The delicate carving on the side of the sarcophagus depicts Zeus, in the guise of a swan, graphically forcing himself on the Spartan queen Leda. The scene is one of the best known in ancient Greek mythology, so its appearance on a sarcophagus should be no surprise. This particular sarcophagus, however, comes not from ancient Greece or Rome, but from the vast Jewish burial chambers at Beth She’arim, in Lower Galilee. Numerous rabbis and their relatives were buried here in the third and fourth centuries C.E. Even Rabbi Judah the Prince, the compiler of the great Jewish code of law, the Mishnah, is interred here. What is such a blatantly sexual image, featuring the head of the pagan pantheon, doing in this Jewish cemetery?

A closer look at the sarcophagus tells a second story: The images of Leda and the swan were intentionally damaged in late antiquity. Apparently a visitor to Beth She’arim could no longer bear the presence of this three-dimensional pagan image and took a hammer to it. The sarcophagus was then turned against the wall so that the offensive image all but disappeared.

The relief depicting Leda and the swan was not alone in its fate. Between the late seventh and ninth centuries C.E., a wave of iconoclasm swept across Palestine. Synagogue mosaics, which previously had often been rich in representations of people, mythic figures and animals, were frequently defaced.

Who destroyed these ancient Jewish images? How does their creation and subsequent destruction reflect a shifting understanding of the biblical prohibition known in Jewish tradition as the second commandment:(a) “You shall not make for yourself a sculptured image, or any likeness of what is in the heavens above, or on the earth below, or in the waters under the earth. You shall not bow down to them or serve them” (Exodus 20:4–5)?

Scholars have long been fascinated by iconoclasm, a religious community’s deliberate destruction of images that had previously been found acceptable.(b) But they have been prone to a curious blind spot about the subject: They have tended to shift blame for iconoclasm away from the particular community they identify with. Specialists in Christian art, for example, long blamed Jews and Muslims for the destruction of Christian images in the eighth and ninth centuries C.E.(1)

Scholars of Islamic art have attributed the Muslim aversion to depicting humans or animals in religious contexts to Jewish and Christian influence.(2) Can you guess whom scholars of Judaism have tended to blame for synagogue iconoclasm?

Three explanations were put forth early in the 20th century to account for the destruction of images in ancient synagogues. The Palestinian Jewish scholar Samuel Klein suggested that Christian zealots defiled the art in the process of destroying the synagogues themselves.(3) This view, however, fails to explain why only human images were destroyed and very few animal images, and why Hebrew inscriptions and religious symbols such as menorahs were left untouched.

Another Jewish scholar, archaeologist Samuel Yeivin, blamed the Muslim conquerors of Palestine in the seventh century C.E.,(4) though there is no real evidence to support this view, either.

A more subtle theory was advanced by Carl Watzinger, the German archaeologist who in the early years of this century conducted the first systematic excavations of Palestinian synagogues. He argued that iconoclasm within synagogues reflected an internal shift in Jewish religious orientation; the Jews were responsible for despoiling art within their own places of worship.(5) I believe Watzinger was right, although for the wrong reasons.(6)

Judaism has never had a uniform attitude toward art.(c) The Bible itself includes not only prohibitions on images but also loving descriptions of the lavish decoration of King Solomon’s Temple (1 Kings 6–7) and a detailed account of the ornate construction of the Tabernacle and its vessels (Exodus 25–27).

During the late Second Temple period (first century B.C.E. and first century C.E.), however, Jewish art was aniconic, perhaps reflecting a strict approach to the second commandment. Mosaic floors in Jerusalem homes were decorated only in geometric patterns. Newly minted coins lacked the images of rulers that were common throughout the Roman world. In the second century C.E., however, that attitude changed. The rabbis of the second through the early fifth centuries discussed, and had varying opinions on, all sorts of art: Does the second commandment ban all images or just the worship of mages? they asked. In other words, is the law banning “sculptured images” (Exodus 20:4) dependent on the subsequent law “You shall not bow down to them or worship them” (Exodus 20:5)? Were only three-dimensional representations forbidden by the second commandment, or did the prohibition extend to two-dimensional images as well? Were images of human beings inappropriate but not representations of animals or plants? Was art made by Jews restricted but not art made by Gentiles?(7)

The debate would continue for centuries.

Some Jewish sages vehemently opposed any artistic reference to Greco-Roman motifs, even when it came to the portraits of rulers (whom the Romans worshiped as divine) on coins. Rabbi Nahum son of Simai, called “Nahum of the Holy of Holies,” a Palestinian rabbi of the third century C.E., was praised “because he never looked upon the [idolatrous] image on a coin.”(8) When Nahum died, all the images that were in his town were covered with mats out of respect. No matter how much Nahum is praised, however, it seems clear that his views were in the minority and that such extreme aniconism was rare. Indeed, if his had been the common attitude, why would he have been singled out for praise?

