Bible and Beyond
Immerse yourself in the ongoing Sepphoris mikveh debate

Click here to view the original article.

Scholars have been arguing for some time about the purpose of several plaster-clad stepped pools in the ancient Galilean settlement of Sepphoris, just 4 miles northwest of Nazareth. Not long ago, BAR published the opposing views of two prominent scholars. Duke University professor Eric Meyers, who has extensively excavated at Sepphoris, insisted that the Sepphoris pools were indeed ritual baths, known as mikva’ot (singular- mikveh)—but Hanan Eshel, of Bar-Ilan University in Israel, contended that the pools were only used for routine bathing.a

The debate matters because a mikveh definitely establishes a Jewish presence in a community—and many mikva’ot means many Jews. The mikveh question is key to determining if, in the first century C.E., the inhabitants of Sepphoris were predominantly Jewish, or not.b

I believe Meyers is correct- These water installations in Sepphoris were mikva’ot.

On a recent informal visit to the site, I counted 22 stepped and plastered water installations (rather than the “dozen or so” mentioned by Eshel) in the northwestern excavation area, in the upper part the city. Even more installations might be hidden in underground cellars. I have marked these 22 installations on a plan of this area of the site that was published in stages by Meyers with Carol L. Meyers and Kenneth G. Hoglund in 1995 and 1997.1 I have also inserted on the map three stepless small installations (unnumbered), which might have served as bathtubs.

The installations are coated with the gray plaster that is characteristic of cisterns and aqueducts as well as mikva’ot of the early Roman (late Second Temple) and Byzantine periods. Some of these installations at Sepphoris have as many as seven layers of plaster, a fact mentioned neither by Meyers nor Eshel. Multiple coats of plaster were often applied to mikva’ot, as has been observed in Jerusalem, Jericho, Qumran and elsewhere.

Why so many coatings? The answer might be found in the detailed religious regulations recorded in the Mishnah, the first great rabbinic work (edited in about 200 C.E.), which devotes an entire tractate to mikva’ot. The Mishnah describes the mikveh as “a pool of water containing 40 seahs [60–250 gallons] in which they immerse,”2 (Mikva’ot 1.7), and prescribes how to stop up a leak, should the water of mikveh “creep out” (Mikva’ot 5.5). This regulation indicates that leaks occurred often enough to be of some concern to mikveh users—thus the application of so many layers of plaster.

Another important rule pertaining to mikva’ot requires that the mikveh contain “living water,” rather than “drawn water.” Indeed, both Meyers and Eshel use the term “living water” (mayyim h\ayyim) to describe the quality of water that is required in a mikveh. But this term is more complex than they suggest.

“Living water” is a Biblical term that is used only in a discussion of the purification of a zav, a person suffering from a genital discharge (due most likely to venereal disease). According to Leviticus 15-13, seven days after the discharge ceases, the zav must, among other things, be immersed in “living water” to become ritually clean again. Rabbinic authorities describe “living water” as water that flows directly from a natural spring, as opposed to rainwater gathered in a man-made basin. Unfortunately, both Eshel and Meyers incorrectly include in their definition of “living water” both water that flows directly from a natural spring and rainwater that flows directly into a pool. Such rainwater does indeed qualify for a mikveh, but it is not “living water.”

The first several verses of Mishnah Mikva’ot rank six types of water in ascending order of excellence. The water of a mikveh is ranked at the third level of excellence (Mikva’ot 1.7); “living water” is rated at level six, the very highest grade of excellence (Mikva’ot 1.8). This difference in rank clearly shows that the water of a mikveh need not be at the highest level.

This distinction is important because all 22 of the stepped water installations that I examined at Sepphoris are located well above the aqueduct that brought water from the nearby spring. Hence, they could not be filled directly with “living water”—rather, they were filled with rainwater that flowed in by gravity from roofs and courtyards.3

Most of the Sepphoris installations were built near cisterns. This is also true of mikva’ot in Jerusalem that were found in private houses. Some mikva’ot in Jericho, however, were fed by an aqueduct from a spring and did not require a nearby cistern, as they could be quite easily filled and refilled with living water.

