Saturday, March 25, 2017

模倣子 What do Memetic States Look Like?

Memetic Index 

I have noticed, and it has come up in discussions, that a good example of a collection of memetic states is needed, and that I don't really seem to have one yet. I need an "elevator pitch" for the concept of memetic states and how they relate to MIAOs and the memes that influence them. I think I probably need to incorporate the concepts of ideomemetic systems and MIAOplexes in this example.

I need some kind of scenario, a collection of MIAOs, some memes (including immunomemes), and a collection of transition matrices, including (example) ideomemetic transition matrices for various hypothetical individuals that make up the cohort of the memetic fabric (or subsection of fabric) under consideration.

I'm interested in considering the Blue Shirt Tuesday Doughnut scenario. It's simple, and I have done actual experimentation on this very memeplex, with results  that have led directly to a number of memetic theories. I could also contrast this with the Koffee Klatch scenario, especially if I wanted to bring in the concept of residual memetic debt, which seems to be very useful in illumination various "cheating" behaviors (1). The Prime Pizza Thursday scenario is another good example of how lack of residual memetic debt seems to curtail cheating, and offers solid experimental proof (2). The doughnut scenario especially illuminated the concept of memetic polarization.

I hope to put forward my diagrams for memetic systems as well. I have been using them in my comics (6), but I need to make them more rigorous, and bring them more into line with the actual transition matrices they try to represent.

Example of Notation with Blue Shirt Tuesday
As I mentioned, the Blue Shirt Tuesday memeplex had as objective to measure the level of difficulty involved in infecting a chosen cohort with a novel memetic system that was self-enforcing through enactment of simple, very clear immunomemes. The experiment consisted of offering a libidinal bribe for participation in the memeplex in the form of free doughnuts on every Tuesday, on condition that "participants" be wearing a blue shirt. Questions were:
  1. Would people participate, i.e., eat the doughnuts?
  2. Would there be cheating?
  3. Would people self-police, i.e., bully or socially shame those who tried to take doughnuts without wearing a blue shirt?
All of these proved to be true, to a surprising degree, and the memeplex took hold very quickly (5), and became universally known throughout the cohort, although not all members actually ate the doughnuts (3). This let to the discovery of memetic fabric polarization, among other things.

One thing which was probably significant about the Blue Shirt Tuesday Memeplex is that the "rules" for bullying opportunities, or more technically, the immunomemes which could be deployed and the contexts (memetic states) in which the could be deployed were very clear and simple. This is probably a significant contrast with the Koffee Klatch Memeplex (see below) and an important lesson for any would-be memetic designer, i.e., if the immunomemeplex is not designed to be clear and simple, and specifically focused on managing the desired behavior, in this case, shaming people out of eating doughnuts when their shirt is not blue, then the system will not converge to the desired mode of operation. The bullies must know:
  1. exactly which immunomemes are at their disposal
  2. under which circumstances (states) they may deploy them
  3. the immunomemes must be unassailable (4)
If any of these properties are not present, adequate social shaming to enforce the desired function of the memeplex may not be possible.

Anyway, with that, on to the notation.
This is a notation system I am trying to finalize, and one which I have been using in one form or another in my comics (6) for a number of years now. The "clouds" represent memetic states, the "balls" are MIAOs, and the arrows (and other symbols) represent memes. I am experimenting with the idea that MIAOs, or rather, MIAOplexes, determine the memetic inventory of a given state. Note the "capitalization" scheme I'm using:
  1. MIAO (all caps)
  2. State (capitalized, actually CamelCase)
  3. meme! (lowercase with exclamation mark)
  4. unmeme! (strike-through or overbar for a meme not enacted) 

I'll try to express the above system in these terms. There are several things that need to be expressed:

  1. the description of each state in terms of the MIAOs which are in play
  2. the memes which are available in each state
  3. which memes may be deployed in response to which memes (bullying opportunities)
  4. state transitions predicated by meme deployment (if any) (7)
So we can try to describe the system like this:

State [ SALIENT MIAOs ] available memes!, meme! => response opportunities!

