Memetics is also a bunch of hooey

Well, for it to be dogma, it would have to be in principle unfalsifiable, as opposed to merely currently – i.e. unfalsifiability ought to be inherent in the concept of memetics, which I don’t think it is: if we could map brain states to ideas/memes (if such a thing is possible), we could observe their evolution, and thus, if it doesn’t happen via the selection process described by memetics, falsify it. Just that we’re currently unable to do so (or do something similar) doesn’t make memetics in principle unfalsifiable. We’d probably need more research to really find out if and how memetics is falsifiable, but well, that’s just an argument for continuing the development of its underlying ideas.

On the other hand, the logical argument behind meme evolution – i.e. that imperfect copying with selection automatically leads to an evolutionary process – is in itself falsifiable: find a way to have something being copied with changes that affect its likelihood of being copied, yet have it not evolve, or contest the assumption that there is such a thing as information that’s being copied imperfectly in what we call ‘memes’, and I think you’d basically be able to shoot the whole thing down.

jackdavinci writes:

> It’s pretty clear that the reason you think memetics is hooey is because you
> haven’t actually bothered to learn anything about memetics. Your OP is full of
> misconceptions, exaggerations, and outright falsities regarding the subject.
> Maybe you can repurpose your strawman to scare away some crows.

O.K., then give us a complete explanation of the theory of memetics so that we can understand it better. Then we can discuss it. Otherwise, all your saying is that we’re too stupid to understand it.

This is exactly the point at issue. I don’t believe that, at this time, memetics is even falsifiable in principle.

And, further, if we could define an experiment that could falsify it, but which doesn’t have to be feasible to do today, then I’d be quite happy to accept that Memetics is a legitimate scientific hypothesis.

And herein lies the problem. The vagueness of the concept means we cannot even imagine an experiment, irrespective of any practical difficulties.

If memetics is falsifiable?
I stand by what I said; if it isn’t falsifiable, it’s at best useless, and it has yet to be shown to be falsifiable.

I dispute that memetics follows from the basic concepts of imperfect reproduction and selection.

Firstly, in what sense are ideas really copied? Does any of the neuronal structure in my brain have to match the structure of another person having the idea? How different can an idea be before it’s no longer valid to consider it an imperfect copy?

Secondly, how does selection apply to ideas? Conscious beings can consider ideas to be good and bad, but that depends on a complex interplay of genetics, past experiences and current environment (including social relationships). And clearly, this is a moving goalpost – this selection mechanism must always change from selection to selection.
Whereas with natural selection, changes to genes have a limited scope in how much they can affect the environment.

As for “contest the assumption that there is such a thing as information that’s being copied imperfectly in what we call ‘memes’”, how could I do that when the meme concept is so vague? Of course the brain in some sense stores ideas, catchy tunes, techniques…does that necessarily validate Memetics?

I don’t know whether there’s an actual physical/biological representation in the form of a repeatable pattern that stays consistent in different environments – i.e. brains – meaning that if the two of us held the same idea, it’d be represented the same way in our respective brains; and in fact, based on the little I know about how brains work, I’d highly doubt it. I’m not even sure if the same idea would be represented in the same way in the same brain all the time.

But that’s not necessary for the copying of information at all – last weeks ‘Heroes’ episode stays the same whether I convert it to avi, flv, mpeg or put it on video tape (and in a certain sense, even if I re-enact it with hand puppets).

As to when an idea becomes so different that it has to be considered distinct – you’re right in pointing out the ambiguity and arbitrariness here, but compare it to speciation in biological evolution – it’s quite easy to point out two different species, but finding the exact point at which their hereditary trees split? Basically impossible.

As a measure of distinctness of ideas, you could probably look at their ‘goals’, which at least for some classes of ideas will give you a good approximation of how similar they are or aren’t – take a simple instruction, for example for building a paper flyer: if you’ve get the same flyer, it’s the same basic idea; if you get something different, but still similar enough to fly, i.e. to accomplish the same goal, it’s at least a related idea; if what you built doesn’t fly at all, it’s a different idea (even thought it might now acquire a purpose of its own – perhaps it just looks really pretty, and people start using it as a decorative item).
Obviously, that doesn’t work as neatly and intuitively for all ideas, and you risk getting into ‘just so’-story territory trying to think up ‘goals’ for more abstract ideas, like hairstyles and such, but that doesn’t automatically mean it isn’t a worthwhile exercise.