Yet, in Mikva’ot 2.4, the Mishnah stipulates that an empty mikveh may not be filled from a cistern, because the water in a mikveh cannot be “drawn water.”4 The religious principle undergirding this requirement is that the water of a mikveh should come “by the hand of heaven.” Water drawn with a vessel from a cistern, however, is considered to be brought “by the hand of man” and hence would not qualify. The “living water” of a mikveh must come by direct flow from a natural source, such as a spring or lake; rainwater that flows directly into the mikveh from a roof or courtyard is also usable, even though it is not “living water.” This type of gravity-fed mikveh was typical in the hill country of Galilee and Judea, where most Jews in the land of Israel lived during the first century C.E.

At first glance, this seems to be a puzzling contradiction. Obviously, so many mikva’ot were constructed near cisterns because they were replenished by cistern water. But water taken from a cistern is “drawn water”—by the hand of man. How can this practice have been permitted under Jewish law? The answer is that mikveh water had the power to purify. That, after all, was why people immersed themselves in it. Most utensils and vessels could also be ritually purified by immersion in a mikveh. It stands to reason that drawn water could also be purified by the pure, “living” water of the mikveh. As long as at least 40 seahs of pure water was present in the mikveh, drawn water added to it in small quantities would be purified. This could be repeated as required. (This practice is no longer maintained today.)

In rare cases, an adjacent small pool has been found beside the mikveh, connected to it by a pipe with a plug that could be opened and closed. This side compartment, effectively a mikveh itself, is called an otzer.5 It would be filled either with “living water” or with rainwater that flowed in from a roof or courtyard. Eshel argues that the absence of even one otzer at Sepphoris makes it doubtful that the water installations there are mikva’ot. But there is no religious requirement that a mikveh have an otzer, as even he recognizes. In fact, at hill country sites such as Jerusalem, the mikveh and otzer combination is actually quite infrequent. There is no reason to expect this combination at Sepphoris.

Instead, we often find mikva’ot in pairs, side by side. We must remember that where the mikveh was not served by a continuously flowing spring (as in the vast majority of cases), the pure water had to last from the end of one rainy season, in around March, until the beginning of the next rainy season, about October. If one of a pair of mikva’ot no longer had enough pure water, the other could be used. What’s more, the waters of a pure mikveh could be used to purify the water in the other. Rabbinic literature describes how a moveable pipe or assemblage of pipes could be used for this purpose (Tosefta mikva’ot, 5.5). Pairs of mikva’ot are frequently found in Jerusalem. At Sepphoris I saw five or six pairs or groups that can easily be identified- numbers 4, 5 and 6; 7 and 9; 14 and 15; 17 and 18; 19 and 20; 21 and 22 on the site map.

Another difference between bathtubs and mikva’ot that helps us distinguish between the two is size- Bathtubs were small installations, much smaller than those in Sepphoris. The reason is simple. A bathtub was used for routine bathing, and not for occasional ritual purification. After people cleaned the dirt from themselves in a bathtub the dirty bathwater was discarded into the gutter or the street. In ancient times such a large quantity of water as 40 seahs would not be wasted on a bath. The water of a mikveh, on the other hand, ordinarily is not poured out after use. It remains in the pool for a much longer period, usually from one rainy season to the next.

For this reason many mikva’ot were located in dark basements, preventing the penetration of light and the growth of algae in the water. A later regulation required users of a mikveh to bathe first in a regular bathtub, undoubtedly to preserve the cleanliness of the standing water in the mikveh for as long as possible.

Also, bathtubs had at most one or two steps made for sitting, and never a longer staircase. Among the water installations at Sepphoris that are not mikva’ot are some very small water receptacles with one or two steps. Some of these could have been used as bathtubs.

Different regions produced different types of mikva’ot, just as they produced different types of pottery. In Sepphoris, three of the mikva’ot (19, 20 and 21 on the site map) are quite large (or at least larger than neighboring installations) and are rectangular in shape, with steps along their entire width. Their steps alternate with wider and narrower treads. At least one installation has a small auxiliary step at the bottom. The excavators hesitantly identify this installation as a mikveh.6
There is no doubt in my mind that all three are mikva’ot. They are the “Jerusalem type” of mikveh (also found at Qumran). This kind of mikveh occasionally has a double entrance and/or partitioned staircase and so is quite different from the “Jericho type” mikveh, which in most cases has a staircase on the narrow side of the structure. (In the bathhouses in Jericho, however, are some mikva’ot of the “Jerusalem type”).

