© Koninklijke Brill NV, Leiden, 2009

ZUTOT 6.1

Also available online – brill.nl/zuto

DID SHMU’EL BEN NATHAN AND NATHAN HANOVER

EXAGGERATE? ESTIMATES OF JEWISH CASUALTIES IN

THE UKRAINE DURING THE COSSACK REVOLT IN 1648

Jits van Straten

Independent scholar, Bennekom, The Netherlands

The Cossack revolt of 1648 led by Chmielnicki cost the lives of many
Jews in the Ukraine. There exist two contemporary reports about this
event: Tit ha-yaveyn (ca. 1650) by Shmu’el Feibesh ben Nathan Feydel
from Vienna

1

and Yeveyn metsula (1653) by Nathan Nute ben Moshe

Hanover.

2

Ben Nathan describes how the Cossacks went from place to place

and tells us how many people they killed.

3

This is no more than a list

of casualties. In addition to the Jews killed by Chmielnicki, he also
provides the number of casualties during the campaign of the Swedish
king against the Polish king, which took place shortly after 1648. Finally,
he estimates the total number of ba‘ale batim (‘house-owners’) killed by
Chmielnicki and the Swedish king at 600,070.

Hanover’s report reads more like a story. He mentions fewer place-

names, but describes the situation in more detail. His report concludes
with the estimate of over 235,000 casualties, a 110,000 of which are
made up by poor people.

4

This event is also mentioned briefly in a letter by contemporary

author Menasseh ben Israel who only reports the total number of
Jewish casualties and the number of surviving Jews, without telling us
how he obtained this information. In his famous pamphlet addressed
to Oliver Cromwell, Ben Israel writes:

1

S.F. ben Nathan Feydel, Tit ha-yaveyn [The Slimy Clay], 2nd edition (Odessa 1889).

Ben Nathan’s report was possibly published for the first time in 1650, see Y. Vinograd,

Otsar ha-sefer ha-‘ivri [Thesaurus of the Hebrew book], catalogued under Venice 1290

( Jerusalem 1993) 270. The title is taken from psalm 40:3.

2

N.N. ben Moshe Hanover, Yeveyn metsula [The Bottomless Mire] (Venice 1653).

The title is taken from psalm 69:3.

3

Nathan Feydel, Tit ha-yaveyn, 9–19.

4

Hanover, Yeveyn metsula, [9] and [15].

76

jits van straten

‘But yet a greater number of Jews are found in the Kingdome of Poland,

Prussin and Lethuania [. . .] There is in this place such infinite number of

Jews, that although the Cosaques in the late warres have killed of them

above one hundred and fourescore thousand [180,000], yet it is sustained

that they are at this day as innumerable as those were that came out of

Egypt’ [600,000].

5

Turning to modern authors, Weinryb is of the opinion that the above-
mentioned figures are all gross exaggerations: ‘those in Tit ha-Yeveyn
[Yiddish for yaveyn] are patently absurd.’

6

In addition, he writes that

the amount ‘670,000 (or 600,070?)’ includes women and children.
This, however, cannot be found in Ben Nathan’s text. As for Hanover’s
Yeveyn metsula, Weinryb distinguishes the number of people killed—over
80,000—from the number who died in epidemics—41,000 or 141,000.
Together, the number of casualties adds up to 121,000 or 221,000.
Weinryb considers 40,000 to 50,000 casualties out of a total Jewish
population of 160,000 to 250,000 a ‘reasonable estimate’.

7

He also

quotes the 180,000 casualties of Ben Israel, but considers the number
of 600,000 surviving Polish Jews unrealistic.

The most recent estimate is provided by Stampfer.

8

Initially, he sug-

gests that between 6,000 and 14,000 Jews were killed,

9

but concludes

in the end that the total number of casualties could not have exceeded
20,000. According to Stampfer, Ben Nathan and Hanover estimate the
number of casualties at 670,000 and 80,000 respectively, figures that
are at odds with those mentioned in the original sources. He also refers
to an estimate by the Council of the Lithuanian Jewish community in
1650 of ‘several tens of thousands.’ He does not mention Ben Israel’s
estimates.

It will be clear that the variation between the contemporary and

modern estimates is very significant. I will therefore take a closer look
at how the estimates by Ben Nathan and Hanover were arrived at.

5

M. ben Israel, ‘To his Highnesse the Lord Protector of the Commonwealth of

England, Scotland and Ireland. The Humble Addresses of Menasseh ben Israel, a

Divine, and Doctor of Physick, in Behalfe of the Jewish Nation’ (1651) 7.

