This post assumes familiarity with David Deutsch's (DD) original idea of static and rational memes in The Beginning of Infinity (BoI). (Summary.)
DD's theory says there are two types of memes with two different replication strategies. Static memes replicate by suppressing criticism of themselves (and sometimes also of other ideas). Rational memes replicate by being useful.
I propose instead that there are three types of memes.
- Rational, useful memes.
- Static memes that suppress criticism and creativity.
- Static companion memes that do not suppress creativity or criticism themselves but are adapted to replicate in an environment where they are suppressed.
The additional category is a companion meme which requires criticism suppression but relies on other (static) memes doing it.
I think static companion memes have a variety of replication strategies rather than being defined by one. One possibility is being highly adapted to appear useful to people who aren't critically thinking.
In a fully static society, not all memes have to follow the replication strategy of suppressing criticism of themselves. The reason is the reach of knowledge (another concept explained in BoI). Some static memes do (partially or fully) general purpose criticism suppression rather than only suppressing criticism of themselves.
I think there is basically a core of some powerful static memes which are very effective at suppressing criticism in general. Once those exist, other memes don't need to suppress criticism of themselves because criticism isn't happening anyway. So they can evolve to compete for replication bandwidth in other ways.
A further complication, which blurs the categorization of memes into two or three types, is that ideas can be like code libraries from programming which provide callable APIs. In order to suppress criticism of itself, idea B can call a library function provided by idea A. However, if the host has idea B without having idea A, then that function call doesn't work and B fails to suppress criticism of itself. In this case, much of the knowledge of criticism suppression is outsourced, however idea B is able to actively suppress criticism of itself in the right environment.
Another complication is that there need not be a black and white dividing line between static memes and static companion memes. A meme can do some of each: it can be adapted partly to suppress criticism and partly adapted to do something else (such as appearing useful or good to non-critical thinkers). I'd guess that many memes are mixed because the core of criticism suppressing memes remove some but not all of the selection pressure on other memes to optimize for criticism suppression. That allows them to adapt for other purposes too, and some may entirely lose their criticism suppression. This other adaptation would be to better compete with other memes for replication bandwidth in a static society environment. Once criticism is largely suppressed, the amount the typical meme replicates won't have much to do with how well that memes suppresses criticism.
Update:
Memes replicate in two different ways. Within a mind and between minds.
Memes must replicate between minds to last over time.
Critical Rationalism says we learn by doing evolution of ideas within our mind.
Do static memes replicate within a mind? That sounds potentially bad because they'd make progress and change, not stay static. But it depends on what selection pressure they're being exposed to. If a static meme could control the selection then it could use within-mind replication to get more optimized. This would be different than suppressing criticism (DD's idea of static memes). It'd be changing the nature of the criticism instead.
If static memes simply suppress criticism, they can't get more adapted by within-mind evolution. But if they could instead control the types of criticism, then they can benefit from within-mind evolution.
So I'm thinking static memes do within-mind evolution in some cases while keeping control over the selection (criticism). I think that's a significant way static meme theory is incorrect.
Messages (42)
Example
>Static companion memes that do not suppress creativity or criticism themselves but are adapted to replicate in an environment where they are suppressed.
What would be a real world example of one?
#14799 Don't know, it's a theoretical category, and there are the complications I mentioned, and precisely specifying any type of meme is hard.
#14799 I can give a hypothetical example. Maybe horoscopes seem really great and useful to people full of static memes who don't think critically about anything, but the horoscope meme itself doesn't suppress criticism.
And you can insert anything else you think is bad instead of horoscopes to get another example that could at least hypothetically work.
#14801
Are they "neutral" memes that are false?
> Some static memes do (partially or fully) general purpose criticism suppression rather than only suppressing criticism of themselves.
How does this square with static memes evolving in the direction of only causing fine-grained damage to the holder's mind? If some static memes cause general purpose criticism suppression then how does the holder live long enough to be able to enact as much as possible the meme's behaviors?
#14803 Static societies live very inefficiently and ineffectively! They're bad at it!
