Introduction
The rock tomb of Qizqapan is located in the northeast of Iraq, in the Kurdistan region of Iraq, 60 km to the northwest of Sulaimaniyah, alongside a road leading to Dukan dam (Fig. 1). In the portico of the tomb are two Ionian style half-columns, one on each side of the funeral chamber entrance. A ritual scene is carved between them, showing two persons praying towards a fire altar. Above the praying scene and to the sides of the capitals, three smaller symbols are carved (Figs. 2, 3). The rock-cut tomb of Qizqapan was first described by C.J. Edmonds in 1934. He did not put forward a hypothesis about the dating of the tomb's construction but only pointed out that it was similar in terms of its general plan to the rock-cut tombs of the Median and Achaemenid periods previously described by Herzfeld and De Morgan (Edmonds Reference Edmonds1934: 184). Subsequently, several scholars suggested dates for Qizqapan, primarily based on its architectural ornaments. The suggested dates are relative and cover four eras, as follows.
1. Median Period
In 1941, Herzfeld described Qizqapan as a late Median tomb and indicated that its construction dated to 600 to 550 B.C. (Herzfeld Reference Herzfeld1941: 204). Subsequently, researchers such as Ghirshman (Reference Ghirshman1963: 88) and Diakonoff (Reference Diakonoff and Keshavarz1998: 375) accepted Herzfeld's Median dating for Qizqapan. In addition, Diakonoff assumed that Qizqapan is the tomb of Cyaxares, the third Median king in the narrative of Herodotus. For several decades following their suggestions, Qizqapan has been described as a Median rock-cut tomb in Iranian academic archaeology books (see Mollazadeh Reference Mollazadeh2014: 358-341; Sarfaraz and Firozmandi Reference Sarfaraz and Firozmandi Shireh-Jini2002: 70-68). Following Diakonoff, some Iranian researchers believe that Qizqapan is the tomb of Cyaxares (Sarfaraz and Firozmandi Reference Sarfaraz and Firozmandi Shireh-Jini2002: 69). The Median dating has also been accepted by some Iraqi researchers, who propose that the depicted scene marks the end of the military conflicts between the Medes, with their great king Cyaxares, and the Lydians, with their king Alyattes (Amin Reference Amin2018). A simple search for “Cyaxares tomb” in different languages on the internet shows how frequently Qizqapan has been identified as the tomb of the famous Median king. Generally, the most important reasons for attribution of Qizqapan to the Median period are as follows.
A. Like most other rock tombs attributed to the Median period, the Qizqapan tomb is situated in the geographical region of the ancient Median empire (Herzfeld Reference Herzfeld1941: 200).
B. According to Herzfeld and his followers, some of Qizqapan's elements or motifs, such as the column capitals and their palmettes, are more ‘primitive’ and thus are older than examples from the classical and Achaemenid periods (ibid: 204).
C. Some features from the reliefs of Qizqapan, such as the costume, are unquestionably analogous to those of the Medes (ibid: 205).
2. Achaemenid Period
Kantor was one of the first scholars to identify Qizqapan as an Achaemenid rock-cut tomb (Kantor Reference Kantor1957: 17). Subsequently, von Gall studied Qizqapan comprehensively and dated it to the period between the end of the fifth century B.C. and the late Achaemenid dynasty (von Gall Reference Von Gall1988). Recently, Bahrani described the Qizqapan tomb iconography as Median and dated it to the sixth-fifth century B.C., contemporary with the Achaemenid era (Bahrani Reference Bahrani2017: 300–302). The most important reasons for attribution of the Achaemenid date are as follows.
A. According to Porada, Qizqapan and Da-u-Dukhtar have large close-set volutes that are reminiscent of Greek fifth-century B.C. capitals (Porada Reference Porada1965: 138).
B. According to von Gall, the ceiling of Qizqapan is influenced by Paphlagonian rock tombs in the north of Turkey, which are dated to the Achaemenid period.
C. According to von Gall, the presence of a worshipper or a king by a fire-altar is a very common scene in Achaemenid art (von Gall Reference Von Gall1988: 563).
3. Hellenistic Period (Seleucids and the Frataraka Rulers)
Stronach, mainly based on an iconographic comparison between elements from Qizqapan and motifs from the Frataraka coins, believes that Qizqapan was constructed by local rulers who, for well over a century after the fall of the Achaemenid Empire, still looked back to the original model of Darius’ tomb (Stronach Reference Stronach1966: 221). Other scholars, such as Herrmann (Reference Herrmann1977) and Haerinck (Reference Haerinck and Curtis1997), also agree with this dating. The most significant reasons for which followers of a post-Achaemenid date attribute Qizqapan to the Hellenistic period are as follows.
