Professional Documents
Culture Documents
Fluctuations of Fish Deliveries at Deir El-Medina
Fluctuations of Fish Deliveries at Deir El-Medina
A Statistical
Analysis
Author(s): Jean-Christophe Antoine
Source: Studien zur Altägyptischen Kultur, Bd. 35 (2006), pp. 25-41
Published by: Helmut Buske Verlag GmbH
Stable URL: http://www.jstor.org/stable/25157769
Accessed: 16-12-2017 12:45 UTC
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms
Helmut Buske Verlag GmbH is collaborating with JSTOR to digitize, preserve and extend
access to Studien zur Altägyptischen Kultur
This content downloaded from 195.43.22.134 on Sat, 16 Dec 2017 12:45:14 UTC
All use subject to http://about.jstor.org/terms
Fluctuations offish deliveries at Deir el-Medina in the Twentieth Dynasty.
A statistical analysis
Jean-Christophe Antoine
Abstract
The series of papyrus and ostraca from Deir el-Medina recording fish deliveries during the Twentieth
Dynasty constitute a large amount of homogeneous material suitable for a statistical analysis. We studied
186 data corresponding to a quantity offish expressed in deben and delivered by a fisherman on a given
day to the village of the royal tomb workers. An uni- and multivariate analysis showed that fish deliveries
fluctuated according to the size of the workmen gang. In addition, the Nile flood was probably responsible
for intra-annual seasonal variations. A striking finding was a significant reduction of the amount offish
delivered at the end of the reign of Ramesses DC and under Ramesses X comparatively to the time of
Ramesses III and IV. A series of abnormal Nile floods rather than social or economical factors may account
for this phenomenon.
Introduction
Written documents providing a large amount of sufficiently homogeneous data suitable for
a statistical analysis are rare in Egyptology.1 The group of Twentieth Dynasty ostraca and
papyrus from Deir el-Medina recording fish deliveries fulfil all these quality criteria and
may be of great interest for such a study all the more so since the general social and
historical setting surrounding these documents has been extensively studied.2 As since
antiquity fish is with bread the basis of the Egyptian diet,3 any fluctuation of its production
may result in important social and health consequences. A statistical analyse may thus
provide interesting data on the causes and consequences of the local fish production at
western Thebes in the Twelfth century BC.
Material and methods
The fishermen who provided the Royal Tomb workmen with fish belonged to the smdt and
were usually attached to one ?side" of the gang.4 They were on duty on a rotation system
and, three to four times a month, they brought their production to the village enclosure5
1 The best-known exception is the Wilbour papyrus. For this, see the pioneer work of S. Katary, Land
Tenure in the Ramesside Period, 1989 and the preface by J.J. Janssen in the same book (p. xv).
2 L. A. Christophe, Le ravitaillement en poissons des artisans de la necropole thebaine a la fin du regne
de Ramses III, in: BIFAO 65, 1967, 177-199; J. Cerny, A community of workmen at Thebes in the
ramesside period, BdE 50,1970,183-190; D. Valbelle, Les ouvriers de la Tombe. Deir el-Medineh a
l'epoque ramesside, BdE 96, 1985, 126-136; W. Helck, Materialien zur Wirtschaftsgeschichte des
Neuen Reiches. Part 5,1965; J.J. Janssen, Village Varia. Ten Studies on the History and Administration
of Deir el-Medina, Egyptologische Uitgaven 11,1997, 37-54.
3 D.J. Brewer/R.F. Friedman, Fish and Fishing in Ancient Egypt, 1989,16; W.J. Darby/P. Ghalioungui/
L. Grivetti, Food: The Gift of Osiris, 1977, vol. I, 337-383.
4 Christophe, in: BIFAO 65,1967,185; Janssen, Village Varia, 39.
5 Cerny, A community of workmen, 188-189; Christophe, in: BIFAO 65, 1967,179-180; Janssen, in:
Village Varia, 39-40.
This content downloaded from 195.43.22.134 on Sat, 16 Dec 2017 12:45:14 UTC
All use subject to http://about.jstor.org/terms
26 J.-C. Antoine SAK 35
where it was weighted and registered before being distributed to the gang.6 For this study,
we reviewed documents recording fish deliveries and dated with confidence from the
Twentieth Dynasty.7 The weights of Deir el-Medina may give further information on the
subject,8 but as it will be seen later, they were probably not employed for the operation
studied here. As a whole, documents recording fish deliveries fall into three categories. The
first corresponds to the daily entries of the Necropolis Journal or its abbreviated equivalent
on ostraca.9 The second are notes of work mainly on ostraca recording the activity of one
or several fishermen over generally a few days.10 The last category assesses fish supplies
over a longer period.11 Only the first two categories were suitable for the study.
