The Meme Machine by Susan Blackmore

This episode and both voices are AI-generated. Eleanor is an AI-generated interpretation of Susan Blackmore's published arguments, not Susan Blackmore; our fictional sponsor is PocketKettle Mini, and details may be hallucinated or wrong, so please double-check anything important.

Today we're briefing The Meme Machine, Blackmore's ambitious claim that culture is not merely something humans do. It may be an evolutionary system with its own interests, which is a fairly rude thing for an idea to say about its hosts.

The book's central proposal is that unusually capable human imitation created a second kind of replicator: the meme. A meme is copyable cultural information—an instruction, practice, story, skill, or idea—that persists through variation, selection, and retention.

And crucially, its popularity does not certify it as true, kind, useful, or good for human reproduction.

Exactly. The question becomes: what gets copied onward through limited attention, memory, conversation, institutions, and media? That perspective can make cultural success look very different from human benefit.

First, we'll get the actual argument straight: memes, imitation, and the proposed feedback loop between culture and genes. Then we'll pressure-test it through language, fertility, altruism, religion, and the machinery that makes copying easier.

Finally, we'll take up the book's most unsettling extension: the self as a memeplex, rather than a little commander inside the head. We'll also separate the framework's useful insight from the parts that remain speculative or contested.

Good. Let's begin with the term that has been flattened into pictures of distracted cats: what does Blackmore mean by a meme?

Let’s put the internet-joke meaning of “meme” aside. What is the book’s actual claim?

My argument in the book is that human culture may contain a second evolutionary replicator: the meme. A meme is copyable cultural information, such as a skill, practice, tune, story, or idea, that is transmitted through imitation.

Second replicator means something stronger than “people influence one another,” yes?

Much stronger. For Darwinian evolution, there must be variation, selection, and retention: copies differ, some copies get transmitted more than others, and enough of their information survives to make further copying possible.

And that sequence can happen outside biology?

That is the Universal Darwinism premise. The logic of evolution is not restricted to DNA or carbon-based organisms; it can operate wherever a replicating system has those three properties.

But genes are physical sequences with known machinery. Are memes meant to be tiny cultural genes floating around in hats?

No. The book does not claim that memes and genes have identical bodies or mechanisms. A meme may be information distributed across brains, speech, writing, and other media, provided it is preserved well enough to be copied again.

So the unit can be fuzzy. A recipe, perhaps, rather than one perfectly bounded particle?

Exactly. Its boundaries are often uncertain, which is a real problem for the theory. But a recipe can retain a recognizable structure while different cooks alter ingredients, quantities, or wording.

Give me the simplest example of variation without cultural chaos.

Think of an urban legend retold by many people. Details change, but a recognizable story persists because versions that are memorable and tellable gain another chance to be passed along.

And the winning version is not necessarily the accurate one.

That distinction is central. Memetic success means successful onward copying, not truth, moral worth, usefulness to an individual, or benefit to the carrier’s genes.

That is the book’s slightly chilly shift in viewpoint: stop asking whether an idea helps us, and ask whether it helps itself spread.

Yes, though “itself” is shorthand, not a claim that ideas have intentions. The meme’s-eye question is: among limited attention, memory, and communication opportunities, which cultural variants gain access and reproduce?

What kinds of traits might win that contest?

Memorability, emotional force, apparent usefulness, repetition, and easy communication can all matter. But the framework asks us not to assume beforehand that popularity reflects a benefit to people.

Single ideas rarely travel alone, though. They arrive with rules, stories, rituals, and a request to rearrange your weekend.

That is what the book calls a memeplex. It is a mutually reinforcing package of memes—ideas, practices, texts, norms, and institutions—that helps its components be retained and transmitted together.

So a memeplex is more like an ecosystem of cultural instructions than a single slogan.

That is a good way to put it. The components can support one another: a text preserves a doctrine, a norm encourages repetition, and a practice gives people occasions to pass both on.

Now, why make imitation the hinge of the whole story? Animals learn socially all the time.

The book distinguishes loose social learning from true imitation. An animal may be drawn to the same food, place, or result as another animal without copying the detailed actions that produced it.

What counts as true imitation here?

It means reproducing the behavior or procedure itself with enough fidelity that complex cultural information can accumulate. Seeing someone make a difficult tool and learning the sequence of actions is more relevant than merely noticing that tools are useful.

