You may read this text sequentially, as you would a printed article, or scroll backward and forward as you would in an electronic text. At the end of each section there is a link that allows you to return to the outline if you wish.
The very peak of the northernmost hill of the site we believe to be Shikhin lies at 188 m. above sea level. Since some of the antiquities extend to the next hill 600 m. to the south south-west, it is possible that the village, which was known in Talmudic times for its pottery-making, extended roughly 750 m. from north to south and about 500 m. east to west. But in the absence of excavation we do not know how much of this area was occupied in which periods. (See Map 2: Topographic Map of Shikhin) [You may read on by scrolling, or return to the outline now.]
In 1881 the Palestine Exploration Fund brought out The Survey of Western Palestine. In that work "Tell Bedeiwiyeh" was mentioned at the west end of the Beit Netofa valley, but was not identified with any site in the Bible, Josephus, or rabbinic writings.
In 1896 Schlatter advanced the linguistic argument that the Asaphon of Josephus must represent the Hebrew place name tsprn, which is found in Josh 13.27. He added as an aside the observation that the Asochis of Josephus must represent the Hebrew place name Shikhin.
In 1905 W. Oehler identified the Asochis of Josephus with Tell el-Bedawiye at the west end of the Bet Netofa valley (which he identified with the "Plain of Asochis" in Josephus) north of Sepphoris. Oehler's argument was simply that Asochis must lie in the west end of the valley, that Asochis cannot be Kefr Manda, which is known to the rabbis, and that the tell lay close to Sepphoris.
In 1907 Peter Thomsen published Loca Sancta. In one place he identifies the Josephan city of Asochis with modern Tell el-Bedawiye. On the other hand (and in another place) he mentions that the identification of Shikhin in Galilee is unknown.
Probably the scholar who treated the talmudic texts relating to Shikhin most exhaustively was Samuel Klein. As early as 1909 Klein published a thorough review of the rabbinical and Josephan references in his work on the twenty-four priestly courses. Klein argued for the identification of Asochis, Shikhin, and Tell el-Bedawiye on the basis of four traditions about Shikhin: (A) The story of the fire at Shikhin that elicited the response from the "Soldaten aus der Akropolis von Sepphoris" suggests that the two localities were side by side. (B) Sepphoris and Shikhin are mentioned in a tosefta as examples of neighboring cities (t Meila 2.9). (C) That R. Jose of Sepphoris reports on several interesting details of Shikhin further suggests that the two localities are next to one another. (D) A tosephta implies that Shikhin and Ruma are two Sabbath journeys or 4,000 cubits apart (t Erub 3.17). Klein believed that this settles the question, for he believed that distance places the investigator at Tell el- Bedawiye. This great scholar maintained this position in his subsequent publications until his death in 1940.
Little that is new appeared in topographical research to advance the argument after 1909. W.F. Albright repeated the identification of Shikhin with Asochis and rejected the identification with Tell el-Bedawiye in 1923, but thought it lay in the vicinity of the tell.  Subsequently, Dalman used his extensive knowledge of Semitic languages and of the texts with his equally detailed knowledge of the topography, but he did not advance the discussion.  Szczepanski repeated this three-way identification in 1928, but with a question mark.  Avi-Yonah accepted the identification of Asochis, Shikhin, and Tell el-Bedawiye in his various editions of the Map of Roman Palestine and The Holy Land from the Persian to the Arab Conquest, beginning in 1931, but culminating in his Gazetteer of Roman Palestine in 1976. Abel followed Klein in detail in 1933. Press in 1955 accepted the identification of Shikhin with Asochis and mentions that Klein identified Shikhin with Tell el-Bedawiye, but that Saarisalo advanced his own hypothesis.
Saarisalo's hypothesis reappears in a recent re-issue of Klein. Bagatti did not treat ancient Shikhin or Asochis in his work on Galilee, since neither were associated with ancient Christian settlement. Miller and Safrai collected most of the talmudic texts again in 1984 and 1985, especially as they related to Sepphoris, but Miller did not attempt a new identification for Shikhin. On the other hand Safrai identified Shikhin with H. Ruma. The identification of Shikhin with Tel Bedawiyeh has remained normative as late as 1989 in the Tübingen Atlas des vorderen Orients. The exception was Saarisalo, who made a major new proposal for the identification of Asochis in 1929. Saarisalo followed Klein in arguing that the Greek place name Asochis is doubtless to be identified with the talmudic place name Shikhin.
Furthermore he argued from Josephus that Asochis is to be found in the western end of the Sahl el-Battof, today's Bet Netofa Valley. There are only two possibilities there for Asochis: Kefar Mandi, a known talmudic site, and Tell el-Bedawiye, the preferred site. But Saarisalo noted that there are no Roman potsherds to be found at the tell. In fact, he posits a gap at Tell el-Bedawiye from the Early Iron to the Arab I periods. Saarisalo proposed that ancient Asochis be identified with a Roman ruin he found during the summer of 1928 at the west end of the valley. Saarisalo speaks of a Roman town two kilometers north of Tell el-Bedawiye and slightly over one kilometer southwest of Kefr Mandi. He suggests that the local place name be Khirbet el-Lon, since the whole district is called by the Arabs "Ard el-Lon." He describes the ruin as situated on a low, rocky ridge, just north of the beginning of the Wadi Bedawiye. The area of debris is 270 m. long from WSW to ENE. This would place the site at about map ref 1727 2453. No site appears at this location in either the Survey of Western Palestine (as Saarisalo noted, p. 35), or in the Yalqut ha-Pirsumim, the official list of antiquities sites in modern Israel, nor on the 1:10,000 maps of the region published by the Survey of Israel. Since this putative site is north of Tell Bedawiye, it is surely too far from Sepphoris to be a likely candidate for Asochis.