I don't think it's fine-grained damage. I talk with people and find they are irrational about wide varieties of stuff, including ideas they'd never heard of before and which is pretty far removed from direct criticism of any static meme as far I can tell. And I'd expect powerful static memes to have reach, partly because of the importance of dealing with indirect criticism/threats (e.g. background knowledge that can lead to an argument against the meme, or even the first 6 steps in a 7 part argument against the meme – since those 6 steps don't constitute a direct criticism of the meme, suppression would indicate some reach to the suppression, and I think suppression of that kind of thing is crucial).
There's also the well known idea that people dislike criticism itself. There are some real issues behind that concept even if it's not fully correct in our mixed society.
And if static memes were leaving people with critical thinking in general, minus a bunch of exceptions, then I think static societies would have considerably more innovation than they do. There'd be some innovations invented that don't contradict the static memes.
People can live without critical thinking (at a high level) because the static memeplex dictates a lifestyle for them to live. This unfortunately has significant similarities to animals which also don't live by critical thinking.
Maybe critical thinking remains significantly active at a low, unconscious level. Could a static society member use critical thinking to e.g. resolve ambiguity in a sentence I say, not according to some preset cultural default expectation but by understanding what sort of thing I would mean? What sort of small scale problem solving could they do – is it like an animal relying on algorithms or can they deal to some extent with the situation going off script in a way that pre-existing knowledge doesn't address? I don't think anyone's figured out all the details.
It makes sense that static memes can have reach so that they do general purpose criticism suppression and not just criticism suppression in regard to themselves. As you say, if this were not the case, then we should expect to see considerably more innovation than we do because the holders' minds would not have been disabled in all areas.
It also makes sense that people in static societies can live without high-level critical thinking and that they become not dissimilar to animals enacting algorithms.
But I think there is disagreement with Deutsch and your post doesn't mention that.
Just in case I misrepresented him, here is the exact quote from BoI that I was thinking of:
> The overarching selection pressure on memes is towards being faithfully replicated. But, within that, there is also pressure to do as little damage to the holder's mind as possible, because that mind is what the human uses to be long-lived enough to be able to enact the meme's behaviours as much as possible. This pushes memes in the direction of causing a *finely tuned* compulsion in the holder's mind: ideally, this would be just the inability to refrain from enacting that particular meme (or memeplex). Thus, for example, long-lived religions typically cause fear of specific supernatural entities, but they do not cause general fearfulness or gullibility, because that would both harm the holders in general and make them more susceptible to rival memes. So the evolutionary pressure is for the psychological damage to be confined to a relatively narrow area of the recipients' thinking, but to be deeply entrenched, so that the recipients find themselves facing a large emotional cost if they subsequently consider deviating from the meme's prescribed behaviours.
He mentions here that susceptibility to rival memes is another reason that a static meme limits the damage it does to the holder. But a static meme that can do general purpose criticism suppression has sophisticated adapted knowledge to prevent rivals getting a toe-hold. It is the product of generations of variants that had a slight edge doing this so it is going to be extremely difficult for a variant of itself or another meme to begin to compete.
> But a static meme that can do general purpose criticism suppression has sophisticated adapted knowledge to prevent rivals getting a toe-hold.
Yes but it need not have sophisticated knowledge to prevent *non-rivals* from getting a toe-hold.
So suppose some static companion meme gets in. Then what happens between it and its rivals? The environment is very low criticism so I don't think they would necessarily compete by being best able to suppress criticism of themselves. They could compete by e.g. seeming good in some way, as well as directly attacking each other.
You might say *all* memes are rivals because they compete for bandwidth to replicate. But if some important meme is replicating reliably, it might already be getting the bandwidth it needs and the memes that sometimes don't replicate due to limited bandwidth are other ones. The important meme doesn't need all the bandwidth so doesn't try to drive everything else out.
Also even if there was no space for new memes, an existing static meme could lose some or all of its criticism suppression knowledge over time if there was no longer selection pressure for that individual meme to suppress criticism because other memes were suppressing enough criticism.
