 |
The second Egypt-based paper I wrote for my degree. Another part of my MA involved a course called "Sources and Methods in Ancient History", it included a lecture on papyri and I ended up choosing it for one of my essays.
Again, the footnotes have been removed by the posting of it but I've added the bibliography at the end (I highly recommend the book by Rowlandson).
WORD OF WARNING: In Microsoft Word it is 10 pages long!
What are the Problems in Writing the History of the Family from Egyptian Texts?
In the study of Greek and Roman Egypt we have a decent amount of written evidence that has come down to us in the form of papyri, private letters, official documents, census returns, these things have survived and helped us build up a picture of life in Egypt. We know about administration, private squabbles, gossip and the movement of armies to name just a few aspects. But the papyri also create a lot of problems. In this essay I will look at the history of the family, the problems we face when trying to build this history, and the reasons for some of these problems.
The first problem is the condition of the evidence. In the dry climate of Egypt papyrus survives, but rarely will it survive in good condition, often coming to us in fragments that then have to be pieced together before being translated and released. Census records can be so badly damaged that all we have a small corner, or the age of someone on the record but no name. 131-Ox-6 is a perfect example of this, all we know is that there are four people on the record, the declarant’s name begins with “Plou” but the rest of the name and their age is lost, while we have three other ages (62, 68 and 7 cool but no names so no indication of gender or possibly relationship between them. Bagnall adds that after the fourth entry the text is so damaged that we cannot get any more information from it even though there may have been more members of the household listed . Another example is 159-Ar-5 , which has a long list of twenty two family members, of which we have five names and six ages. Luckily most of what has been preserved is the relationship between the members of the family, which also gives us the gender of the household members without needing the names. They also have a list of five slaves, of which we have four names and one age. But Bagnall points out that some of the families mentioned could be renters, and the status of the so-called slaves is not fully clear either , which just goes to show how a damaged census can give us so much information but leave us with plenty more questions that cannot be answered.
Not only that, Bagnall points out that in the case of papyri from Oxyrhynchus, a lot of it is simply sitting in storage waiting to be edited, and a lot of lists and registers are also waiting to be translated due to editorial choice and Rowlandson says similar about texts from Antaiopolis . There may be key texts sitting in archives, which had not been translated because they do not appear to be important enough. Even with the texts we do have, those that are fragmentary give us an incomplete picture, P.Oxy. XII 1547 (see handout) tells us about the writer, his wife and two of her brothers, but the end is missing so we do not know if they have children or if the wife has any more brothers or a sister.
Sometimes the excavation alone can be a big problem. In his book on soldiers in Roman Egypt Alston discusses the village of Karanis, which was excavated between 1928 and 1935 . By the time archaeologists managed to organise an excavation an Italian company was taking ancient rubbish called sebakh and selling it as fertiliser, the Egyptian government required the archaeologists to continue providing the company with the sebakh, if they wished to carry out their excavations . This meant not only that the excavation was rushed and bits of it would have fallen below modern scientific standards , but also some vital remains (including fragments of papyri) may have been missed and handed over to the fertiliser company. Archaeologists could find half a text while the other half is in a field somewhere, helping crops grow.
We also have problems with the locations of the evidence. Some times only a single papyrus text will be discovered, but at other times a large collection of texts, called an archive, may be found. However in the late 19th to early 20th centuries, people that discovered archives would often separate them and then sell them on to various scholars and agents of museums and private collectors, by separating them the sellers sold more pieces for more money but this was to the detriment of our understanding of the texts. This means that we are unsure of where the texts were found, thus removing a great deal of our understanding of the context of what we have, and by separating the archives we have pieces of texts scattered throughout various collections, one example of this is “archive of Apollonius the strategos and Aline” used by Rowlandson, who says that an archive of over 200 texts was dug up in Hermopolis and sold by dealers to various institutions . Separating texts also makes it difficult to work out if any are collected, there are many common names, such as Thoonis, which repeatedly crop up in our sources, but if these names appear in one text in Berlin, one in London and two more in Boston, it is difficult to tell if they were once from the same author.