Going back to the – arguably limited – example of goals, think again about the paper flyer: when the building instructions are passed around, someone’s gonna get it wrong somewhere, most likely leading to a non-functional flyer. However, if it’s possible at all, chances are that somebody’s flyer will in fact end up accomplishing those goals better than the original version, essentially competing with its predecessor and, ultimately, succeeding over it, which is pretty much the definition of selection. Consciousness isn’t strictly necessary for that, it’s a perfectly mechanistic process.

Of course, the example is extremely simple, and the notion of ‘goals’ probably a naive one, but since it works so well in this toy model of memetic competition, I don’t see much harm in checking whether or not this holds for more complex memetic systems, up to and including our own consciousness. It’s just the logical generalization to make, in my opinion – the mechanism as described appears to apply to a certain subset of ideas, and there’s nothing fundamental and obvious that prevents it from applying to all kinds of ideas – all kinds of information, really – so we probably should expect that it does, and try to find out whether that expectation is validated.

Regarding falsifiability, I guess I’m just too much of a Feyerabend man to worry inordinately much about that; besides, if we were not to study a subject before its falsifiability was proven, how would we ever prove its falsifiability in the first place? If that means you don’t want to call memetics science, then well, you’re free to do that, of course. Personally, I do see some potential in it, so I do think it’s a worthwhile avenue to pursue; that it might ultimately turn out to be as scientific as phrenology doesn’t detract from that at all.

You seem to be assuming that when humans “consider” ideas to be good or bad that there is some non-deterministic process occurring, that the human actually causes the decision to go one way or another, as opposed to the human just following the laws of physics, much like a virus does.

That’s a huge assumption, and based on what we know about the universe (at this point in time), probably not the best assumption to make.

So, the paper flyer counts as one meme?
Even though, depending on the experiences of the person making the flyer, it may constitute more than one new idea?

If I choose to make my plane out of a different material, is that a mutation of the paper flyer meme?
But if that were the case, surely every meme that is related to constructing physical objects would have a “try other materials” meme. Or memes.

I don’t know where you found this straw man. But, to be clear, I see no reason to suppose that any brain processes are indeterministic.

Having said that, I find attempts to reduce consciousness down to constituent physical processes to be misguided. Consciousness is not merely the aggregate sum of many neurons firing; it is a higher-level process, that we don’t understand yet, that requires complex brain structures.
I don’t want to dwell on this point though, because this could kick off a very long thread of its own…
But suffice it to say, I don’t think the way that the brain consciously values ideas is directly analogous to survival of the fittest natural selection.

Not the paper flyer itself, which is actually totally irrelevant to the meme – it merely provides the vector for reproduction (similar to the way a living being does that for genes). And whether it’s exactly one meme, or what’s been called a ‘memeplex’, is hard to say, but, going again back to the genetic analogy, that’s similar to a given trait in an organism’s phenotype – it’s not immediately clear whether ‘blue eyes’ are due to a single gene or a couple of them acting in conjunction, yet the observation of the inheritance of blue eyes is strong evidence in favour of genetics.

I’m not sure what you’re getting at here – seems to me that would have to do more with the definition of idea than anything else; you may well say that the instructions for making this particular kind of flyer out of this particular kind of material are a single idea, which you could either have previously encountered or not, making it one new idea for every single individual encountering it regardless of prior history.

I’d say yes, mainly based on thinking that ‘choice’ is something that happens due to memetic competition; i.e. it’s not something from a different framework acting on the memetic environment, but a part and emergent property of that environment.

Or ‘try different materials’ may be a wholly separate meme that can be combined with a number of construction-related memes.

And the hope behind memetics is that it might at least help to uncover how those high level processes work.

But it might well be the case that the conscious evaluation of ideas is merely a post facto rationalization of a more fundamental underlying process, something hinted at for instance by Benjamin Libet’s research. Memetic competition is at least one suggestion for an underlying mechanism.

You misunderstand. Natural selection has been criticized as something that happens more or less by definition. Independent of mimetics, how would you falsify natural selection as a principle? The same type of thing could be used to falsify mimetics - obviously not the exact same thing. So, mimetics is as useful as natural selection to help us understand.

It’s true that ideas popping up independently is absurd given our knowledge of the world, but that doesn’t make this an invalid falsification criterion. It just shows that the principle corresponds well with our prior knowledge. Some discoveries just involve recognizing a pattern in what everyone knows, and don’t require the collection of new information. These memes quickly gets the “that’s obvious” treatment after being recognized.

