The new pseudoscience of memes

A review of Thought Contagion by Aaron Lynch

I snapped this book up as soon as I saw it. I love the idea of memes, which Richard Dawkins introduced in The Selfish Gene, and which Lynch says he independently invented. And who wouldn't? The notion that ideas can reproduce-- like genes-- not for our benefit but for theirs is simultaneously chilling and fascinating, and I was eager to see what twenty years of thought on the subject had produced

Very little, it turns out. Lynch's book is deeply disappointing; what memetics has chiefly generated is a new way to blather about society, sex, and politics, without rigor and without the slightest need to make sense.

Lynch worries a bit in the introduction about this reaction; he pleads for a little patience-- he wants to show what sort of exciting ideas memetics can come up with, not get bogged down in factual nitpicking.

What he doesn't see is that the real problem is not just that he gets facts wrong. It's that he's developed memetics into a scheme for generating factless scenarios. By excluding rigorous analysis, testing, and verification from his methodology, he's simply refined an ability to tell just-so stories about social behaviors.

Generally you can take Lynch's stories, and with further analysis come to completely opposite conclusions. What this should tell us is that this sort of memetics has no value at all, except to reinforce our prejudices with what sounds like science.

I'll give some examples-- many, many examples-- and then offer some thoughts on why memetics has caught on despite its essential vacuousness.


The citations inblue italics below are from Lynch's book, at the page numbers specified. The boldfacing is mine, to indicate topics discussed.

Although men can sometimes pass their genes along by copulating with and deserting women, they never pass their memes along this way. (48)

They don't? Not even by infecting the mother with them? Isn't it, for instance, a fairly effective way to pass on the meme for lovin' and leavin' women?

Polygyny memes fare much worse at proselytism. When men look for a mate, they prefer to see competing men take as few women as possible. So regardless of how many wives they want for themselves, they would rather tell unrelated men to limit themselves to one. (49)

Odd, then, how polygyny spread so successfully in Africa, in the Islamic world, and China.

Would polygynous men really feel a motivation to keep their ideals to themselves?

The dolls [representing heroes] help boys fantasize about going out in the world and conquering, a role that broadly resembles what they must do to attract a mate and provide for children later on. (57)

In what way does farming or factory or office work resemble "conquering"? And how exactly does going "kapow! you're dead" with an action figure resemble attracting a mate? And while we're at it, when historically did boys start playing with hero dolls?

Prominent examples... are the personality memes of 'girlish helplessness' and 'manly responsibility.' (61)

I hope Dave Sim never reads the exposition that follows-- not least because it never occurs to Lynch that our society is also filled with memes of womanly practicality and manly irresponsibility.

If... the [gay rights] movement lets most homosexually inclined people live homosexual lives, their average fertility rate comes thundering down... Progay beliefs replicate less proselytically at that point.... the movement dwindles, especially when the life span of prior adherents has passed. (83)

Lynch is predicting a cycle of tolerance and oppression of gays. Is there any evidence that such a cycle exists?

The existence of homosexuality is a challenge for both memetics and genetics; it seems so obviously a counter-reproductive strategy. Lynch doesn't really do worse than the evolutionary biologists here; but he doesn't do any better, either. The fact that homosexuality is so widespread, and is found in the animal kingdom as well, suggests that any prediction that homosexuality will decline is probably wrong.

Frankly, the scientific, unbiased study of homosexuality is so new that pundits should probably drop the subject for a generation. People's thoughts are still too full of moralisms, stereotypes, false assumptions, and cultural baggage-- we don't even know what the facts are, much less what the explanations are. We don't know if there's a "gay gene", or a spectrum of orientations; we can't agree on who is a homosexual; we have too little experience with "progay beliefs" to predict their future.

