Stones, Scripts, and Scholars

The professional historian of antiquity has often looked with a skeptical eye upon the evidence of traditions, such as those incorporated in Homer, Herodotus, and the Old Testament, and transmitted from generation to generation. The relatively young discipline of archaeology has tempered this skepticism by uncovering materials and inscriptions that have at various points confirmed the traditions.

A great many of the traditions, however, still lack corroborative external evidence. Scholars often assume that these unconfirmed traditions must remain suspect. This attitude may be one of commendable reserve, but it is also one that is based partly on an argument from silence—an argument that is precarious because of the accidents of overlap in areas of evidence, and because of the fragmentary nature of the archaeological evidence.

Overlapping Areas Of Evidence

The non-traditional and therefore contemporary evidence may be divided into two categories: (1) material remains, and (2) inscriptional evidence. The first category could include a subcategory of artistic evidence for areas such as Greece. But for Israelite history this kind of evidence is scanty. Inscriptional evidence would include such documents as royal inscriptions, letters, treaties, and contracts, of which relatively few have been found in Palestine in contrast to the thousands of texts from Egypt, Mesopotamia, Anatolia, and Greece.

Plotting the sources of our evidence for ancient history as overlapping circles reveals that there are theoretically seven possible combinations: three in which one source stands alone, three in which two sources overlap, and an ideal situation in which the three sources overlap.

That each of these combinations may occur may be seen from a chart of the evidence for plants and animals prepared by Emily Vermeule, Greece in the Bronze Age (1964). In her list, inscriptions are represented by Linear B texts, and tradition by Homer. (1) The apple is attested alone in Homer; (2) the mint alone in Linear B; (3) the almond alone by excavations; (4) the pear by both Homer and excavations; (5) the cypress by both Homer and Linear B; (6) the coriander by both Linear B and excavations; and (7) linen by all three sources.

The implication of this random distribution is that just as an object may be attested alone by excavations or alone by inscriptions, it may very often stand alone in the traditions without any necessary reflection upon its authenticity. Yet scholars have often assumed that the overlap of traditions with either inscriptional or material evidence is not only desirable but necessary.

Now, until the decipherment of Linear B by Michael Ventris in 1952, there was no corroborating evidence from inscriptions for the Homeric traditions. Even with the decipherment some scholars are troubled that Homer does not mention the frescoes of the Mycenaeans, and that the epics do not reflect the bureaucracy of the Linear B texts. But one should not expect laundry lists in epics, any more than he should look for stock quotations in poetry.

In a collection of hundreds of Sumerian proverbs, there is not a single reference to law or to painting. Yet no one doubts that these proverbs are an intimate reflection of their times. Some scholars have been convinced that the traditional Sumerian King List is not reliable since few rulers appear both in the List and in inscriptions. But Jacobsen has shown that there are good reasons for the omission of the rulers of Lagash, for example, from the List (T. Jacobsen, The Sumerian King List [1939], p. 180).

Critics were convinced that the Book of Daniel was inaccurate since it referred not to Nabonidus, the last Neo-Babylonian king, but to Belshazzar. But cuneiform documents published since 1924 have brought to light the extraordinary and unpredictable exile of Nabonidus from Babylon to Arabia, so that the “kingship” of Babylon was left to Belshazzar his son.

Often it is assumed that the historicity of a person is suspect unless corroborated by inscriptional evidence (e.g., Darius the Mede in the book of Daniel). Attempts to identify a person in a tradition with someone in the inscriptions often founder on the lack of overlapping evidence. Before the discovery of cuneiform documents that identified Belshazzar as the son of Nabonidus, some declared his name a pure invention, and others tried to identify him with Evil-Merodach or Neriglissar. Tatnai (Ezra 5:3, 6) was mistakenly identified with the satrap over Babylon, Ushtannu, until in 1944 Olmstead called attention to a text where Ta-at-tan-ni is mentioned as a governor subordinate to the satrap (A. Olmstead, Journal of Near Eastern Studies, 3 [1944], p. 46).