In most documents, fish is designated under the general appellation of rm ?fish" and
measured in deben. In others, it is named by species and quantities are given without any
unit of measure suggesting that it was counted by specimen.12 In a last group, fish
designated as rm is recorded without indication of a unit of measure.13 Several arguments
suggest that the deben should be intended in these cases. First, the deben is the most
frequent unit.14 Second, in several texts, the deben is not systematically mentioned
throughout implying that when it is not, it is intended. Last, a comparison of the quantities
of fish (rm) expressed in deben with those without any unit did not show statistical
difference concerning their mean numerical values.15 Hence, these cases were considered
as expressed in deben and kept for the study. Conversely, those counting fish by specimen
were discarded.
This content downloaded from 195.43.22.134 on Sat, 16 Dec 2017 12:45:14 UTC
All use subject to http://about.jstor.org/terms
2006 Fluctuations offish deliveries at Deir el-Medina 27
The most difficult point was to determine whether fish was fresh or dried when
delivered.16 O. DM 140 shows that both forms may be found in a same delivery and that in
the ancient Egyptian mind there was no difference in nature between them as they were
added in the total amount of delivered fish.17 O. DM 584 gives a proportion of 2/3-1/3 for
each form,18but whether this is representative or not is unknown. Since as a mean each
fisherman came to the village every two weeks, most fish probably needed to be dried before
being delivered.19 On the statistical point of view, the discrepancy introduced by these
probable small quantities of fresh fish is reduced by their expected even distribution.
To be kept for the statistical analysis, the data should fulfil several criteria: they should
be dated with confidence from a definite reign,20 they should preserve at least the date of both
the season and month, and the quantity offish delivered by a fisherman should preserve at
least the hundreds.21 Using these criteria, 186 entries were entered in the statistical analysis
and 20 were rejected as being incomplete.22 Each data corresponded to a quantity offish
expressed in deben and delivered by one fisherman on a given day.23
In a second step, three explanatory variables that may have influence on the fishermen's
production were attributed to each data. The first one was the strength of the workmen crew
at the date of the delivery. It is known with precision for several years.24 When lacking, the
16 It is generally admitted that for conservation reasons fish was usually dressed and dried or salted as soon
as cached. See Brewer/Friedman, Fish and fishing, 12; Darby e.a., Food, 369-370; Valbelle, Ouvriers,
272-274; Janssen, Village Varia, 49.
17 Kitchen, Ram. Inscr. V, 484-485.
18 Janssen, Village Varia, 44. The document is attributed to the reign of Ramesses III.
19 The mean time separating two deliveries for a single fisherman calculated from 83 data was 15.65 days.
Extremes 1 to 45 days, mean 15.65 +_9.45 days, median 14. These values are minimal since we only
included in this calculation cases in which a fisherman is mentioned at least two times in the same
document.
20 The attribution of P. Greg to Ramesses III is debated. J.J. Janssen dates it from Sethos II (Janssen,
Village Varia, 111-129). This papyrus represents only two entries.
21 When several fishermen were involved in fish delivery on a same day, the mean value of the amount of
fish delivered by the fishermen was entered in the computer.
22 A total of 276 individual fish deliveries were recorded corresponding to 187 day of deliveries as in
several instances more than one fisherman came to the Village on the same day. One very high outlying
value was excluded leaving 186 data in the study. This outlier fell on the period of maximum fishing yield
so that ist suppression did not affect the conclusions obtained from the analysis.
23 As underlined by Janssen, Village Varia, 40, it is sometimes difficult to determine when two names are
given side by side if they should be understood as ?fisherman X (son of) Y" or as ?fishermen X and Y".
However, comparison of the documents between them permits in most case to clarify the situation.