In the pages that follow we will examine the Josephan and rabbinic texts on Shikhin and relate them to our archaeological data. Our objective will be to recapitulate some of the scholarship about this important Talmudic site, but more to the point, to show that in all likelihood we have discovered the actual site of ancient Shikhin/Asochis. [You may read on by scrolling, or return to the outline now.]
In A 13.338, Josephus mentions that Sepphoris lies "a little distance from" (mikron apothen) Asochis. In fact, the two sites lie so close together that Asochis is usually located in relation to Sepphoris whenever it is mentioned (cf. B.233, 384). In the passage at hand, how are we to understand the phrase "a little distance from" (mikron apothen)? The exact phrase occurs only twice in Josephus, here and at A 4.79 where it means "a short distance outside" the Israelite camp. The adjective "little" (mikros) can be used alone of space to express being a short distance from something (as at A 8.206), but Josephus' more characteristic usage is to employ it in a temporal or emotional sense to express nearness to a disaster or a "close call" (e.g., A. 13.106; 14.279,356). Clearly the phrase "a little distance from" (mikron apothen) indicates that Josephus located Asochis close by Sepphoris. Moreover, Josephus knew this site first-hand, having visited it himself (cf. V.384).
From another passage (V.233) we learn that to get to Asochis one must descend from Sepphoris. Here Josephus' narrative describes how delegates of his enemies go first to the village of Japha, southwest of Nazareth, then to Sepphoris, and then descend to Asochis. Thus Asochis should lie just below Sepphoris, on the other side away from Nazareth, exactly where our survey site is located.
Lastly, it should be noted that Asochis is called a "city" (polis) whenever Josephus mentions it. Since our identification of Asochis/Shikhin, places the city adjacent to the large "city" (polis) of Sepphoris, in what sense may Asochis be said to be a polis? Here a digression on the meaning of the term polis in Josephus' writings seems necessary. That Josephus works primarily with two dominant terms to designate urban life in Palestine becomes clear in his description of Galilee where he specifically mentions "cities" (poleis) and "villages (komai) (V.235; B 3.43). While numerous places are specifically termed "cities" (poleis), in a more general sense the term designates those places which are distinct from the countryside (e.g., A 11.28) or a village (B 4.127; A 18.28; 20.130). In the course of his historical narratives, Josephus often uses the term "cities" (poleis) with no particular specificity of meaning (B1.316; 2.365; 3.63), even when he lists specific cities (B 1.156;165-66; 2.97; 2.629). In its loosest usage, the term appears in rhetorical statements which allude to "every city" (e.g., B 1.614;2.109, 125) or "all the cities" (B 2.279; 7.96).
An earlier study had indicated that Josephus thought of "city" (polis) in the Hellenistic sense of a city properly organized along Greek political lines, especially when describing Jerusalem as a polis. Traces of this understanding appear about other cities, e.g., when we hear mention of the "Council" (boule) of Tiberias (V.169) or of the political independence of Sepphoris (autopolis A 18.27) or in the numerous references where polis = the citizen body in its collective sense (e.g., "states": B 1.242,428,474).
Furthermore, when a city is designated a "capital" city, it also carries this political understanding, especially in relation to Jerusalem (B 1.339,433, etc.), which is also most frequently referred to as a "city" in Josephus' writings. Indeed, Jerusalem especially can be termed the city (e polis, B 1.229). Moreover, the term metropolis can be applied to Jerusalem in a spiritual/political sense (B7.375), but that same term designates other political capitals (cf. Shechem at A 11.340; Gadera, B 4.413).
Similarly, the rule of a city over surrounding villages is sometimes noted (e.g., Julias at A 20.159) or the [implied] suzerainty of a city over surrounding villages is alluded to (Sepphoris: V.346) or a specified territory is mentioned (B 2.252; 4.443, 444, 452).  All of these carry something of the political meanings of the term (polis) in the Hellenistic political vocabulary.
But when Josephus uses the term polis with some observable precision, he seems to have thought that population size was one of the primary factors that demarcated a "city" (polis) from a "village" (kome). Thus he can say of Batanea that it is "a village not inferior in size to a city" (A 17.23; cf. A 20.130 of Lydda). Therefore when Herod Antipas raises the village of Bethsaida on the Sea of Galilee to the status of a polis, he does so by adding residents and strengthening its fortifications (A 18.28). It is the size of the population, not the fortifications, that is significant here; for both cities and villages were fortified in the Galilee (cf. B 4.127)--this in addition to the numerous fortresses (phrouria) that Josephus mentions (B 1.56, 57, 316; 3.34 [Sepphoris itself]; 4.446, etc.). Thus population size, rather than possessing a wall or fortifications, seems to have been what determined the difference between a city and a village.
However, size can be a rather imprecise factor to use in selecting urban terminology, and the designation "city" for a site may still rest on political charters which we do not know about from our received sources. Josephus seems to use population size as his benchmark for when to call a place a city. But exactly because size is an imprecise category, we can read in Josephus about small cities and large villages. Josephus even has a term to designate the "small city," what we might call today the "town" (cf. B 3.20): polichne.. He uses this term of Giscala (B 4.84), Engaddi (B 4.402), and Hebron (B 4.529). Less frequently (three times) he uses the related term "small town" (polichnion). The small towns along the coast between Joppa and Dora before Caesarea was built, he designates with a hapax legomenon: polismatia.... (A 15.333).