Even if we do not know the exact location of texts when they were dug up, items such as official documents may tell the reader where they were composed or sent to. The main documents that do these are census records which record where a family is living at the time. But we run in to problems here due to the random nature of text survival. Bagnall and Frier have a table on census returns with data on households, which shows that a large number of texts (144 of 233) come from the Arsinoite area, the second largest collection is only 30 texts from Oxyrhynchite, but nothing survives from Antaiopolite, and Berenike and Great Oasis only have one surviving census text each . This means that our evidence in terms of demography is heavily biased in favour of one area. Bagnall also says that the returns are more urban, with 49 percent of the texts coming from nome metropoleis (where the other 51 percent come from it not mentioned). And it is not just a problem with geographical areas, chronologically speaking we have very little evidence from most of the first century, our evidence is mainly from 117 AD to 215 AD, at which point the amount of evidence trails off again . So our evidence from census records is heavily biased towards second century urban areas, especially Arsinoite, while records for families living in villages or in an earlier period are scant at best. It is difficult to get an idea of how people lived, and to make estimates on things such as common family size and age at death and how these may have changed over the years, when the evidence is so focused on one small area. It also means that if a family moved we do not know about it, one census record placing a family at Fayum for example does not mean they were still there fourteen years later, making family history difficult.
Even when the census papyri are in decent condition and give us plenty of information about various families, we still come across problems with accuracy and omissions.
One big inaccuracy is age. The census was taken every fourteen years, and fourteen was the age at which one had to start paying poll tax. Bagnell says that there were two ways for families to avoid this, by either concealing young males or declare their age as younger than fourteen . He suggests that concealment was more common in villages while in urban areas families were more likely to falsify ages . This is probably because in urban areas there were more people around, someone might inform officials that a family had a son, whereas very few people would be argue against age.
Bagnall also points out that it is common for people’s ages to be divisible by five, and uses the “Whipple Index” to work out if Egyptians did this regularly . Populations in poor countries tend to do this more than others, but ancient Egypt is relatively low on the scale, suggesting that ages may have been more accurate than we realise . There are still census returns that do this though, 131-Ox-1 lists fourteen members of the household, ages for nine of them are visible and only one is not divisible by five, that of a three year old boy whose relationship to the main family is not known . If we had the ages for the other six members of the household we may have known if the other ages were also rounded, but it is odd that only one young person on the list has an age that is not rounded. Either one member of the household was better at counting than the others, or they decided he looked younger than five but was clearly not a newborn so chose an age in the middle.
We also have omissions, family members that are missing from one census but appear in a later one with an age that suggests they should have been registered in the previous census. One example used by Bagnall is from Hermopolis, in 187-Hm-1 there is a family that consists of the father aged 47, his wife aged 51, three sons aged between 21 and 8, and a daughter called Isidoros aged 0, but in the census records 28 years later, 215-Hm-1/2 the family now consists of two of the sons, and a “full sister and wife” of one of the sons, called Dioskorous who is apparently 30 years old . She should have been two years old and thus mentioned in the first census, but is not. Bagnall adds that this is not uncommon, and points out that other metropolitan declarants may have failed to report girls under the age of 5 . The reasons for this are unclear, one suggestion is that girls were not valued as highly boys . I think one reason may be due to the high rate of infant mortality, especially among girls, the parents may have seen no point in registering a daughter until she was past a certain age where her continued survival would be more likely. But again it makes writing family history difficult. Personal letters, which I will discuss in greater detail later, are more reliable when it comes to such information as they tend to involve the mother or father informing a family member of the new arrival, for example one woman writes to her mother telling her she has just had a baby . This child may not show up on a census record (if we have one for this family) until it was older but among family at least it’s birth has been acknowledged.