24 See Valbelle, Ouvriers, 103-105. There is no known data for year 22 to 24 of Ramesses III. The
estimated size was then 40 as this value is the most frequent in the last part of the reign of that king. The
same value was also retained for the 21 entries dated from Ramesses III and the 9 dated from the end
of the reign of Ramesses III and the first years of Ramesses IV that have lost the quantum of the year
since the vast majority of the documentation from Deir el-Medina concerning Ramesses III is posterior
to year 22. Between year 1 and 3 of Ramesses X, Valbelle, Ouvriers, 105, gives 32+x men. This was
corrected into 63 as an entry of the Turin Necropolis Journal of year 3 of this king indicates that 30
copper tools were delivered to one side of the crew (Rt 1.7-1.8 Kitchen; Ram. Inscr. VI, 688,1). This
corresponds better to the usual size of the gang under Ramesses IX.
This content downloaded from 195.43.22.134 on Sat, 16 Dec 2017 12:45:14 UTC
All use subject to http://about.jstor.org/terms
28 J.-C. Antoine SAK 35
nearest known value was retained. Then, the strength values were grouped into three
categories corresponding to different ranges of the gang size: < 50 persons, 51 to 60 persons
and > 60 persons.25 The second variable was the date in the year. For this, the Egyptian dates
were transformed into their corresponding dates in the Gregorian calendar26 and, for each
entry, the date of the corresponding ideal Egyptian season was attributed.27 The last variable
was chronological. The 186 data covered 23 years corresponding to two distinct historical
periods separated by a gap of nearly 45 years. The first one totalling 126 data corresponded
to the reigns of Ramesses III and IV and was designated as the RIII-RIV period. The second,
namely the RIX-RX period, gathered 60 data from the reigns of Ramesses IX and X.
The data and their corresponding explanatory variables were entered into a computer table
using the Excel software. Statistics were performed with the Statview software. As the
distribution offish deliveries expressed in deben was not normal, non-parametric tests were
used.28 These included the Man-Whitney test and the Kruskal-Wallis test where appropriate.
The chi-2 test was used for the study of frequencies. For each of these tests, a value of
p<0.05 was considered as significant.29
Results
Descriptive statistics
Fish deliveries ranged from 40 to 2000 deben with a mean value of 413.07+ 271.20 deben
and a median of 352 deben. Comparatively, the weighs of Deir el-Medina measured from
0.91 to 24.5 deben, mean 7.86 ? 5.13, median 6.81 deben showing that these objects were
not used for weighting the fish deliveries studied here. As documents preserving fish
deliveries over a long period are rare, it was difficult to estimate the total amount of fish
supplied by the fishermen each month. The available data gave a number of 1 to 8 deliveries
per month (median 4) and a number of fishermen on duty ranging from 1 to 10 (median 4).30
This content downloaded from 195.43.22.134 on Sat, 16 Dec 2017 12:45:14 UTC
All use subject to http://about.jstor.org/terms
2006 Fluctuations offish deliveries at Deir el-Medina 29
A rough estimation suggests that about 3600 to 6400 deben offish may have been delivered
monthly.31 This compares with the quota of 5520 deben estimated from O. Gardiner 192 for
a crew of 40 workmen under Ramesses III.32 In other documents, the amount actually
delivered varied from 680 to 7390 deben per month.33
respectively.
31 By counting 3 or 4 deliveries per month for a fisherman and 3 or 4 fishermen delivering fish.
32 Janssen, Village Varia, 43. Estimated from O. Gardiner 192.
33 On these values, see Janssen, Village Varia, 43-47.
34 p<0.0001 when comparing gang size > 60 to gang size 51-60; p=0.0001 when comparing gang size <60
to gang size < or = 50 and p=0.09 when comparing gang size < or = 50 to gang size 51-60 by the
ANOVAtest.
35 P. Turin 1891, Rt 1-12 (Kitchen, Ram. Inscr. VI, 76).
36 Valbelle, Ouvriers, table III, p 104.
This content downloaded from 195.43.22.134 on Sat, 16 Dec 2017 12:45:14 UTC
All use subject to http://about.jstor.org/terms
30 J.-C. Antoine SAK 35
using fish delivery as the dependent variable.37 The results showed that both the period
and the gang strength were significant independent explanatory variables (p = 0.0001 each
of them).
A multiple linear regression taking the period as the dependant variable was similarly
performed. The strength of the gang and the value of fish deliveries were independent
explanatory variables while the season was not.38 Hence, in order of smoothing the
respective effects of these variables when comparing the two periods a weighting offish
deliveries by the strength of the gang was performed. Following this correction, the value
offish delivery in deben per member of the gang was significantly lower during the RIX
X period than in the RIII-RIV period: mean 5.19 ? 3.10 versus 9.79 ? 6.49 (p < 0.0001).
