I hope rather to give a brief

(and comprehensible) summary of the scientific

side of the issue of Ptolemaic incest. Before embarking on

it, however,

it would be well to quote

Scheidel's own remark at the end of some seventeen pages of painstaking mathematical, scien-

tific, and statistical analysis:

All in all, the available evidence from Roman Egypt cannot be taken to refute the model of increasing

inbreeding depression and its potentially disastrous consequences for the offspring of brother-sister

matings. At the same time, neither the census returns nor any other sources I am aware of offer any

indications of unusually elevated levels of infant mortality and severe physical or mental handicaps

among the inbreeding families of Roman Egypt.51

46 Ogden (1999)  67-116  on  the Ptolemies.  Walter

Scheidel's expansive study (1996a) deals with the genet-

ic  question

in  detail, but is  primarily

focused  on  the

Roman-period marriages among commoners, rather than

the Hellenistic dynasty.

47 Ogden (1999) 86

(on Ptolemy Eupator) and 94

(on

the two sons of Ptolemy

IX and Kleopatra Selene).

48 Green

(1990) 554.

49 Grant

(1972) 27.

50 Scheidel

(1996a).  The marriages are attested in

census returns; see n. 14 above.

51 Scheidel

(1996a) 28

(my emphasis). 10  SHEILA L. AGER

Such a remark is cautionary, and adumbrates my own conclusions about the Ptolemies, who did

not form part of Scheidel's

study.

Sexually  reproducing organisms

such as human beings  replicate

the

species  through a

constant recombination of  genetic material in each generation,

the  offspring incorporating

randomly half of the father's genes and half of the mother's. As the number of genes and the

number of  combinations are virtually endless, each human is  genetically unique.  The chief

benefit of  sexual reproduction as a means of propagation, at the level  of  the species,

is  the

opportunities

it provides

for

species

fitness and enhancement

through

the process of evolution

and natural selection: organisms with 'bad' genetic material are

likely

to die young or be unable

to reproduce ('reproductive death'), while organisms with

'good' genetic material will survive to

pass on those genes

to their own offspring.

The classification of genetic material here as

'good' or 'bad' is to some extent a

subjective

one, and is certainly simplistic. Whether a gene

is good or bad depends

to an extent on its com-

bination with other genes, on the nature of the organism, and on the environment within which

the organism

lives.  Briefly (and again simplistically) put,

'bad' genetic material would be that

which kills, disables, or prevents

from reproducing

its bearer within the bearer's natural context.

A more objective classification of genes

is their designation as dominant or recessive, or rather

the designation of  alternative alleles

(variants) of the same gene as dominant or recessive.  A

dominant allele of a gene will 'mask' the recessive allele of the same gene.  Thus, 'sex allows a

beneficial  allele  inherited from one  parent

to  "complement", or mask, a  deleterious allele

inherited from the other parent'.52 The potential danger of human

inbreeding

-  for

instance,

in

a

sibling marriage

-  lies in the increased opportunities

it provides

for recessive alleles of  the

same gene

to match up and become manifested in the organism. Since brothers and sisters share

much more genetic material than unrelated individuals, they have a much higher

likelihood of

carrying

the same recessive alleles.  Furthermore, recessive genes on the whole stand a higher

chance than dominant genes of being noxious in some way (even though

the majority of reces-

sives are in fact more or less neutral). Precisely because

they are recessive, they have not been

'cleansed' from the breeding population

in the same way

that a dominant gene would be. A dom-

inant gene

that had a

significantly damaging effect would tend to kill off its bearers, or otherwise

prevent

them from reproducing, and hence would wipe

itself out; deleterious recessives, on the

other hand, can lurk forever.53

On the face of it, then, significant inbreeding,

such as was practised at several points

in the

Ptolemaic

line, would seem to have been

likely

to produce genetic complications. But

'likely'

is not the same as 'inevitable'. Combinations of genetic material are random, not determined by

a genetically (or divinely) mandated programme of punishment

for offenders against

the taboo.

52 Durham

(1991) 297.  For a fuller discussion of the

genetics  of  breeding  and

inbreeding,

see  Scheidel

(1996a)  15-28;  Durham

(1991)  296-309;  Vogel  and

Motulsky (1997) 549-69.