And Blackmore’s proposal is that humans crossed this threshold unusually far.

Yes. Once highly capable imitation appeared, copied practices could accumulate across people and generations, creating a new evolutionary process alongside genetic evolution.

That gets us to the large claim: culture did not just decorate an already evolved human brain. It helped remake the conditions under which brains evolved.

That is memetic driving. Useful or contagious cultural practices create a memetic environment in which individuals better able to imitate, learn, communicate, and perhaps attractively display those abilities may have reproductive advantages.

Walk me through the causal chain slowly, because this is where a metaphor can put on a lab coat.

First, imitation permits cultural variants to proliferate. Second, a growing stock of transmissible skills and conventions makes copying ability more valuable; third, genetic selection may favor capacities that improve participation in that cultural system.

So genes build better copiers because memes have made copying consequential.

Potentially, yes. Better imitators can acquire and transmit more culture, while better learners and speakers may thrive in a world increasingly structured by transmitted information.

The book connects that to our unusually large and expensive brains.

It offers memetic driving as a possible contribution to human brain enlargement and language. Human brains are metabolically costly, slow to develop, and carry obstetric costs, so the book argues that familiar explanations such as foraging, toolmaking, and social intelligence may not fully settle the puzzle.

What is the evidence offered, rather than the desired ending?

Difficult stone-tool manufacture is treated as suggestive evidence that imitation was operating by roughly 2.5 million years ago. But the archaeological evidence is indirect, and earlier perishable practices would not leave fossils.

Language leaves even less of a fossil trail, inconveniently.

Quite. The proposal is that language improves cultural replication through discrete words, grammar, vocal transmission, and eventually external records, making memes more faithful, more numerous, and longer-lived.

In practice, language gives an instruction a better chance of surviving the trip between brains.

Precisely. It permits complex instructions to be copied with more structure than many gestures alone, and it expands the opportunities for cultural variants to compete.

The book even treats a great deal of talking—gossip, conspicuous verbal skill, perhaps attraction to articulate people—as part of this memetic niche.

It does, as hypotheses. If speech carries memes, then talkativeness and preferences for capable speakers might be influenced not only by direct genetic advantages, but by selection pressures generated within culture.

That sounds like ordinary gene–culture coevolution. Where does Blackmore insist on a sharper break?

The key break is independence of interests. Gene-centered accounts explain learned behavior when learning ultimately helps genes reproduce, but memetic evolution allows cultural items to spread even when they do not serve genetic fitness.

And the Baldwin effect is the contrast case?

Yes. In a Baldwin-style account, learning changes which genes are favored because learned flexibility helps organisms survive and reproduce; genes remain the ultimate beneficiaries. Memetic driving proposes that copied culture can itself impose selection pressures without that guarantee.

So culture can become more than an employee of reproduction. It may become a demanding co-tenant with its own copying agenda.

That is the book’s intended provocation. Memes may initially exploit human preferences shaped by genes, yet then spread in directions that genes would not select for.

We will need concrete cases for that, because otherwise every human oddity can be declared a meme in a trench coat.

Agreed. The framework earns its keep only if it distinguishes transmission dynamics from genetic, psychological, social, and economic explanations, rather than relabeling whatever is already common.

Before we test it, give the listener the cleanest one-sentence map of the argument.

Unusually powerful human imitation may have launched a Darwinian process of cultural replication; those replicating cultural systems then altered human environments and possibly genetic selection, while spreading according to transmissibility rather than our truth, welfare, or reproductive interests.

And the status of that map?

It is a forceful, wide-ranging, and contested proposal, not a consensus theory of cultural evolution. The next question is whether its mechanisms explain actual cultural patterns better than the alternatives.

We have the basic claim: copied culture can evolve on its own terms. Now make it earn its keep—how does a meme spread without simply riding on family reproduction?

The key distinction is vertical versus horizontal transmission. Vertically, parents pass practices to children; horizontally, peers, schools, media, and now communication technologies pass them across unrelated people.

So a cultural item can gain hosts who do not share its previous host's genes. That is the escape hatch from ordinary genetic advantage?