37 This method is based on the assumption that a linear relationship exists between different continuous
variables. To perform this analysis, it was necessary to transform the categorical variables into numeric
variables by attributing a discrete value to each of them: 1 to 3 or 1 or 2 according to the number of
categories in each variable. The result is significant (p<0.0001), counted on 152 complete data.
38 p <0.0001. Counted on 152 complete data.
This content downloaded from 195.43.22.134 on Sat, 16 Dec 2017 12:45:14 UTC
All use subject to http://about.jstor.org/terms
2006 Fluctuations of fish deliveries at Deir el-Medina 31
Discussion
The validity of results
Inherently to the nature of the sources, the analysis was retrospective and thus did not escape
the possibility of statistical bias. Thus, we have first to consider how the results are valid and
representative. As a whole, the assumptions used for the study were few. The main one
concerned the repartition between dry and fresh fish expected to be smoothed by the fact that
most fish was probably dried when delivered to the Village. A rough estimation suggests that
if the 186 data may represent 3.3 to 4.4% of all the records offish delivery produced along
the Twentieth Dynasty, this proportion reaches 16.8-22.5% when considering only the RIII
RIV and RIX-RX periods.39 The sample is thus representative of these two periods. Another
point to be underlined is the significance of a fish delivery. It is the production of a single
man over a period at the end of which he judged his catch suitable for delivery. His decision
to deliver fish was probably the result of a compromise between the availability offish in the
Nile River and the urge imposed by the administration to deliver regularly a fixed quota.
This content downloaded from 195.43.22.134 on Sat, 16 Dec 2017 12:45:14 UTC
All use subject to http://about.jstor.org/terms
32 J.-C. Antoine SAK 35
Fishing methods
Ancient Egyptians fishermen had several methods of fishing at their disposal,45 but we know
almost nothing about those used on the west bank of Thebes. Our results show that fish
production was closely linked to the annual Nile flood. In the RIII-IV period, fishing yield
is maximal when the river is returning to its bed. At this time, fishes trapped in shallow pools
abandoned by the receding flood are easily cached even by hand. This traditional way of
fishing is attested in Egypt since prehistoric times.46 The weights of Deir el-Medina bearing
the name offish species brought to the Village support these results. Among the identified
genus, the most frequently mentioned is Tilapia nilotica (wd) which lives in shallow water
near the riverbank and is frequently snared in the receding water pools. Conversely, Mugil
cephalus (r_Vr_w), which is the most important commercial fish in Egypt today and prefers
deep water, is seldom mentioned.47
45 See Brewer/Friedman, Fish and fishing, 21^6 and for the level of technology and skill needed by each
of these methods see Darby e.a., Food, I, 339-355.
46 Brewer/Friedman, Fish and Fishing, 5; Darby e.a., Food, 337-340.
47 See Valbelle, Catalogue des poids, 24-25. Fifty-two weights bear a fish name corresponding to 54
mentions. Among the 13 mentioned species, five have been identified. Beside those indicated above, there
are the M-fish (Mormyrus kannume), the whc and sir fishes (Synodontis schall Bloch-Schneider) and
the wf-fish (another species of Tilapia). All of these live in deep waters. The wd fish occurs 17 times,
followed by the d_\s-fish (11 times) which has not yet been identified. Other species represent from one
to six occurrences each. The cd/cdw fish (a Mugil?) occurs only two times. On the life habits of the
identified species, see Brewer/Friedman, Fish and Fishing, 48-81. On the identification of species, see
inter allia Valbelle, Ouvriers, 272-274 and the review in Janssen, Village Varia, 49-53.
48 When expressed in deben/crew members, the ratio is 1.89 indicating a reduction by almost a factor 2.
49 See note 41.
50 II Vs 1-7, Kitchen, Ram. Inscr. VI, 581-590. It is sometimes difficult to reconstruct the calculations as
the text contains many lacunae in addition to probable scribal errors. Roughly, data suggests that in year
15, 9686 deben of fish were delivered when 11412 were expected, these were 2358 in year 16for4722
deben expected. An entry for the same years, gives 9612 and 4000 further deben of missing fish. Year
17 is much damaged, but dues are also mentioned. On this text, see Janssen, Village Varia, 45-46 and
Helck, Materialien V, 827-828.