A statement from the Jerusalem Talmud that has been preserved only in a manuscript fragment discovered in the Cairo Genizah(d) reveals the opinion of two influential rabbis, who (apparently after the fact) seem to have allowed art on the walls and floors of Jewish public buildings: “In the days of Rabbi Johanan [third century C.E.], they permitted images on the walls, and he did not stop them. In the days of Rabbi Abun [fourth century C.E.], they permitted images on mosaics, and he did not stop them.”(9)

An important interpretation of the biblical prohibition is expressed in an Aramaic paraphrase of Leviticus 26:1, found in Targum Pseudo-Jonathan.(e) The passage in Leviticus reads: “You shall not make idols for yourselves, nor set up for yourselves carved images or pillars, nor place a figured stone in your land to bow down upon it, for I the Lord am your God.”(f) But the Aramaic paraphrase significantly expands on the biblical text in order to distinguish between idols and other more acceptable images: “You shall not make idols for yourselves, or set up for yourselves carved images or pillars, or place figured stones in your land, but a pavement figured with images and likenesses you may make on the floor of miqdasheikhon. And do not bow down to it, for I am the Lord your God” (Targum Pseudo-Jonathan).

A “pavement figured with images and likenesses” is a decorated mosaic floor. The term miqdasheikhon is a little harder to understand, however, because of difficulties with translating the Aramaic. It has been variously rendered as “your synagogues” and as a reference to the Jerusalem Temple. I prefer the latter reading,(10) which would have offered a strong precedent for the use of mosaic carpets in synagogues: Just as mosaics “figured with images and likenesses” were once permissible in the Temple (Miqdash), this passage seems to suggest, they now legitimately exist within the “small temple” (miqdash me’at), the synagogue—as long as they are not bowed down to. This text suggests that the acceptability of decorative mosaics in synagogues was a debated subject in Byzantine and early Islamic Palestine.

The broad spectrum of rabbinic attitudes toward visual images is also reflected in the archaeological record. The sarcophagus carving featuring Leda and the swan is just one of many diverse images found in the third- to fourth-century C.E. rabbinic necropolis at Beth She’arim (see the sidebar to this article).(11) Sarcophagi bearing images of an eagle, bulls’ heads and a fragmented marble amazon were discovered in catacomb 20, which houses the remains of Rabbi Joshua and his family and Miriam daughter of Rabbi Jonathan. The sarcophagus of Gamaliel son of Rabbi Eliezer, also from catacomb 20, is more modestly decorated with floral and animal motifs, as well as architectural and geometric designs, but no human figures. Catacomb 14, identified as the tomb of Rabbi Judah the Prince, contains no images. Evidently those responsible for this catacomb held a more conservative view of representational art.(12)

The archaeological evidence from the fourth century C.E. demonstrates a wide range of attitudes toward synagogue art. The synagogue at Hammath Tiberias, on the western shore of the Sea of Galilee, which served the aristocratic circle around the patriarch (a member of the urban elite, who, in the time of the Roman Empire, would have been regarded as a leader of the Jewish community in Palestine and the Diaspora), has a mosaic that goes far beyond conservative rabbinic conceptions. At the center of this beautifully preserved mosaic floor is the boyish sun god Helios, portrayed with rays of light extending from his halo. Surrounding Helios is a large wheel with the 12 signs of the zodiac, depicted in colorful detail and labeled in Hebrew. Among them is a nude (and uncircumcised!) Libra (see photo of mosaic floor from synagogue at Hammath Tiberias). Remember, this image was not transferred to the synagogue from a non-Jewish context; it was commissioned, and perhaps even produced, by Jews. Just above the zodiac panel lies an image of a Torah shrine flanked by two menorahs.

The juxtaposition would be jarring if it were not so common. The Hammath Tiberias mosaic is the earliest known example in a series of synagogue mosaics—from Sepphoris, Beit Alpha, Naaran, Khirbet Susiya and Huseifa—that pair zodiacs with traditional Jewish imagery. Subtle variations among these mosaics, however, indicate slight differences in various communities’ attitudes toward the images.

The recently excavated zodiac mosaic from Sepphoris, also in the Galilee, indicates that this community was less willing than the one at Hammath Tiberias to depict pagan deities.g Their fifth-century C.E. zodiac mosaic replaces the sun god with a sun disk, and the symbols for the months are more modestly dressed.