Out [ ] enter!
In [BLUE, DONUT] exit!, eat! => shame! praise! shame!

So here we see that the eat! meme offers the response/bullying opportunities of shaming the eater (if they don't have a blue shirt), praising them for taking a doughnut while wearing a blue shirt, i.e., participating in the memeplex, or just not doing anything (10). I have a notational issue to do with which memes are allowed to be triggered by eat!, i.e., is this predicated solely on the MIAOs in play, or do I need a special chart? I think I need more states, and to go back to the idea of a state providing a memetic inventory at each step, which means there would be an "In with blue shirt" state and an "In without blue shirt" state, and the eat! action meme would provide different opportunities in each.

MIAO LIST: attached memes!

DONUT: eat!
BLUE, DONUT: eat! shame! shame! praise! enter! exit!

I have another problem here, too. There are some memes which are arguably inherently associated with given MIAOs, but other memes which are not, or which are perhaps only relevant to a given state or states, e.g., enter! and exit!, in other words, there is a kind of synergistic effect, where a MIAOplex may have memes associated with it which the individual MIAOs comprising it do not.

Let's try to restate the memetic state description to address some of these issues, before we remake the visual diagram. Let's try leaving out the Out state since it doesn't really add anything.

Blue [ BLUE, DONUT ] eat! => praise! shame!
NoBlue [ BLUE, DONUT ] eat! => shame!

Nota Bene: there are no real "state transitions" in this case, or in any of the other memeplexes I'm going to discuss here.  This is something of a shortfall, since memes causing the cohort to change states is an important theoretical factor I'm working on. Anyway, baby steps. Another issue is that although having more states clears things up a lot, there is still the issue of how to decide whether a cohort member deploys praise! or shame! in the Blue state. Since deployment does not, in principle (11), lead to a state transition, there are no race conditions ("Jinx situations")

The idea of MIAOs and anti-MIAOs may seem a bit odd. In this case we're really talking about a blue shirt versus a shirt of any other color. There are a lot of strange dichotomies that people accept. For example, vanilla and chocolate, apples and oranges, etc., are, in the US anyway, thought to be opposite, or are so used rhetorically at least. So, does the following equation hold true? CHOCOLATE = VANILLA.

Weakness of the Koffee Klatch
I've discussed elsewhere the Prime Pizza Thursday and Koffee Klatch memeplexes. I think that PPT is quite similar to BST in that the immunomemes and their deployment contexts are quite clear, and the experimental evidence is that they work very efficiently as artificial memeplexes. However, Koffee Klatch failed, and I'm curious if there's a good way to analyze this notationally, or was the system "flawed" in some kind of esoteric way about which I don't have enough information to examine further?

There were at least a couple of things that we wanted cohort members to do:
  1. contribute money to the kitty
  2. bring coffee
  3. don't drink without contributing in some way

This failed across the board, since nobody brought coffee, nobody contributed, and the coffee always seemed to disappear immediately with the only two members barely even getting any. Clearly a flop! Why did the system fail to prevent all this, and fail to support its ostensible objectives?

The memes we wanted to see were along the lines of contribute!, drink!, bring-coffee!, shame!, praise! for starters. There was a meme which we didn't want to see, which was cheat!, and ended up being the only one the cohort exhibited. In fact, there may not have been a cohort, per se, and that may have been a big part of the problem (13).

Were there actual MIAOs associated with the memeplex? This may also be a huge issue. We named the Koffee Klatch after a beloved former coworker, but many people may not have known the reference. Unlike with blue shirts, doughnuts, prime numbers, and pizza, there was no coherent MIAO to tie these memes together. This lack may be rather hard to quantify, however, but maybe a kind of memetic transactional analysis will be possible.