The weak point in many a genetic or memetic analysis of homosexuality, I think, is the assumption that it discourages reproduction. It sure seems like it would, doesn't it? And yet in the real world, gays and lesbians often have children. The theorists forget that humans indulge in a whole lot more sex than childbirth. For the species to reproduce, each of us needs to have an average of one viable offspring. That's nothing compared to the thousands of times we're going to have sex.

Historically, most gays and lesbians (whatever exactly it means to project these categories into the past) were probably required to participate in heterosexual marriages, and reproduced about as well as hets. And today, you can hardly fling a brick in the lesbian subculture without injuring a lesbian mother. And the natural father may well be a gay friend.... whose partner, an older bear, has a couple of children from a previous marriage...

The discussion of "proselytizing" gays is also rather disturbing. I don't know any gay friends who were talked into becoming gay. They describe it as discovering or realizing something about themselves.

Actively proselytizing men also try harder with attractive partners, so one expects disproportionate fractions of both attractive men living homosexual lives and attractive women living heterosexual lives. After the effect causes selective depletion, remaining populations of actively heterosexual men and homosexual women should contain proportionately fewer attractive people-- at least to men's eyes." (85)

Oh. My. God. Lynch is apparentely saying that dykes are ugly (and gay men are beautiful), and explaining why. Argh! Does he know any lesbians? It's just sad to see someone coming up with a pseudo-scientific basis to locker-room b.s.

We had might as well wonder why, if men like attractive women, we don't end up with all women being attractive. It may help to remember Lake Wobegon, where all the children are above average... Half of all women must, by definition, be of less than average attractiveness. If the cuter half reproduces more (and the reader might, for practice, come up with an argument or two why this might not happen), then the standard of cuteness will get more rigorous, leaving half the female population still below average.

What does all that have to do with sexual orientation? Nothing at all, very likely. Men cannot conspire to keep all the hot women heterosexual any more than they can conspire to make all women attractive. Furthermore, looking good (itself a cultural construct) is only one factor in choice of mate, even for men; class is surely many times more important.

Although hips much wider than waist are also distinctively female, they make subtler candidates [than breasts] for fetish fashion statements and take more words to describe. (85)

Then why did ancient societies have a hip fetish? Jeez, for this guy life is 20C America.

So women's body-fat percentage now correlates with their age... By thus favoring young women, the lean-partner preference out replicates the fat-partner preference in modern, well-fed societies. (87)

But young women (and young people in general) are increasingly fat today!

The masturbation taboo earlier illustrated quantity parental replication. Taboo adherents [were pressured toward] more frequent vaginal sex. (90)

And so on, a whole paragraph which fails to confront the fact that this taboo is a spectacular failure. Do all the benefits Lynch claim for the taboo have any relevance if it is not in fact observed?

The memes against birth control... offer the clearest examples of the quantity parental effect. By raising extra babies, followers of these memes can outpopulate nonhosts across various times and places. (91)

Lynch has forgotten Ecology 101. Why don't all animals follow the rabbit strategy? Because increased parental investment is often a better strategy than increased fertility.

This error-- the notion that populations with high birth rates will outreproduce those that use birth control-- undermines Niven & Pournelle's The Mote in God's Eye as well. Reality, as usual, is more complicated. High birth rates almost invariably correlate with high death rates... and with some shocking attitudes toward human life.

Read Marvin Harris on what happened before effective birth control. Executive summary: untrammeled reproduction was never a good strategy, except for populating new territories. What people have always done, as opposed to what Lynch and Niven think they should do, is control their fertility, by hook or by crook-- to the point of widespread infanticide, if safe abortion is not possible.

Memes specifying vaginal intercourse for all sex also gain moral prevalence by making adherents have children. Memes directly forbidding oral and anal intercourse increase the reproductive advantage still more by eliminating shades of doubt about how to act. (93)

Or perhaps they decrease the reproductive advantage by making sex less fun and thus less frequent. Or perhaps oral and anal sex were understandably deprecated in times of poor hygiene.