The attitude toward representational art continued to develop during the sixth to the eighth centuries C.E. The sixth-century synagogue mosaic at Beit Alpha, in Lower Galilee, includes both a zodiac wheel with Helios in his chariot (see photo of mosaic floor from synagogue at Beth Alpha) and a biblical scene (the binding of Isaac). The latter scene shows Abraham, knife in hand, with Isaac, the sacrificial victim, to his right. At the top of the panel, a small hand—representing God—reaches down from the heavens, stopping Abraham in the nick of time.

At Ein Gedi, on the western shore of the Dead Sea, we find much starker decoration: A sixth-century C.E. mosaic panel in the narthex (entry hall) of the synagogue simply lists the names of the months and the zodiac signs, but includes no images. The synagogue at Rehov, 5 miles southeast of Beit Alpha, has no zodiac. Instead, a 29-line inscription from rabbinic literature graces the narthex of the synagogue.

In the city of Beth-Shean, 3 miles east of Beit Alpha, we can see how a single community struggled to create mosaics that were appropriate to their particular settings. A public room, known as “the House of Kyrios Leontis,” and a study house or synagogue both open onto the same courtyard at Beth-Shean. The public room floor is enlivened with mosaics showing two scenes from Homer’s Odyssey and one Nile river god. The mosaic in the study house is, in comparison, rather tame, depicting a menorah and what scholars call an “inhabited scroll”—a vine that winds around images of animals but not, at least in this case, a single human or mythological figure.

All of this changes in the eighth century C.E., when a more severe aesthetic takes hold in Palestine. The often lovely depictions of people and animals that had once graced synagogue floors fell out of fashion. One of the latest extant synagogue mosaics was laid in Jericho in this period (see photo of mosaic floor from synagogue at Jericho): At the center appears a small roundel featuring a menorah flanked by a ram’s horn (shofar) and a palm frond (lulav). Beneath these images is a simple Hebrew inscription from Psalm 128: “Peace upon Israel” (Shalom al Yisrael). The remainder of the floor is covered in a basic geometric pattern of alternating diamonds and heart-shaped motifs. A contemporaneous synagogue mosaic from Tiberias bears images of palm frond bundles (lulavim), but, again, no humans or animals.

That is not the only change witnessed in this period. For it is at this time that iconoclasm begins to surface in synagogues. While new synagogues were built with less and less representational art, older synagogues were also brought into line. Art that did not conform to the new aesthetic was altered to render it acceptable.

The changes to mosaic floors and sculptured reliefs were executed with extreme care. Only those parts of the image that were deemed offensive under the new stricter standards were removed—and no more. That this process involved careful excision rather than haphazard destruction is our strongest evidence that the iconoclasm was carried out by Jews and not by others.(13)

Examples of this “retroactive sensibility” can be found at Kefar Baram and Capernaum in Upper Galilee, where in both cases the wreaths carved into the lintel of the entryway still remain, but the winged Nikes (Greek goddesses of victory) that originally flanked the wreaths were carefully removed (see photos and and drawing, above). A synagogue screen discovered in Tiberias was also altered: Two birds still stand at either side of a menorah, but their heads have been chipped off. A mosaic from the Meroth synagogue, in Upper Galilee, shows a Roman soldier surrounded by his helmet, shield and sword (see photo of mosaic floor from synagogue at Meroth). But the soldier’s eyes have been painstakingly excised—a well-known form of iconoclastic defacement.(h)

The most dramatic example of iconoclasm comes from the fifth- to sixth-century C.E. synagogue at Naaran, a Jewish village just north of Jericho. Here the zodiac wheel remains visible, but all of the symbols have been gouged out (see photos of mosaic floor from synagogue at Naaran). The individual tiles have been removed so carefully, however, that we can still pick out the slender-waisted figure of Virgo, though only the outline of her torso remains. The legs of Scorpio, the scorpion, remain, though his body is no more. At the center of the zodiac, the wheels of Helios’s chariot are visible, as are the rays of light that once emanated from his face. But the face itself is gone. Elsewhere in the mosaic floor, depictions of Daniel in the lions’ den, the signs of the seasons and the birds that once inhabited an elaborate scrolling vine are similarly defaced yet still recognizable. Whoever did the damage, however, seems to have shown an insider’s respect for some elements of the mosaic: Images of the menorah and of the Torah shrine seem to have remained unscathed, as do almost all of the Hebrew and Aramaic inscriptions (even those labeling the mythological figures).