53 The risk of  various degrees of

inbreeding can be

expressed mathematically through

the

'relationship coef-

ficient'

(r), which is an expression of degree of related-

ness

through

the potential of shared genetic material, and

which  can  be  employed

to  determine the

'inbreeding

coefficient'

(F), a mathematical expression of the degree

of

inbreeding between two persons with a relatedness of

r.  The formula to  determine F  in a population where

there is  no  built-up inbreeding (e.g.  between

siblings

whose parents

themselves are unrelated)

is F = r/2. Thus,

since the

relationship coefficient of

siblings

is  0.5,  the

inbreeding coefficient for the potential offspring of such

a couple would be 0.25.  The formula is more complicated

when  there is  already a  certain amount of  generalized

inbreeding

in the population;

see Durham

(1991) 300-1;

Vogel and Motulsky (1997) 550-2.

A closed population

that has been

inbreeding regularly

for a

long

time has a

tendency

to cleanse the lethal reces-

sives from the gene pool, and can reach a point of equi-

librium where

inbreeding no

longer presents

the  same

dangers as it does to a normally outbreeding population

(see, e.g., Reddy (1992)).

It is

impossible

to

say whether

this

tendency had any effect on  the Ptolemies, beyond

pointing out that there may have been enough 'foreign'

genes brought

in from time to time

(e.g. Kleopatra I)

to

offset  the  possible  benefits  of  continual and  rigorous

inbreeding (since  a  whole  new  set  of  lethal recessive

genes  might  be

imported).  See  Moore

(1992)  930;

Scheidel

(1997). INCEST AND THE PTOLEMAIC DYNASTY  11

There have been numerous anthropological and medical studies of

inbreeding groups carried out

in the last four or five decades,

in India,

in Japan,

in Czechoslovakia,

in the United States and

elsewhere.54 The majority of these studies tend to focus on first-cousin matings,

since none of

the societies involved actually

sanctions anything closer. The number of studies on true nuclear-

family

incestuous unions, and their offspring, remains therefore relatively low, which means that

we must be cautious about relying on them as

statistically representative.5s While these studies

do seem to bear out some of our expectations about genetic damage

-  higher rates of malfor-

mations,

infant mortality, mental disabilities -  there are still many unanswered questions and

contradictory findings.56 The empirical has not always borne out the theoretical. In some cases,

the

sample

size is probably

too small to be

statistically significant.57

There is also at times a question of appreciable bias in some of these studies.58 There may

not be an appropriate corrective taken for socio-economic or other factors. As Eva Seemanova

points out in her

study of  161 offspring of nuclear-family

incest published

in 1971, '[u]ndoubt-

edly,

the parents of the children of incest are not a representative sample of the general popula-

tion'.59 Factors of general health, access to education, family income, parental mental ability,

and so on, could have a clear

impact on the offspring of incestuous and non-incestuous unions

alike.60 The bias in some studies particularly comes into play when it is a matter of estimation

of  the

impact of

inbreeding on

intelligence levels, perhaps because the measurement of  these

qualities can be so

subjective. Some decades ago,

for

instance, a

study of

intelligence

levels was

carried out among

the highly

inbred population of the remote island of Tristan da Cunha in the

south Atlantic.61 The

study was based on a scale that recognized only one 'normal' category, no

'above-average' category, and five categories below normal, ranging through

'slow cerebration'

and

'very

slow cerebration', down to

'low-grade mental defective'. Clearly

the

investigator went

there expecting

to  find subnormal mental activity, and not unnaturally

found what he was

looking

for.62 Furthermore, even when the potential

for a skewed

sample

is recognized, and a

corrective is applied,

the end result is that we are still left with little to no

statistically significant

data on the viability of  the offspring of incestuous parents

from a well-to-do, well-educated,

hygienic  environment with  access  to  all  that wealth and position have to  offer

(which

the

Ptolemies of course were, at least by

the standards of their own era).

Nevertheless, there do  seem  to be  clear indicators that genetic problems

-  diseases and

deformities, some severe enough

to cause death -  are more

likely

to arise in a closely inbreeding

54 Among

these studies are Reed and Reed

(1965) 62-

4;  Cook and Hanslip (1966); Adams and Neel

(1967);

Adams et al.

(1967); Seemanovyi (1971); Schull and Neel

(1972);  Bashi

(1977);  Chakraborty and  Chakravarti

(1977); Ansari and Sinha

(1978); Lindelius

(1980); Baird

and McGillivray (1982); Al-Awadi et al.