A. The Ionic capitals of Qizqapan are comparable to some Seleucid capitals, for example, in Khurha (Stronach Reference Stronach1966: 221).
B. At least one of Bagadat's coins illustrates the same empty hanging sleeve of the left-hand Qizqapan figure. In addition, several other coins bear the rare pleated skirt or pleated trouser leg also seen on the left-hand Qizqapan figure (ibid.: 221).
C. The bow of the Qizqapan worshippers can be seen on the Frataraka coins (ibid.: 221).
4. Parthian Period
A small number of scholars, such as Hopkins, have suggested a Parthian date for Qizqapan. He believes that the form of the Qizqapan capitals is more relevant to the Parthian than to the Achaemenid period (Hopkins Reference Hopkins1942: 406).
Fig. 1 Location of the rock-cut tomb of Qizqapan in the Kurdistan region of Iraq and locations of other relevant sites
Fig. 2 Qizqapan portico and its reliefs (photo by Akam Omar Ahmad Qaradaghi)
Fig. 3 Qizqapan portico and its reliefs (drawing by Naser Aminikhah)
Qizqapan Dating, Based on Its Components
The Columned Portico and Carved Ceiling
Before the Achaemenid period, and aside from the so-called Median rock-cut tombs, which have uncertain chronology, no rock-cut tomb with a columned portico has been found among the numerous Urartian tombs in Iran, Turkey and Armenia (Piotrovskii and Khatib Shahidi Reference Piotrovskii, Khatib Shahidi and Khatib Shahidi2004). Based on our current information, the tomb of the Achaemenid king Darius I (522-486 B.C.) is the oldest known rock-cut tomb in the region with a columned portico and an accurate date. Even after the Achaemenid period, the building of rock-cut tombs with columned porticos continued in the Near East. Some examples include the rock-cut tombs of Kaunos, Amyntas, Myra, and Maziköy in Turkey, the Tomb of Zachariah and Queen Helena in Jerusalem, and tombs at Petra in Jordan (Fedak Reference Fedak1990: figs. 39, 120, 122, 125, 204, 215, 229, 234, 235). Consequently, taking the comparable tombs into account, the suggestion of a later date than the Median era seems acceptable for Qizqapan. Ghirshman, influenced by Herodotus’ report of the short domination of the Scythians over Media and Mannea during the mid-seventh century B.C. (1.73; trans. Godley Reference Godley1975), believed that the ceiling pattern of Qizqapan had been adopted from the wooden roofs of Scythian tombs (Ghirshman Reference Ghirshman1963: 88). Based on the historical text of Herodotus (4.71; trans. Godley Reference Godley1975) and archaeological evidence (Ivantchik Reference Ivantchik2011: 83), one form of Scythian burial was in the shape of a hut with wooden ceilings, buried beneath a heap of soil. Nonetheless, a Scythian rock-cut tomb has yet to be reported. Hence, it is clear that in Qizqapan and the Zagros region, a different tradition of burial than the Scythian tradition can be seen.
Von Gall believes that the ceiling of Qizqapan (Fig. 4) was influenced by the rock-cut tombs of Paphlagonia, which date to the fifth-fourth centuries B.C. (von Gall Reference Von Gall1988: 577). It must be noted that the western Iranian rock-cut tombs situated east of Qizqapan (including the rock-cut tombs of the Kermanshah province) either do not have a carved ceiling at all or simply show a profile of ceiling beams similar to those of the Achaemenid royal tombs. On the other hand, carved ceilings similar to that of the Qizqapan tomb, regardless of the arrangement of their vertical and horizontal beams, can be seen in many of the rock-cut tombs of Turkey belonging to the Achaemenid, Hellenistic, and Roman periods, such as Donalar, Salarkoy, Terelik, Kastamonu, Aşaği Güneyköy, Sakkale, Gerdek Boğazi (Summerer and von Kienlin Reference Summerer, Von Kienlin, Nieling and Rehm2010: 207–208), Limyra (Vernet Reference Vernet2017: fig. 1), Gerdek Kaya (Fedak Reference Fedak1990: 346, fig. 124), Pinara, and Mira, and also the rock-cut tombs of Telmissos and Antiphellos, and the Lycian tombs in the Kibyratis (fourth century B.C.) (Gay and Corsten Reference Gay and Corsten2006: figs. 2.3.7, 12.16, 17). Therefore, due to the absence of this decorative pattern in the eastern rock-cut tombs of western Iran, such as Dukkān-i Dāūd, Dira, Ravānsar, Sahneh, Eshāqvand, Barnāj, and Fakhrika, it can be concluded that its origin may lie within the northern regions of Turkey and the Black Sea, extending south-east to Qizqapan. Since the rock-cut tombs of Turkey with this ceiling pattern mainly belong to the Achaemenid and Hellenistic periods, it is probable that Qizqapan belongs to these periods as well.