This failure to consider alternative effects really ruins memetics as science. Instead of being a tool for telling us what can happen to an idea and what ideas will prevail, it becomes a crutch for moralism, a tool-- rather like deconstructionism-- for offering facile, factless explanations of why things are as they are.

An indicator of its historical force, the taboo against sex talk apparently affects the wording of the standard marriage vows. (94)

When Lynch does show an awareness of history, he still forgets to check what's actually in it. Here, he might discover that this particular taboo is rather modern; a few centuries ago, C.S. Lewis reminds us, a quite respectable woman could speak in ways that only a "completely abandoned" one could in later times.

Almost any belief in a powerful spirit motivates conversion of family and friends. (98)

Does it? Some beliefs don't proselytize much. And perhaps some beliefs are like systems for beating the lottery-- best kept to oneself.

While we're at it, once a system of belief has spread throughout society, there isn't much point in proselytism; so wouldn't the memes for proselytism disappear through non-use? (Genes can be stored unused for future activation; but the most useless parts of a belief system are going to be subject to erosion.)

In two of the simplest examples of self-propagating faith, the Genesis commandments to "Be fruitful and multiply" and "Populate the earth abundantly and multiply in it" stimulate their own proliferation....

10,000 years ago: ...New terms like add and multiply enter the lexicon. (100)

Lynch is attempting to tie these Biblical phrases to the development of accounting.

Oh my. A quick glance at the concordance shows that the verb in the Hebrew is râbâh, a primitive root meaning "increase" or "grow"; it's translated by many English words, including enlarge, excel, exceeding, full of, be great, grow up, heap, increase, be long, much, multiply, nourish, plenty.

This is only the silliest link in a long chain of reasoning whose chief characteristic is its unreality. Do people need the encouragement of memes to reproduce? Didn't our genes take care of that adequately? Historically, given such factors as the lack of birth control, have ideologies favoring reproduction had any effect at all on fertility?

About the only supporting argument you could make is to drag in the Amish, who carry on a pre-modern reproductive rate into modern times, and thus have increased their numbers tenfold in the last century. That might be an effective strategy for them today, but in previous centuries it would arguably have had no effect at all: lots of childbearing was countered by lots of infant and child mortality, and belief probably had little to do with either rate.

By the way, some relevant factoids (from Salon, I think): China has 118 boys for every 100 girls under age 5; in the northern states of India, the boy to girl birth ratio is 1.25 to 1. What's happening to the girls? Abortion and infanticide. If any population in the world should be the playground of high-fertility memes, it's these two most populous countries in the world. Yet they turn out to be dominated by some pretty strong anti-fertility memes. Some deep reflection on this fact would keep Lynch (and others) from making some very foolish statements about which fertility-related genes and memes will prosper best.

And the call for a weekly Lord's day sets aside time to spread and preserve the faith. (102)

This in a discussion of the Ten Commandments, the foundation of a non-proselytizing religion. But besides that: has any sect anywhere dedicated the sabbath to evangelism?

The specific commandment against adultery likewise flourishes through parental transmission. By preventing broken families and absentee parenthood, the taboo improves its own retransmission..." (103; emphasis mine)

Surely we can do better than this level of analysis: the existence of a law implies its fulfilment? You could better argue the reverse. (Who has more abortions, the US or Sweden?)

Laws against eating shellfish, pork, and other parasite-laden animals may reduce morality rates, thus propagating the movement. (105)

So, cows have no diseases? Apparently Lynch is not even aware of Marvin Harris's competing, and more plausible, explanations. (Harris, in fact, is a pioneer in explaining beliefs as not arbitrary signs of culture but fomentors of ecological health-- and doing hard research to see if his explanations hold water. It's what memetics should be like.)

On the congregation level, 'Love your neighbor' can help prevent dropouts. By suppressing serious conflict, the meme stops churches from splitting up and members from going away mad. (109)

Jeez, Lynch, you don't know anyting at all about churches, do you? Churches roil with conflict-- and do split up, the members dispersing angrily.