In my opinion, the larger mikva’ot in Sepphoris almost surely date to the Second Temple period (ending in 70 C.E., with the Roman destruction of the Temple)—but we can expect a more confident and closer date when Meyers, Meyers and Hoglund publish their final excavation report on Sepphoris.

All the other mikva’ot marked on the Sepphoris site map are smaller than those three “Jerusalem type” mikva’ot. It seems that they have enough distinctive characteristics to be called “Sepphoris-type” mikva’ot. Not only are they smaller and have smaller stairs, but they have a less regular outline and a staircase that usually makes a turn near the bottom. But the Sepphoris mikva’ot also have some features in common with the mikva’ot of Jerusalem- Both have multiple layers of expertly finished plaster, including the rounding of corners and step edges—and both occasionally use a small auxiliary step. Both types also are generally built close to water cisterns.

In short, the similarities between the installations in Sepphoris and in Jerusalem indicate that if not all of the Sepphoris installations served as mikva’ot, most of them certainly did.

Both Eshel and Meyers refer to a famous incident in the 1960s when the flamboyant Israeli archaeologist and general, Yigael Yadin, called in a group of rabbis to confirm that water installations he had excavated at Masada met all the necessary requirements for a valid mikveh. At Masada, Herod the Great’s desert/mountain retreat and fortress (that later was occupied by the last holdouts in the Great Jewish Revolt against Rome), there were more than 20 mikva’ot, not one of which had an otzer, the adjacent reserve pool. But Yadin did not show any of these 20 to the rabbis. Instead he brought them to look at the two that did each have an otzer. The rabbis, not surprisingly, ruled that each of these two conformed to the requirements for a valid mikveh.

An even smaller proportion of otzer-mikveh pairs were revealed in excavations in the Jewish Quarter of Jerusalem, led by the late Nahman Avigad- Only one otzer was found among more than 60 mikva’ot. If a valid mikveh required an otzer, there would have been almost no “kosher” mikva’ot in the Jewish Quarter. Indeed, I know of only two mikveh-otzer combinations among the more than 150 installations in Jerusalem from the Second Temple period.

Although Yadin had a flair for showmanship, modern rabbis should not have the last word in determining the validity of an ancient mikveh, whether or not it has an otzer. The rabbis’ perspective is shaped by two millennia of religious teaching, much of which postdates the installations themselves. It cannot substitute for sound archaeological method, bolstered by the evidence at hand.

a. See Hanan Eshel and Eric M. Meyers, “The Pools of Sepphoris—Ritual Baths or Bathtubs?” BAR 26-04.

b. See, for example, Mark Chancey and Eric M. Meyers, “How Jewish Was Sepphoris in Jesus’ Time?” BAR 26-04.

1. Eric M. Meyers, Carol L. Meyers and Kenneth G. Hoglund, “Sepphoris (S\ippori), 1994,” Israel Exploration Journal 45, no. 1 (1995), p. 69. See also Eric M. Meyers, “Sepphoris on the Eve of the Great Revolt (67–68 C.E.)- Archaeology and Josephus,” in Galilee Through the Centuries, Confluence and Cultures, ed. Eric M. Meyers (Winona Lake, IN- Eisenbrauns, 1999), p. 109–122.

2. As Meyers points out, estimates of 40 seahs vary from 60 to 250 gallons. The quantity of water in a seah varied from region to region, according to rabbinic literature; a Jerusalem seah was different from a desert seah (Mishnah Terumah 7.1).

3. According to Eshel, “Living water … must come directly from a river or a spring or from rainwater that flows into the pool.” According to Meyers, “Rainwater … was a major source of living water.” Both Eshel and Meyers incorrectly regard rainwater as “living water.” BAR 26-04.

4. See, e.g., Torat Kohanim (Sifra), Shemini, 9. Several passages in the Mishnah examine the problem of how much drawn water disqualifies a mikveh that has less than the required 40 seahs of “living water” from a spring or lake. See, e.g., Mishnah Mikva’ot 2.4.

5. The term otzar (related to otzer) was used in the late Second Temple period for a storehouse (e.g., an otzar of wood, etc.) and was not applied to mikva’ot until the Middle Ages.

6. See Meyers, Meyers and Hoglund, “Sepphoris (S|ippori), 1996,” Israel Exploration Journal 47, no. 3–4 (1997), p. 267, fig. 2.