(1986); Reddy

(1992).  For

surveys and summaries of these studies, see

(inter alia) May (1979); Bittles

(1981); Arens

(1986)  16-

24; Durham

(1991) 305-9; Scheidel

(1996a) 20-2; Vogel

and Motulsky (1997) 566-7.

55 Scheidel

(1996a) 20 puts

it well when he

says

'in

general,

the availability of empirical data might be said to

be

inversely correlated with the

intensity of

inbreeding'.

56 See  Vogel  and Motulsky (1997)  562-3;  Leavitt

(1990) 974-5; Bittles et al.

(1991); Bittles

(2005).

57 E.g. Adams and Neel

(1967) (the basis of Adams et

al.

(1967)); Baird and McGillivray (1982).

58  See  Bittles

((1981),  ap.  van den Berghe (1983)

103-4 and Bittles

(2005)); Baird and McGillivray (1982);

Durham

(1991)  307; Scheidel

(1996a) 20-1; Vogel and

Motulsky (1997) 557.

59 Seemanovy (1971) 118

(still

the

single most

impor-

tant

study of nuclear-family incest); cf  Reed and Reed

(1965) 63 and Baird and McGillivray (1982) 857.

60  See  the  comments made by  Cook  and Hanslip

(1966) 95,  98; Schull and Neel

(1972) 425-6; Lindelius

(1980) 190; Bittles

(1981, 2005).

61 Roberts

(1967), reporting on a

study done by L.

Woolley

in 1942; see Arens

(1986) 20 for criticism of the

subjectivity of the Tristan da Cunha

study. For criticism

of  the methodologies of the standard studies, as applied

to

IQ, see Kamin

(1980);

for another problematic study of

the  effects  of  incest  on  mental ability,

see  Jancar and

Johnston

(1990).

62 In  his  1967  paper, Roberts acknowledged

the

methodological

flaws of the 1942

survey, but in 1992 he

reiterated the  'mental dullness'  of  the  inhabitants of

Tristan da Cunha, only a few pages after he had examined

the poor educational

system on the

island,

itself

surely a

socioeconomic factor bound to affect the apparent

intel-

ligence

levels of the islanders. 12  SHEILA L. AGER

population. But

they do not

inevitably arise in all  offspring of

incest -  that is  the

important

point.63

In

Seemanovi's study, a very significant percentage (roughly 50%) of the 161 children

exhibited the negative effects of

inbreeding: they either died or had some major defect.  Yet it is

worthwhile

stating

the obvious here: 50% of the children survived without defect.  In any case,

we must also ask the question, are we

justified

in

finding evidence of such

inbreeding depression

in what we know of the Ptolemies?  Scheidel was unable to demonstrate

inbreeding depression

among

the incestuous commoners of Roman Egypt

from the available sources; are the sources

on the Ptolemies any more helpful

in this regard?

Certainly

there is nothing

to

suggest

that the dynasty

suffered from mental defect in any way

that we  can detect, though we  might want to qualify Ptolemy XI's ill-advised murder of  the

beloved Kleopatra Berenike as a

singular

instance of

'very slow cerebration'. Moreover, the

Ptolemies and the Kleopatras really do not seem to have sustained a noticeable reduction in

fertility, certainly not by comparison with other dynasties

such as the Antigonids or the woefully

infertile -  but  not  inbred -  Attalids.64  Except

for  the  marriage of  Arsinoe  and Ptolemy

Philadelphos, and the putative marriages of  Kleopatra VII to her younger brothers, virtually

every

incestuous Ptolemaic marriage resulted in offspring.65 Some of these children died young,

to be  sure, but infant mortality among

the Ptolemies does not seem to have been any more

demonstrably pronounced

than in any other

family

in the ancient world.

Ray Bixler's objections are overdrawn. In an article designed

to debunk the extent of royal

sibling

incest that actually went on in the royal

families of Ptolemaic Egypt,

Inca Peru and old

Hawaii, he rightly draws critical attention to the naive generalizations made by Ruffer and

others who claimed

(a)

that the Ptolemaic rulers were all the product of

incest, and

(b)

that all

that

inbreeding never hurt them

(which

is not quite

the same

thing as arguing

that we cannot con-

clusively demonstrate that

inbreeding hurt

them).66 But Bixler's protests are themselves based

on generalizations and outdated historical reconstructions. He focuses only on

sibling marriage,

failing

to note that the offspring of  an uncle and inbred niece  would  also have had a high

inbreeding coefficient;67 he does not note the production of children other than those who even-

tually

inherited the throne; he traces the rulers only through

the male

line,

in a dynasty which

produced

such powerfully assertive and effective female co-regents (and daughters of

sibling

incest) as Kleopatra III, Kleopatra Berenike, Berenike IV, and Kleopatra VII; and

finally, he

places

too much emphasis on

strenuously arguing

that these marriages entailed little or no sexual

attraction. Few would claim that

they did, but that is not the same as claiming

that

they entailed

no sexual activity.

