How big did Noah's family grow before Babel, and how much of its ancestry still came from the ark?

1The question

Genesis records very long lives for the first generations after the Flood. This page explores one proposal: that those long lives were inherited, so that the people who carried them would have left descendants, and perhaps skeletons and DNA, in the ancient world.

This page asks two questions. How big was the core family of Noah's descendants before it scattered from Babel? And what would a fair test of ancient skeletons need, if the aim were to look for ancestry from the ark?

The page answers with a computer model, built one step at a time. Each step says what Genesis records, what the book's reading adds, and what the model assumes.

2The starting point

Genesis says that eight people came off the ark: Noah, his wife, his three sons Shem, Ham and Japheth, and their wives.

This model follows the book A Timeline of Origins. The book argues that the Flood destroyed the old world in which Noah lived. It also argues that the world Genesis describes after the Flood, with people such as the descendants of Cain, was already inhabited. This page calls those people the wider population. The model takes the book's reading as its starting point. It does not set out to prove it.

Genesis 11 shows lifespans falling generation by generation. The proposal explored here treats long life as something inherited from the old world, which this page calls the ancient world. In the proposal, long life thins out as the family marries people from the wider population. The model makes no judgement about the worth of anyone in either group.

Dates follow this project's own chronology, which uses the ages in the Greek translation of the Old Testament (the Septuagint). Genesis itself gives no dates in years BC. On that chronology the family leaves the ark in about 2903 BC. Peleg is born in 2372 BC and dies in 2033 BC. Abraham is born in 1771 BC.

Age at death in Genesis, in years, using the Septuagint figures this project follows

3How many people

Genesis 10 names sixteen grandsons of Noah. It also names 36 of their sons. That is an average of 2.25 sons for each of the sixteen grandsons.

The model treats each of those names as one son. Many of the names are peoples or places, so this is a simplification. For men whose sons Genesis does not list, the model uses the same average of about two and a quarter sons. It assumes that each couple has as many daughters as sons.

Genesis 10 is a list of names. It is not a census of every birth or of every child who survived. The model uses those counts as a starting assumption for family size. The counts alone do not tell us how many children every couple had or how many survived.

Under these choices, the model gives about 66 people in the family by 2703 BC, two centuries after the ark. That size is an output of the choices, not an observed historical figure. One of those choices is that every unnamed child survives to the age at which the model lets them have children. The model has no infant or child mortality, and the family's size depends on that assumption.

The first part of the top chart, enlarged: the family's first 200 years Showing: the central setting

4The slow clock: late sexual maturity

In Genesis 11 the fathers in the first generations are about 130 years old when the next man in the line is born. Later that age falls sharply.

Age of each father when the next man in the line was born, from the birth years the model uses

The model assumes that people with a high share of ancient-world ancestry reach sexual maturity, the age at which a person can first have children, late: at about 130. As the share falls, sexual maturity comes sooner, at about 70 and then at about 30. The table shows the steps the model uses. They are choices made so that the Genesis birth years fit.

Age at sexual maturity by potential lifespan, the lifespan assigned by the formula before its 29-year tolerance

One consequence follows. A family that marries people from the wider population has children sooner, so it grows faster, even while its share of ancient-world ancestry falls.

5Long life and the share of ancient-world ancestry

Each person in the model has a share of ancient-world ancestry. This share is an assigned quantity. Each of the eight people on the ark is given 1. Each person from the wider population is given 0. A child is given the average of the two parents' shares. So a share of 1 means that every ancestor in the model came off the ark, and a share of 0 means that none did. The share is an average: it is the share a child is expected to inherit. Real children vary around it, so a person given one-eighth could have inherited more or less than one-eighth of their DNA from the eight people on the ark.

The share is not a count of distinct ancestors, and it is not measured DNA. The model simulates no DNA sequences and no particular genetic variants for long life.

