Meme Pools and Sexual Reproduction (was Lynch's Hypothesis)

Timothy Perper/Martha Cornog (
Sat, 21 Jun 1997 13:38:06 -0500

Message-Id: <>
Date: Sat, 21 Jun 1997 13:38:06 -0500
From: (Timothy Perper/Martha Cornog)
Subject: Meme Pools and Sexual Reproduction (was Lynch's Hypothesis)

Although not exactly brief, the following may illustrate what meme pools
look like in concrete detail when we encounter one from the data side. We
think the resulting methodological and interpretational problems are worth
describing. (Again, this is from both of us.)

Although we start talking about masturbation, we very quickly encounter a
great many more topics. And that is the point: virtually any starting
question is a stream that leads to the meme pool itself.

1.) Deleting various digressions to cut to the heart of the matter in a
previous thread (from 6/20/97):

>>... let's say we have population #1 with (male) masturbation
>>rate of X1 and impregnation rate Y1, and population #2 with masturbation
>>rate X2 and impregnation rate Y2. If X1 > X2, what is the predicted
>>difference between Y1 and Y2?
>>1. Y1 > Y2. The more masturbation, the higher the impregnation rate.
>>2. Y1 = Y2. Impregnation rate is independent of masturbation rate.
>>3. Y1 < Y2. The more masturbation, the lower the impregnation rate.
>>Which of these three best represents Lynch's hypothesis?
Aaron Lynch>
>None of these three represent my hypothesis, because populations 1 and 2
>are not defined with respect to age.

EXACTLY!! Your caution is well placed. The effect of masturbation on
impregnation rates could very well depend on the man's age, and in an
earlier posting we mentioned some reasons why.

But we need to know more than just age. We also need to know marital
status. For example, the low impregnation sample might have fewer married
men in it.

Yet that isn't enough either. Religion can affect both masturbation and
impregnation rates -- if, for example, one sample had more devout,
practicing Roman Catholic men than the other, we might expect that they had
lower masturbation rates and larger families, for those are desired goals
for devout Catholics. So we need to know about religious beliefs of the

Then too we need to know something about the women. How religious are they?
What is their age range? What sorts of contraception are they using, if

Thus we learn a crucial fact: the effect of a given belief -- a meme, if
you will -- depends on context, including other memes. And so the example
shows that whereas one *can* study a meme in isolation, its effects cannot
be predicted without knowing the rest of the social-cultural-memetic
background of the sample.

Obvious? Well, in theory perhaps so. But in actual research, identifying
these interrelated factors is *not* obvious at all. We need to know our
subject and its literature in systematic detail: theoretical memetics is
not sufficient.

2.) Moreover, the example illustrates another point. If we in fact were
to obtain such appropriately matched samples, we would be working within
well known methodological paradigms of social psychology and sociology.
Indeed, both fields have some quite powerful methods for disentangling
various influences statistically, e.g., by use of analyses of covariance.
Accordingly, the study of memes is, if not indistinguishable from existing
social and behavioral sciences, then is strongly dependent on them for
*methods* of analysis and interpretation of numerical data.

Yet what if the two samples came from different periods in history, where,
by good fortune, we had comparable statistics? Could we infer something
about the historical *changes* in the relation of masturbation and

No. History does not affect ONLY masturbation and pregnancy. So we may
not assume that "other things are equal," because they are NOT equal. And
so we need historiography!

Do these comments "reduce" memetics to historical or statistical sociology?
No. Memetics has its own theoretical framework (or is developing one or
more) that differs from those of the traditional social, behavioral, and
historical sciences. But memetics cannot *ignore* these sciences, because
they provide access to the list of factors that interact with what we
study. They are, so to speak, the geology that surrounds the meme pool and
gives it shape.

3.) Well, so far so good -- and so let us now assume that we have a
properly chosen and analyzed sample from the same society at the same
period, and find that low masturbation rates are associated with higher
impregnation rates (as Lynch suggests we should find). Would this result
demonstrate that higher impregnation rates are caused by memes for reduced

The answer is No. All that has been demonstrated is a *correlation* -- and
correlation is not causation. To see why, again consider a devout,
practicing Roman Catholic man. He has lower masturbation rates AND a
larger family. Does the first "cause" the second memetically?

Again the answer is No. The reason is that lower masturbation rates AND
larger families themselves may both arise from the same underlying and
coherent belief (or meme) system, e.g., of Catholicism. Lower masturbation
rates and larger families share a common basis in Roman Catholic doctrine,
and the two could emerge from this memetic substrate without either one
*causing* the other.

To see how this works in concrete reality, consider contraception. The
man's masturbation frequency will obviously be without effect on
impregnation if his wife is consistently using contraception. So
contraception -- ot the lack of it -- is an intervening variable between
masturbation rate and impregnation.

Of course, a devout and conservative Catholic woman would NOT use
contraception. However -- and it is the crux -- that is not because her
husband does not masturbate. It is because SHE accepts the teachings of
the magisterium about raising a large family. A set of apparently
different behavior patterns, which one might be tempted to say were
independent memes, actually arise from a common core of belief that links
them together into a *web* of reciprocal causes and effects.

These too can be analyzed using known social scientific methods, such as
structural analysis or pathway analysis, but they are also part and parcel
of historiography and ethnography.

Similar arguments hold for *any* set of intervening variables.

4.) Above, we said that even if lower masturbation rates might perhaps be
correlated with higher impregnation rates, the first does not "cause" the
second. Now this comment needs BIOLOGICAL elaboration. Bear with us for
some details, because they are very illuminating.