As far as Ptolemaic

fertility goes,

it is true that Ptolemy

IV and Arsino  III only had one child

-  but then both of them died at a relatively young age.  Furthermore, Ptolemy

IV was notorious

for a sex life that did not include his sister-wife, and it is very easy

to

suppose that, apart

from

siring his heir, he

largely ignored her.  It is worth noting as well that their child, Ptolemy V,

is

specifically attested in the ancient record as being very fit and athletic.68 He too died young, but

63

'[The] effects of

inbreeding are probabilistic, not

deterministic': Durham

(1991) 301.

64 Ogden (1999) believes the Ptolemaic

sibling mar-

riages

to  have  been

'virtually

infertile'

(67),  whether

through

failure to conceive, stillbirth, or post-natal/pre-

reproductive mortality  of  inbred  offspring.  But  see

Scheidel

(1997) 367: 'In general,

there is no evidence that

inbreeding

increases the incidence of

sterility

... or the

incidence of fetal death'; and Vogel and Motulsky (1997)

562-3.

65 In the case of Kleopatra

it is almost certain, and in

the case of

Arsinoi possible,

that these marriages entailed

no sexual activity;

the question of their

fertility

is there-

fore moot. The nineteen-day marriage of Ptolemy XI and

Kleopatra Berenike seems hardly worth mentioning

in

this context.

66 Bixler

(1982a; see also 1982b); Ruffer

(1921) 341-

56.

67 The

inbreeding coefficient

(F)  of  an uncle-niece

pair when there is no

family background of

inbreeding

is

0.125; but for Ptolemy VIII and Kleopatra III, multiply

related as

they were, F would have been 0.25,

the same

as for a brother-sister or parent-child pair (this figure was

calculated using Wright's method of

'path coefficients';

see Vogel and Motulsky (1997) 551-2 and n.53 above).

68

Polyb. 22.3. INCEST AND THE PTOLEMAIC DYNASTY  13

not before

siring

three healthy children, and his death came as enough of a

surprise

to give rise

to the

supposition

that he had been murdered.69 The marriages of Kleopatra II to both her brothers

were quite remarkably fruitful, especially when we consider that she may have been over 40

when she bore Memphites

to her much-despised brother Ptolemy VIII.  If Memphites

failed to

survive

long enough

to inherit the throne, his genes were hardly

to blame

(except

insofar as

they

linked him to a

spectacularly bad choice of a

father).  Kleopatra III, herself the child of  full

brother-sister incest, and married to her uncle on both sides, seems to have had no trouble pro-

ducing

five healthy babies in quick succession, all of whom went on to produce children of their

own, and none of whom seem to have been in any need of

'hybrid vigour'.70 And if Ptolemy

IX

could not get an heir from a Ptolemaic sister-bride,

it was

surely because his marriages kept

ending

in divorce, not because

they were infertile

(they were not).

In his argument

for Ptolemaic

'genetic compromise', Ogden makes much of the

(arguably)

freakish corpulence of Ptolemy VIII, who was known popularly as

'Physkon' or

'Pot-belly'.71

To the Romans ... he was as ludicrous a

figure as he was a cruel one to his fellow-citizens. He had an

ugly face, and was short in stature; and he had a distended belly more like an animal's than a man's.

The repulsiveness of his appearance was heightened by his dress, which was exceedingly fine-spun

to

the point of

transparency,  just as if he had some motive for putting on display what a decent man should

have made every effort to conceal.72

Justin's description of him is fleshed out

(as

it were) by John Whitehome's

comment: 'Justin ...

invites us to

imagine Ptolemy VIII dressed in a sort of

see-through nightdress as he dragged his

great weight, puffing and panting,

towards his brother's widow ...'73  But while Physkon may

have been gross,

in more ways

than one, he was not himself the child of

incest;

in fact, since

his parents were only

third cousins, he probably had a share in some of  the freshest genetic

material that had been in the

family

for some time.