It is the proposed escape hatch. If transmission mainly follows kin and increases descendants, gene-centered explanations may do much of the work; broad horizontal copying lets a practice spread even when it does not increase its carriers' reproduction.

Give me the uncomfortable practical examples, but do not turn every child-free adult into proof of a rogue idea parasite.

Quite right. Contraception, voluntary celibacy, adoption, and career-centered childlessness show that culturally common choices need not straightforwardly maximize genetic output, but none alone demonstrates independent memetic selection.

The Fore funerary-cannibalism case is sharper because a transmitted practice had grave health costs for participants. Still, ecological, social, institutional, and historical explanations remain possible alongside a memetic one.

Then what specifically does the memetic account add, rather than just saying culture sometimes beats biology in a bar fight?

My argument in the book is that the cultural item itself becomes a candidate for explanation. Ask what features give it access to attention, memory, teaching, prestige, and repeated performance, rather than asking only why its human carriers benefit.

And this is where you separate yourself from the Baldwin effect.

Yes. In a Baldwin-effect story, learning changes which genes prosper, but learning is ultimately retained because it serves genetic success.

Memetic driving makes the stronger proposal that accumulated cultural practices alter selection pressures in ways with no guarantee of genetic benefit. Genes may then be selected to cope with, learn, or reproduce the culturally created environment.

That sounds grand enough to swallow human evolution whole. Where do brains and language enter the machinery?

Once imitation allowed useful practices to accumulate, better imitators and learners would have had access to a growing cultural store. That could favor more capable brains, despite their heavy metabolic and developmental costs.

Language then makes copying more exact, abundant, and durable. Discrete words and grammar can preserve instructions better than repeatedly watching someone improvise a complex action.

Instruction-copying is doing a lot of work there. What does it mean in practice?

An instruction can reproduce a complex product through many people: a recipe, a tool-making procedure, a ritual, or a technical method. The product may vary, but the transferable directions can persist and generate more copies.

And external media turns minds from the only storage devices into a much bigger copying factory.

That is the trajectory the book emphasizes. Writing, printing, digital storage, and networked communication increase retention, reach, and replication speed, so cultural evolution need not depend on fragile memory alone.

Does the book say the Internet proves memetics?

No. It illustrates copying machinery with remarkable fidelity and fecundity, but it does not settle whether the copied units meet every requirement of a Darwinian replicator theory.

Back to early humans: difficult stone tools are presented as a clue that imitation was already operating. How firm is that clue?

It is indirect. Complex stone-tool manufacture is treated as suggestive evidence for imitation by roughly 2.5 million years ago, while earlier practices in wood, gesture, or social routine would leave little or no archaeological trace.

Language does not fossilize either. So the enlarged-brain account, and claims about selection for articulate partners, remain hypotheses about a possible memetic niche rather than recovered evolutionary history.

What about talk itself? Gossip and conspicuous speech can look like ordinary social bonding, status competition, or mating behavior.

They can, and those alternatives matter. The memetic angle asks whether frequent speech also gives speech-borne memes more chances to reproduce, and whether preferences for fluent speakers can amplify that process.

Fine. Now altruism: the book's “altruism trick” sounds suspiciously like kindness has been recruited into marketing.

The sequence is simple. Altruistic people may become liked or popular; popular people become more influential models; copying them can spread both their generosity and arbitrary beliefs, habits, or symbols attached to them.

Why not stop at reputation and reciprocity? Generosity earns trust, trust earns cooperation, end of story.

Those are established kinds of evolutionary explanation, and the book does not deny kin selection or reciprocal altruism. Its additional hypothesis is that copying admired people can transmit a larger cultural package, including elements with no obvious payoff to copier or model.

But the critical bridge is weakly evidenced in the book: whether altruism reliably produces imitation was little studied. Without that link, the trick is an intriguing mechanism, not a demonstrated general explanation.

So generosity might spread because it is useful, because it is admired, because institutions reward it, or because several forces overlap. The meme lens is not a monopoly license.

Exactly. Its value depends on generating predictions that distinguish those routes, rather than merely renaming any successful norm a meme.

Religion is the larger test case. What makes a belief system a memeplex in this account?

A durable religious memeplex can combine existential answers, emotionally forceful ideas, ritual, identity markers, authority, sacred texts, institutions, and incentives to recruit others. Costly practices can also mark commitment and bind participation.

