Estimating the Population of Ancient Jerusalem

010

Despite its obvious importance, the number of ancient Jerusalem’s inhabitants is a subject that is often ignored.

Until recently, writers who did deal with the matter based their estimates on ancient literary sources, which, however, are generally considered to be untrustworthy.¹ Even Josephus, the first century Jewish historian, who is usually quite exact with figures, is unrealistic when it comes to population figures. For example, he tells us that there were 204 villages in the Galilee, of which the smallest had 15,000 inhabitants (Jewish War, III, 3, 2, pp. 587f.)—an obvious impossibility.

Josephus’ population figures for Jerusalem are equally unreliable. The Romans, he tells us, left 1,100,000 Jerusalemites dead when they captured and destroyed the city in 70 A.D. Josephus’ figure for Jerusalem’s population at the beginning of the 3rd century B.C. is 120,000 (Contra Apion, I, 197, pp. 242f.) Here Josephus relies on the authority of Hecataeus, a Greek historian and ethnographer. This, too, as we shall see, is an exaggeration.

With the accumulation of modern archaeological data, we now have a far more reliable method of estimating the population of ancient cities, including Jerusalem. This requires only a reliable estimate of the size of the city and an understanding of the density of population in ancient urban areas generally.

Extensive excavation in Jerusalem, particularly during the last 20 years, has now given us a clear idea of most of the city’s fluctuating boundaries during its long existence. Archaeology also provides us with a reasonable estimate of urban population density in ancient times—the second essential tool for estimating the population of ancient cities. We shall examine both of these factors, the second first.

Based on the work of a large number of scholars working independently, we can say that the population density of ancient cities was about 160–200 persons per acre.

Henry Frankfort and Phineas Delougaz found that there were on the average 20 houses per acre in three Mesopotamian cities which they studied (Ur, Eshnunna and Khafajeh). If the average household numbered six to ten people (including children and slaves), this comes to approximately 160 people per square acre (8 × 20).² John Garstang estimated that ancient towns he studied had 200–240 people per acre.³ H. Packer suggests that Ostia, the well-excavated harbor town for the city of Rome, had almost 160 persons per acre.⁴

So 160–200 persons per acre seems a fair and independently corroborated number.

Interestingly enough, ancient towns which have survived to this day contain populations of almost the same density as archaeologists have estimated for ancient towns which have been destroyed. The population density of Aleppo and Damascus is about 160 persons per acre.⁵ In 1918, the population density of the Old City of Jerusalem was about 200 persons per acre.⁶ The same number is given for Lebanese Tripoli.⁷

This too corroborates our estimates of 160 to 200 persons per acre and is in striking contrast to modern American cities which are far less densely populated. An American city the size of Lebanese Tripoli (180,000) would be a half to a quarter as dense as this ancient town. (An American city of 180,000 would have a population density of 50 to 80 persons per acre).⁸

If we can agree on the population density of ancient Jerusalem as 160–200 persons per acre, we 011need only determine the number of acres within the city during its long history to arrive at a reliable population estimate.

It is far easier to understand the changes in Jerusalem’s boundaries by looking at maps than reading descriptions, so I shall keep my remarks to a minimum and rely primarily on maps.

• The first map shows Jebusite and Davidic Jerusalem. King David captured the city about 1,000 B.C. The solid black line represents the present Old City wall. The rectangle in the lower right of the Old City is the existing remains of Herod’s Temple Mount on which the Dome of the Rock now sits. The shaded area south of the present Old City is where Jebusite and Davidic Jerusalem was located:

The Jebusite and Davidic city sat on the eastern and smaller of two ridges that extend below the present Old City. (The eastern ridge is now known as the Ophel, the western ridge as Mount Zion.) The two ridges are separated by a valley. Kathleen Kenyon’s excavations between 1961 and 1967 located the eastern wall of the Jebusite city lower on the eastern slope than was formerly supposed. This meant that the city was larger than previously thought. But even so, it was a bare 12 acres at most.⁹ Although the northern wall has not been located, where I have placed it on the map cannot be far off the mark.