51 Kitchen, Ram. Inscr. VI, 581, 7.
This content downloaded from 195.43.22.134 on Sat, 16 Dec 2017 12:45:14 UTC
All use subject to http://about.jstor.org/terms
2006 Fluctuations of fish deliveries at Deir el-Medina 33
abbreviated entries correspond to the same operation.52 The regular meaning of kiy is
? sycamore-fig". Here, it has tentatively been translated as ?roe" by Helck and Janssen
although the latter recognises that there is no other example of this meaning.53 Interpreting
this passage in a context offish shortage permits to keep the usual meaning of the word and
explains the strange operation performed here. A similar process is possibly attested in P.
Turin 2075 dated from year 19 of Ramesses IX and year 1 of Ramesses X where corn was
exchanged for fish. Large amount of missing fish (circa 5010 deben for 11100 delivered) are
also mentioned in this text.54
Fish deficits are not specific of the RIX-X period. They also occurred under Ramesses III
and IV. If we assume as proposed by J.J. Janssen a quota of 5520 deben per month, a
monthly deficit of 240 deben can be deduced from O. DM 142 for IV iht of year 26 of
Ramesses III and 710 deben from O. DM 394 for II prt of year 29.55 O. DM 143 indicates
a more severe shortage of 3480 deben in IV Smw of an unknown year of probably Ramesses
III.56 O. DM 41 and 42 suggest a deficit for year 1 of Ramsses IV.57 However, comparatively
to the time of Ramesses IX and X, the shortage recorded in these documents is usually minor
and is seemingly not imbedded into a pluriannual trend.
Several hypotheses may explain the situation at the time of Ramesses IX and X. As the
fishermen organisation was fundamentally unchanged at this date, administrative reasons are
unlikely. Social perturbations well illustrated in the Theban area all along this period may
have prevented the fishermen to exert their activity normally.58 However, several arguments
point for a role of climatic factors. Thus, the annual fluctuation offish deliveries observed
under the Ramesses III-IV period is absent under Ramesses IX and X (figure 4).
Furthermore, the lowest values offish delivery occurred between February and May when
the yield was maximal under Ramesses III and IV. Another argument can be drawn from an
entry of the Turin Necropolis Journal of year 15 of Ramesses IX stating59: ,J Smw
52 Vs B2a: 1; B2a:2; B2a:3; B2a:8; B2a: 10; B2a: 12: Kitchen, Ram. Inscr. VI, 581-583. The exchange value
varies from one entry to another.
53 Helck, Materialien V,823; Janssen, Village Varia, 53.
54 Kitchen, Ram. Inscr. VI, 653-655 and 682-684. The passage is as follow: ?(Rt 3:l)Khonsumose,
following the fisherman Bakentjaao and the fisherman Amenemope. He (3:2) took the quota (iw.fSd
blkw). Coming under his hand: 110 deben offish. (3:3) Coming under his hand: 8 khar of corn making
3200 deben offish" (Kitchen, Ram. Inscr. VI, 682).
55 Janssen, Village Varia, 43?44.
56 Janssen, Village Varia, 43. The ostracon is dated to the end of the reign of Ramesses III. O. DM 237
which mention a huge deficit of 24,700 deben for two fishermen over six months is to be dated from the
Nineteenth Dynasty according to Janssen, Village Varia, 40.
57 O. DM 42: 2180 deben delivered for 2/3 of the month. O. DM 41: 680 deben delivered for one month.
O. DM 721 from the same year could indicate a compensation. In line 6 and 7 of this document, we
read: ?Boutefgereg (?): two dliw garments, making four pieces of cloth (?), making 3000 deben offish".
58 The process of compensation by the fishermen themselves offish deficit with an equivalent quantity of
sycamore figs or grain is more likely indicative of a shortage of fish availability than of deteriorated
working conditions which would have prevented any fisherman activity including the search for a
compensatory food.