Iconoclasts at the synagogue of Khirbet Susiya in the Hebron Hills were more thorough. Here they removed not only the offending figures but an entire round mosaic—presumably a zodiac wheel—and put a crude geometric mosaic in its place at the center of the synagogue floor. The somewhat imprecise iconoclasts at work here left one side of the original mosaic partially undisturbed. Here we can see a figure of Daniel in the orans position (with arms raised in prayer) flanked by two lions. The last two letters of Daniel’s name, alef and lamed, are all that remains of the accompanying Hebrew inscription.

The marble screens that surrounded the main bema, or raised platform where the liturgy was enacted, in this same synagogue were also subject to intense iconoclastic activity. Images of animals were removed, while Jewish symbols such as menorahs were left intact.

And of course at Beth She’arim, we have the sarcophagus relief of Leda and the Swan, now missing their faces.

This is Jewish iconoclasm at work: Imagery that is no longer acceptable is fine-tuned into respectability, while cherished religious symbols and texts are preserved intact. By altering the images in their synagogues, the Jewish communities at Khirbet Susiya, Naaran and numerous other sites were able to continue using their synagogues for centuries afterward without being disturbed by now-forbidden images.

What spurred this pervasive change in Jewish attitudes? It is not impossible that some Jews were influenced by what they perceived as Christian excesses and idolatry. Jewish antagonism toward Christian images was expressed as early as the sixth century in a poem by the Jewish liturgical poet Yannai. The poem rails against Christians, who “rejoice in statues of human figures, / Who cleave to the dead over the living, / Who become excited and turn aside to lies,…/ Who prostrate and pray to a bush [a reference to the cross].”14 However, the changes to Jewish synagogues date not to the Byzantine Christian era but to the early centuries of Arab domination in the Holy Land.

In 638 C.E., the Muslim caliph Omar, Commander of the Faithful, entered Jerusalem on foot. The Byzantine period in the East had ended, and almost 1,300 years of Arab domination were about to begin.

It is the coming of Islam, more than anything else, that led to iconoclasm in synagogues. But this does not mean that the Muslims were responsible for the damage done.

By the late seventh century C.E., the Jewish community in Palestine found itself no longer in the image-rich milieu of the Roman-Byzantine world, but within the orbit of a culture that eschewed the use of images in religious settings. The Muslim rulers of Palestine were greeted much more positively by Palestinian Jews than were their Byzantine Christian predecessors. With rabbinic attitudes toward images markedly mixed to begin with, this new atmosphere offered little reason to maintain the vibrant visual vocabulary of past centuries. Christian communities in Palestine seem to have gone through a similar process. In fact, the changing attitude toward representational art that we have traced in the Jewish community points to the general truth that every community of faith in every age responds to and interacts with the prevailing trends of the broader cultural context. While each faith community reacts in its own unique ways, no religion is isolated from the general currents of its time and place.

a. On the numbering of this commandment as the first or second in Jewish and Christian tradition, see Ronald Youngblood, “Counting the Ten Commandments,” BR 10:06.

b. Archaeologist Robert Schick has explored the effects of eighth-century iconoclasm on Christian churches in Palestine. See “The Image Destroyers,” AO 02:05.

c. On the changing interpretation of the second commandment in Jewish tradition over time, see Victor Hurowitz, “Did King Solomon Violate the Second Commandment?” BR 10:05.

d. According to Jewish law, old books and ritual objects bearing the name of God may never be destroyed and should be buried in consecrated ground. A synagogue’s genizah served as a temporary storage place until the writings could be properly buried. The Ben Ezra synagogue in Old Cairo is the most famous genizah yet discovered.

e. A Targum is an Aramaic paraphrase, often with expansions, of the Hebrew Bible. Pseudo-Jonathan, the most expansive of the Targums to the Pentateuch, is roughly twice the length of the original Hebrew text. It was redacted in the eighth or ninth century C.E.

f. This same passage still troubles scholars today. For a new understanding of the “figured stone” (even maskit), see Victor Hurowitz, “Wish Upon a Stone: Discovering the Idolatry of the Even Maskit,” BR 15:05.

g. See Zeev Weiss, “The Sepphoris Synagogue Mosaic,” BAR 26:05.

h. See Emmanuel Damati and Zvi Ilan, “The Synagogue at Meroth: Does It Fix Israel’s Northern Border in Second Temple Times?” BAR 15:02.