Anti-doubt defenses matter because they protect transmission when claims meet criticism. The point is not that every religious participant has one motive, or that religion has no social benefit, beauty, or insight.

And a transmission explanation does not tell us whether its claims are true. Important, because “it spreads well” is a terrible fact-checking method.

Precisely. The same analysis is applied to paranormal, New Age, and alternative-health beliefs: vivid experiences, promised care or control, truth claims, community, and resistance to disconfirmation can aid circulation, without proving such beliefs false.

Could the same suspicious lens dissolve science too? Laboratories, journals, credentials, and textbooks also look wonderfully replicable.

Science is also a memeplex in this framework. The book's asymmetrical judgment is that science is superior at its best because it builds in testable prediction, criticism, and correction instead of making claims immune to challenge.

“At its best” is carrying the sober part of that sentence. Institutions can preserve errors as well as correct them.

They can. Memetics explains neither truth nor error automatically; Blackmore's preference for science rests on methods that permit correction, not on its popularity or institutional durability.

Let me test the full causal picture. Horizontal transmission and better copying technology create opportunities; language sharpens copying; admired models and organized institutions steer attention; then packages defend and reproduce themselves.

That is the operational core. It connects personal imitation, social prestige, material media, and institutional storage into a competition for finite channels of memory and communication.

But a package spreads because it is useful, coercively imposed, emotionally satisfying, profitable, or historically lucky as often as because it has some abstract replication advantage.

Yes, and those need not be rivals in every case. A memetic explanation should specify how usefulness, coercion, emotion, profit, or luck become transmission advantages, then show why that account predicts more than a conventional psychological or social explanation.

So the framework is strongest as a change in what we inspect: not just what people choose, but what systems of copying make likely. Next, we should ask whether that shift produces a real science or a very adaptable story.

We have treated memes as competitors for attention, memory, and copying channels. The book ends by aiming that machinery at the person doing the copying, which is either clarifying or a tiny bit rude.

My argument in the book is that the ordinary self is an especially powerful memeplex, or selfplex. It is a cluster of stories, concepts, memories, and habits organized around the sense that there is one continuous inner owner.

So this is not merely saying that identity is socially shaped. It is saying the feeling of being a single captain may itself be a cultural construction with replication value?

That is the stronger proposal. When ideas become mine—my beliefs, my group, my possessions, my character—I am more likely to defend them, explain them, argue for them, and pass them on.

The selfplex, then, gives memes a very committed public-relations department. But what makes it more than a clever metaphor?

The book combines the memetic proposal with a naturalistic challenge to a central inner controller. Brain injury, stimulation, and drugs can alter self-experience, which makes an independently fixed inner owner difficult to locate.

That establishes dependence on brains. It does not yet establish that there is no self, does it?

No, and that distinction matters. Neural dependence strongly pressures dualist pictures in which an immaterial self simply directs the brain, but it does not by itself settle every theory of consciousness or agency.

What is the next step in the book's case?

It draws on distributed, parallel accounts of brain function and Dennett's critique of the Cartesian Theatre. There may be no single place where a fully formed experience is displayed to an inner observer, and no central executive that initiates every action.

In practical terms, many processes are happening, and the brain tells a coherent story afterward?

Roughly. The claim is not that nothing happens or nobody acts, but that action and experience arise from interacting brain processes, genes, memes, and environments rather than from a separate commander inside the system.

And the readiness-potential experiments enter here, because they appear to put neural preparation before the reported conscious decision.

Libet-style findings are consistent with that ordering: detectable neural activity can precede a person's reported awareness of intending to move. The book treats this as support for the idea that conscious intention may often follow preparation rather than cause it.

Consistent with is doing useful work there. Those experiments do not close the free-will department forever.

They do not. Their interpretation remains contested, and the book does not offer a complete theory of subjective experience; it advances a provocative account of how the self-story may be constructed and maintained.

Why call this self-story a memeplex rather than simply a brain-generated narrative?

Because the book adds a selection claim. A self-concept can bind together language, social norms, ownership, reputation, autobiographical memory, and beliefs about choice in ways that make attached cultural material easier to protect and retransmit.

That could explain why a disagreement about an idea so quickly feels like someone stepped on your foot.