Re inhibiting rivals, if a meme can inhibit all rivals (including the null hypothesis), so a person has no choice/alternatives, then that might be enough if the person isn't creative enough to create new rivals.
My big picture view, which I've had for a long time, is I think once there is a static society environment, some memes have advantages within that environment such as seeming good to that kind of person. I think there's selection pressure for that and actually it can be a big deal when there are ~zero dynamic memes anywhere to be found. Attacking rivals is another trait static memes can evolve for. It's like how an animal can have selection pressure on multiple traits like sharp claws, strong muscles, good eyesight, etc., and those can all evolve at the same time, and this disagrees with the concept of static memes being totally defined as memes that evolved to suppress criticism, to oppose reason, to be anti-rational.
I don't think this is a major innovation or change to DD's view but I think it's some ideas worth thinking about which DD didn't explain. I think meme theory has lots of room for additional development.
Basically I think BoI oversimplifies when it says:
> There are only two basic strategies of meme replication: to help prospective holders or to disable the holders’ critical faculties.
And that *two basic strategies* point is a theme DD has emphasized. I think that's a crucial starting point for understanding what a static society as a whole is like, but I think some individual memes can do well in a static society with other traits, not all the selection pressure is related to that basic strategy (sometimes it may even be ~none).
re your quote:
> Thus, for example, long-lived religions typically cause fear of specific supernatural entities, but they do not cause general fearfulness or gullibility,
I think long-lived religions *do* cause substantial amounts of general gullibility and contribute to people being fooled by stuff outside their religion like cold reading, spoon bending, horoscopes, etc.
so yeah some disagreement with DD but just as refinements to static meme theory which i broadly think is great.
> They could compete by e.g. seeming good in some way, as well as directly attacking each other.
Would one way of seeming good be to actually be good? I guess this mostly doesn't happen because people's standards are so low. And I guess if it is actually good it would not persist long in that form because there is little error correction happening to maintain it.
> Would one way of seeming good be to actually be good?
Not necessarily. Depends how divorced from reality people's judgments in a static society are (I think a *lot*). So I was thinking more like whatever dumb standards of judgment people have, it evolves to fit that, so not very reality oriented evolution.
You say that static companion memes benefit from the core group of general criticism suppressing static memes. Is this relationship symbiotic? Do that core groups of memes gain any benefit from static companion memes?
#14813 I guess the benefit would be if the static companion memes had more than zero effectiveness at keeping the person alive.
The type of static meme that Deutsch described which suppresses criticism mostly in regard to itself would be rare right? Maybe they once weren't, but once the core group of general criticism-suppressing memes evolved the specialised memes would not have selection pressure on them to suppress criticism and so go extinct or become companion memes.
#14815 Yeah that's what I was thinking. Some core memes suppressing lots of criticism, and then most memes in a static society don't have selection pressure on them to do a lot of criticism suppression.
Though, mentioning again, this is a simplified model. E.g. the API issue in the OP blurs the issues.
Thanks for answering my questions. No more for the moment. I think your idea is worthwhile and a good refinement. Hope you get more discussion.
#14818 Thanks for the questions, I think they helped clarify things.
Updated the blog post with a new idea about static memes.
> Do static memes replicate within a mind?
When we recall to mind an idea from our memory do we replicate the idea in the act of thinking about it? I guess so, as recall is confabulation. That must mean static memes can replicate within a mind.
> If static memes simply suppress criticism, they can't get more adapted by within-mind evolution. But if they could instead control the types of criticism, then they can benefit from within-mind evolution.
The criticism would help to find errors in the replication strategy of the meme right? It would help the meme shut down criticism of its payload.
> Do static memes replicate within a mind? That sounds potentially bad because they'd make progress and change, not stay static. But it depends on what selection pressure they're being exposed to. If a static meme could control the selection then it could use within-mind replication to get more optimized. This would be different than suppressing criticism (DD's idea of static memes). It'd be changing the nature of the criticism instead.