59 Vs B 2a: 1. Kitchen, Ram. Inscr. VI, 581-582.
This content downloaded from 195.43.22.134 on Sat, 16 Dec 2017 12:45:14 UTC
All use subject to http://about.jstor.org/terms
34 j.-c. Antoine SAK 35
10+x.300+ xfish.oipe,60 making 110 deben. Pentawer 3A *A */16 oipe making 610
+ x.brought from the great inundation61". In the Gregorian calendar, the date of this
entry corresponds to the end of January when the Nile is expected to have returned to his
regular bed. Mention of a still inundating Nile at this time suggests a delayed or excessive
and prolonged flood.62
Fish deliveries as a tentative proxy of the Nile flood levels fluctuation during the Twentieth
Dynasty
Studies of the flood levels recorded at Roda Island between the seventh and fifteenth century
AD have revealed several abrupt changes and separate wet and dry episodes of the Nile
regimen on a time scale spanning from 30 to 160 years. Interestingly, these episodes are
closely linked to reports of starvation, plagues, desert tribe raids, or important historical
events.63
In the Twentieth Dynasty, fluctuations offish deliveries can be detected over several years
for which there is a reasonable number of available data. The curve suggests a fish shortage
between 1110-1103 BC (from year 15 of Ramesses IX to year 3 of Ramesses X). It also
shows that if in the RIII-RIV period fishing usually gave a yield above the mean, in at least
two years, namely 1154 BC (year 28 of Ramesses III) and 1148 BC (year 3 of Ramesses
IV). An abnormal situation can also be suspected for year 1 and 5 of Ramesses IV. These
results are tentative as there is frequently a few available data for each year. However, if, as
hypothesised here, fish deliveries were indicative of the fluctuations of the Nile flood level,
we may expect that the effect of an abnormal inundation on agricultural production and food
availability is reflected elsewhere in the documentation. Table 3 shows that the index
generally fits well with the social events recorded in the Theban Necropolis. Interestingly,
years with the lowest fishing index correspond to the report of hunger or shortage offish,
grain, or vegetables in the same year or the next one.64
60 K. A. Kitchen places here a word sfd (?) that gives no sense. One would expect rather a quantity as in
the similar entries in vs B2a:8 and 10.
61 hrpy rl (great inundation) is the usual phrase to designate the Nile flood. The adjective rl has probably
no particular significance regarding the eighth of the flood and should be distinguished from wr which
probably designates an exceptionally high flood. See J.J. Jansen, The day the inundation began, in: JNES
46, 1987, 129-136.
62 Whether this mention of an abnormally prolonged flood is indicative of a pluriannual trend is unknown.
Studies of the fluctuation of the level of several African lakes rather suggest that the beginning of a dry
period coincided with the end of the Twentieth Dynasty. Occasional excessive floods may be interspers
ed within a period of low inundations. See K.W. Butzer, Early Hydraulic Civilization in Egypt, 1976.
63 See J. Jiang/R. Mendelssohn/F. Schwig/K. Fraedrich. Coherency Detection of Multiscale Abrupt Chan
ges in Historic Nile Flood Levels', in: Geophysical research letters 29, 2002,112-1-4 and R. Said, The
River Nile, Geology, Hydrology and Utilization, 1993, 320.
64 It should be remembered that the Necropolis Journal or its abbreviated equivalent on ostraca in which
we usually found the recording of these events are not available for each year and that, when present,
they do not cover all a given year. Indications of hunger are recorded in year 9 (O. DM 571) and 13
(Turin Necropolis Journal, year, 13 Vs 1,4: Kitchen, Ram. Inscr. VI, 563) of Ramesses IX for which
there is no or to few data concerning fish delivery and year 16 (P. Turin 1884, Rt 1,19: Kitchen, Ram.
Inscr. VI, 646) of the same king. Starvation is also mentioned in year 1 of the Renaissance Era (P. BM
This content downloaded from 195.43.22.134 on Sat, 16 Dec 2017 12:45:14 UTC
All use subject to http://about.jstor.org/terms
2006 Fluctuations offish deliveries at Deir el-Medina 35
Conclusion
By its size and homogeneity, the series of documents recording fish deliveries at Deir el
Medina during the Twentieth Dynasty fulfils the criteria required for a statistical analysis.
Although the conclusions of this study may be modified by the discovery of new documents,
the results fit in with a well-illustrated historical and social framework giving us the
opportunity to confront statistics to evidences of a quite different nature. Fish delivery
illustrates several aspects of the social and economical life of the community of the Tomb.
In a broader point of view, they may be an indirect reflect of the fluctuations of the Nile flood
level. A major finding of this study is the collapse offish delivery under Ramesses IX and
X leading to the hypothesis that this period was marked by a series of abnormal floods.