Yes. If a claim is woven into identity, defending it can feel like self-preservation, while also giving the claim more airtime and more opportunities to copy.

Does this leave us as helpless broadcasting equipment, with a nervous system attached?

The book rejects that fatalistic reading. It suggests that noticing thoughts as events, rather than treating each as an expression of a solid owner, can weaken automatic identification and create more room for attention and response.

That is where Buddhist no-self thought becomes relevant. Is the recommendation basically meditation?

Present-moment attention is offered as a practical route, not as medical advice or a demonstrated universal cure. The point is to reduce compulsive attachment to the selfplex while retaining compassion and moral concern.

So no permanent soul is required for kindness. That may be the book's most emotionally consequential claim.

It is consequential, but it remains partly philosophical and introspective. The book argues that morality need not depend on a separate inner agent, not that science has conclusively dissolved every meaningful account of responsibility.

Let us zoom back out. After brains, language, religion, altruism, and the self, what is the strongest idea that survives even if the full memetic theory fails?

The durable insight is that cultural success need not track truth, human welfare, or genetic fitness. Beliefs and practices can spread because they exploit attention, fit institutions, recruit advocates, or travel well through available media.

That is a useful warning against reading popularity as a merit badge. But do we need literal Darwinian replicators to make it?

Not necessarily. Psychology, sociology, history, economics, and ordinary network effects can all explain transmission patterns; the distinct memetic ambition is to treat cultural items as a second evolving replicator with its own selection pressures.

And that ambition carries some very stubborn unresolved questions. What exactly is the unit that replicates?

Precisely. Meme boundaries are uncertain: is the meme a sentence, a tune, a practice, an instruction, or a changing family of versions, and how is it physically stored and copied across brains and media?

Genes have comparatively clear material machinery. With memes, fidelity seems squishier, which is a technical term I am prepared to defend.

It is the right concern. The book distinguishes the abstract requirements of variation, selection, and retention from genes' particular biology, but a robust theory still needs to show sufficient fidelity, inheritance, and identifiable replication mechanisms.

An urban legend can remain recognizable through altered retellings. Is that enough retention?

It shows that some cultural information persists amid variation, which is suggestive. It does not alone demonstrate that memes are discrete Darwinian units, or tell us which version, feature, or transmission rule is being selected.

There is also a danger of explanations that always arrive after the party. A practice spread, so we say it was good at spreading.

That is the major methodological problem. A transmission story becomes scientifically stronger only when it makes discriminating predictions that beat genetic, psychological, social, ecological, economic, and historical alternatives.

Give me the standard we should apply. Suppose a fertility-reducing practice becomes widespread.

That fact alone does not prove memetic selection against genetic interests. One would need evidence about the routes of transmission, the incentives and constraints involved, competing explanations, and whether traits that enhance copying predict spread better than those alternatives.

The same restraint applies to religion, paranormal belief, and alternative health claims. Explaining why an idea travels does not tell us whether it is true.

Exactly. The book's analysis of ritual, authority, identity, texts, recruitment, and resistance to doubt explains possible transmission advantages; it neither verifies nor falsifies the belief's content.

And the comparison with science is not that science is meme-free. It is that at its best, science builds correction and testable prediction into its transmission system.

That is the book's asymmetrical judgment. Science is also a memeplex, but its norms can expose claims to revision rather than principally shielding them from challenge.

Where should listeners be most skeptical of the book's big applications?

Treat the accounts of brain enlargement, language evolution, articulate mate preferences, and media-driven fertility change as hypotheses, not confirmed histories. The archaeological evidence is indirect, language does not fossilize, and several forecasts and empirical discussions are historically situated.

The altruism mechanism also needs its bridge: generous people become popular models, then people copy them and their attached package. Was that bridge firmly demonstrated?

No. The book presents the altruism-to-popularity-to-imitation sequence as a promising mechanism, while acknowledging that direct evidence for the crucial imitation link was limited.

And the self chapter inherits the same calibration: intriguing synthesis, contested experiments, no finished account of consciousness. It is not a warrant to announce that choice has been deleted.

That is a fair verdict. Its value is to make the familiar self less automatic and to ask how cultural narratives participate in agency, rather than to deliver a final laboratory proof that persons are unreal.

Who should read the full book closely?