Thinking about this further: When a recipient acquires a static meme there are lots of rounds of variation and criticism happening in the recipient's mind. There is evolution of ideas within a mind. The static meme has set up the environment so that the criticism is not directed towards finding errors in the meme's payload but is directed towards recreating the meme as faithfully as possible. It is controlling the types of criticism.
So how is what you are saying different to Deutsch?
>> This would be different than suppressing criticism (DD's idea of static memes).
> So how is what you are saying different to Deutsch?
?
#14906
>> This would be different than suppressing criticism
(DD's idea of static memes).
DD's idea is that a static meme suppresses criticism of its payload. It doesn't suppress other criticism and, in particular, within-mind criticism of how well the holder is conforming to it. So when you elaborate DD's idea it seems similar to the idea that the meme is controlling the types of criticism it receives.
Something went wrong with the quote. Sorry.
#14907 Where does DD say that? The word "payload" isn't even in BoI.
Memes evolving within a single mind
> Memes replicate in two different ways. Within a mind and between minds.
How does meme evolution work within a single mind? Is it due to memory changing details each time the idea is remembered, like #14891 suggested? In this usage, is meme synonymous with idea?
DD explains memes here as having to do with behaviour and spreading to other minds, but maybe this is too narrow a conception of a replicator?: https://twitter.com/DavidDeutschOxf/status/1191276695023751168?s=20
> Present in a mind, enacted as behaviour, consequently recreated in another mind, enacted as behaviour.
> How does meme evolution work within a single mind? Is it due to memory changing details each time the idea is remembered
No, the primary way is *thinking consists of making copies of ideas with variations and doing criticism/selection*. Making a bunch of copies is a core part of thinking.
#14909 Payload is my term. DD says a static meme "disables the recipient's critical faculties in regard to itself". Which puts it a bit awkwardly I think.
"Bad" by what standard?
Elliot Temple wrote:
> Do static memes replicate within a mind? That sounds potentially bad because they'd make progress and change, not stay static.
"Bad" by what standard? I can think of two different interpretations that kinda make sense.
*If static memes made progress instead of staying static, it would be potentially bad for* ...
(1) **people**, because it might mean that static memes would get better at controlling people and keeping them static.
(2) **the theory of static memes**, because it might somehow contradict the theory.
I meant it'd be bad for the meme's replication strategy (staticity).
Within-mind static meme evolution
Here's how I now see the problem:
All memes have a *payload* and a *replication strategy*. The payload is the idea represented by the meme. In the case of a static meme, the payload does not need to have any truth content. The replication strategy is the knowledge that the meme uses to pass itself on. This must have truth content for the meme to be successful.
In a static meme, the replication strategy prevents criticism of the truth-content of the payload. The strategy pushes the recipient to replicate the payload faithfully. This requires creativity but not creativity directed at truth-seeking.
The knowledge in the replication-strategy must also be passed on to the recipient. There are two ways to do this. Either the replication-strategy prevents criticism of the knowledge it contains. Or it allows criticism of this knowledge.
Criticism would be beneficial to the truth content of the replication strategy. It would help to eliminate errors and improve the strategy. But, it carries a risk that progress would happen because new knowledge is being created. But if criticism about the replication strategy is suppressed, then a mechanism for detecting errors is not being used and the meme might be outcompeted by other variants that can somehow use that mechanism.
Is that an accurate summary of the problem?
static memes & staticity
#14926
My paraphrase of what curi wrote above:
*If static memes made progress instead of staying static, it would be potentially bad for their replication strategy (staticity).*
Staticity of what? The static meme, the host, both, or something else?
curi wrote in the main article above:
> Static memes replicate by suppressing criticism of themselves (and sometimes also of other ideas).
In what way would static memes making progress be bad for their replication strategy? It seems to me that making progress, for a static meme, could mean getting *better* at suppressing their hosts' criticism of them, or adapting so that they continue to be able to replicate in a slowly changing environment.
> Staticity of what?
First, static meme itself. Second, the environment that its adapted to (which includes traits of the host like the hosts other ideas)
> In what way would static memes making progress be bad for their replication strategy?