Although climatic factors were probably not the only cause, their consequences may have a
heavy influence on the social perturbations that occurred in Thebes at this time.
EA 10052, Vs 11:8: Kitchen, Ram. Inscr. VI, 791). This Era is usually considered as following year 19
of Ramesses XI. It as recently been suggested that year 19 may belong to the reign of Ramesses IX: see
A. Thijs, Reconsidering the end of the Twentieth Dynasty. Part I: The fisherman Pnekhtemope and the
date of BM 10054, in: GM 167, 1998, 95-109 and Part II, in: GM 168, 1999, 83-99.
This content downloaded from 195.43.22.134 on Sat, 16 Dec 2017 12:45:14 UTC
All use subject to http://about.jstor.org/terms
36 J.-C.Antoine SAK 35
Appendix
This appendix gives the list of the 66 documents used for the statistical analysis and classified according
to the reign. Each document furnished from 1 to 27 data, mean 2.8, median 1. They are schematically
arranged by chronological order.
Ramesses III:
P. Greg; O. DM 140; O. DM 137; O. DM 139; O. DM 32; O. Petrie 50; O. Glasgow 67; O. Glasgow
76; O. DM 148; O. DM 142; O. Berlin 12629; O. Turin 57153; O. DM 33; O. DM 427; O. Berlin
12627; O. DM 34; O. Cairo 25683; O. DM 394; O. DM 855; O. DM 153; O. Berlin 14666; O. DM 154;
O.DM158;O.DM155;O.DM36;O.DM153;O.DM38;O.DM39+174;O.DM281;O.DM653;
O. DM 247; O. DM 80; O. Berlin 11272; O. Brussels E 3214; O. Michaelides 33; O. Cairo 25697; O.
DM 78; O. Leipzig 14; O. DM 689; O. Turin 57125.
Ramesses IV:
O. Berlin 12631; O. DM 726; O. DM 40; O. DM 41; O. DM 42; O. DM 43; O. DM 47; O. DM 44; O.
DM 45; O. DM 46; O. Gardiner 113; O. DM 694; O. DM 401; O. DM 196; O. DM 395. O. Berlin
14156; O.Gardiner 253.
Ramesses IX:
P. Turin 1881; Turin Necropolis Journal of year 17; P. Turin 2075.
Ramesses X:
P. Turin 2975; Turin Necropolis Journal of year 3.
This content downloaded from 195.43.22.134 on Sat, 16 Dec 2017 12:45:14 UTC
All use subject to http://about.jstor.org/terms
2006 Fluctuations of fish deliveries at Deir el-Medina 37
Table 1
Table 1
Evolution offish deliveries in deben (mean + standard deviation) from the end of the reign of Ramesses
III to the first years of Ramesses IV before and after the increase of the gang strength to 120 men. The right
hand column gives the mean value ? standard deviation offish deliveries weighted by the gang strength.
The values of ?p" by the Mann-Whitney test are given.
Table 2
Fishing index, king, and reign years, number of preserved data per year, and events recorded in the same
year in the Royal Tomb documentation. An asterisk indicates a year for which the Necropolis Journal or
its abbreviated form on ostraca is preserved at least in part. A cross indicates a year for which there are
documents recording events at Deir el-Medina other than the Necropolis Journal.65
65 For a chronological record of all the events occurring at Deir el-Medina see Valbelle, Ouvriers, 49-54
and Janssen, Village Varia, 147-173.