Read it closely if you want a bold, cross-disciplinary defense of Universal Darwinism applied to culture, or if you are working through cultural evolution, consciousness, identity, language, or Buddhist no-self questions. The book is most rewarding when you want to examine an ambitious framework, including where it overreaches.

Who should consult only selected stretches, rather than marrying the entire meme machine?

Readers chiefly interested in language and brain evolution, religion and belief systems, or agency and selfhood can profit from those focused discussions while keeping the core replicator claim in view. They should also bring more current research, especially to the empirical controversies.

And who has probably learned enough here to move on without guilt?

Anyone who mainly needed the central lens and its caveats can reasonably stop here: cultural spread is not a verdict on truth, goodness, or genetic benefit, and calling culture Darwinian requires much more evidence than vivid examples.

What would the complete book add beyond this briefing, then, besides several opportunities to argue with it in the margins?

It adds the sustained conceptual case, extended examples, interdisciplinary comparisons, proposed tests, detailed applications, and the full path from imitation to the selfplex. It does not conceal a missing conclusion or supply certainty; it gives you more material for judging whether memetics deserves to become a real explanatory science.

Let’s land this without pretending we solved culture between two microphone checks. What is the book’s central move in one clean sentence?

My argument in the book is that unusually capable human imitation may have created a second evolutionary replicator: memes. Cultural items can then compete to be copied, with success measured by transmission rather than truth, human welfare, or genetic fitness.

And that is bigger than saying culture influences us, which nobody needed a book to discover.

Yes. The stronger claim is Darwinian: if copyable information varies, persists sufficiently, and is selected for onward copying, culture can evolve in its own right. Genes and memes need not have identical physical machinery for that abstract logic to apply.

The practical lens was: stop asking only whether an idea helps people, and ask how it captures attention, memory, and communication channels.

Exactly. A memeplex is a mutually reinforcing package of ideas, rituals, texts, norms, and institutions that improves those copying opportunities. That helps explain persistence, but it never establishes that the package is true.

Then comes the audacious bridge: memes do not merely ride on big human brains; they may have helped select for them.

That is memetic driving. Accumulating practices could favor better imitators, learners, speakers, and perhaps preferred mates, contributing to costly brains and language; but the book presents this as a testable evolutionary hypothesis, not settled prehistory.

The cases made the tension vivid: peers, schools, media, writing, and digital networks can spread practices without asking whether their carriers produce more children.

Right. Contraception, celibacy, adoption, career-centered childlessness, and harmful inherited practices show why cultural prevalence cannot automatically be read as genetic advantage. But none of those cases, alone, proves an independently selected meme.

And the same restraint applies to altruism and religion: a generous person may become a model, while rituals, identity, authority, and anti-doubt defenses can stabilize belief systems. That is a transmission account, not an automatic debunking machine.

Science, too, is a memeplex in this framework. The book’s important asymmetry is that science is better at its best because it builds in testable prediction and correction, rather than treating resistance to challenge as a virtue.

At the far end, the book turns the lens on the person doing all this copying.

The selfplex treats the ordinary sense of a continuous inner owner as a powerful narrative cluster, shaped in a biological brain and useful for defending identity-linked beliefs. Brain dependence, distributed processing, and readiness-potential studies pressure the idea of a central inner controller, but they do not conclusively settle consciousness or agency.

So the durable insight is modestly disruptive: culture may optimize for spread, not for us. The weak point is that a clever spread story can be fitted after the fact unless it beats genetic, psychological, social, economic, and historical rivals with real predictions.

That is the right verdict. Meme boundaries, copying fidelity, storage, inheritance, and the evidence for key causal links remain unresolved, so memetics is a provocative framework rather than a confirmed science.

Who gets the most from the complete book?

Read it closely if you want a bold, cross-disciplinary case connecting Universal Darwinism to culture, language, religion, consciousness, and Buddhist no-self thought. The full book adds sustained argument, detailed examples, rival comparisons, and proposed tests—not a hidden conclusion or a final certainty.

And who can economize?

Consult selected parts if your interest is mainly language and brain evolution, belief systems, or agency and selfhood. If you chiefly wanted the central model and its cautions, this briefing is enough: treat transmission as something to explain, not as a stamp of truth, goodness, or inevitability.