6

B.D. Weinryb, The Jews of Poland: a Social and Economic History of the Jewish Community

in Poland from 1100 to 1800 (Philadelphia 1973) 194.

7

Ibid., 197.

8

S. Stampfer, ‘What actually happened to the Jews of Ukraine in 1648?,’ Jewish

History 17 (2003) 207–227.

9

Ibid., note 5.

did shmu’el ben nathan and nathan hanover exaggerate

?

77

Let us first consider the estimate by Ben Nathan. Ba‘ale batim are not
just house-owners, they are members of the Jewish community, men
who had enough money to pay for its membership fee. In Amsterdam
the term ba‘ale batim was also used in this way. As this group consists
only of men, the total number of casualties must have been even higher
than Ben Nathan’s estimate of 600,070. Such a number of Jews killed
is simply impossible as it implies a Jewish population of more than two
million. Therefore, I first took a look at each location where something
had happened to the Jews first. In addition to ba‘ale batim, Ben Nathan
sometimes also mentions the total number of casualties. It appears that
in some places all Jews were killed, but in other places the number of
casualties falls into less well-defined categories like, ‘almost all,’ ‘a few,’
‘part of,’ and, ‘almost all and many converted.’ For statistic purposes
the latter three categories are useless, but fortunately only refer to less
than 3,000 people. The numbers of killed ba‘ale batim are arranged in
Table 1 according to the categories ‘all killed’ and ‘almost all killed.’

Leaving aside the matter of how to interpret these data, one thing

is immediately clear: the number of casualties among ba‘ale batim is
nowhere near 600,070. Could this figure be a printing mistake? All
but for two numbers Ben Nathan provides are rounded off to tens,
hundreds or thousands. This can only mean that his numbers are
approximations. The distribution of his numbers indicates thus. Of the
262 Jewish communities destroyed, 143 have one hundred or more ba‘ale
batim. It is remarkable that of these communities, 139 have rounded
hundreds (from 100–800), four have 150, 350 or 650 and none have
anything in between.

But more questions remain. Why does Ben Nathan exclude women?

Were they really killed? He only mentions women on two occasions: ‘of
6,000 people killed only some women remained’

10

and, ‘20,000 people

10

Ben Nathan, Tit ha-yaveyn, 9.

Table 1. The number of casualties at the hands of Chmielnicki and the

Swedish king among ba‘ale batim according to Shmu’el ben Nathan

In places where all were

killed

In places where almost all

were killed

Chmielnicki

24,210

52,272

Swedish king

5,190

2,980

Total

29,400

55,252

78

jits van straten

killed including women and children.’

11

Why does he mostly mention

ba‘ale batim? Was there no information available to him about the poor,
or were they looked down upon to such an extent that they were not
worth discussing? Is it possible that, because of the horrible situation,
Ben Nathan had decided to apply the term ba‘ale batim to all married
men, regardless of their economical status? It is impossible to assess the
number of casualties exactly without answers to these questions.

In most cases, Ben Nathan mentions that ‘almost all’ were killed.

I will assume the position that this amounts to anywhere between 75
and one hundred per cent. The total number of ba‘ale batim killed by
Chmielnicki would then be somewhere between 63,000 (24,210 +
39,204) and 76,000 (24,210 + 52,272). Including the ba‘ale batim killed
by the Swedish king, the total number of casualties would add up to
between 71,000 and 85,000. In addition to ba‘ale batim, Ben Nathan
also mentions 70,000 killed unspecified persons. This would bring the
total number of reported killings by Chmielnicki somewhere between
133,000 and 146,000. These figures also include Jews killed outside the
Ukraine. Excluding the number of casualties outside of the Ukraine,
we arrive at a rough assessment of the number of Jews killed in the
Ukraine, 100,000 up to 115,000.

Turning to the story by Hanover, it appears that in this case too the

text does not provide a clear-cut estimate of the number of casualties.
On two occasions, Hanover mentions ‘several thousands’ and ‘tens of
thousands’. I decided to use the lowest possible numbers: 3,000 for
the former (there is a separate word in Hebrew for 2,000: alpayim that
he could have used) and 20,000 for the latter. These numbers may be
too low, but they at least account for the minimum of casualties by
Hanover’s account. I did not distinguish between those casualties that
were caused by acts of violence and by starvation and/or the plague,
as all three circumstances were directly related to the revolt.