Here’s a way of falsifying mimetics without burrowing into the brain.
We first need a metric. For mimetics, a reasonable one would be the number of times an idea is copied. That’s roughly equivalent to the number of people holding the idea, which is measurable in principle.
Mimetics, as I understand it, states that there should be a multi-peaked distribution for this metric, where ideas that are successful will reach lots of people, while ideas that are not should die out quickly. If mimetics were not true, we’d expect to see some other distribution. I’m not exactly sure if a Gaussian or exponential distribution makes more sense - but it would be different. All we have to do is to take some polls on a wide range of ideas, but we already know that there is a two headed distribution.

For doubters what distribution would you expect to see if mimetics were not true?

Have you actually read the thread? I’m suspecting not. I never said any such thing about viruses, so I’m not sure what the heck you’re talking about.

No. Because you get into the whole nonsense about memes “competing”, which, again, I do not accept. They compete only in the sense that an individual has exposure to multiple ideas, and makes decisions as to which will be maintained and passed on, and which will not. It is forced competition, so to speak.

As for “learning by imitation”, it seems to me that memes are here being defined as “any behavior or action that is mimicable”. That doesn’t seem like a very useful definition.

So how about a definition for meme that we can work from?

Well I ralready ebutted a lot of things Darwin Finch said. But looking back I see that DF wasn’t the actual OPer. I’ll try and respond to that directly.

The word wasn’t invented to be fancy. Dawkins was making a point about how any system with components similar to genes would be subject to similar processes. He suggested that one such system might be the communication of ideas, and came up with the term ‘meme’ to indicate the idea equivalent of the gene. And a meme isn’t just an idea. Ideas are more comparable to DNA. Like genes, memes are the smallest heritable unit of transmission.

No one in the actual field of memetics has ever suggested that memes have agency. Any use of anthropomorphic words is merely metaphor. Genes “replicate” too but that does not suggest that they have intention or will.

Cite? No one in the field would suggest memes are actually like organisms except perhaps in an informal essay to laypeople using analogies. A better analogy is probably comparing memes to programs and the mind to a CPU.

Of course it’s an environment. An environment is just the set of components within a given area. Is it like an ecological environment? In some ways yes, in some ways not. Ecological environments are hardly “entirely passive”.

Again a meme is the smallest reproducible component of an idea. In genetics, it’s clear what happens at the smallest level of genes to an extent (they make proteins - although there are other functions that aren’t as well understood yet) but it isn’t clear exactly what genes correspond to the making of larger systems such as an arm. And to an extent it isn’t possible to, since larger structures are impossible to enitely separate from the body systems that support them.

Good questions. I’m not sure if they have been entirely solved yet. But even if they haven’t, so what? It’s a relatively new discipline. You can’t expect every question to be answered with five minutes of the start of a new avenue of exploration. Especially since it’s closely related to the subject of brain science, which although having been studied far longer is still somewhat in it’s infancy. And because, not too many people want to have their brains examined too closely while it’s still functional, so we have to figure out less invasive methods of examination.

Memetics does make some fairly specific and clear claims, many of which are comparable to genetics, evolution, and natural selection - and others that are more related to brain science. But memetics isn’t just a theory, it’s also a subject of inquiry. I think you are confusing the two. The basic tenets of memetics are obviously true, glaringly so. The debate and study comes in the nuances of the mechanisms and principles involved.

No one can measure the exact frequency of a gene either. What’s your point?

Wow, harsh man lol. Of course there are results. Most of the results, obviously, are in the area of computer simulations, because we don’t have a really good way of examining the brain yet. But better diagnostic tools are being created each year, both mechanical/electrical and biological, and so each year we get more answers. And again, I think you are confusing theory with area of inquiry. Whether any particular tenet of a memetic theory turns out to be true or false, the questions are still there, and still need to be answered. A particular aspect of memetic theory may be proven or disproven, but memetics as an area of inquiry cannot be false or true.

I think here all you’re proposing is a distribution for information propagation, not the underlying mechanism for that distribution. I, for one, would have no particular distribution pattern in mind at all when examining the validity of memes.

As for falsifying natural selection, since NS is, in essence, but a logical syllogism, one need only demonstrate that one or more of the premises that lead to the conclusion are false.

To wit:

  1. If individuals in a population vary with respect to specific traits, and those variations are heritable, and
  2. If those variations confer differential survivability and reproduction, in a given environment, amongst individuals of the population, then
    C) The frequency of those favored variations in a population will increase.

So, as far as falsifying natural selection, one need only demonstrate that traits are not heritable, or that the variation is not heritable, or that if heritable, no differential survivability / reproduction is conferred by such variations. If everyone has an equal chance, for example, then natural selection cannot operate. If there is no variation, then NS cannot operate. If variations are not heritable, nor caused by an underlying heritable mechanism, then NS cannot operate.