Thus Josephus has three basic urban categories: "city" (polis) "town" (polichne), and "village" (kome); and size seems to be the most usual way of determining which is which (cf. B 4.44 of Vespasian's activities; cf. 4.438). Since an imprecise and informal category like size is how Josephus characterizes a polis, we should not be surprised that a place designated a "small city" in one place, can be called simply "a city" elsewhere (e.g. Giscala at B 4.104; Hebron at B 4.530), or that a site called a "city' one place can be called a "village" elsewhere (e.g., Garis: cf. B 3.129 to V.395). The latter heightens our suspicion that at a number of points polis simply means "small city" or "town" (e. g. Hebron, A 12.353). It is this last usage that Josephus must be reflecting when he calls Asochis a polis: he means it has a large enough size to be a "town," though it clearly cannot be a polis in the same sense that near-by Sepphoris is one. When we first walked the perimeter of the site with the regional inspector of the Lower Galilee, Ariel Berman, we were struck by the size of the then putative Shikhin/Asochis and the density of the pottery at the base of the hill and in the surrounding cultivated fields. Though we had not then located the kilns, we knew we were not examining the typical small out-lying village site.
This view that the term polis in Josephus designates towns of widely varying size is further reinforced by the care with which Josephus points out which cities and villages are the largest (megista in a region (e.g., V.123, 230, 232; B 1.86; 3.34). The degree to which a site is fortified seems (as we saw above) to be a somewhat different and distinct category from that of a site's size--that is, the "strongest" city (krateotate: B 2.111; 3.34, ochurotata [the most frequent term]: A 3.111; 14.113; 18.379; B2.568, etc.) is a slightly different category than that of the "largest" city. Asochis, then, as we said, should be understood, in Josephus' parlance, to be a "town" or small city directly adjacent to and downhill from Sepphoris, to be found on the far side of Sepphoris from Japha. [You may read on by scrolling, or return to the outline now.]
Also a certain Nehemiah, "a man of Kefar Shikhin," is alluded to here and there among the Midrashim, Tosefta, and in the Yerushalmi.
For the purposes of location and identification of the site of Shikhin, it is important to note that Shikhin and Sepphoris were close to one another in the rabbinic citations. Tosefta Meila 2.9, cited above, names Sepphoris and Shikhin as examples of neighbors: "And he said to him, "...one from Shikhin and one from Sepphoris...'" But they were closer than merely in the same region. The two localities play a major role in a well known story about a fire in Shikhin to which soldiers of the "castra" (fortress) of Sepphoris responded:
"It happened that a fire broke out in the courtyard of Joseph ben Simai of Shikhin, and the men of the castra (qtstrh of Sepphoris came to extinguish it. But he did not permit them. ... Nevertheless, on the night following the Sabbath he (Joseph) sent to each of them (the men of the "castra") a sela [tetradrachm], and to their prefect he sent fifty dinarin [denarii]." (t Shabbat 13.9, y Shabbat 16d, b, 121a and y Nedarin 4.9G.).In other words, Shikhin lies quite near Sepphoris, for (1) Shikhin was so close the soldiers could see the fire, get to it, put it out (had that been allowed), and return in one day and (2) the soldiers must have had orders to control such fires, else they would not have tried to put it out. That implies that Shikhin lies within the immediate administration of the city of Sepphoris. (3) The soldiers went down to Shikhin, which places Shikhin at a lower altitude, perhaps simply below the hill of Sepphoris. Note that this is in accord with Josephus, wherein the delegation travels down to Shikhin from Sepphoris (A V. 233, see above).
Furthermore that the Sepphoreans would ask R. Jose ben Halafta of Sepphoris about a situation in a cave of Shikhin suggests that Shikhin lay in the vicinity of R. Jose's city and within his aegis. That is, the writers assume that R. Jose's link with Sepphoris is probably historical.
We have mentioned that Klein has estimated the distance between Sepphoris and Shikhin as two Sabbath day's journeys:
"R. Judah related: It once happened that the Memel and Gorion families at Ruma [rumah distributed dried figs and dried grapes to the poor in a time of dearth, and the poor men of Kefar Shikhin and Kefar Hananyah used to come and wait [on the Sabbath eve] at their Sabbath limit until dusk, and on the following day got up early and proceeded to their destination." (b Erubin 51b; t Erubin 3.17; y Erubin 4.22a(m))
By using an 'eruv, therefore, the poor of Shikhin extended their permitted Sabbath walking distance so as to be able to reach Ruma (likely the rouma of Josephus, B3.7.21--233) and return home. This places Shikhin at a distance greater than 2,000 cubits but less than 4,000 cubits from Ruma (cf. t Erubin 4.10), modern Kh. Ruma, at map ref. 178 244. This distance is about 2.24 kilometers, if the cubit is 56 cm, or 2.14 kilometers, if the cubit is 53.34 cm, or 1.83 kilometers, if the cubit is 45.75 cm. If we take this text literally, we still cannot measure the distance from ancient Ruma and our survey site, since we do not know the location of the outermost structures considered to belong to these settlements. Nevertheless the northern slopes of the survey site, now cultivated fields with large quantities of pottery sherds (see below, section VII), are no more than 2.4 km. from Kh. er-Ruma. It may be conjectured, therefore, that the effective distance between the settlements for purposes of an 'erub was even smaller, and could well have been less than 4,000 cubits. [You may read on by scrolling, or return to the outline now.]