How does Ben Nathan’s range of casualties (100,000 to 115,000)

relate to Hanover’s estimate? Firstly, Hanover mentions fewer places
where Jews were killed. Secondly, he mostly reports casualties in
general terms and only refers specifically to ba‘ale batim in one case.

12

His number of unspecified casualties is 130,300. Contrary to Ben
Nathan, Hanover mentions the number of poor people twice: 10,000

11

Ibid., 14.

12

Hanover, Yeveyn metsula, 10.

did shmu’el ben nathan and nathan hanover exaggerate

?

79

and 100,000.

13

All together, this adds up to 240,000 casualties. The

110,000 poor people may be the result of guesswork, as the poor were
not registered as such anywhere.

The question about the number of casualties among women and

children remains. The following quotes from Hanover’s report show
that women and children were not necessarily killed:

‘in all places where the massacres took place, the women were usually

left alive’

14

‘everywhere where they [the Cossacks] had caused a massacre among

them, a few hundred little boys and girls remained behind who had been

baptized and the Jews took them away from the non-Jews by force and,

after inquiry and investigation, wrote down on a tag to which family each

one belonged, and they hung the tag around their neck.’

15

Still, Hanover’s numbers may be somewhat on the low side. When
the casualties in Poland and Lithuania are excluded, the total number
of casualties in the Ukraine is 195,000. One should bear in mind
that the number of casualties by both Hanover and Ben Nathan are
estimates.

Now, a remarkable change has taken place. The highest estimate

is no longer reported by Ben Nathan, but by Hanover, whose story is
described in the introduction of Abyss of Despair, the 1950 English edi-
tion of Yeveyn metsula, as ‘by far the most popular and authentic’

16

of the

two. A reason for the book’s popularity is probably the way in which
Hanover tells the story: not only with all kinds of gory details, but also
with some romanticising. Although the truthfulness of his account can
be called into question, it certainly seems to appeal to readers. The
number of Jews killed in the Ukraine according to Ben Nathan and
Hanover then becomes 100,000–115,000 and 195,000, respectively.

How realistic are the Stampfer’s relatively low estimates? He first

tries to estimate the size of the Jewish population before the revolt took
place. He states that it is not only impossible to find the exact number
of Ukrainian Jews in that period, but that we also do not even know
how great the non-Jewish Ukrainian population was in 1648.

13

Ibid., 9 and 15.

14

Ibid., 7. My translation.

15

Ibid., 18. My translation.

16

N.N. Hanover, Abyss of Despair (Yeven mezulah); the famous 17th century chronicle depicting

Jewish life in Russia and Poland during the Chmielnicki massacres of 1648–1649 (New York

1950) 7.

80

jits van straten

Trying to make the best of it, he uses two methods to obtain what

he calls a ‘reasonable figure’: a survey of the number of houses and the
number of seats in the local synagogue. As it is impossible to establish
both how accurate the survey of the number of houses is and how many
people lived in one house on average, it seems that this method is based
on too many assumptions that cannot be proven and, therefore, does
not suffice. Stampfer’s argument that the number of synagogue seats
is somewhat proportional to the number of Jews in the community,
does not seem to make sense either. When the seats for women and the
standing room for the non-residents Stampfer mentions are included, his
argument still does not.

17

If anything, one might think that the number

of seats is proportional to the number of ba‘ale batim. After all, they
were the ones who could pay for a seat. In reality, however, the ba‘ale
batim may have owned more than one seat—in Amsterdam some ba‘ale
batim had more than four seats. Furthermore, the number of seats does
not relate at all to the number of poor people, who Stampfer puts in
back of the synagogue. The number of poor people usually exceeds
the number of well-to-do people by far. For these reasons, the number
of seats in a synagogue does not equal the total number of Jews in a
community. In Amsterdam, for example, there were in total 1700 seats
for men and 800 seats for women in the four Ashkenazi synagogues in
1800, while the Ashkenazi Jewish population was about 25,000.

18

Application of the two methods mentioned above results in a popula-

tion of 40,000 Jews in the Ukraine before the revolt. With an estimated
22,000 survivors, Stampfer concludes that the number of casualties was
somewhere in the vicinity of 18,000. In order to determine the size of
the Jewish population after the revolt, Stampfer used the high annual
average compound growth rates (henceforth called annual growth
rates) of the Jewish community between 1650 and 1900

19

in combina-

tion with the size of the Jewish population in the Ukraine in 1764.