What we want is something like this:

CoffeeStation [ MEMBER ] contribute!Contributed, drink!Drank
CoffeeStation [ MEMBER ] contribute!Contributed, drink!Cheat
Drank [ ]
Cheat [ ] shame!Shame
Contributed [ ] praise!, shame!
Shame [ ]

So a big problem is that there are no clear bullying opportunities or bullying behaviors for people to do if they perceive somebody cheating. Plus, there is no "marking" of members and non-members,  i.e., the MEMBER MIAO is not distinguishable, unlike with the BST memeplex, or when somebody drops a chit in for PPT. It's hard to be certain, but it's clear that there is little for people to work with when it comes to enforcing this system through social bullying.  People want clearly-defined bullying opportunities, and what they are allowed to do to express them (14).

A Stronger Blue Shirt Tuesday
As I note in (11), there are no state transitions in BST, but there could easily be if there were "punishments" associated with cheating. Using my new notation, let's imagine:

Blue [ BLUE, DONUT ] eat!BlueAte       state transition
BlueAte [ BLUE, DONUT ] praise!, shame! (12)
NoBlue [ BLUE, DONUT ] eat!NoBlueAte
NoBlueAte [ BLUE, DONUT ] shame!Shame
Shame [ BLUE, DONUT ] spank!, dunce-cap!DunceCap, ostracize!Ostracize
DunceCap [ DUNCE, DUNCECAP ] shame!
Ostracize [ PARIAH ] talk!, acknowledge!
Note that the decision to shame (or anything else) is always optional, hence shaming leads us into the state of choosing from a list of possible shaming memes. This seems like extra steps, but it could also accommodate the situation where one person initiates the shaming, i.e., calls somebody out on it, and then others actually attack, rather like inciting of a mob, which seems to be the way that things really work, even with a single individual, i.e., an individual gives themself [sic] permission to bully, and then acts on it themself [sic]. One question is what happens to the system state after a spank! meme deployment? Do we just go back to NoBlue? Likewise Ostracize and DunceCap are somehow "created" as states, but the cohort may transition to another state when the pariah or dunce leaves the immediate vicinity. The concept of state transition remains, at least somewhat, rather loosey-goosey, and whether this is a problem of resolution or of inadequate modeling remains to be seen.

There are still a lot of questions. I have a notational system which basically represents the states, MIAOs, memes, and state transitions of a simple memeplex. One problem is how to represent memetic loops, that is, we assume that enactment of a meme is in anticipation of a response or reward (negative or positive), and so is there a kind of "return" to the original meme at the end of whatever "chain" of memes? Also, how does this system tie into ideo/endomemeplexes, resolve race conditions or "Jinx events"? These simple examples don't really have these, but it's out there. Another issue is how do MIAOs and MIAOplexes relate to the system. They are important, but how to they influence the memetic inventory of a state, how crucial are they to the definition of a state (if at all), and so forth. Also, while this new notation clearly still points toward a series of matrices to describe the system, which matrices exactly do we need? Another issue is "terminal nodes" or terminal meme deployments, such as spank! and praise! above. They lead nowhere. Also, there are "available states" which are created, but which may only exist when a "punished/awarded person" appears on the scene. Do we have "temporary MIAOs" for them, or what?

Residual memetic debt still does not seem to be clearly represented in this system. It could be related to ideomemeplexes, i.e., a concept of the individual expectation of memetic reward. This could drive deployment decisions, including things like masochism, or the ability of some to consider negative the memetic outcomes which others might consider positive, e.g., being a scapegoat. Of course, BST probably does not have RMD, which is part of the success of its design, while Koffee Klatch very likely does (it's where I got the idea for RMD in the first place). So if perceived reward differs from real reward, this could be a motivation for cheating. Cheating is still a problem (1) in need of general explanation.

(1) "Cheating" behaviors tend to be of great interest in many areas of psychological and sociological research. Some explanations for same might be considered to be "wanting."

(2) The number of chits collected seemed to equal the number of pizza slices quite closely, indicating a high degree of non-cheating participation.

(3) See other essay for description of some of the behaviors around this memeplex.

(4) unassailability is a basic property of immunomemes, meaning that there is little or no risk to the deployer of same of being counterattacked. The need for unassailability may be less if they deployer has high status, or is a memetic nexus (which may be one and the same thing -- I'm still working on this idea).