Now, as a high-level process, I have no problem with ideas being subject to NS. Surely, ideas vary, and those variations will have differential “survivability”, in that some will persist and others will not, based on certain properties. No problems there. Where I do have problems is when that is linked directly to organic evolution, wherein we get such theories as humans were hard-wired for religion (that is, religious beliefs directly affected our biological, rather than cultural, evolution, for example), etc. I do not subscribe to the adaptationist school of thought, therefore I do not see all behaviors (nor ideas) as necessarily adaptive. It is, in fact, for many of the reasons that I reject evolutionary psychology that I reject meme theory (and this is fundamentally why I disagree with Dawkins on both memes and selfish genes).

That implies, it seems to me, that the choosing/decision making process is in some way external to the memetic environment, doesn’t it? Because, otherwise, it’s just a component of it affecting the likelihood of the propagation of a meme, which would mean the competition isn’t ‘forced’ at all (though I’m not exactly clear on why ‘forced’ competition would not result in memes being subject to the evolutionary algorithm, either).
This, then, seems to me to delve into the supernatural – essentially relegating consciousness into a ‘black box’ inaccessible to scientific reasoning (because again, I cannot see why else it wouldn’t merely be part of the environment a meme exists in that determines its chances for propagation).

What’s wrong with ‘a copyable piece of information’? True, that covers anything from the Declaration of Independence to the letter q to giving someone the finger, but why exactly is that bad? There’s no need to identify a kind of ‘atom of thought’ or whatever to have this definition be meaningful; it clearly distinguishes things that are memes from things that are not memes, at least as far as I can tell.

Well, yes. And that’s what I’ve been arguing. The memes themselves play no part in their success. I’ve asked a few times already how one can describe the action of meme transfer without resorting to analogy; that is, without saying they “want” this, or “try” that. How does an idea compete with other ideas, independent of the individuals housing those ideas?

By “forced”, I mean that there really is no competition, in the sense required for natural selection, at all! The competition occurs because individuals have to make choices. You can’t vote for both McCain AND Obama, after all (well, not legally, anyway).

But you have your own black box: “we don’t have the technology yet to examine memes”. So regardless who’s right, at this point, the brain is a big black box for all concerned.

It’s bad because then “meme” is nothing more than a synonym for “information” (all information is copyable, after all), in which case you might just as well stick with information theory. Memetics would therefore bring nothing new to the proverbial table, yes?

But that’s exactly what memeticists are striving to do (see jackdavinci’s posts, for example)!

I must have mis-interpreted your point that I had quoted.

I both disagree and I agree. I disagree that it can’t/shouldn’t be reduced.

I agree that it’s a thread in itself, but it’s absolutely critical to my point of view.

If the brain processes (high level or low level) arive at the same output given the input and mental state, then we can eliminate any discussion of “consciously valuing ideas” and think only in terms of mental state and application of input.

If the combination of mental state and input (meme) results in meme propagation, then that meme can be considered to be “more fit” than a meme that does not result in propagation.

This seems directly analagous to a physical environment and a sequence of amino acids. If the combination of the environment and the sequence of amino acids result in the propagation of the sequence of amino acids then that sequence can be considered to be “more fit” than a sequence that does not result in propagation.

I used the term “more fit” but really, all that is going on in either situation is that the combination of existing environment and some input introduced to that environment, may or may not cause the input to be replicated. We tend to be interested in those inputs that get replicated (in either situation) and less interested in those that don’t get replicated.

At different levels of abstraction there are clearly differences between gene propagation and meme propagation but both seem to simply rely on the combination of the environment and input into that environment.

They don’t, obviously. But you might as well ask: how does a gene compete with other genes, independent of the lifeforms ‘housing’ them? The genes themselves play no part in their success; the resulting phenotype does.

Take the paper flyer, for example: you want one that flies as far as possible, thus, instructions for creating better fliers are more likely reproduced than those leading to an inferior design, and there you’ve got competing memes being selected, none of which want anything or try any funny stuff.
The phrase ‘you want one’ above, by the way, is merely part of the description of the environment; it’s an artefact of us being used to thinking of ourselves as the actors rather than the background. Equivalently phrased, you might say that water ‘wants’ lifeforms well adapted to swimming.

All the more reason to utilize all attack vectors to crack that black box open!

