yield the floor" to the other in some way. That is to say, one speaker must make some kind of gesture to hand the floor over, and one must make the gesture to ask for the floor. In more advanced cases, one speaker may prompt the other, asking something like "What do you think?" or "What do you have to say?" or such.

This could be modeled by alliance and immunomemes. For instance:

agent-b wants to start talking, and is resisted by agent-a, to stay in State A where agent-a is still talking, or yields the floor to agent-b (null! meme), or encourages agent-a to keep speaking

StateA.agent-b.meme-b! agent-a.null! => StateB    b request, a lets it happen
StateA.agent-b.meme-b! agent-a.meme-a! => StateA     b requests, a resists

StateA.agent-b.meme-b! agent-a.meme-b! => StateB     b requests, a encourages

StateA.agent-a.meme-b! agent-b.null! => StateB    agent-a offers, b passively accepts
StateA.agent-a.meme-b! agent-b.meme-b! => StateB    b actively accepts a's offer
StateA.agent-a.meme-b! agent-b.meme-a! => StateA     agent-a offers, B refuses

Table 2. Deployment Descriptors

Six deployment descriptors (7). Is this a complete picture? A state change requires a deployment of the request, by either agent, two states, then the other agent can either passively acquiesce (null!), resist the transition, or actively accept / encourage it.

The deployment of any of each of these three memes brings the total number of possibilities to six.

If we leave out StateA.meme-a! (7) for now, we end up with only StateA.meme-b! and StateB.meme-a! and we can think of each of these as producing a "compelled" state (8) where the other agent, or the "targeted" agent, must react immediately in one of one or more clearly specified ways. This is what immunomemes and alliance memes do.

fig. 3. State Transition Diagram for "Conversation"

Here we see how once the two compelled states, which I've named StateAB and StateBB, are reached, the agent whose memetic deployment invokes those states is now subject to what we could call "bullying opportunities" or "alliance opportunities" and we can see how they are symmetrical, and how the other agent is "compelled" to react, or, the other agent is "given the opportunity" to react. This is the sort of thing we see in the Triangular Baseball Model, that is, that the agent who is reacting does not have this opportunity to do so until the other agent has taken their action, and that whether the reaction is "beneficial" to the initiating agent is unclear from the network structure of the memeplex alone (9).

Immunomemetic Notation

As I started to do in Triangular Baseball, I use a shorthand to remove the compelled states to do with immunomemetic (and alliance meme) transitions to just be a series of memes that are interdependent and interrelated by putting these group memes into the memetic transition matrices. This starts to imply a causal relationship, which may be imposed artificially. In principle, however, a given deployment causes a "bullying opportunity" or an immunomemetic deployment opportunity, which then sends the system to a given "other" state, different to the one "intended" by the initial deployer.

State A	agent-a	agent-b
meme-a!	Compelled State AB b.b! => B b.null! => B b.a! => A
meme-b!		Compelled State BB a.b! => B a.null! => B a.a! => A

State B	agent-a	agent-b
meme-a!	Compelled State AA b.null! => A b.a! => A b.b! => B
meme-b!		Compelled State BA a.null! => A a.a! => A a.b! => B

fig. 4. Transition Matrices, Immunomemetic Notation

This is problematic, since how does one ascertain what the initial deployer "intended" and what would've happened had the bully not deployed their immunomeme? This would appear to come back to the problem of defining what a state is.

In figure 4 above, the helper meme transitions in italics can be thought of as immunomemetic, since they defeat the deployer's "intent" and prevent the transition that would've otherwise happened as a result of the initial deployment. In simpler terms, agent-b "wants" to go from State A to State B, and when agent-a deploys meme-a!, this transition is thwarted, so to speak. In practical cases, this "intent" may not be so clear-cut (14).

Summary & Conclusions

The problem of how to define a memetic state is still troubling, since states are ultimately 

not measurable in the way that memes are. Memes are "well-marked" (11) and so may be identified by observation, reported in memetic hacking sessions, and take place in a time series (12) which is also observable.

There may be an "infinite cascade" of possible different meme deployments which lead to a given state. This may be a way to define a state, and perhaps even the n-space in which memetic systems function. The probability of a series of memes being deployed in sequence, and by which agents, may be a way of defining said space, and may even give rise to ideas such as "memetic momentum" (13) which could lead to predicting the behavior of large organic systems.

The memetic matrix model may be problematic in terms of its use as a tool for analyzing real systems, or large systems. It relies on the artificial idea of a "state" and perhaps needs to rely on ideas such as the probability of certain memetic sequences, but this remains to be explored.

_________________________________

(1) A submemeplectic skin, or "a skin," derives from the concept of a submemeplex, which is a memeplex within a larger memetic environment, or a supermemeplex. I've written about this in at least one previous essay. It is a theoretical concept which I am still exploring and working to define. Think of it as a boundary or bubble around a faction, cultural subgroup, organization, cult, etc., where on the "inside" there are different memetic interactions, presumably much "richer" and more numerous, between agents. Inside and outside, and at the boundary, the immunomemetic and alliance meme profile are presumably interestingly different, and in which ways remains to be explored.

(2) All stable (sub)memeplexes contain conservative pathways, typically involving things like immunomemes (see elsewhere) that maintain its stability over time, prevent the memeplex from changing, e.g., new memes being introduced willy-nilly, agents behaving in novel ways, engaging with agents not inured of the memeplex, etc. Theoretically a stable, isolated submemeplex has a powerful and effective immunomemeplex that may or may not be concentrated in the skin (this remains to be seen). Careful study might reveal how memes may be introduced across the skin, for example, allowing a submemeplex to be diluted, contaminated, or dismantled, or the skin itself might be manipulated or adulterated so as to blunt the ability of the submemeplex to exert influence outside itself.