Another problem is the death of a person while the census is taking place, Bagnall notes that declarants were permitted to file returns up to a year after the census year, but did not declare persons who had died in the mean time, listing only 3 census records which do this . This is only really a problem when it comes to collecting statistics of ages at death, if a person appears on one census and is not there 14 years later on the next one, that presumably means they died at some time in those 14 years, but even with their age known from the last census we cannot know at what point they died, it may have been months after the previous census, or 5 years later. We can make an educated guess on statistics about age at death, but they may far more off the mark than people think. We have a similar problem in some of our letters and other texts, which creates another problem with writing family history. In the archive of the weaver Tryphon, whom I will discuss in more detail later, we find his mother and brother Onnophris are no longer mentioned after 44AD . But we do not know if they died, and if so when and why, or if they moved to another part of Egypt.
The wording of the censuses also causes a problem, namely the word “oikia”. Bagnall says that the census was organised by household, using the word oikia, but there are many different means to this word . An oikia can be a family of blood relatives, a household of a family and their slaves, or a building, in which one or several families or households might live. Bagnall says that even the officials taking the census might be confused, as the census was used to keep track of households, men eligible to pay tax, and all habitable land, and cites five examples of married couples registered in different returns due to the property they or their extended families own . He also points out the confusion over non-kin lodgers, who share a space in the home but live separately, and might be listed as part of the owner’s household, or be registered separately . One text from Arsinoe is used as an example as one cannot fully ascertain if we have one large household or several smaller ones listed as residents at the property .
Our second lot of evidence comes in the form of letters. These can be personal ones sent between family members, or more official ones in forms of appeals from one person in trouble to a government official that they believe will be able to help them.
But again we have problems with these documents as well. Along with the splitting up of archives, as already discussed, we have the problem of selective survival of texts. This is more of a problem when it comes to official appeals. For example Rowlandson has several letters from a temple to Serapis, written by a man named Ptolemaios on behalf of twin girls, Tages (also known as Thaues) and Taous. The first one she discusses gives us a very interesting family story, the mother of the girls, named Nephoris, left their father and moved in with another man, Phillipos. After an attempted murder the father, Hargynouti, fled to Herakleopolis and died there, his brothers collected his body but from the sounds of it they could not afford a funeral. Nephoris took her husband’s possessions, and presumably his property as she is apparently receiving rent, and threw her daughters out. The girls took sanctuary in the Serapeum, and took part in the mourning of the Apis bull when it died. Friends of their mother persuaded them to accept her son (presumably a half brother of their) as their attendant, but he stole the few belongings they had at the Serapeum and a few other items and then returned to Nephoris . The text breaks off there, but a second version was found in different handwriting and with corrections added, in much better condition than the first version. The twins appealed for the king Ptolemy to persuade other officials to withhold items from their mother and persuade her to return the property she has stolen . We know from another letter written by a colleague of Ptolemaios, called Hermais, that the twins had another sister named Tathemis, who was brought to the Serapeum and left under the guardianship of Hermais himself. Nephoris showed up again claiming to need money for various wedding preparations for Tathemis, but once she was given the money she disappeared again . Rowlandson adds that twins continued to petition the king and queen for help with other problems that cropped up . Clearly we can see that letters give us a more personal history of a family, but still miss information that we get from the census records. We do not know the age of any of the family members mentioned, or the girls themselves. We do not know Nephoris’ side of the story, or anything about the man she moved in with, nor do we know if they had any children together. We also presumably do not know if either of the twins left the sanctuary of the Serapeum and married, information that we might get from the census, and we have no replies from any government official on behalf of Ptolemy either responding positively to their requests for help or saying there was nothing he could do.
We can use other petition letters to build up an idea of a family tree, something that we can also do using census records. Rowlandson uses letters to put together several family trees, for example the family of Dryton and Apollonia alias Senmonthis and the family of Tryphon and Thamounis/Thamounion . Generally these family trees come from official petitions, but at one point we get the minutes from an appeal to the prefect by Tryphon, over a problem involving his baby son and a missing slave child . This sort of personal problem is not something we would get from the census records, which are there for purely formal reasons. This type of evidence brings the history of a family to life, we no longer know their names and how they are all related to each other, we also know how they lived, that at one point they were in financial difficulties and so the wife, Saraeus, became a wet-nurse, but was then accused of stealing the baby after her own son died. A census might go as far as to include the name of the child, and possibly the occupation of the mother, but never the accusations thrown at her.