The same thing happens in the movements of the far Left, which doesn't have this meme, except perhaps in the attentuated form that the workers are all brothers. I think there's a lesson about ideologically saturated people there... but it's not that "Love your neighbor" causes love.

Crosses that graphically depict a dying Jesus work especially well at reminding believers of the immense sacrifice the Savior made for them... Crosses and cross-display memes thus become the most common signs of Christianity around the world. (111)

Typical shallowness; Lynch seizes on anything that seems to point his way, ignoring the complications and contradictions. E.g. Catholics put Jesus on their crosses; Evangelicals-- who are much more interested in proselytism-- do not.

It's also rather naive to expect that an explicit or transparent symbol works "especially well". He'll have trouble, for instance, when he tries to explain why the second-most popular sign of Christianity is the fish. (Is there a semiotician in the house?)

The battle casualties also threaten the well-being of war widows and their children. Islam limits these secondary losses by allowing each man to have up to four wives. This keeps the women fed and bearing children while saving youngsters from orphanhood. (128)

Well, unless the polygamist dies, at which point polygamy would multiply the suffering caused by a death in battle.

Is there any tradition in Muslim society of remarrying war widows? Do more warlike Muslim nations have more polygamy? Is the effect even noticeable, when the vast majority of Muslim men are too poor to afford multiple wives?

Yet the very phonomenon of evolution by natural selection easily propels religions past the minor challenges raised by scientific ideas. Even evolutionism loses popular ground to divine creationism in modern times. (133)

Shamefully, Lynch finishes a full chapter on religion without once mentioning atheism. It's a skanky idea-analysis that won't shine the spotlight on the wielder's own ideas.

One of the actual insights of memetics is that a meme complex will act to discourage conversion to competing ideas. Religions teach that the practitioners of other religions are going to hell and that to question doctrine is a sin-- or, more cleverly and just as effectively, that they already incorporate any truth found in other religions. Communism teaches that all other belief systems are "ideologies"... a technical term meaning that they're diseased and incorrect. And atheism fits right into the mold: atheists believe that all religions are false, and that people are foolish or evil to believe in them.

Breast-feeding has a natural contraceptive effect... The biological mechanism provides birth spacing needed to prevent babies from starving under primitive conditions. (137)

Lynch finally realizes that a factor besides high fertility might be important! The proper thing would be to go back and rethink all those arguments that memes encouraging high fertility would propagate fastest.

The more often adherents diet, the more often they retransmit the memes. By failing to change long-term habits, a dietary meme set can actually induce those episodes of recurrent dieting-- and recurrent meme dissemination. The 'defective' memes thereby grab a proselytic edge over more permanent weight-control methods. (139)

This is one of Lynch's few interesting insights! This is one of the relatively few cases where it's useful to talk about "memes" at all, as opposed to "good ideas". There is no mystery, after all, when ideas spread that are useful, or which are taken as such. The interesting case is why false or useless ideas spread.

This is an example: memes for bad diets could spread better than memes for good ones, because they're bad: because the same host can be reinfected, bad diet memes have more opportunity to propagate.

Still, I distrust Lynch so much by this point that I want proof even for his reasonable ideas. Do false diets really spread better than effective ones? A false diet, after all, eventually provokes the "Oh no here it comes again" response. And wouldn't the logic imply that bad solutions to problems should propagate better than good solutions, in general?

So shouldn't we expect that, say, ineffective treatments for malaria should outpropagate the use of quinine? Shouldn't bad treatments for alcoholism outperform AA? Perhaps not. There's something to be said for better performance after all.

In humans, thought contagion gives viruses a unique set of obstacles and manipulation potentials that favor virulent forms of sexually transmitted disease. This may well explain why STDs strike our own species so harshly. (146)

Lynch needs to read Jared Diamond. Biologically, one of the most striking facts about our species is its high population density. And species with high population density are the ideal hosts for disease.