Although it is surely correct that Asochis and Shikhin are linguistically related, it does not follow that Asochis/Shikhin must be identified with Tell el-Bedawiye. Klein noticed that the name appears to mean "pit" (shich). There are no traces of an ancient potter's pit at or near Tell el-Bedawiye, but the old maps of British Palestine of our (then unnamed) survey site reveal a large pit on the north west side of the northern hill (see Map 4). Our informants at Kibbutz Ha-Solelim told us that they filled the pit to simplify their agricultural operations in that field. We interpret this feature as the pit from which the potters mined the clay for the kilns. Our surface survey reveals that no ancient features are to be found in this pit.
The Commanding, near presence of Tell el-Badawiye suggests that there is some connection between the potter's village of Shikhin and the nearby tell. One might speculate that the tell was a strategic and fortified military post in various periods, and the village simply bore the Hebrew version of the name of a main topographical feature. On the other hand, perhaps Shikhin functioned as the support village for a contingent of soldiers camped on the tell. The village housed the soldiers' dependents and eventually turned to the pottery industry. Perhaps this situation would be roughly analogous to that of "Greater Herodium" as the support village for the fortress of Herodium. Finally, the finding of pottery "wasters" and shards at our site during foot reconnaissance survey in 1988, and the discovery that these wasters and shards are identical by neutron activation analysis with one of the main pottery groups found at Sepphoris, seems to clinch the matter. The wasters indicate the presence of a pottery industry. That the majority of the storage jars of Galilee come from this site, and that nearly 15% of the pottery repertory of Sepphoris, as ascertained by neutron activation analysis, is likewise made here suggests strongly that the survey site can be none other than Shikhin, the village cited by Jose ben Halafta of Sepphoris so often in the rabbinic sources. [You may read on by scrolling, or return to the outline now.]
The surface survey of the site revealed modest remains of ancient structures. There are architectural fragments distributed about the site. Several of these fragments are now in re-use in the many modern terrace walls on the hill; others have tumbled down the slopes (especially on the east) and lie at the base of the hill close to fields presently under cultivation. Still others lie in the numerous piles of stones which have been gathered to make the area more suitable for agricultural use. Occasionally large pieces of ancient plaster were observed on the surface. Cuttings in bedrock, noted throughout the site, show that the hill was extensively quarried for building stones. These outcroppings are also frequently incorporated into buildings and other structures, a situation similar to that at nearby Sepphoris. Today this entire area is the location of a flourishing olive orchard.
Indicator shards were gathered from the surface of the site. A team of volunteers was organized so that the entire area would be systematically explored in strips from north to south. A count of the shards collected is provided in the following table:
The pottery distribution indicates a long history of activity at this site, ranging from the late Iron Age through the Byzantine period. Indeed, the heavy concentration of pottery from the Roman period (109 of 140 shards collected) suggests that the site flourished from the first through the fourth centuries. Six wasters were found in the northwest portion of the site not far from the pit which appears on the 1924 maps and which is referred to above as a location from which clay was likely excavated for the manufacture of pottery. As we have suggested, this is evidence for a kiln in this immediate vicinity.
Fifty-eight features of archaeological interest were mapped in relation to known survey points using a TOPCON GTS-3B Total Station. This sophisticated electronic surveying instrument provided a high degree of accuracy (an error factor of less than 1 in 10,000) in the location of these features. The features included thirty-two cisterns in various stages of deterioration, twelve solution cavities apparently used in antiquity (for water storage, dry storage, or as lime kilns), seven features in cut bedrock (foundations for walls, thresholds, door posts, and water channels), three stone fragments (one threshold or doorpost and two basalt grinders in re-use in modern terrace walls), one small fragment of an olive press in reuse to close a cistern, one large screw-type olive press, one ancient wall visible at the surface, and one small aqueduct with several capping stones in place.
A detailed list of these features, with numbers which allow their location on Map 3, follows:
88001 A survey point from earlier surveying activity, consisting of a round headed spike driver into quarried bedrock near the pit at the northwest of the site. This was used as a primary point for the location of some of the features described in this survey.
88002 A bell-shaped cistern with a capstone having an opening ca. 0.3 x 0.3 m. square. The depth is estimated to be greater than 4 m.
88003 A fragment of a doorpost or threshold in re-use in a later terrace wall. The dimensions of the fragment are ca. 0.5 x 0.4 x 0.2 m.
88004 A cistern without capstone presently blocked with a large boulder. The mouth of the cistern is ca. 0.6 m in diameter.
88005 A large solution cavity in bedrock, open from the northwest and the southwest. The opening is ca. 2.3 m wide and ca. 0.7 m high. This chamber was plastered in antiquity; fragments of the plaster survive.
88006 A possible building foundation on quarried bedrock. The walls range from ca. 0.8 to 1 m thick. Preserved length is ca. 2 m from north to south by 10 m east to west.
88007 A cistern blocked with a large boulder. The mouth of the cistern is eroded to a diameter of ca. 1 m. A fragment of a capstone is wedged into the opening to support the boulder.
88008 A fragment of an olive press ca. 1.5 x 0.7 x O.35 m now in use to block the opening to a cistern with an eroded mouth ca. 0.7 m in diameter.
88009 A cistern. The bedrock seems to have been cut to receive a capstone. Present diameter of the mouth is ca. 0.7 m. This cistern is filled with stone and debris.
88010 A large bell-shaped cistern with a mouth ca. 1 m in diameter. The depth of the shaft to the bell is ca. 0.8 m. The cistern is filled with debris to ca. 1.8 m of the surface.
88011 A bell-shaped cistern with a square opening ca. 1 m, now badly eroded. The cistern is filled with debris to ca. 2 m of the surface.
88012 A large cistern with a badly eroded mouth greater than 1 m in diameter. A tree is growing from one side of the cistern. The depth of the cistern is estimated to be greater than 6 m.