Consider the physiology of fertilization. To summarize some material posted
in response to Hans-Cees Speel on 6/18/97, "...the probabilty of pregnancy
is nearly 0 from 8 to 2 days before ovulation; it then rises to about 0.5
from 2 days before ovulation to approximately 1 day after ovulation; and
then drops back to nearly zero from approximately 2 days after ovulation to
+10 days after ovulation. ..."

: ___
The pattern looks like this ______/ \________ and is of course is an
estimate, which, however, is convenient for illustrative purposes. Now let
us ask about the effects of masturbation on impregnation.

It is immediately obvious that if the man NEVER masturbates but has ONLY
non-contracepting intercourse with his wife, then he will in all likelihood
have more offspring than anyone else. However, this possibility can in
practice be set aside because most men DO masturbate, and we need to deal
with the real world, where men do masturbate, at least sometimes. We wish
to understand human society, not exceptional cases.

A digression: we quote from Frederick M. Sturgis, MD, writing in 1907: "I
feel as Berger does in his _Vorlesungen_, that 99 per cent of young men and
women masturbate occasionally, while the hundredth conceals the truth.
(Archiv. fur Psychiatrie, 1876)." (Sturgis, 1907, page 398.)

Now we must think. Assume that the couple is trying to have a child and is
having intercourse every two days. If we were omniscient, we could look at
the couple and know with certainty that she DID become pregnant when they
had intercourse on the 14th day of her cycle. We would also then know with
certainty that if he had masturbated instead, she would not have become
pregnant, at least during that cycle.

But neither we nor the couple are omniscient. If he happens to masturbate
on her 8th cycle day, and at no other time, then there is no reproductive
cost: she cannot become pregnant on the 8th cycle day -- nor after her peak
fertile period. Thus we learn that an anti-masturbation meme is
biologically pro-conceptive ONLY during her fertile period -- at any other
time it makes no difference what HE does sexually. He can be celibate, he
can masturbate, he can sex with other women -- none of these things alter
the inability of his wife to become pregnant except around ovulation. The
system does not consist of self-moving monadic memes each causing its own
unitary effect: meme systems are intrinsically interactive.

Now consider something else. Say he only masturbates and does NOT have
intercourse with her for say 3 years -- that certainly would reduce the
number of children, would it not?

Perhaps not. If she is pregnant, gives birth, and breastfeeds, then she
has a very low probability of conceiving a second child for the entire 3
years -- 9 months pregnant plus a bit over two years breastfeeding
(lactational sterility is discussed in Stuart-Macadam and Dettwyler, 1995).
The nature of pregnancy and lactation causes the purely *reproductive*
cost of male masturbation to dwindle.

One might object that the cost does not reach zero, and that is all that
counts. Biologically, perhaps so, but that is far from clear
*memetically*. If reproductive advantage ALONE drives the maintenance and
spread of this meme, it hardly pays the man reproductively not to "spill
his seed" for 3 years! And so we ecounter yet ANOTHER complexity: the
physiology of pregnancy and lactation INTERACT with the man's sexual
motivations, urges, and needs. If he can have intercourse with his wife
without impregnating her, why not masturbate? Reproductively the effect is
the same during those years -- zero. In fact, it is zero for all times
except her fertile period. So, reproduction by itself is insufficient to
understand the phenomenon.

In addition, we need to understand that *memetically* masturbation may well
have perceived *spiritual* costs: fear of loss of God's grace. And for
the religious person in the Western tradition, the same concern exists for
all forms of sexual behavior.

So, it turns out that we also need to consider the interactions of
physiology, psychology, and religion to understand the memetics of human
reproduction. The meme pool is becoming increasingly complex, but in
identifiable ways.

5.) How, then, can we understand the original data sets we described,
which showed that both impregnation and masturbation rates were low, then
rose, and then fell with the man's age? The simplest hypothesis is of
course to say that the desire to masturbate and to have intercourse are
both driven by a common, underlying causal condition that we might call
"sex drive" or "libido" or something like that. (Indeed, it is widely held
that "libido" varies with age, that is, has its own developmental
psychological trajectory.) So, for a man with a steady sexual partner, if
masturbation frequency changes with age, then intercourse frequencies would
change in parallel, because one causal mechanism -- sex drive -- is common
to both.

Now, a critic can argue that "sex drive" is a folk explanation. Foucault
and his deconstructionist followers assert that the notion of sexuality
itself is modern. Perhaps so, but we are looking at memetics -- and
therefore we are in part *studying* folk explanations!

Sex researchers no longer attribute sexual drives only -- or even primarily
-- to biology, but see them also emerging from sexual experience and sexual
scripting, among other processes. And so we have yet another encompassing
context into which we must place human reproductive behavior.

6.) And now a conclusion. When we examine masturbation, we cannot examine
one meme in isolation. Instead, a single meme is only an entry point into
the complex totality. One ends up studying an interconnected system of
sexual and reproductive beliefs, behavior, experience, and preferences, of
which masturbation is only a very small part, all enmeshed in religion, in
contraception and its history, in changes in men's and women's notion of
the ideal family, in shifts in sexual expectations of ourselves and others
-- and we have not even mentioned the "demographic transition"!

We suggest that when memetics studies mixtures as complex as this -- and
ALL human social phenomena are as complicated! -- our understanding must
match the richness of the system we study. In brief, it's time to jump
into the meme pool.


Stuart-Macadam, Patricia and Katherine A. Dettwyler, editors 1995
Breastfeeding: Biocultural Perspectives. NY: Aldine de Gruyter.

Sturgis, Frederick M. 1907 The comparative prevalence of masturbation in
males and females. American Journal of Dermatology and Genito Urinary
Diseases, September, pages 396-400.

This was distributed via the memetics list associated with the
Journal of Memetics - Evolutionary Models of Information Transmission
For information about the journal and the list (e.g. unsubscribing)