Accordingly, the population of Jerusalem during this period cannot have been much more than 2400 (12 × 200), and probably numbered somewhat less, say 2000.

• King Solomon (c. 965 B.C.–928 B.C.) expanded the Davidic city northward to include the Temple Mount area. There he built the First Temple and his royal residence:

The Solomonic city covered almost 32 acres. We cannot be far off if we estimate the population of Solomonic Jerusalem at 4500–5000.

It is unlikely that there was any major expansion of the city during the reigns of Solomon’s successors in the 9th and 8th centuries. Enlargement on the south or east was not feasible because of the steep valleys—which gave the city a large measure of natural protection on these sides. West of the Temple Mount area, Professor Benjamin Mazar has recently discovered tombs from the 9th and 8th centuries and has concluded that this area was then a cemetery.¹⁰ Thus, this area must have been outside the city walls at that time. (Only royal burials were permitted within the city.) The area north and northwest of the Temple Mount has not been tested archaeologically, because it is covered with “modern” buildings. But any substantial expansion of the city during the 9th and 8th centuries B.C. seems unlikely—the Bible itself indicates that the building activities of such Judean kings as Uzziah (769–733 B.C.) and his co-regent Jotham (758–743 B.C.) were confined to the city as it then existed (see 2 Chronicles 26:9, 2 Chronicles 26:15; 2 Chronicles 27:3).

• The next major enlargement of the city of which we have archaeological knowledge occurred at the very end of the 8th century, apparently during the reign of King Hezekiah. (727–698 B.C.). From Professor Avigad’s excavations in the Jewish Quarter, we learn that at the end of the 8th century, the city mushroomed—extending to the area that would later be called the Upper City. The principal evidence for this expansion is the so-called Broad Wall (between 21 and 23 feet wide) which apparently formed part of the western wall of the expanded city. Beneath this wall were found late 8th century houses, indicating that the western hill or upper city had been inhabited even before the late 8th century 012wall marked the expanded boundary of the city.¹¹ Probably the city accreted on the west with unwalled suburbs until a new city wall was built.¹²

Incidentally, this considerable and quite abrupt growth of the city is consistent with the establishment of new settlements in the Judean mountains at about the same time.¹³ In an earlier BAR article (“‘Digging Up Jerusalem’—A Critique,” BAR 01:03), I have suggested that this growth was probably the result of the arrival in Jerusalem of Israelite refugees who fled south after the fall of the northern kingdom to the Assyrians in 721 B.C., as well as refugees from the eastward migration of the Judahite population which occurred when Sennacherib in 701 B.C. punished rebellious Hezekiah by ceding a sizable part of Western Judah to the Philistine city states.

In estimating the population of Jerusalem during this period we shall take into account only the area within the walls, but the fact that suburbs were pushing outwards indicates higher density inside the walled city. The walled area was about 125 acres which, at 200 persons per acre, comes to about 25,000 persons within the city walls.

• The ancient city did not change much during the seventh century B.C., even in the reigns of Manasseh (698–642 B.C.) and Josiah (639–609 B.C.). Then in 586 B.C. the Babylonians captured and destroyed the city. Jerusalem remained unwalled and largely uninhabited until the restoration during the Persian Period.

The returnees had no need for a large city, however. Relatively few came back in the early years of the restoration. Professor Avigad’s excavations in the Jewish Quarter (on the western hill across from the Temple Mount) uncovered no remains from this so-called Persian period. I found a similar situation on Mount Zion (the western ridge below the present south wall of the Old City)—no evidence of occupation during the Persian period.

Kathleen Kenyon discovered the eastern wall of Nehemiah’s Jerusalem higher up on the eastern ridge (below the southern wall of the Old City) than the Jebusite-Davidic wall, indicating a narrowing of the city’s girth during the Persian Period.

This is what the city looked like then:

013

Nehemiah’s Jerusalem covered about 30 acres, a little smaller than Solomon’s Jerusalem because the eastern wall was a little higher up on the ridge. Its population was about 4,500.