Scholars have questioned the authenticity of the Sanballat mentioned by Josephus as a contemporary of Alexander the Great. Many thought this was a mistaken reference to the Sanballat who lived in the time of Nehemiah (c. 445 B.c.). The discovery in 1962 of the Aramaic papyri (dated 375–335 B.C.) now makes clear that there were three Sanballats—one in Nehemiah’s time, a second at the time of the papyri, and a third in the time of Alexander (F. Cross, The Biblical Archaeologist, 26 [1963], p. 120).

The reference in Luke 3:1 to Lysanias, a tetrarch of Abilene in the time of John the Baptist (A.D. 27), was considered an error since the only known Lysanias of the area was one executed by Cleopatra in 36 B.c. But then the publication of a Greek inscription from Abila proved that there was a Lysanias who was a tetrarch between A.D. 14 and 29 (F. F. Bruce, “Archaeological Confirmation of the New Testament,” in Revelation and the Bible, ed. Carl F. H. Henry [1958], p. 327).

If we had had to depend upon inscriptional evidence to prove the historicity of Pontius Pilate, we would have had to wait until 1961, when the first epigraphical documentation concerning him was discovered at Caesarea (J. Vardaman, Journal of Biblical Literature, 81 [1962], p. 70).

The Fragmentary Nature Of The Evidence

Historians of antiquity in using the archaeological evidence have very often failed to realize how slight is the evidence at our disposal. It would not be exaggerating to point out that what we have is but one fraction of a second fraction of a third fraction of a fourth fraction of a fifth fraction of the possible evidence.

First of all, only a fraction of what is made or what is written ever survives. Only about one-tenth of the works of the three greatest Greek dramatists—Aeschylus, Euripides, and Sophocles—have come down to us. Of all the Greek lyric poets who wrote in the seventh and sixth centuries B.C., we have manuscripts only for Theognis and Pindar, and just fragments for the rest.

Although we know from clay sealings that papyri were used, none has been recovered from the Minoan-Mycenaean period in the Aegean, and from the Seleucid period in lower Mesopotamia. Wax tablets were also used for writing, but the only examples of writing on wax recovered are isolated finds from Nimrud, Gordion, Fayyum, and Marsiliana (in Etruria) (M. Mallowan, Nimrud and Its Remains [1966], pp. 152 ff.).

Inscriptions listing the twenty-four courses of the priesthood probably hung in hundreds of synagogues in Palestine. Thus far, only fragments of two such inscriptions have been recovered—one found at Ascalon in the 1920s, and fragments from Caesarea in the 1960s. In a fragment from Caesarea (dated to the third and fourth centuries A.D.), the name “Nazareth” appears. “This is the only time so far that the name ‘Nazareth’ has been found in an inscription, in particular in a Hebrew inscription; it is also the earliest occurrence of the name in Hebrew” (M. Avi-Yonah, The Teacher’s Yoke, ed. J. Vardaman [1964], p. 48). The next occurrence of the name in a Hebrew inscription is in a Genizah piyyutim (liturgical poem) fragment of the eleventh century!

In the second place, only a fraction of the available sites have been surveyed. “All in all, well over three hundred Mycenaean sites are known, and it is probable that this number would be quadrupled if all Greece were carefully explored for evidence” (A. Samuel, The Mycenaeans in History [1966], p. 101).

In Mesopotamia, Agade, the capital of the Akkadian kingdom, and Washukani, the capital of Mitanni, have not yet been positively identified. In Anatolia it was only in 1956 that Derbe, a site visited by the Apostle Paul, was discovered.