88013 A cistern carved into a bedrock shelf created by quarrying. The cistern is now filled to the surface with debris. The mouth diameter appears to be ca. 0.5 m.
88014 A large bell-shaped cistern partially closed with a boulder. The mouth diameter is ca. 0.3 m widening rapidly into the bell shape. The present depth to debris on the bottom is ca. 7 m.
88015 A possible grinder or cistern capstone fallen away from secondary use in a modern terrace wall (which appears to incorporate ca. 4 m of an earlier wall. The stone is cube, ca. 0.4 m in each dimension. The center hole is ca. 0.3 m narrowing to 0.15 m.
88016 A possible Byzantine wall (judging from the construction technique). This wall is 2 rows wide with a rubble core. The width is ca. 1.1 m. The preserved length is ca. 25 m. An angle iron point from some earlier survey was observed at the east end of the wall.
88017 A large cistern with a badly eroded mouth greater than 2.5 m in diameter. A tree growing from the cistern makes access difficult. The depth appears to be greater than 6 m.
88018 A bell-shaped cistern cut into bedrock. The circular shaft, ca. 1 m in diameter is ca. 0.9 m. in depth to the bell. The depth appears to be ca. 6 m.
88019 A cistern with a square-cut opening now partially eroded. The cistern is filled with debris to ca. 1 m of the surface.
88020 Quarried bedrock, possibly used as the foundation of a building, but no distinct pattern was evident.
88021 A storage room ca. 3 x 3 m with an arched doorway cut into bedrock. The chamber is filled with debris to ca. 1 m of the ceiling. Modern bottles lie on top of earlier debris.
88022 Cut bedrock, apparently for the construction of a doorway. The opening is ca. 1.4 m wide. The preserved height is ca. 0.4 m and the preserved length of quarried bedrock exposed is ca. 8.5 m.
88023 A cistern now blocked with a boulder. The mouth appears to be ca. 0.6 m in diameter; the depth greater than 3 m.
88024 A cistern filled to the top with debris. The mouth, now badly eroded, appears to be ca 1.1 in diameter.
88025 A collapsed cistern or lime kiln (although no liming is now evident). The opening is now ca. 3 m in diameter.
88026 A cistern with a nicely cut shaft, now filled to the top with debris. The neck opening is ca. 0.55 m in diameter.
88027 Quarried bedrock, possibly for the foundation of a building. The bedrock forms a "wall" ca. 0.5 m wide and is preserved to a length of ca. 6.2 m.
88028 A possible structure or entrance to an underground chamber. Collapsed earth beneath a modern terrace wall reveals a possible lintel or portion of a structure.
88029 Probable aqueduct carved into bedrock and running approximately from north to south. The channel is U-shaped in section and has a preserved depth of ca. 0.2 m and a uniform width of 0.4 m. The preserved length is 5.4 m now showing at the surface.
88030 A deep cistern exposed by recent harrowing around an olive tree. The cistern is typically bell-shaped with a neck opening of ca. 0.3 m, depth to bell of ca. 1 m. Depth appears to be greater than 4 m.
88031 A cave or solution cavity in bedrock.
88032 A cistern partially blocked with stones. The neck is ca. 0.6 m in diameter, the depth to the bell is ca. 0.5 m. Overall depth to debris is estimated to be greater than 3 m. One blocking stone might be a fragment of a threshold.
88033 A large, circular collapse, perhaps from a cistern or lime kiln. The diameter is ca. 3.2 m and the depth ca. ca. 1.5 m.
88034 A circular collapse, perhaps from a cistern or lime kiln. The diameter is ca. 1 x 1.5 m and the depth ca. 1 m.
88035  A cistern with a badly eroded mouth. A tree growing from this cistern has caused considerable erosion. The diameter of the mouth is ca. 1.5 and the depth to the bell is ca. 0.5 m. The overall depth could not be determined.
88036 A very large cistern with a fig tree growing from its mouth. The present diameter of the mouth is greater than 4 m. The depth is also greater than 4 m. A shaft high on the east northeast wall may connect this cistern to 88037.
88037 A very large bell-shaped cistern ca. 6 m in diameter widening to the south. The northern wall has collapsed. An opening greater than 3 m in diameter appears to be collapsed and not the original neck. A shaft high up on the west northwest wall may connect this cistern to 88036.
88038 A ramp-like depression, ca. 5 m wide and 5 m long leads to a solution cavity the opening of which is ca. 5 m wide.
88039 A cistern blocked with stones. The neck is eroded to a present diameter of ca. 1.6 m; the depth is estimated at ca. 5m.
88340 A cistern with an opening originally ca. 0.8 m in diameter but now badly eroded on the west. This cistern is nicely carved from bedrock and is now filled to ca. 1.1 of the surface with debris.
88041 A cistern with an opening ca. 0.8 m. The mouth is blocked with stones, and the cistern is filled with debris to ca. 2.3 m of the surface.
88042 A large solution cavity with an opening ca. 10 m wide to the east and 2 openings to the west.
88043 Cuttings in bedrock above a solution cavity. The cuttings in bedrock consist of a circular depression, 0.7 m in diameter and 0.15 m deep. Adjacent to the south southwest is a cup-like, conical depression ca. C.3 m in diameter and 0.18 m in depth. A modern Israeli survey point is located adjacent to these features.
88044 A solution cavity ca. 4 m wide with a very shallow chamber.
88045 A large cistern greater than 2 m in diameter and greater than 8 m deep.
88046 A screw-type olive press carved from bedrock. The press would be ca. 1.7 m wide with post holes on each side. This press is broken (and not preserved) on the east side.