Letters can also show us the dynamics of the family unit. Letters between the family members of Apollonios the Strategos are a good example of this, from these we know that the daughter of Apollonios and his wife Aline lived with her paternal grandmother Eudaimonis in Hermopolis . Rowlandson also adds that Aline seems to have lived with Apollonios with their other children . This is something that we would probably only find out about if their census returns showed the property they were registered at, but we face the problem mentioned before of whether “oikos” meant family or building.
One letter from Eudaimonis to Aline assures her that her daughter is doing all her schoolwork. We also know that Aline is pregnant again as Eudaimonis says she prays it will go well and a son will be born. Whether or not they did have a son, or subsequent children, is something a census record might tell us if we do not have any more letters.
In an archive from the late 3rd century AD we have a group of letters from Paniskos to his wife Ploutogenia. From these we can see that they were living in separate cities, Paniskos tries several times to persuade his wife to come and live with him and bring their daughter and a list of items, including his shield, jars of wine and his lances . But she does not move to be with him, nor does it appear that she even sent his items to him. He admonishes her for instead moving to a property belonging to either her or her family . Again this is something we would not know about if we only had census records, we would probably know about the other property but not necessarily who lives there, as we saw before people could be registered at several different properties without us knowing which one was being used the family home.
There is one other possibility that could explain this odd naming within the family. In China a type of marriage known as sim-pua was quite common in certain parts of the country . When a poor family had a daughter they would give her away or sell her to another family, who raised her as an adopted daughter in law, once she was older she would be married to their son . There were many reasons for this, a wedding was a big social occasion that cost a lot of money, whereas if your daughter in law was already living with your family it would drastically reduce the financial drain . It was also a way of making sure the daughter in law did not cause trouble in the family, she was less likely to argue with her mother in law if she saw her more as her own mother, likewise the mother in law did not feel threatened by a strange woman entering the house, she had already dominated the girl for years and a wedding would not change this, and in some cases she may view the girl has her own daughter rather than an in-law . It also meant that if there was conflict between the two women, the girl had no other family to run to for help, reducing outside interference in family affairs . While this type of marriage is only seen in east Asia, mainly China and Taiwan, it could be that a similar situation occurred in ancient Egypt. With a high mortality rate and a decent age gap between husband and wife, a woman could end up raising children from both hers and her new husband’s first marriage. It would take a while for step-siblings to get along and view each other as brother and sister, and marrying them could make sense for the family wanting to keep property in the family line without worrying about any problems arising from a fully incestuous marriage (such as birth defects). Even if their parents were not married, children in families who shared buildings and farm land together would grow up having a close relationship, and both families would view a marriage as strengthening their ties of friendship. It is only really census records (and the occasional wedding party invitation or marriage contract) that tell us if we have a fully incestuous marriage.
One problem stems from sibling-marriage. This was common among Egyptians until Roman citizenship was granted to them in 212 AD, although Rowlandson says that there were some families that were still doing this towards the end of the third century . This can be heightened by the problem of common names, in his article on close-kin marriage Shaw discusses a book by Goody which said that incestuous relationships were a big part of Mediterranean society . Goody’s explanation for this was the banning of such marriages by the Christian church, forcing people to marry outside the family group, and it was based on a few literary texts and some changes in the law by a Christian emperor . Shaw disputes this on the basis that closer examination of evidence does not support this theory . One thing that Shaw says makes determining these relationships difficult is the Roman way of naming, children took the nomen of their fathers, which sons then pass on to their own children, but shared names are not a guarantee of such a marriage, just an indicator that it is a possibility . Not only that, but when slaves were set free they took the nomen of their former master, a male and female slave set free by the same master would share the nomen but would not be blood relations, and it can be difficult to work out if someone is a freed slave as they did not always identify themselves as ex-slaves . Finally, there are some names that were common than others, for example the name Julius . Common Greek names such as Apollodorus and Dionysius can be thrown in to the mix as well, although this problem can be eased when we find some people with both Greek and Egyptian names.