In 1944 the Palestine Gazette listed a total of about 3,000 sites in Cis-Jordan and several hundred in Trans-Jordan. In 1963 the total of known sites increased to about 5,000. The Israeli surveys of 1968, covering the Golan Heights, Samaria, and Judah, have increased this total. Moshe Kochavi, the director of the Judean survey, writes (in a letter of November, 1968): “Our Survey surveyed about 1,200 sites, of which some 20–30 per cent are new sites previously unrecorded. A second phase of the Survey, which is being carried out now, may lead to the same results.… I estimate that not more than one-third of the amount of possible sites were recorded, and a thorough survey is a question of many years (including the yet unsurveyed parts of pre-war Israel).”

One momentous result of the recent survey is that Albright’s identification of Tell Beit Mirsim with biblical Debir will have to be abandoned in favor of the new site of Rabud, excavated by Kochavi in the summer of 1968.

In the third place, only a fraction of the surveyed sites have been excavated. In 1963 Paul Lapp estimated that of a total of 5,000 sites in Palestine there had been scientific excavations at about 150, including twenty-six major excavations. “To be sure, many of the sites on record would not merit extensive excavation, but if only one in four were promising, major excavations have till now been carried out at only 2 per cent of the potential sites” (The Biblical Archaeologist, 26 [1963], p. 122).

Seton Lloyd notes that by 1949 more than 5,000 mounds had been located in Iraq (Mounds of the Near East [1963], p. 99). As of 1962, Beek’s atlas recorded twenty-eight major excavations in Iraq, less than 1 per cent of the total sites (Atlas of Mesopotamia [1962], map 2). When Leonard Woolley wished to excavate in the Amuq plain at the mouth of the Orontes River, he was faced with making a choice among two hundred mounds that dot the plain (A Forgotten Kingdom [1953], p. 20).

Many sites are still occupied so that their excavation is impossible or impractical: e.g., Arbela-Erbil in Iraq, Aleppo-Haleb in Syria, Gaza in Palestine. One of the two mounds of Nineveh, Nebi Yunus, has not been excavated because it is the site of a modern village. At the turn of the century it was possible to move an entire village from the site of an ancient settlement, as at Delphi. Today to excavate the important but occupied site of Thebes in Greece, where soundings have yielded Mesopotamian seals, Linear B texts, and so on, would require at least a million dollars for the expropriation of the land.

In Israel many important and unencumbered mounds still remain unexcavated, e.g., Jezreel, Tell Beersheba, Tell Akko, and Khirbet Muqaneh (possibly Ekron). The last two tells are so extensive that it could cost a million dollars for thorough excavations.

In the fourth place, with the exception of small and short-lived sites such as Qumran and Masada, it is always the case that only a small fraction of any excavated site is actually examined. The wealthy Oriental Institute excavations at Megiddo 1925–34 succeeded in completely removing the top four strata. But this grandiose scheme was abandoned in later seasons and has not been attempted at any site of comparable size. This means that any given excavation may very well miss important finds. (It is embarrassing to report that a cuneiform text of the Gilgamesh epic was found by shepherds in the discarded debris from the excavations at Megiddo.)

The British excavated at Zakro in eastern Crete in 1901. They found houses but missed a palace, which was not found until the excavations begun in 1962. At Ephesus since 1894 the Austrians have found vast remains of the later periods. But nothing of the Bronze Age was found until in 1963 Turkish engineers built a parking lot and found a Mycenaean burial. For decades nothing of the Bronze Age was found at Halicarnassus, until in 1962 George Bass saw a man walking down the street carrying a Mycenaean jar from a nearby village (G. Hanfmann, The Antioch Review [Spring, 1965], p. 42).