88047 Possible cistern with an opening eroded to a diameter of ca. 1.3 m. This chamber is filled to ca. 0.4 m of the surface with debris.
88048 Possible foundation for a building cut from bedrock. There appears to be a collapse of cut stone from upper courses into the interior space defined by walls ca. 0.55 m wide preserved to a length of 1.7 m from east to west and 2.7 m north to south.
88049 A possible collapsed cistern open at ground level. An eroded or collapsed opening ca. 0.65 m exposes a "pit" ca. 1 m deep which may open into a cistern.
88050 A possible cistern or pit filled to ca. 0.2 m of the surface with earth and stones.
88051 A solution cavity in bedrock.
88052 A very large, bell-shaped cistern with a broken capstone. The mouth of the cistern beneath the capstone is greater than 1.5 m and the depth of the cistern, to debris fill, is greater than 5 m.
88053 A possible solution cavity with a very small opening, ca. 0.2 m.
88054 A solution cavity with an opening ca. 0.75 m.
88055 A circular depression which appears to be a collapsed cistern with an eroded mouth now ca. 1.3 m in diameter. This possible cistern is filled with stone and debris to ca. 0.75 m of the surface.
88056 Possible drain or aqueduct of irregular width. Preserved length is ca. 1 m.
88057 A bell-shaped cistern adjacent a pile of stones. The neck has a diameter of ca. 0.6 m. It was not possible to estimate the depth.
88058 Circular depression over an underground cavity, perhaps a cistern. The diameter is 0.4 m and the depth greater than 2.5 m.
88059 A small basalt grinder, ca. 0.3 m in diameter, in reuse in a modern terrace wall.
As we have suggested above, a kiln site may be tentatively identified on the northwest slope of the north hill included in this survey. A study of the attached map (Map 3: Survey of Shikhin) shows that the concentration of features and cisterns tends to be along the summit of the north hill and on its eastern slope. Evidence of ancient occupation (primarily cisterns and solution cavities used in antiquity) diminishes but continues onto the northern slopes of the second, south hill.
The discovery of a kiln in this region of lower Galilee is of great importance and we propose to excavate of this site in the near future. [You may read on by scrolling, or return to the outline now.]
One class of evidence that has not been dealt with in the past in the identification of ancient Shikhin is the role of this town in pottery manufacture. Only two Galilean settlements, as noted, are mentioned in Rabbinic literature as centers for pottery manufacture in the Roman period. This information is significant because much of Rabbinic literature was formulated in Roman Galilee. The first of these centers, at Kefar Hanania, has been the subject of an earlier study. This study showed that this settlement was the principal supplier of common cooking ware to the Galilee and also marketed its ware to the Golan from the latter part of the first century BCE through the early fifth century CE. That study involved the analysis by neutron activation of a large number of common pottery vessels from 17 sites in the Galilee and Golan. These data also showed that the common storage jars used in the Galilee during the Roman period were not made at Kefar Hanania. Employing this evidence and subsequent analytical data obtained at the Hebrew University in Jerusalem and the Lawrence Berkeley Laboratory of the University of California, this survey appears to have located that manufacturing center.
An important component of work on local trade in common pottery in the Galilee and Golan is the quantification of pottery collections from a series of excavated site in these regions. These counts provide an estimate of the relative quantities of each of the vessel forms found at these sites. One of the quantified collections was from Sepphoris. All of the pottery (excluding lamps) was counted that the University of South Florida Excavations at Sepphoris recovered from two areas near the present fortress (Field I, areas 3 and 6). These areas were selected because they contained remains of residential structures and subterranean cavities dating from the Early Roman through the Late Roman Periods. These quantitative data showed that two bowl types of distinct form were unusually abundant at the site. Counts of collections from other excavated sites showed that bowls of these two types, referred to here for convenience as serving bowls and bell bowls, are relatively uncommon outside the central hill area of lower Galilee. The large quantity of these specialty bowls at Sepphoris seemed to be best explained by hypothesizing that they were made in or near that city. This supposition was tested by neutron activation analysis of six examples of these bowls from Sepphoris, three of each type. The analysis showed that all six bowls matched in chemical composition, supporting our initial hypothesis.
Other analyzed vessel types share the same composition as these bowls. Among these vessels were twenty-two storage jar pieces of a single class excavated at Sepphoris, Hammath Tiberias, Tabgha, Capernaum, and Horvat Hazon in the Lower Galilee, at Meiron, and Nabratein in the Upper Galilee, and at Susita, Gamla, Ein Nashut, and Dabiya in the Golan (Map 4: Sites in the Galilee and Golan with analyzed pottery belonging to the Shikhin provenance group). All the analyzed storage jars of this class except one matched in chemical composition. The composition of the last storage jar was quite similar, and it is likely that it too was made in the same vicinity.  The storage jars of this class are the common storage jars found at Galilean sites in contexts dating from the late first century BCE until about the mid third century CE; they also occur at Gamla and other sites in the Golan.
Rabbinic sources provide significant information for locating the site of manufacture of these common Early and Middle Roman Galilean storage jars. As mentioned, only two Galilean settlements are referred to as centers for pottery manufacture, Kefar Hanania and Shikhin. However, whereas Kefar Hanania is specifically noted as a center for the manufacture of several utility vessels, Shikhin is specifically noted only for its manufacture of storage jars. The Shikhin storage jar was so well known by the mid second century that the measure of its volume could be proposed as a standard for halachic purposes (t Terumot 7.14; y Terumot 8.6, 45d).