Depends on how they changed. Without a ton of control over the process, could change to be less adapted as a static meme, there could be innovation, etc.
#14930 I don't think separating payload and replication strategy makes sense. Why would a replicator have two parts like that? What is the selection pressure on the part which is separate from how it replicates? What creates that part or makes it matter? I think the replication strategy is the core of the meme (which is a replicator), that's what it is and does.
Do you have examples of what you have in mind re two part memes?
#14940 I wasn't thinking that the payload and replication strategy are independent. Among other things, the replication strategy has true knowledge about the payload. The payload is required by the replication strategy. Substitute a completely different payload and the meme won't work. But the payload doesn't necessarily have knowledge about the replication strategy. The payload may have no truth-content whatsoever. The replication strategy is the machinery that gets the meme (payload+replication strategy) replicated.
Example of a payload: when you learn the "Happy Birthday" meme, the payload is the overt content of the song.
In some memes, I imagine the distinction between payload and replication strategy is blurred. Perhaps that is the case with all memes and it doesn't make sense to make the distinction, as you say. ionno.
Thoughts?
Question about static memes replicating when they pass from one mind to another
curi wrote in the original essay (with his comment #14926 spliced in):
> Do static memes replicate within a mind? That sounds potentially bad [for the meme's replication strategy (staticity)] because they'd make progress and change, not stay static.
IIUC, all memes -- including static memes -- replicate when they pass from one mind to another, a process which involves variation and selection, and therefore progress and change. Is that *also* potentially bad for the static meme's replication strategy?
Some thoughts/questions about within-mind evolution.
When we do within-mind evolution of ideas, there is some population of ideas in flight. We select ideas by critically examining each idea for flaws. If no flaw is found in an idea, fine, that is a candidate solution to our problem. Typically, however, every idea in our initial population of ideas will have flaws. So we need to decide which ideas to carry forward and work on and which to leave behind. Whichever ideas we choose, they are not viable solutions to our problem, they are just works-in-progress.
In biological evolution, works-in-progress do not get carried forward. For genes to become adapted to a niche they must cause the construction of a viable phenotype which must survive long enough to reproduce. And they must succeed at doing this time and again. If an organism is not viable or cannot reproduce then the genes it carries die. So, in biological evolution, the population of genes always contains some true replicators (or the species dies).
So there is a different between within-mind evolution of ideas and biological evolution. During within-mind evolution, work-in-progress ideas can be all we have and these don't seem to cause their own replication. None can survive in the problem niche.
However, in the discussion of curi's map generator, self-replication was identified as an important feature of evolution.
Perhaps the work-in-progress ideas do cause their own replication by having certain features that other ideas lack? They might have more explanatory power, or be more concise, or something. This sounds like critical preferences, but that might be OK when dealing with intermediate ideas.
Perhaps we ask: given that no idea completely solves our problem what shall we do now? Then we attempt to solve that (simpler) problem. That would lead to more within-mind evolution, however, to solve the new problem. And to regress. Which is not good.
Thoughts?
> When we do within-mind evolution of ideas, there is some population of ideas in flight. We select ideas by critically examining each idea for flaws. If no flaw is found in an idea, fine, that is a candidate solution to our problem. Typically, however, every idea in our initial population of ideas will have flaws. So we need to decide which ideas to carry forward and work on and which to leave behind.
Not necessarily. We could generate more ideas for the first generation.
Also for how to decide which to carry forward, if you do that: by decision criteria you don't have a criticism of. One way to do this is to carry forward ideas which solve some other problems besides target problem (especially less ambitious versions of the target problem).
> Perhaps the work-in-progress ideas do cause their own replication by having certain features that other ideas lack? They might have more explanatory power, or be more concise, or something.
sure, criteria along those lines can be considered and sometimes used.