The lifespan formula is a proposal being explored on this page. It is not an established scientific law. In the formula, a person's potential lifespan, the lifespan assigned by the formula before its 29-year tolerance, rises with the share. It runs from about 86 years for a share of 0 to about 565 years for a share of 1, which is Arphaxad's age in Genesis. Potential lifespan is not a statistical expected lifespan. Noah lived to 950 and Shem to 600, longer than the top of this range. Noah and Shem are not in this calculation, because each of them lived a number of years in the ancient world, and those years would have lengthened their lives.

The formula must allow a person to reach the age recorded in Genesis, within a tolerance of 29 years. A person may die much earlier than the formula's lifespan. Where a recorded age is more than the tolerance below the potential lifespan, the chosen arrangement assumes an early death. Those early deaths are assumptions of the arrangement, not diagnoses of a cause.

The average share among Noah's descendants in Shinar, year by year Showing: the central setting

6What Genesis 11 tells us about marriages

Genesis gives each man's age at death. The formula must allow each man to reach his recorded age, within the tolerance of 29 years. So the recorded ages limit the shares that a man's parents could have had. They do not fix those shares on their own. The page chooses one arrangement of parents that fits every recorded age, and a mother's share is part of that chosen arrangement. It is not something the ages show independently.

Genesis records none of these marriages. It names very few of the wives, and it never says where their families came from. So the program searches for an arrangement that uses as few marriages to people from the wider population as it can find.

For Cainan the recorded age is 460 and the formula gives 565 years. So this arrangement assumes that he died early, about 105 years below the lifespan the formula allows.

This page follows the Greek text, which includes Cainan between Arphaxad and Shelah. The Hebrew text leaves him out.

The buttons at the top of the first figure use this line's marriages to propose a setting for the other branches of the family, the ones Genesis does not follow name by name. They offer a central setting, a lower setting and a higher setting. The note under the buttons says what the setting is and how it was reached.

The line from Arphaxad to Isaac. For each person, the top bar is the age at death in Genesis and the bottom bar is the potential lifespan the formula allows for their share.

7When the family split

Genesis 10:25 says that the earth was divided in the days of Peleg. The model assumes that the family stays together in Shinar until after Peleg's birth. Shinar is the Bible's name for southern Mesopotamia, in what is now southern Iraq, where Babel stood. So in the model no sizeable part of the family goes to the lands named in Genesis 10 before 2372 BC.

The model also assumes that the men named in Genesis 10 lead their own families to their own lands during Peleg's lifetime. There are two cases to note. The men in the line from Arphaxad to Peleg stay in Shinar. Nimrod's departure date is set by a separate rule, 150 years after his birth. For the other men, each one leaves early enough to arrive alive and to live in his new land for some years. Noah's grandsons have the least time left, so they leave within a few decades of Peleg's birth. Later generations leave over the following centuries, once they are old enough to have children and before the scattering from Babel. Where Genesis names a man but this project cannot place his land, he still leaves, and the page says that his destination is not known.

When a man named in Genesis 10 leaves Shinar, he takes his wives, his unmarried children and the descendant families that go with him. Each descendant family has a one-in-four chance of staying behind in Shinar, except the families that the fitted arrangement of the line to Abraham needs, which always stay. So about a quarter of the other families stay.

Some men are named in Genesis 10, but nobody can say where their land was. In the model they still leave Shinar, and on the map they fade out. A son who is born after his father has left is born in the new land and is followed there, within the limits described in section 8.

How the book reads Babel

The book A Timeline of Origins reads Babel in a particular way. What follows is the book's argument, not an established fact.

Genesis 11:1 sets the background: the whole land had one language and the same words. The book reads this as the world before the Flood, the world of Adam through Noah. It never describes the world after the Flood, which, the book holds, already had other languages.

Genesis 11:2 begins a much later scene. It places these people in the east. The Hebrew word, miqqedem, can mean in the east, from the east or in ancient times. The same word is used of Eden in Genesis 2:8. Here it means east of where most of the Bible's story is set. The people settled on a plain in Shinar. The book reads them as people from the old world, descendants of Adam, who still spoke the original language. They lived among peoples who did not. The model follows Noah's descendants among them.