As mentioned, the analyzed storage jars discussed above are the common storage jars found in Galilean context of Roman date. This common occurrence of one class of storage jar is the pattern we would expect for jars attested in the literary sources as being so well known. The sampled jars of this class share the same provenance, and most of the analyzed jars are well dated by their archaeological contexts to the first and second centuries. The common provenance of these jars, their wide distribution, and their chronology are all consistent with the literary reference to Shikhin as a well known center for the manufacture of storage jars by the mid-second century. It is likely, therefore, that the storage jars of this class are the storage jars of Shikhin.
Based on the archaeometric evidence outlined above, it was thus learned that a group of pottery, including bowls of distinctive form and the common storage jars of Roman Galilee, was made somewhere in the vicinity of Sepphoris. The literary sources gave reason to believe that the place of manufacture was Shikhin.
Three examples of ceramic waste ("wasters") were recovered from the site discussed above during the 1988 survey. Two of the wasters were found within a radius of several meters. Such ceramic waste is considered one of the best indicators of local pottery manufacture. Since wasters were not marketed, the only explanation for their recovery at a site is that pottery was made there.
One of the wasters seems to be an example of the serving bowls mentioned above. In Photo X-1, this waster is shown beside a large fragment of one of these bowls. One sees that, although the waster is misshapen and is fused to globs of fired clay, both pieces have a deep groove of similar diameter and a higher adjacent rim. The lower and inner sides of the waster are vitrified. The Second waster is a deformed storage jar handle (Photo X-2). This piece has a warped body, and pronounced bloating and vitrification of the handle are evident. The third waster, of nondescript form, is apparently doubled up (Photo X-3). This piece is extensively bloated and is vitrified on the exterior.
The wasters were analyzed by neutron activation, and the data for each of the three pieces are shown in Table 2, columns 3-5. A comparison of the chemical composition of the wasters with that of the pottery group, which includes the specialty bowls and common Galilean storage jars (Column 2, called Group 1), shows that the values all agree closely. For the apparent serving bowl waster (Column 3), one sees that all values except Cr lie within one sigma of the pottery group, while Cr is 2.1 standard deviations below the group value. For what would be considered a valid statistical match, we would expect about eleven of the sixteen elements to deviate from the group mean (M) by less than 1. standard deviation, about fifteen elements should lie within 2. standard deviations, and about one should deviate by greater than 2 standard deviations.. For the deformed handle of the storage jar (Column 4), we note that thirteen elements lie within 1 standard deviation of the group, and all sixteen are within 2. standard deviations. Finally, for the waster of nondescript form (Column 5), eight values are within 1 standard deviation., fourteen lie within 2. standard deviations, and two elements are between 2 and 3 standard deviations.. Although the values for the last piece do not seem to match as closely with Group 1 as do those of the previous two wasters, it should be noted that the clay used to make Group 1 pottery was diluted in the process of manufacture. One cause of this dilution was apparently the addition of calcite. When the composition of the third waster is decreased by 3% to normalize its clay content to that of Group 1 (Column 6), it is seen that thirteen elements are within l standard deviation of Group 1, fifteen are within 2 standard deviations, and one value lies between 2 and 3 standard deviation.
This match in composition between the pottery group (Group 1), on the one hand, and the three wasters on the other, is substantive for the provenance of the pottery group at this manufacturing center. Two additional vessels recovered during the survey, one a serving bowl and the other one of the common Galilean storage jars, were also analyzed. The compositions of these vessels also matched that of the pottery group and are included in that group.
It should be noted that the composition of the pottery group discussed here (Group 1) differs markedly from that of the Kefar Hanania Group, the other important Galilean pottery provenance group of the Roman period. In order to provide perspective on these differences, the compositions for the 183 pieces of the Kefar Hanania group are shown in Table 2, Column 7. One sees that elements Ca, Hf, Sm, Ta, and Th cover a range in Group 1 (Column 2), but there is no overlap between this range and that for the Kefar Hanania Group.
From the foregoing analytical evidence it seems well established that the common storage jars of Roman Galilee and other pottery vessels were made at the surveyed site identified as Shikhin. Recent analyses have demonstrated moreover, that this provenance group also includes common storage jars of Late Hellenistic date.  Thus the location of the surveyed site, its role as a pottery manufacturing center, its principal product, and the period of its production are all consistent with what is known about Shikhin from the literary sources.
The pottery repertory from the survey is limited to sherds picked up from the area between the two hills, not from the north side of the north hill. Although hundreds of sherds were visible on the north, they were not sampled, for they had undergone severe disturbance by modern plowing. Here we also present sherds similar to the "wasters" treated in Section VI of this paper. The forms that are to be found in survey in no way reflect the full range of forms found at any site by excavation. On the other hand it is significant that there were no surprises in the pottery of the Roman period, as is treated here.
Fig. 1:7: Rim of a jar with thickened and inverted rim, but with a small eversion on the exterior lip. It is red (2.5 YR 5/7) on the interior and exterior with a 100% grey core. It resembles an Early Roman jar rim from Meiron.
Fig. 1:36: Base of a small juglet. The base is finely tooled, with a groove in the ring base. In general form and fabric it is an Early Roman sherd, very lightly ribbed on the outside.
Fig. 1:14: The sherd is from the string-cut base of a juglet. It is reddish-brown on the exterior and interior with a 100% grey core. In form it resembles Herodian juglets, especially Lapp's form 31.2 (globular juglets with flat base), but its faint ribbing and generally rougher fabric might place this juglet in the second century. 
Fig. 1:34: In this case we see the ribbed body and less red ware of a Late Roman juglet. The base is formed by turning the pot over and closing the base by hand.
Fig. 1:35: This juglet base is string-cut, as no. 14, but with rather more pronounced ribbing. It appears to be of Early Roman date.