Those who use nootropics [memory-enhancing drugs] would presumably retain their memes longer than those who memetically reject the drugs. In particular, they would remember their reasons for using nootropic drugs longer than those opposed to nootropics would remember their reasons. (152)

Are you pulling our legs? Habit takes over when memory fails. Or do you think teetotallers become alkies once they forget their reasons for not drinking?

Explicit abortion taboos spread themselves through childbirth too. Christians who specifically regarded abortion as sinful outreplicated Christians with the more permissive abortion memes. (165)

Many of Lynch's errors come from taking ideas at face value: e.g. an anti-abortion meme will discourage abortion. He needs to start looking at what the memes make you do, not what they say. The abortion rate is highest among Catholics-- that should tell him something.

If widespread abortion bans take hold again, then restored reproduction and proselytic vigor will again shift to the pro-choice side. That, in turn, will lead back to renewed abortion rights and consequent trends back to banning abortion. (168)

So Lynch is predicting an acceptance/rejection cycle on abortion. It's true that the legalization of abortion led to anti-abortion campaigns; but I think Lynch is mistaking reaction for cyclical recurrence.

Abortion used to be too risky to be acceptable; now it's widespread. This is merely part of an enormous change in human affairs: technology increasingly gives us the ability to control things that were once givens of the human condition. This change is not going to go away due to mere storms in the ideosphere.

Two things will limit the anti-abortion movement: first, there are limits to how far anti-abortion memes can travel-- the religious right has become more visible, but not really larger. And second, the anti-abortion forces are riddled with unadmitted defections anyway.

Besides, do societies really go on for centuries flip-flopping on such a highly charged issue? I think it's a lot more likely that a near consensus will be achieved in the next century or so. (Not necessarily in favor of abortion in its present form; I suspect that contraception will become so easy that abortion will become a moot point. And then, once it no longer matters, the battle could go either way.)

Since parents have enormous influence over a child's primary and even secondary language, the Spanish language can replicate more often in Hispanic Americans than does the English language. (168)

Only they don't, and it doesn't. Once again Lynch doesn't even make an effort to see if his mental eructations have any relation to reality. In fact the biggest influence on a child's primary language, as can be attested by generations of immigrants, is the child's peers. Spanish is losing ground to English among Latinos; the main reason Spanish survives is that it's continually refreshed by new waves of immigration.

Men whose memes lead them to death or to several years of foreign fighting can also have fewer children than their dovish counterparts. (170)

Right, as shown by the WWII generation, which returned from several years of foreign fighting... to have the biggest generation of children in American history. Uh...

Thought contagion theory can thus become a uniquely self-replicating belief system, and an object for its own discourse. (176)

Yes on the first part; on the second part-- the sooner the better. Surely one of the (few) actual insights of memetics is that the truth of a theory can have little to do with its propagation. Memetics should ask why the idea of memetics is spreading.

I'll offer an idea: memetics is a neat idea-- and neat ideas spread quickly through the ideosphere. They spread whether they're true or false, and even whether they're believed or not: if they're intriguing or charming enough, people will repeat them just for the fun of it.

And Lynch has added another arrow to its quiver: memetics is now a great way to bludgeon beliefs one doesn't like. Notice how memeticists always analyze religion as a meme, never atheism, though it can fruitfully be applied to both. I predict that we'll see a good deal of this rhetorical technique in the near future: suggestions that one's opponents' beliefs, unlike one's own, are nothing but memes.

Lynch claims in his introduction that there are some mathematical foundations to his work-- although he never produces any in his book. It'd be nice if he trusted the general science reader to have a modicum of intelligence. But even if he did, I have to ask, would math really improve the theory? Lynch is busy proving things that aren't so, reinforcing his prejudices, and passing off just-so stories as science. Would we be happier to know that he had a highly mathematical method for producing bad science?

--M.R.


Back to Metaverse