On this reading, Babel is a later scene in the world after the Flood. In it, the family's language was confused and the family scattered. The book argues that the language was Sumerian. The book places Babel within Peleg's days and allows a wide range. This site places it near the end of his life, between 2033 and 2004 BC, which fits the fall of Ur.

Genesis 10:5, 10:20 and 10:31 give the clans of chapter 10 their own languages and lands. On this reading that is no contradiction, because the people already living in the world after the Flood spoke other languages.

Genesis says that in Peleg's days the earth was divided (10:25), and that the nations of chapter 10 went out into their lands after the Flood (10:32). Genesis gives no year for either. It does not say whether the dividing and the scattering from Babel (11:8 to 9) are the same thing.

This site's chronology ends Peleg's days in 2033 BC, close to the fall of Ur in 2004 BC. The historian Georges Roux called that fall the end of the Sumerian nation.

The model treats the dividing as the whole of Peleg's lifetime. In those years the families named in Genesis 10 leave Shinar one by one. The model treats the scattering from Babel as the last part of the dividing. So in this model the families leave Shinar during Peleg's days, before the scattering from Babel. Both choices are the model's assumptions. A comparison setting, not yet on this page, would place Babel early in Peleg's days.

What happens in the model

The model does not choose the households that leave by their ancestry. Each household carries whatever share its own branch of the family had.

The top chart again: the line is the median, the middle result, of how many people there were, and the pale blue band is the middle 80 percent of runs. The gold band marks the years in which households leave. The thin dotted lines mark Peleg's birth, the scattering from Babel and the departure of Terah's household. Showing: the central setting

The households that left Shinar. In the model, a household is the man who leads it, his wives, his unmarried children, and the descendant families that go with him. The descendant families that stay behind are not part of it.

Each household is listed by the man who led it and the land he went to. The blue bar shows how many people each household had when it left. The gold bar shows their average share of ancient-world ancestry, where a full bar means 100 percent. Showing: the central setting

8When they arrive in new places

In a new land, the model lets a household's descendants marry both within the household's own descendants and with local people, and some of the husbands are local men. The program tries to find a partner for each person, and it does not assume that everyone marries. The table below is a separate illustration, not the regional simulation. In it every marriage is to a local person, and each generation that does so halves the share of ancient-world ancestry. In that illustration the very long potential lifespans last only for the first few generations after arrival.

A separate illustration of how the share thins

Loading the map's population file.

Where does the model place people with at least half their ancestry from the ark?

This question is about living people in the model. For each region, the page divides the number of living people whose assigned share of ancestry from the ark is at least half by the number of people living there. The assigned share is defined in section 5. It is not measured DNA. Dividing by the living population does not give the probability that an excavated skeleton belongs to such a person, because the timing of death, burial, preservation, excavation and DNA recognition all differ between people and between regions.

9What a fair test of ancient skeletons would need

Section 8 gives shares of living people in the model. It does not say what a study of excavated skeletons would find. A fair test would need the following.

• It would compare early, intermediate and later evidence in each region. Descendants need time to grow in number, to move between places, and to die and be buried. A study of only the first century after a household arrived would miss later descendants.
• It would say, before the study, how many skeletons from which places and dates could reasonably contain such a person, given the growth, mixing and movement that the history allows.
• A result that finds nothing would count against descendants only after the study has shown that it had a reasonable chance of detecting them. An assumed frequency of a marker is not enough.
• It would need something to look for. No genetic marker has been identified for the centuries of life proposed here, and bones show a person's age at death only roughly.

Ancient DNA from southern Mesopotamia would also help, because that is where the model places the family before Babel.