Fig. 1:30: A rim and partial collar of a Roman jar is red (2.5YR6/8) on the interior and exterior with a 99% grey core.
Fig. 1:8: This rim of a Late Roman jar has a slight groove near the exterior lip of the rim and is the same color on the exterior surface as it is within the fabric (5YR5/1 or grey). It resembles a type of Late Roman jar found at Meiron.
Fig. 1:18 A rim of a large cooking pot with an outside diameter of about 31 cm. It has a pronounced rolled rim with a ridge that forms a groove just below the rim. The red ware with many sparkling inclusions has no core. In general form and fabric this form resembles 8th century Iron IIC cooking pots. It also resembles the 7th century cooking pots from Ta'anach.
Fig. 1:38 A rim of a jar with outside diameter about 11.4 cm., of grey ware with 100% core. Inclusions are few, large, black. In general this resembles Iron Age wares, such as the large jars from Ta'anach, periods IB (12th cent.) and IIB (10th cent.), though the Ta'anach jars have outside diameters at the rim of about 20 to 23 cm.
Fig. 2:5. Cooking pot with rounded body, thin strap handles, short neck, and a single groove inside the rounded rim. This Early Roman cooking pot is well known over most of the Galilee. For example, it is called Form A4 at Capernaum, where it is wellrepresented in the Early Roman stratification. David Adan-Bayewitz calls it form 4A in the Khirbet Hananya repertory.
Fig 2:23 and 33. These sherds represent the "bowl with everted lip" from Khirbet Shema and Meiron, or Adan-Bayewitz' form 3A in the Khirbet Hananya repertory. At Capernaum this is form A17, "Tegami a labbro sporgente." Loffreda dates this pot at Capernaum from 63 BCE to 135 CE.
Fig 2:3. This handle of a small juglet is less than three centimeters long. The attachment at the top has a tight enough curvature to suggest that it was attached to a narrow mouthed juglet. This may be Capernaum form AlOa, usually less than 10 cm high, and with a collar at the top. This form dates 63 BCE to 135 CE at Capernaum. 
Fig. 2:31, a handle similar to number 3, but in a redder ware and larger. The curvature of the sherd at the top suggests a neck less than one centimeter in internal diameter, therefore a sherd of a juglet. This may also be Capernaum form AlOa of 63 BCE to 135 CE.
Fig. 2:19, a jar or amphora handle. Its form is quite similar to Capernaum jars of class A, or likely the Shikhin jar. Wall thickness at the handle attachment is only 4-5 mm. A second handle from the survey (not illustrated), no. 25, is redder and slightly wider by three millimeters. On the other hand the thickness of the wall at the attachment is 3-4 mm. These two sherds give some hint of the possible variability of this jar at Shikhin.
The pottery from Shikhin, at least as we have sampled it here, reveals a small occupation in the eighth and perhaps seventh centuries BCE, a small presence in the Persian period, and a possibly greater presence in the Hellenistic Periods I and II. It is during the Early Roman period that the number and types of sherds suddenly expands dramatically (see Table 1 and Fig. 3: Pottery Counts from the Survey of Shikhin.) Although this is a small sample, its importance cannot be gainsaid. The sudden rise in counts is similar in contour of the pottery counts for the entire site of Sepphoris, as shown in Fig. 5, though without the additional peak in the Byzantine I period.
Thus it seems reasonable to hypothesize that the survey site may have been occupied in generally the same periods as Sepphoris, except that so far we find no Arab period potsherds there. Of course only excavation will tell the tale, but at this venture it would be well to sink a probe at the survey site, surely ancient Shikhin, and test the hypothesis that its occupation indeed parallels that of Sepphoris. [You may read on by scrolling, or return to the outline now.]
Although it is surely correct that Asochis and Shikhin are linguistically related, it does not follow that Asochis/Shikhin must be identified with Tell el-Bedawiye. Klein noticed that the name appears to mean "pit" (shich). There are no traces of an ancient potter's pit at or near Tell el-Bedawiye, but the old maps of British Palestine of our (then unnamed) survey site reveal a large pit on the north west side of the northern hill (see Map 3). Our informants at Kibbutz Ha-Solelim told us that they filled the pit to simplify their agricultural operations in that field. We interpret this feature as the pit from which the potters mined the clay for the kilns. Our surface survey reveals that no ancient features are to be found in this pit.
The commanding, near presence of Tell el-Badawiye suggests that there is some connection between the potter's village of Shikhin and the nearby tell. One might speculate that the tell was a strategic and fortified military post in various periods, and the village simply bore the Hebrew version of the name of a main topographical feature. On the other hand, perhaps Shikhin functioned as the support village for a contingent of soldiers camped on the tell. The village housed the soldiers' dependents and eventually turned to the pottery industry. Perhaps this situation would be roughly analogous to that of "Greater Herodium" as the support village for the fortress of Herodium. Finally, the finding of pottery "wasters" and sherds at our site during foot reconnaissance survey in 1988, and the discovery that these wasters and sherds are identical by neutron activation analysis with one of the main pottery groups found at Sepphoris, seems to clinch the matter. The wasters indicate the presence of a pottery industry. That the majority of the storage jars of Galilee come from this site, and that nearly 15% of the pottery repertory of Sepphoris, as ascertained by neutron activation analysis, suggest strongly that the survey site can be none other than Shikhin, the village cited with Jose ben Halafta of Sepphoris so often in the rabbinic sources.
[You may return to the outline now.]
Thomas R. W. Longstaff <firstname.lastname@example.org>
Return to the Archaeolgy Web Page