1. Most recently Robert Schick, The Christian Communities of Palestine from Byzantine to Islamic Rule: A Historical and Archaeological Study (Princeton: Darwin, 1995), pp. 180–224. The bibliography on iconoclasm is huge. See the following basic studies and the bibliographies cited by each: Peter Brown, “A Dark-Age Crisis: Aspects of the Iconographic Controversy,” English Historical Review 88 (1973), pp. 1–34; Daniel J. Sahas, Icon and Logos: Sources in Eighth-Century Iconoclasm (Toronto: University of Toronto, 1986), pp. 16–18; Sidney H. Griffith, Theodore Abu Qurrah, A Treatise on the Veneration of the Holy Icons (Leuven: Peeters, 1997), pp. 1–27.

2. See Keppel Archibald Cameron Creswell, Early Muslim Architecture (Oxford: Clarendon, 1932), pp. 269–271; Oleg Grabar, The Formation of Islamic Art (New Haven: Yale Univ. Press, 1987), esp. pp. 43–98.

3. Samuel Klein, The History of the Jewish Settlement in Palestine (Jerusalem: Mitzpeh, 1935), pp. 36–37 (Hebrew). I discuss evidence for the Christian destruction of synagogues in my “Non-Jews in the Synagogues of Palestine: Rabbinic and Archaeological Perspectives,” in Jews, Christians and Polytheists in the Ancient Synagogue: Cultural Interaction During the Greco-Roman Period, ed. Steven Fine (London: Routledge, 1999), pp. 233–236.

4. Samuel Yeivin, “Excavations in the Land of Israel 1925 Season),” Zion: Yedyiot ha-Hevra ha-Eretz-Yisraelit le-Historia ve-Etnographia 2 (1930), p. 15.

5. Carl Watzinger, “Die Antiken Synagogen Galiläas: Neue Ausgrabungen und Forschungen,” Der Morgen 6 (1930), pp. 362–363.

6. Carl Watzinger, following H.H. Kitchener, theorized that the synagogues in Galilee had been constructed on behalf of the Jews by Roman authorities during the Severan dynasty. When this dynasty fell into political turmoil in the mid-third century, the Jews took this opportunity to remove offensive Roman images from their synagogues. The notion of Roman imperial sponsorship of synagogues reflects the misconception, common in the 19th and early 20th centuries, that art and Judaism are incompatible.

7. The best discussion of rabbinic attitudes toward art is still Joseph Baumgarten, “Art in the Synagogue: Some Talmudic Views,” Judaism 6 (1970), pp. 196–206.

8. y. Meg. 1:11, 72b; y. Sanh. 10:5, 29c; y. Abod. Zar. 3:1, 42c; b. Pesah. 104a; b. Abod. Zar. 50a; Eccl. Rab. 9:10. Arthur Marmorstein, The Doctrine of Merits in Old Rabbinic Literature and the Old Rabbinic Doctrine of God (New York: Ktav, 1968), pp. 215–216.

9. Jerusalem Talmud, Abod. Zar. 42b.

10. The problem with reading miqdasheikhon as “your synagogues” is that the term “temple” or “small temple” is never used in rabbinic literature to refer to synagogues without being paired explicitly with the word “synagogue.”

11. See Lee I. Levine, “The Finds from Beth-Shearim and Their Importance for the Study of Talmudic History,” Eretz Israel 18 (1985), p. 275 (Hebrew).

12. Nahman Avigad, Beth She’arim (New Brunswick: Rutgers Univ. Press, 1976), vol. 3, pp. 42–65. See Tessa Rajak’s excellent and strongly revisionist reassessment of scholarly interpretation of Beth She’arim, “The Rabbinic Dead and the Diaspora Dead at Beth She’arim,” in The Talmud Yerushalmi and Graeco-Roman Culture I, ed. Peter Schäfer (Tübingen: Mohr Siebeck, 1998), pp. 349–366.

13. See Michael Avi-Yonah, Oriental Art in Roman Palestine (Rome: Centro di studi semitici, Istituto di studi del vicino Oriente, 1961), p. 42, reprinted in Art in Ancient Palestine, ed. Hannah Katzenstein and Yoram Tsafrir (Jerusalem: Magnes, 1981), p. 159; Schick, Christian Communities of Palestine, p. 193.

14. The Liturgical Poetry of Rabbi Yannai, ed. Zvi M. Rabinovitz (Jerusalem: Bialik Institute, 1985–1987), vol. 2, pp. 221–222.

This article is based on Steven Fine’s “Iconoclasm and the Art of Late Antique Palestinian Synagogues,” which will appear in the forthcoming book From Dura to Sepphoris: Studies in Jewish Art and Society in Late Antiquity, ed. Lee I. Levine and Zeev Weiss (Ann Arbor, MI: Journal of Roman Archaeology Supplementary Series, 2001).