This content downloaded from 195.43.22.134 on Sat, 16 Dec 2017 12:45:14 UTC
All use subject to http://about.jstor.org/terms
u> GO U)
1154 Ramesses I I, year 281 0.702 5 Enemy incursion.1 Strike in year 29. "Hunger. There is no fish and no grain in the granary".2
1108 Ramesses IX, year 17* 0.954 3 Compensation of fish by sycamore figs.6 Workmen in hunger. Grain shortage.7
1104 Ramesses X, year 3* 0.947 7 The authorities have difficulties to provide the workmen grain salary.9
1110 Ramesses IX, year 15* 0.479 5 Compensation of fish by sycamore figs. Abnormal flood recorded.5
BC index of data
1118 Ramesses X, year 7 1.189 3 <* 1105 Ramesses X, year 1 -2 0.682 16 1103 Ramesses X, year 3* 0.635 14
1157 Ramesses III, year 25* 1.147 7 1156 Ramesses III, year 26f 1.429 13 1152 Ramesses I I, year 30* 0.941 3 1151 Ramesses I I, year 31* 1.298 7
Table 2
This content downloaded from 195.43.22.134 on Sat, 16 Dec 2017 12:45:14 UTC
All use subject to http://about.jstor.org/terms
OOOn p sr ^.5" i
&
are in hunger "because there is no wood, no vegetables, and no fish". This situation fits with a low fishing index as it is the case with year 3 of g*
4. Several documents suggest that year 3 of Ramesses IV was very troublesome. O. Sidney, Nicholson Museum, Inv R.97: Kitchen, Ram. Inscr. VI, g
Ramesses IV or with year 1, which is also marked by a probable fish deficit as indicated above. P. DM 24 (Kitchen, Ram. Inscr. VI, 134-135) dated from <d
Vil age Varia, 163-164, dates this ostracon from the early years of Ramesses IV and possibly year 1. In this ostracon, the workmen declare that they g*5 interprets this document as an indication of shortage of supplies at this date (Vil age Varia, 164). Valbelle, Ouvriers, 197, note 9, attributes O. Cairo O
1. 0. DM 35, 9 (Kitchen, Ram. inscr. V, 520). On II Peret, 24, the chiefs Medjai report, "the enemy is descending". This enemy is probably some desert IV iht 10 of year 3 relates an unusual delivery of goods including fish in the exceptional presence of the Vizier and the High Priest of Amun. J.J. Jans en ?
clothes, no fat, no fish, and no vegetables". In Rt 3,3 the vizier To explains that he is unable to provide the workmen with their grain ration as "it happens
1. Turin Necropolis Journal of year 17. In Rt B1:8-15 : Kitchen, Ram. Inscr. VI, 571-572, it is stated several times that the "crew of the Tomb is idle (wsf)
2. Turin Strike Papyrus, Gardiner, RAD, 45-58. In Rt 2,3-2,4, the workmen declare that they are on strike "because of hunger and thirst as there is no
151-152. C.J. Eyre, in J. Ruffle/G.A. Gaballa/K.A. Kitchen, in: Studies in Honour of H.W. Fairman. Glimpses of Ancient Egypt, 1979, 80-91. Janssen, ^
9. Turin Necropolis Journal. Year 3 of Ramesses X. Rt 3,24-4,3 : Kitchen, Ram. Inscr. VI, 694 and Valbelle, Ouvriers, 216-217.
3. On the possibly troubled first year of Ramesses IV, see Janssen, Village Varia, 161-162. & 25533 to year 3 of Ramesses IV. This document among other things relates a strike. 2
because they are hungry", "their being not the ration of I pn" or "hungry and weak".
5. Turin Necropolis Journal of year 17. See above note 61. ft 6. Turin Necropolis Journal of year 17. See above note 61. g
dweller tribe.
This content downloaded from 195.43.22.134 on Sat, 16 Dec 2017 12:45:14 UTC
All use subject to http://about.jstor.org/terms
40 J.-C.Antoine SAK 35
Figure 1
Error Bars: ? 1 Standard Error(s)
16 j A ] | /{ I
^ I I A \sk I/'asJ \ I "*" R^m
= 8 H txj/ A ^J/ a $ r "^~ RIX"RX
The intra-annual evolution offish deliveries in deben per crew member in the two historical periods.
A, monthly evolution. The mean number of data preserved per month is 8.9 (7-29) for the RIII-RIV period
and 5.45 (1-9, except July for which there is no available data) for the RJX-RX period.
B, values are grouped into three equal fishing periods of four months.
* indicates an intra-historical period and
t an inter-historical period statistical significance (p <0.05).
In A and B, the line with open circles indicates the ideal monthly variation of the Nile level in an arbitrary
unit.
This content downloaded from 195.43.22.134 on Sat, 16 Dec 2017 12:45:14 UTC
All use subject to http://about.jstor.org/terms
2006 Fluctuations of fish deliveries at Deir el-Medina 41
Figure 2
Year BC
Evolution of the fishing index with years and reigns. An index > 1 indicates a fish delivery above the
general mean and an index <1 indicates a fish delivery inferior to the general mean. The year dates BC are
indicative only.
This content downloaded from 195.43.22.134 on Sat, 16 Dec 2017 12:45:14 UTC
All use subject to http://about.jstor.org/terms