20

This led to a Jewish population in 1650 of about 45,000. In order to
justify the small difference between the size before and after the revolt
in view of the approximate 18,000 casualties, he decided that not all

17

Stampfer, ‘What actually happened’, 212.

18

J.F. van Agt, Synagogen in Amsterdam (The Hague 1974) 26, 64, 66, 70.

19

As suggested in S. DellaPergola, ‘Some fundamentals of Jewish demographic

history,’ in S. DellaPergola and J. Even, eds, Papers in Jewish Demography (Jerusalem

2001) 11–33, esp. 21.

20

R. Mahler, Toldot ha-yehudim be-polin [History of the Jews in Poland] (Tel-Aviv

1946) 234. C. McEvedy, C. Jones and R. Jones, Atlas of World Population History

(Harmondsworth/New York 1978).

did shmu’el ben nathan and nathan hanover exaggerate

?

81

Jews in the region in 1764 were descendants of survivors. However,
recently it has been shown that the number of Jews in Eastern Europe
between the years 1170 and 1900 not only lacks a factual basis, it can
also only be explained by annual growth rates that are implausible
for those periods in Europe, being two to four times as high as those
of the total population.

21

In 1897 about forty percent of the Jews in

European Russia lived in the Ukraine.

22

Mahler estimated that in 1764

Ukrainian Jews formed 44 percent of the total number of Jews in the
Polish-Lithuanian Commonwealth.

23

Assuming that his percentage is

correct, his absolute numbers are not, this would yield 330,000 Jews.
Departing from Stampfer’s 45,000 Jews in 1650, this would yield more
than a sevenfold increase of the Jewish population over 114 years, an
annual growth rate of 1.8 percent! In view of the general annual growth
rate of 0.45 percent, this is not plausible either.

Stampfer’s estimates of the size of the Jewish population both before

and after the revolt are thus incorrect. By using a realistic annual growth
rate based on the existing 1,3 per cent annual growth rate of the Jewish
population of Posen/Congress Poland during the nineteenth century,
and annual growth rates of the total population of the Polish-Lithuanian
Commonwealth between 1500 and 1800,

24

Van Straten calculated a

Jewish population in Eastern Europe of 830,000 to 1.14 million in
1650.

25

By using Mahler’s 44 percent, this would mean that in 1650,

somewhere between 370,000 and 500,000 Jews lived in the Ukraine.

The size of the Jewish population in 1648 should be roughly the

same as the one in 1650, when taking casualties into account. The
number of Jews who left the region according to Ben Nathan, does not
really change this significantly. To obtain the range of possible sizes of
the Jewish population in the Ukraine in 1648, the lowest and highest
numbers of casualties were added to the number of Jews in 1650 and
the percentages of casualties were calculated (Table 2).

21

See J. van Straten, ‘Early modern Polish Jewry: the Rhineland hypothesis revisited,’

Historical Methods 40 (2007) 39–50.

22

Jewish Encyclopedia 1905, v. 10, 530.

23

Mahler, Toldot ha-yehudim be-polin, 237. He mistakenly writes 42 percent.

24

As shown by McEvedy, Jones and Jones, Atlas; A. Maddison, The World Economy:

A Millennial Perspective (Paris 2001) and Van Straten, ‘Early modern Polish Jewry’

39–50.

25

Van Straten, ‘Early modern Polish Jewry’.

82

jits van straten

Table 2. The possible size of the Jewish population in the Ukraine in 1648,

the lowest and highest Jewish casualties and the corresponding percentages

of the total Jewish population

Possible size of the Jewish population in 1648

Number of Jews killed

100,000

195,000

%

%

470,000

21

565,000

35

600,000

17

695,000

28

The results in table 2 show that the percentage of casualties lies between
17 and 35 percent of the original Jewish population in the Ukraine in
1648, with an average of 25 percent. This is a percentage few historians
will object to. In addition, it gives a fair amount of credibility to the
chroniclers Ben Nathan and Hanover.

Stampfer’s low number of casualties is only valid when it can be

proven firstly, that the estimates by the seventeenth-century Jewish
chroniclers are ten to twentyfold exaggerations, and secondly, that
the applied Jewish annual growth rates between 1500 and 1900 are
plausible.

Summing up, the number of victims as reported by Ben Nathan
and Hanover for the Ukraine are not necessarily gross exaggerations,
since the highest number of casualties reported by the two chroniclers
amounts to ‘only’ 35 percent of the Jewish population in Ukraine in
1648, with an average of 25 percent.