In Mesopotamia only a small, unreliable excavation has been conducted at Bismaya (Adab), and hardly any excavation at the important site of Borsippa. Even at Calah-Nimrud, which was the second major site to be excavated in the Near East by Layard (1845–51), followed by Loftus, Rassam, and George Smith (from 1854 at intervals to 1891), the re-excavations by Mallowan (1949–63) have been able to produce magnificent ivories and important texts bearing on the Old Testament. Of Babylon, Saggs notes: “These extensive ruins, of which, despite Koldewey’s work (1899–1917), only a small proportion has been excavated, have during past centuries been extensively plundered for building materials” (Archaeology and Old Testament Study, hereafter abbreviated A.O.T., ed. D. Winton Thomas [1967], p. 41). Excavations at Babylon are complicated by the fact that the earlier levels of Hammurabi’s time are below the water table.

In Palestine from 1902 to 1904 Sellin excavated nearly a fifth of Tell Ta‘annek (in the early days the work was done with less care and therefore more rapidly), and concluded that there were no more important structures to be found and that the city had never been surrounded by a fortification wall. In 1963 Lapp found both a fortification wall and important structures there.

From soundings made at Hazor in 1928, Garstang concluded that the site was not an important city in the fourteenth and thirteenth centuries because of “the complete absence of Mykenaean specimens” (A.O.T., p. 247). Yadin in his excavations found houses littered with Mycenaean pottery. The site at Hazor comprises an upper city of thirty acres and a lower city of 175 acres. Working with an unusually large staff of more than thirty archaeologists and a crew of more than a hundred laborers, Yadin managed to clear 1/400 of the site—that is 1/1,600 per season from 1955 to 1958. “He has suggested that it would take eight hundred years of about four or five months work (the normal season is three months) per year to clear the entire site” (W. F. Albright, New Horizons in Biblical Research [1966], p. 3).

Some sites in the Near East are even larger than Hazor: “The largest city was undoubtedly Babylon in the Chaldean period; its area covered 2,500 acres. Then follows Nineveh, with 1,850 acres, while Uruk was somewhat smaller, with 1,110 acres. Other cities are much smaller: Hattusha, the Hittite capital, occupied 450 acres; Assur had only 150 acres. Among the royal cities, Dur-Sharrukin was 600 acres; Calah, 800 acres” (A. Oppenheim, Ancient Mesopotamia [1964], p. 140). At Yadin’s estimated rate of progress for Hazor, a complete excavation of Babylon would take at least 8,000 years!

In the fifth place, only a fraction of the materials, and especially the inscriptions, produced by excavations have ever been published. Samuel Kramer estimates that 10 per cent of the 500,000 cuneiform texts recovered have been published.

Nippur was the earliest site to be excavated by the Americans (1889–1900). To a large extent these Sumerian texts have been published, though many unpublished texts remain. But few of the texts from the current (since 1948) University of Chicago excavations at Nippur have been published. More than 16,000 texts have come from Kanish (Kultepe) in eastern Turkey since 1882. Of these texts dated to the Old Assyrian period, about 2,000 have been published. “The main body of texts, excavated by the Turkish Historical Society since 1948, has remained unpublished but for a handful of tablets and is not accessible to scholars” (Oppenheim, p. 397). It is a warning of the incompleteness of our documentation that no text of the Old Assyrian merchants has yet been discovered in Assur proper.

Of the 20,000 documents and letters found at Mari, about 1,300 have been published. Of the Assyrian letters found at Nineveh, about 2,000 are still unpublished in the British Museum. The royal correspondence from Nineveh dates from Sargon II (722–705) to Ashurbanipal (669–633). It is an accident of either survival or discovery that no letter to Sennacherib (705–682) is to be found in this corpus.

The main bulk of the tablets excavated by the University of Chicago 1930–36 at Tell Asmar (Eshnunna) remains unpublished. Many of the texts found at Adab in 1903 and 1904 are unpublished, as are many texts from Babylon.

Material remains, such as Greek pottery found at Near Eastern sites (e.g. Babylon), are sometimes unpublished. The reports on the Beth Shean dig, completed in 1933, have not been fully published, though a work on levels V and IV is forthcoming.