10What this model assumes and cannot tell us

The starting point and the chosen chronology

• The page takes the names, ages and family links in Genesis as the accepted record. It uses the Septuagint figures that this project follows.
• The book's reading of Genesis is taken as given. The model explores that reading and does not prove it.
• Genesis does not date Babel. This model places the scattering from Babel at the end of Peleg's life, which fits the fall of Ur in 2004 BC. A comparison may place it earlier in Peleg's days.
• In this model the families leave Shinar during Peleg's days, before the scattering from Babel. Noah's grandsons leave within a few decades of Peleg's birth, between 2371 and 2338 BC.
• Every man named in Genesis 10 leads his own family to his own land after Peleg's birth and during Peleg's lifetime, with two exceptions. The men in the line from Arphaxad to Peleg stay in Shinar, and Nimrod's departure date is set 150 years after his birth. Each other man leaves early enough to arrive alive and live there for some years. A man whose land this project cannot place still leaves, and the page shows his destination as not known.
• The people who stay in Shinar after the scattering from Babel are a small remnant in the line that leads to Abraham.

The additional biological and demographic choices

• Long life is treated as inherited, in proportion to the assigned share of ancient-world ancestry. The lifespan formula is a proposal being explored, not an established scientific law.
• The ages at sexual maturity, and the number of years in which people can have children, are choices made so that the Genesis ages and births fit.
• The model assumes that most people whom Genesis does not name die at 90 to 100 percent of the potential lifespan the formula allows, and that one in five dies earlier.
• The unnamed children are assumed to survive to the age at which the model lets them start having children. The model has no infant or child mortality.
• The ark couples are exceptions to the fertility rules. Their daughters may be born for decades after the landing. Some lives are extended so that required births or arrivals fit.
• Genesis lists names. It is not a census of every birth or of every child who survived. The model uses the average of 2.25 listed sons per man as a starting assumption. The counts alone do not tell us the number of children born or surviving in every family.
• The program tries to find a partner for each person. It matches each person with the closest relative who is free to marry, preferring a first cousin. Nobody marries a brother or sister, a parent or child, or an aunt, uncle, niece or nephew. A woman who has waited 25 years for a husband becomes a man's second wife.
• In Shinar, people from the wider population join the family only as wives. In the new lands, marriages are both within the household's descendants and with local people, and some husbands are local men.
• When a man named in Genesis 10 leaves Shinar, each descendant family that the fitted arrangement does not need stays behind with the stated chance, about a quarter. This is a setting of the model, not something Genesis says. The years in which households leave are also set by the model.
• To fit the ages in Genesis and the years in which women can have children, the model needs Terah to have three wives: the mothers of Haran, Abraham and Sarah.
• Noah and each of his three sons has one wife, the wife who came off the ark. None of them takes a second wife.
• Hagar, Ishmael's descendants, Rebekah, Leah, Rachel and other later marriages are outside this model.
• The corrected run removes the 1,500-person marriage limit and continues births through 1596 BC, using the same histories for the family and regional figures. It still stops following reproduction below 1/8 assigned ancestry. Section 8 describes the measured effects in 2200 and 2100 BC; the effect in later centuries remains unmeasured.

What the results are

• Each result is what follows from the accepted record, the interpretation and the additional choices together. Other family arrangements or other fertility rules may give different results. This page has not compared its choices with others to show that they are the best possible integration.
• The ranges on this page cover variation among runs under fixed assumptions. They do not cover uncertainty about fertility, childhood survival, chronology or geographical placements.
• The figures are provisional and may change as the model is revised. The book A Timeline of Origins does not give any of these numbers.

If you find a mistake in the reasoning or the figures, the author would be glad to hear of it.

How the map ranks modern peoples

The table lists all 50 peoples included in the calculation, ordered by the middle estimate of their expected percentage of ancestry from the eight people on the ark. These figures are the map's calculation from stated assumptions, not a measurement of anyone's ancestry.

Samaritans and Jewish communities rank near the top because the map follows their ancestry through the Israelite family line and accounts for later marriages outside that line.

Follow Noah's family on the map. Download the table's data.

You can scroll across the table to read every column.

The table is loading.