I’d say that the realization that information evolves is something comparatively new; it’s not an idea unique to memetics, true, but in memetics, we apply that realization to the formation of human cultural tenets and how they influence both our selves and our biology – little as you might like that notion, I see no way around it; culture is just as much of an environment capable of producing selective (genetic) pressure as grasslands or cold climates are (see my earlier allusion to how the adoption of dairy farming in Western cultures has lead to a comparatively widespread ability to absorb lactose).
If you insist, though, the definition from this memetic lexicon includes the caveat that an information pattern is not a meme until it causes someone to replicate it.

I don’t actually think that this interpretation is very common – the meme itself can always be regarded as the smallest unit of information as it pertains to the selection process, but I don’t think there’s a general belief that all ideas ought to be constructed from some sort of fundamental unit-of-thought.

You said “Labeling the tune as a virus-like entity that seeks to reproduce itself” which seems to imply, to me, that the quality which you refer to as “virus-like” is that it “seeks to reproduce itself”. Perhaps I parsed your sentence differently than you intended. But if this is not correct, than what are the qualities of viruses you are claiming are attributed to memes?

But this hypothetical falls down because a meme doesn’t actually map to any brain input, even in an indirect way. What particular sights and sounds can you associate with a meme? None, because it’s conceptual.
Therefore the state system analogy, with memes as inputs, just doesn’t work.

##########################

I’d like to summarise my position, because I don’t want to come across as resistant to the whole idea, or as being anti-Dawkins, who I actually think is a fantastic biologist and writer.

In my opinion, memetics is too vague to be a scientific hypothesis. It makes no clear predictions, and is un-falsifiable.
So it’s perhaps just a metaphor. But it’s a metaphor that I can’t see the use of. What does memetics explain (unambiguously) that can’t be explained without it?

Perhaps at some future time, someone will formulate a more clearly defined version of this concept.

Unambiguously? Nothing; no hypothesis ever does, else it’d be a fact. But it provides, at least in my eyes, probable explanations for a number of issues that are, to date, pretty large open questions:

First of all, it provides a background framework for a dual inheritance theory, i.e. a link between cultural and biological evolution, which has been proposed independently of and I believe also prior to memetics. By identifying the unit of selection on the cultural side of this link, mutual influences are likely to be easier quantifiable than they are at present.

Second, if memes undergo a selection process in an individual’s brain, this sheds light at the question of how exactly we arrive at a decision, and, since decision is the fundamental process of (intelligent) action, how our intelligence works.

Third, and to me, that’s the biggie, because it’s so easily overlooked, yet rather fundamental: if memetics is correct, it might just give us a handle to answer the question of why we are conscious. See, the way it seems to me, we wouldn’t need to be – each and every human action can just as well be undertaken by an unconscious agent. In theory, one can easily envision a table that lists all possible inputs and correlates them with their appropriate outputs; this design can be enhanced, but for the purposes of argument, its mere existence is enough. One could then devise a mechanism that merely looks up the input and gives the appropriate output; I don’t think anybody would argue that this system would be conscious, and yet, it’d be completely able to reproduce any given action a human being can conceivably take.
So, that’s the question: Why are we conscious if it doesn’t seem we need to be?
Now, how does memetics enter the issue? Truth be told, I’m not entirely sure; my reasoning isn’t absolutely sound. Nevertheless, my hunch is that consciousness might be to memes similar as life is to genes – it provides a powerful vector of reproduction. But how, if, as I just pointed out, having or not having consciousness doesn’t affect our actions? Well, it might not affect atomic actions; but it might affect their frequency and grouping, i.e. certain actions become more likely to occur with greater frequency, or in conjunction with other actions, depending on memetic relationships.
This is all very vague, I know, and I don’t really have time right now to elaborate, nor am I even certain my reasoning leads to anything fruitful, but as I said, I believe it’s enough to continue memetics as, as jackdavinci put it, an area of inquiry.

As for ‘not producing any predictions’ – if memetics is indeed correct, then its predictions are basically our whole sphere of experience, which makes it kinda hard to determine any specific – and presently unknown – effects; if those indeed exist, they’d have to be pretty subtle to not be known by now.

And again to address the criticism of being too vaguely defined, take, for instance, a table – everyone knows what it is, but still, a general definition, based on its physical properties, is elusive: the number of legs seems to vary between 0 and n, it can be made out of nearly every material, comes in every imaginable colour and with every decorative pattern imaginable, varies in shape from triangular to circular, and so on and on. It doesn’t seem like a simple, all-encompassing definition of ‘table’ exists, since for nearly every definition I can think of, I can immediately find a counter-example; yet nobody would argue that the label ‘table’ is, therefore, useless.