If one could by an overly optimistic estimate reckon that ¼ of our materials and inscriptions survived, that ¼ of the available sites have been surveyed, that ¼ of those sites have been excavated, that ¼ of the excavated sites have been examined, and that ¼ of the materials and inscriptions excavated have been published, one would still have less than 1/1,000 of the possible evidence (¼×¼×¼×¼×¼). Realistically speaking, the percentage is no doubt even smaller, as suggested in the following example from the Roman world:

In the first three hundred years of the empire there were never less than twenty-five Roman legions, and each legion had five thousand men. The legions were paid three times a year, so that there were 375,000 pay vouchers a year. Multiply that by three hundred, and the result is 112.5 million. Of those, only six and a fragment of a seventh survive [Samuel, The Mycenaeans in History, p. 82].

Problems And Promise

In view of the incompleteness of the excavations and the inadequacy of archaeological experience, some early attempts to associate the excavations with the traditions have proved to be mistaken. Schliemann in his first excavation at Troy thought that he had found “Priam’s” treasures in Level II; he was mistaken by over a thousand years in dating that level to the tradition of the Trojan War. A so-called Jebusite wall found by Macalister in Jerusalem 1923–24 has been redated by Kenyon’s new excavations to the Hellenistic period, a millennium later. Garstang attributed walls found at Jericho to Joshua’s time, but they belonged to the Early Bronze Age, nearly a millennium earlier.

There are complex problems in relating the archaeological evidence of the destruction of sites to the tradition of the Israelite conquest. From data from Bethel, Tell Beit Mirsim, Lachish, and Hazor, Albright has proposed a thirteenth-century date for the conquest that has been widely accepted. But Jericho seems to have been destroyed in the fourteenth century (A.O.T., p. 273), and Ai captured in the twelfth century (J. Callaway, Journal of Biblical Literature, 87 [1968], pp. 312–20).

On the whole, however, it may be safely said that the mass of archaeological evidence has strikingly confirmed the traditions and corrected radical skepticism.

In 1950, H. Lorimer in Homer and the Monuments wished to excise the metal greave and the bronze corslet from the epics, since at that time no known examples had been found in Greece. In 1953 metal greaves were found in Achaea. Then in 1960 at Dendra, not only greaves but also the first Late Bronze corslet were found. In 1963 a second metal corslet from the Mycenaean period was found at Thebes (A. Snodgrass, Early Greek Armour and Weapons [1964], p. 71).

In 1948 G. Hanfmann denied the tradition of an Ionian migration to western Asia Minor in the eleventh century B.C. and claimed that this had not taken place until 800 B.C. (American Journal of Archaeology, 52 [1948], pp. 135–55). The same writer seventeen years later acknowledges that recent finds of Proto-Geometric pottery have now confirmed the traditions of an early migration (The Antioch Review [Spring, 1965], pp. 41–59).

C. Torrey in his Ezra Studies (1910) branded the Aramaic of Ezra a forgery. Among other discoveries that demonstrate the authenticity of Ezra’s Aramaic is a papyrus fragment in Aramaic found in 1942 at Saqqara in Egypt, which is dated to the time of Nebuchadnezzar (J. Bright, The Biblical Archaeologist, 12 [1949], pp. 46–52). Torrey in his Pseudo-Ezekiel and the Original Prophecy (1930) denied the authenticity of the dating of Ezekiel’s prophecies by years of Jehoiachin’s captivity and also Ezekiel’s picture of the material situation of the exiles. Discoveries of jar-stamps in Palestine in the 1930’s and the publication of the ration texts from Babylon by Weidner in 1939 have fully vindicated Ezekiel (W. Albright, The Biblical Archaeologist, 5 [1942], pp. 49–55).

We can agree with D. Winton Thomas, who says (A.O.T., p. xxxii): “Archaeological research will, we may believe, continue steadily to show that the Old Testament narrative is essentially trustworthy.…” And we may add, for the next 8,000 years of excavations!

Have something to add about this? See something we missed? Share your feedback here.