Some times it is easy to work out if a marriage is a sibling one, census records tell us if a married couple are also brother and sister, for example household 159-Ar-11 is the record for the family of Petheus in which we have Petheus whose wife Dideis is also his sister, and their son Isidoros is both brother and husband to Dideis . But in letters it can be less clear, returning to Apollonios and Aline, Rowlandson says that scholars thought the pair of them were in a sibling marriage because they occasionally “refer to each other as ‘brother’ and ‘sister’, while Eudaimonis sometimes called Aline ‘my daughter’ in letters addressed to Aline and Apollonios” . But she dismisses the idea because in letters to Apollonios, Eudaimonis refers to Aline as “your wife” , but this does not entirely convince me. Partly because “your wife” may have been used to avoid confusion if Apollonios had several sisters, and also because in another letter to Eudaimonis says she is praying for the health of Aline and “your brother Apollonios” . But on the other hand, referring to each other as mother, daughter, brother or sister may simply be indicative of a close family unit, if Aline’s own mother had already passed away she may end up closer to her mother-in-law Eudaimonis. The census records would be far more useful here as no doubt they would tell us of the relationship between them, but without them we can only speculate.
Another reason for difficulty in ascertaining if a marriage is an incestuous one is the problem of common names. In his article on close-kin marriage Shaw discusses a book by Goody which said that incestuous relationships were a big part of Mediterranean society . Goody’s explanation for this was the banning of such marriages by the Christian church, forcing people to marry outside the family group, and it was based on a few literary texts and some changes in the law by a Christian emperor . Shaw disputes this on the basis that closer examination of evidence does not support this theory . One thing that Shaw says makes determining these relationships difficult is the Roman way of naming, children took the nomen of their fathers, which sons then pass on to their own children, but shared names are not a guarantee of such a marriage, just an indicator that it is a possibility . Not only that, but when slaves were set free they took the nomen of their former master, a male and female slave set free by the same master would share the nomen but would not be blood relations, and it can be difficult to work out if someone is a freed slave as they did not always identify themselves as ex-slaves . Finally, there are some names that were common than others, for example the name Julius . One letter from a son to his father reveals that they are both named Theon , with an archive (either complete or dispersed) but without a census record it would be difficult to work out which Theon is which.
Another type of source can give us glimpses of family history or legal texts, especially contracts. Thanks to an archive of papyri we know a decent amount about a man named Tryphon, who came from a family of weavers and became one himself . Two of his texts include apprenticeship contracts for his brother Onnophris and later his son Thoonis . The contracts show us a bit about the family dynamic at the time, Brewster states that the contract was drawn up for Onnophris by Tryphon and their mother Thamounis, there is no evidence of a contract for Tryphon himself which suggests he was trained by his father Dionysius , the fact that Onnophris had to be sent to someone else for training suggests that Dionysius was incapacitated at this time and so unable to train his son himself something which would not come up in a census. Furthermore the contract for Thoonis may mean that Tryphon was now unable to train his own son, Brewster says that by 52 AD he was suffering from partial sight loss caused by a cataract . The terms of the contract include who will pay the required taxes and who will cover the expenses of food and clothing for the apprentice , things that are not essential to family history but which may give an indication of wealth. Some contracts may specify if the apprentice is living with his parents or with his teacher, but again this is not particularly important. Another contract involves a slave being sent by his master to learn stenography, the contract stipulates that he will be taught for two years at a cost of one hundred and twenty silver drachmae paid in various instalments depending on whether the education the slave receives is satisfactory . Again this shows the wealth of the family, that they can afford to have a slave apprenticed out for two years (although obviously this would have greatly added to the value of the slave), and also gives an indication as to how the slave’s owner can protect himself against a bad teacher (he will only pay the full amount if the slave can show his skills).