> Perhaps we ask: given that no idea completely solves our problem what shall we do now? Then we attempt to solve that (simpler) problem. That would lead to more within-mind evolution, however, to solve the new problem. And to regress. Which is not good.
yes that's a good thing to do. *it is not a regress* because each time you iterate the problem gets easier, rather than harder or worse. same as with http://fallibleideas.com/avoiding-coercion
#15018 Thanks for your reply. I'm not sure why I wrote the regress is not good after having stated the new problem is simpler than the original problem. The regress will bottom out and fairly quickly, as you explain in your link.
One thing this highlights is that when we do within-mind evolution of ideas we can switch to auxiliary problems to make progress. So when we end up solving our original problem we did so by also solving a succession of simpler problems whose solutions tell us how to proceed in solving the main problem.
> Do static memes replicate within a mind? That sounds potentially bad because they'd make progress and change, not stay static. But it depends on what selection pressure they're being exposed to. If a static meme could control the selection then it could use within-mind replication to get more optimized. This would be different than suppressing criticism (DD's idea of static memes). It'd be changing the nature of the criticism instead.
> If static memes simply suppress criticism, they can't get more adapted by within-mind evolution. But if they could instead control the types of criticism, then they can benefit from within-mind evolution.
> So I'm thinking static memes do within-mind evolution in some cases while keeping control over the selection (criticism). I think that's a significant way static meme theory is incorrect.
This sounds like the idea in BoI re culture evolving and the 180,000 odd years of cultural stasis being due to _creativity devoted to increasing conformation_ with the society. is that not directed in-mind replication? Granted, the _selection_ itself might not be done directly by the meme (the person is "voluntarily" doing it), but the only reason they value it is because of the static meme, so it's still causal.
> But if they could instead control the types of criticism, then they can benefit from within-mind evolution.
What does "type of criticism" mean? Sounds like the meme / you (unconciously) / your brain would need soem sort of teleology or fundamental classification system of criticism, but a) it's a stretch (is it?) to say there's super deep knowledge like which is also ubiquitous but we as ppl still don't really have any theories of (not to mention we'd have to recreate in every human via memes and learning, or would have to be genetic/inborn), but also this presumably is _all_ types of crits that there ever will be, because if it weren't and static memes _encouraged_ too much creativite effort then they risk creating new types of crits, but those crits weren't known about, so how could static-memes possibly be replicated so widely and so faithfully and have all this stuff we don't know yet baked in?
> Granted, the _selection_ itself might not be done directly by the meme (the person is "voluntarily" doing it), but the only reason they value it is because of the static meme, so it's still causal.
Clarification: the _selection_ of ideas to suppress is chosen by the person (who can also, safety for the meme, apply creative effort to coming up with the filters for which ideas to suppress and which to not) because they want it, but the reason they want it is the meme itself.
Also re the opening question I replied to:
> Do static memes replicate within a mind?
If they couldn't at least be generated in a mind they would all have to come from faulty replication or mistakes and luck, but that doesn't feel reasonable when another option is a person trying to suppress another person's rationality (like they were annoying or something). Actually, an idea! need to do some reading on it and things, will post to FI soon if it pans out.
> but also this presumably is _all_ types of crits that there ever will be, because if it weren't and static memes _encouraged_ too much creativite effort then they risk creating new types of crits, but those crits weren't known about, so how could static-memes possibly be replicated so widely and so faithfully and have all this stuff we don't know yet baked in?
Just to follow up / clarify this point: I don't have a problem with static memes having incredible reach and thus blocking new types of crits before they're known about, in the argument I'm quoting I mean to restrict this to
well, I will whether it does or doesn't. don't know which yet.
are rational companion memes impossible?
the idea of static companion memes as a 3rd type of meme seems to imply that rational companion memes are either a 4th type, or impossible. I think it makes more sense that both rational and static companion memes could exist.
Also, are there any companion memes possible that rely on *either* particular rational or static memes? I don't see a reason why variable-esq companion memes wouldn't exist (tho I guess that they'd require *particular* rational/static memes to work with, not arbitrary rational/static memes). I'm not very sure about this last bit, though, and wouldn't be surprised if there was a contradiction in there somewhere. Nothing comes to mind though.