• The next major expansion of Jerusalem occurred during the Hasmonean period (second century B.C.). The line of the wall of the city as it existed at that time—Josephus refers to it as the First Wall—has been located by various expeditions beginning with the excavations of Frederick Bliss on Mount Zion in 1894–1897 and extending to the excavations south of the Citadel which I led beginning in 1973 and Avigad’s excavations in the Jewish Quarter in 1976. Hasmonean Jerusalem is shown on the map:

Hasmonean Jerusalem covered about 165 acres and the population may be estimated at about 30,000–35,000.

• Herod’s Jerusalem (37 B.C.–4 B.C.) was surrounded on the south by Josephus’ First Wall and on the north by what he called the Second Wall. Whether the Second Wall was built by Herod or by one of the Hasmonean Kings preceding him is not certain.¹⁴ The exact course of the Second Wall is also uncertain, both because of the paucity of archaeological remains and the lack of any precise literary description. I believe that the late Avi-Yonah’s reconstruction is the most likely¹⁵ and is shown on the map:

The area enclosed is about 230 acres, but in computing the built-up area the grounds of the Temple Mount area (36 acres) should be deducted because it was not used for habitation. I have not made a specific deduction for earlier periods because only Herod’s huge artificial platform occupied so large and substantial a part of the city. The area of the palaces (the Hasmonean palace and Herod’s palace) and the citadels (both the towers known as the citadel and the Antonia fortress) should not be deduced because they were occupied by the King’s household, administration and soldiers.

I estimate the population of this area to have been almost 40,000.

• In the period between Herod’s death and the Roman destruction of Jerusalem in 70 A.D., Jerusalem expanded considerably towards the north and northwest. (This was the only available direction for expansion, because in other directions the city was surrounded by deep valleys.) Herod Agrippa I (41 A.D.–44 A.D.) started to build the Third Wall on the northern side of the City to defend the new suburbs, but for some reason his work was stopped; the wall was completed hastily only after the First Jewish Revolt broke out in 66 A.D.

The line of the wall Josephus calls the Third Wall is a matter of hot dispute. Edward Robinson, the great 19th century American orientalist first identified the Third Wall with a line about 600 feet north of the Old City’s present northern wall. Substantial 014stretches of this line were unearthed by E. L. Sukenik and L. A. Mayer in 1925–1927 and 1940. However, this position was contested by a leading early 20th century Jerusalem scholar, Father L. H. Vincent of the École Biblique (the French School of Archaeology in Jerusalem). Father Vincent defended the present northern wall of the Old City as the line of the Third Wall. More recently Dame Kathleen Kenyon of the British School of Archaeology has taken up the cudgels for this position. The recent excavations of Sara Ben-Arieh and Ehud Netzer, however, marshalled new evidence for accepting the Sukenik-Mayer line as the Third Wall and for dating the remains of this wall to the first century A.D.¹⁶ Even before the Ben Arieh-Netzer investigations, the Vincent-Kenyon position seemed questionable;¹⁷ now it must be regarded as unacceptable. In the following map, I have followed Robinson’s line to mark the northern boundary of the city in the period just before the Roman destruction:

The area enclosed is about 450 acres. Approximately 80,000 people lived in the city.

• For most later periods our system for estimating Jerusalem’s population will not work. The reason is that between the Roman destruction and the 19th century, the city was very sparsely populated. Thus although we know the area of the city to be about 215 acres, it is difficult to estimate the population accurately.

The exception is the late Byzantine and early Arab periods. In the middle of the 5th century A.D., the Empress Eudocia built a wall enclosing Mount Zion and the city was enlarged to about 300 acres.¹⁸ During Justinian’s reign (527–565 A.D.) and the Omayyad Caliphate (661–750 A.D.) Jerusalem again bustled with activity and habitation. The city then looked like this:

015

The population was about 55,000–60,000.

The next major expansion of Jerusalem was in the late 19th and 20th centuries. We are no longer dealing with ancient Jerusalem.

(For further details see M. Broshi, La population de l’ancienne Jérusalem, Revue Biblique 82 (1975), pp. 5–14.)

Despite its obvious importance, the number of ancient Jerusalem’s inhabitants is a subject that is often ignored.