(5) It's possible that I'm overstating this somewhat -- I don't have exact notes on the early days of the experiment. I did not personally enforce the rules of the memeplex. I recall that early on I was asked by some people whether it was okay to have a doughnut without wearing a blue shirt, or whether their [sic] shirt were blue "enough," and I replied that they could do what they wanted.

(6) See Schrödinger's Catechresis and Running SCAREd for examples.

(7) I have written before that a defining property of a meme is that it causes a state transition. I am taking a closer look at this idea. (8)

(8) I am thinking that it may be more useful to think in terms of closing memetic loops and providing bullying opportunities (or response/resonance opportunities). I need to work out whether there be any notational support for the closing of memetic loops idea. That tends to presuppose a set of one or more "expected" or "hoped-for" responses (9).

(9) a "hoped-for" response by a meme deployer may depend upon an internal economy which might be able to be described in terms of an ideomemeplex (endomemeplex), which should allow for things like one person seeking as desirable something that another person might find painful or otherwise repugnant, or one person's tendency to be a scapegoat or whipping boy, while others have the sense to avoid such situations, and so forth.

(10) As I've written elsewhere, not responding to a person's actions is often a reward in itself. Taking food for free often carries a burden of guilt, so "getting away with" taking it in full view of everybody and having no one object is a reward, a social approval. It's rather like a favorite example of taking all of one's clothes off when at a nudist colony. Normally an anti-social act, but in the special context it elicits no response, which is really an act of approval and is actually a strong memetic reaction.

(11) One could argue that once an individual is shamed for taking a doughnut illegally, the cohort moves into a new state of HavingShamedSomebody. This is perhaps not so interesting unless it allows or disallows a new set of memes, i.e., it changes what people are allowed to do. If one imagined a "crueler" version of the shaming scenario, where there were multiple possible meme choices when somebody cheats, and each went to a different state, i.e., ostracize the cheater for the rest of the day, administer a spanking, make cheater wear a silly hat, the first of which might involve the cohort being in an "ostracize" state for the rest of the day, the hat one would involve bullying the cheater into keeping the hat on, and so forth. How would this sort of thing complicate the diagram? Would there be a different state for each cheater and the punishment they got, or would they overlap? I might want to try this one out. Also, there could be a system of "rewards" where a person got special treatment all day.

(12) You could make the argument for a HavingEaten state, and that that state is where the non-shaming and praising memes become available. This would perhaps be a more consistent view, and would bring us back to the "every meme is a state transition" concept, which is probably much more amenable to matrix representation.

(13) I should mention that the PPT and BST memeplexes were not "advertised" at all. I just put the food out with the simple rules, and they took off over the weeks and months (years?) I was doing them. Koffee Klatch was right next to the existing coffee area, and so may have lacked novelty. Also, unlike the other two, Koffee Klatch may not have developed an actual cohort, i.e., people who knew it existed and saw themselves as members of the club/memeplex. It may well be a sign of the "good design" of PPT and BST that they didn't have to be advertised or explained, i.e., people just naturally started participating. This may be a touchstone of any well-designed memetic system -- more food for thought.

(14) This is perhaps an important message for American politics today. Yes, people and society are racist and misogynist, but the real problem is that they are given permission to express that hatred, those negative behaviors, and worse still, given examples and vocabulary for how to do it. Then it becomes a question of exercising well-known behaviors in a system with a long history of racism and misogyny. As Trevor Noah termed it, there have long been a bunch of racists who have kept it inside, who have held back from expressing themselves, who have now started to do it, and he encouraged them to "keep it inside." The term "emboldening" misses the mark in my view, as expressing racism and misogyny is not an act of boldness so much as sheeplike or robotic behavior while following the "wrong shepherd" or acting out of habit. Political leaders, especially those with a strong political mandate, are in a unique position to create the kinds of social rewards which can engender entirely new systems of behavior, rather than appealing to old archetypes, steeped in old ideas of racism and misogyny.

模倣子  Memetic Essay

No comments:

Post a Comment