In terms of education, not everyone was sent to become an apprentice. One letter from a son named Thonis to his father Arion reveals that the boy has been sent away to a school, presumably one that can teach him better than one local to his family. In the letter the son is complaining that his father has only written back one despite five letters being sent to him, and has not been to visit him to make sure he is being treated well . It also suggests that the father owes the boy’s teacher money for the education, as the teacher has apparently not started teaching the boy yet but is constantly enquiring about when the father is going to visit . Like the apprenticeship contracts that may tell us where the apprentice is living this letter shows that families would occasionally be split up as sons were sent away to get a better education. It can also give an indication of family wealth, along with paying the travel costs and the education costs the family would also have to meet the son’s living expenses. Whoever he lived with would expect a contribution towards feeding and clothing him. The boy might help the family by doing extra work for whoever he was living with but that does not appear to be the case with Thonis.
Finally we have one big legal process that changed families, adoption. A family that did not have children, or at least did not have a surviving son, could adopt a child. Likewise a family that could not afford to raise a child but did not want to expose it could give their child to a family to adopt. We have several contracts dealing with this, the first is between Heracles and Isarion who are giving their two year old son Patermouthis to a man named Horion . Horion declares that he holds Patermouthis as his son and will not send him away or sell him in to slavery as the boy is freeborn, while the biological parents declare that they will not take the boy away from Horion . In another contract a woman puts her daughter up for adoption as she cannot afford to keep her, she states that her husband died, but now the girl is nine years old and the mother cannot look after her any more, if she changes her mind and decides she wants her daughter back she has to pay the adoptive family the costs they incurred from raising her . This not only shows us the desperate poverty that some families were reduced to, but could also explain why some children suddenly appear in the census records. It is also quite interesting that daughters were put up for adoption, in the Greek world it was generally only sons that were adopted, but as we saw, girls were adopted as well.
Overall we can see that when writing the history of the family we face a huge number of problems. Fragments and the separation of archives means we have an incomplete picture, the lack of context due to removal of papyri from areas by people other than archaeologists, and random nature of papyri survival which means some areas have given us very little evidence while others have given more than we can handle. In the case of census returns we have problems due to the original writers lying about ages or leaving out babies until they have grown a bit older, something which can only be discovered if a later return for the same family is discovered. Wording of letters can be a problem, if a woman is referred to as “my wife” in one letter and “my sister” in another letter from the same man, is it an incestuous relationship or simply a term to show how strong their emotional ties were? While we may be able to write a history and family tree of one family it is almost impossible to make generalisations. We might have census records that show that fifty percent of the families in Arsinoe had incestuous marriages but this does not mean that fifty percent of the full population at the time was made up of people in brother-sister marriages. But despite these problems we can piece together a decent amount of family history, movement of families, births and deaths, jobs and arguments with neighbours, which in the end makes all the problems worth it.
Bibliography
Alston, R. (1995)- Soldier and society in Roman Egypt : a social history. Routledge.
Bagnall, R.S. (2007)- Hellenistic and Roman Egypt: Sources and Approaches
Bagnall, R.S. and B.W.Frier (1994)- The Demography of Roman Egypt. Cambridge University Press.
Brewster, E.H. "A Weaver of Oxyrhynchus: Sketch of a Humble Life in Roman Egypt" in Transactions and Proceedings of the American Philological Association, Vol. 58 (1927), pp. 132-154.
Rowlandson, J. (199 cool - Women and Society in Greek and Roman Egypt : a Sourcebook. Cambridge University Press.
Shaw, B.D. and R.P. Saller "Close-Kin Marriage in Roman Society?" in Man, New Series, Vol. 19, No. 3, (Sept 1984), pp. 432-444.
Winter, J.G. (1933)- Life And Letters in the Papyri. University of Michigan Press.
Wolf, A.P. "Adopt a Daughter-in-Law, Marry a Sister: A Chinese Solution to the Problem of the Incest Taboo" in American Anthropologist, New Series, Vol. 70, No. 5 (Oct 196 cool , pp. 864-874.
|
 |
Reply