(3) Memetic tunneling is the idea that if the resonance of two memetic environments (for a given agent) were equivalent, such as within a cult and within a nearby mainstream church outside the cult, a inured agent might be able to "hop" from a place within the cult to the outside church, and still enjoy similar enough access to bible study, religious services, social functions, childcare, spiritual counseling, etc. If enough of the conditions could be attuned, this should be possible, and would tend to happen automatically thereafter. Obviously this happens all the time in the Corporate World, where employees who become dissatisfied, want more money, or have to move for personal reasons, simply change from one company to another, since the environments are all fundamentally similar. The same goes for schools, police departments, etc. Alcoholics Anonymous is a particularly ringing example, where the same simple format of meetings is studiously maintained, and individuals freely go to multiple groups, even in different countries with different languages, and change groups with practically no disruption. The idea that a memetic tunneling opportunity could be established to extract individuals from a cult, fanatic cell, or any other group is hardly far-fetched. The trick is to find what the individuals are "getting" from their membership in the submemeplex, and produce an environment that provides these things as well or better, or with bonuses.

(4) The term "helper meme" is a bit deceptive. In my Triangular Baseball model, and real-life cases such as The Candy Conspiracy, I discovered that immunomemes and alliance memes behave in a very similar way, almost indistinguishable, in fact. This symmetry is broken in ways which I am still working to nail down. One idea is the presence or absence of "rules" (5) and another is whether the initial meme-deploying agent suffers residual memetic debt. One question is the order in which helper memes may be deployed. Also, both immunomemes and alliance memes (ally memes?) seem to involve a (hidden) compelled state. They both seem to have the property of "unassailablility" in that the deployer is guaranteed resonance, or the support of the cohort. Some of these properties to do with timing need to be nailed down. Can both be deployed before or after the "initiator"? For instance, I make it possible for you to apply for a job and be sure to get in, or you tell me you applied, and I pull my strings afterward? Similar, but really two different things. Also, a bully can attack someone seemingly without provocation, or they may bully some action by the victim after it happens.

(5) See also "Free Play" and "Memetic Hell"

(6) One could imagine that the effect of an infinitely diminishing cascade of possible deployments, becoming increasingly immunomemetic, i.e., certain, could cause all of the pathway probabilities to converge to very certain values. This would of course be consistent with the conservative nature of memeplexes, if this idea could be elaborated.

(7) In addition to the meme to change states, e.g., StateA.meme-b! => StateA, one can imagine StateA.meme-a! => StateA, i.e., where one of the agents asserts keeping the system in the same state. This would also provoke a helper meme response from the other agent. We can perhaps look at this addition later.

(8) A compelled state is one in which an agent must immediately choose one or more memes which they must deploy.

(9) We take it as fairly obvious that immunomemes are deleterious to the initiating agent, i.e., rather than asking for help and thus giving a kind and altruistic other agent the chance to be virtuous by helping, an example of an immunomeme might be to give another a chance to embarrass one publicly, and this other taking it. Of course, being publicly embarrassed is a high-ticket service in some circles, and indeed, a child threatened by a parent with spanking for some prespecified misbehavior might nonetheless commit said behavior and thereby force the parent to inflict the spanking, i.e., put the parent into a compelled state, or still yet force the parent to relent (10), which still makes it a compelled state. Being able to put another into a compelled state may be a practical example of power, about which more needs to be written. The point here is that the line, if any, between immunomemes and alliance memes may be murky, or that we have not yet added the properties that distinguish them in our examples so far.

(10) By forcing a parent to relent rather than inflict a threatened punishment, the child reveals that there were other memes at play, about which more could be written.

(11) How "well-marked" a meme is is one of its properties, but in the end a meme that is too badly marked, i.e., not identifiable as a meme by agents in the environment, ceases to be a meme in any practical sense. This is something of an inside joke with "The Blue Man Group" by the way--their longer and more obscure moves become less and less specific, clear, or identifiable (more on this later). In short, memes are inherently "well-marked" to some degree.

(12) Memes happen in time sequence, as opposed to being "caused" by one another, which is a more subjective, attributive judgement, and thus outside of the domain of Macromemetics.

(13) Memetic Momentum is a new term, as yet undefined. We can wonder whether some kind of "memetic energy" can be defined, and whether we can talk about the tendency of a memeplex to gravitate towards one configuration or another.

(14) Determining if a meme is an immunomeme based on the "intent" of the meme it's reacting to may only work in the simplest, and most contrived of cases. It may lead to circular definitions, moreover, since the idea of "intent" (as in "I intend to go from State A to State B and so am deploying meme-x") implies that we a priori know that there is some target state and that whichever meme deployed has some "purpose" all of which are effectively anathema to the science of Macromemetics. All we know is what happened and who did it, and what that may enable others to do, typically based on the "predisposition" of the memetically-linked cohort. Hence quantities such as the change in residual memetic debt or the deployment profile opportunity obtaining at the end of the deployment cascade may be more relevant quantities. In this sense, the distinction between immunomemes and alliance memes may actually be murky (15), i.e., their symmetry may not be clearly broken.

(15) In addition to first-order memetic engineering considerations such as marking and closure, higher-order factors such as changes to residual memetic debt and deployment opportunities should probably be regarded carefully and optimized. For instance, memetic pathways that lead agents that follow them to larger and larger inventories, access to memetic nexuses, lowered risk of incurring residual memetic debt (also something of a marking issue) can be analyzed and optimized. Additionally, pathways in which others who might "oppose" them are at risk of elevated levels of residual memetic debt could be seen as more secure and desireable.