Fig. 4 Ceiling of the Qizqapan portico (drawing by Naser Aminikhah)
The Half-Column Capitals
Based on comparisons for the Qizqapan capitals, researchers such as Herzfeld, Porada, von Gall, Stronach, Herrmann, and Hopkins suggested a variety of dates for this tomb (Herrmann Reference Herrmann1977: 40; Hopkins Reference Hopkins1942: 406; Porada Reference Porada1965: 138; Stronach Reference Stronach1966: 221; von Gall Reference Von Gall1993: 529). Herzfeld believed that there were no truly Ionic capitals in Qizqapan and described how the imposed blocks project far over the upper diameter of the columns to the right and left, but not in front and back, and end, like an early Ionic capital, in two large volutes. Consequently, he believed that the Qizqapan capitals are a Proto-Ionic type, which would date to after 600 B.C. and before 550 B.C. (contemporary with the late Median period) (Herzfeld Reference Herzfeld1941: 203–210). However, Hopkins doubted Herzfeld's idea that “the columns of Qizqapan must be classed as proto-Ionic” (Hopkins Reference Hopkins1942: 406). On the other hand, the lack of true capitals can be seen not only in Qizqapan, which Herzfeld dated to the Median period, but also on some Achaemenid and post-Achaemenid columns and even in examples of contemporary rural architecture in mountainous valleys in different parts of the Near East, including Kurdistan (Herzfeld Reference Herzfeld1941: 210, 242). This implies that the dating of columns based on capitals is of little use when it comes to Qizqapan. However, the form and technical features of Quasi-Ionic capitals dating back to the first half of the first millennium B.C., found in Assyria and neighboring regions (Herzfeld Reference Herzfeld1941: 245, fig. 349), are different from the Qizqapan capitals.
Recent studies show that the combination of convex fasciae and concave scrolls dates capitals to the fifth-second centuries B.C. (Litvinsky and Pichikian Reference Litvinsky and Pichikian1998: 235) and suggest that older Ionic capitals, such as the ones from the Artemis temple at Ephesus (550 B.C.) (British Museum n.d.a) have convex scrolls. Moreover, on the majority of fourth-second century B.C. capitals, the fasciae are enclosed in rectangular fillets (Litvinsky and Pichikian Reference Litvinsky and Pichikian1998: 235). These technical features, which represent a later dating, are illustrated on the Qizqapan capitals as well. Therefore, it can be construed that the Qizqapan capitals were carved in a later period than the Median era (750–550 B.C.).
During the Hellenistic era, the Ionic order was prevalent in a vast area from the Levant to Central Asia. Ionic columns of the Hellenistic period were found in the Oxus temple (Takht-e Sangin) (Litvinsky and Pichikian Reference Litvinsky and Pichikian1998: fig. 4a) and in some areas of Iran such as Khurha, Nahāvand (Rahbar et al. Reference Rahbar, Alibaigi, Haerinck and Overlaet2014: pl. 11), Bard-i Neshāndeh (Ghirshman Reference Ghirshman1976: pl. 18: 171; pl. xxiv: 1–4), Qaleh-Zahāk (Ghandgar et al. Reference Ghandgar, Esmaeili, Rahmatpour and Azarnoush2004: 214: 3, 215: 2; Ghasemi Reference Ghasemi2009: 544 and 588), Bisotun (Luschey Reference Luschey, Kelaiss and Kalmayer1996: 57), and Shiān (Alibaigi et al. Reference Alibaigi, Aliyari and Aminikhah2018), as well as within many of the rock-cut tombs attributed to this period in Jordan, Israel, Syria, northern Arabia and Turkey. Despite the presence of some Greek elements in the art and architecture of the Achaemenid period, thus far no Ionic capital has been found in the royal centers of Susa, Pasargadae, and Persepolis. However, due to its proximity to Turkey, the use of Ionic columns is possible in the western parts of the Achaemenid Empire. Moreover, although Ionic columns from the Parthian period have been identified in places such as the temple of Gareus and Parthian palace in Assur (Colledge Reference Colledge1967: 112), regarding the absence of the Parthian style in the Qizqapan reliefs, it can be stated that a Parthian date for Qizqapan is less likely than the Achaemenid and Seleucid periods. Some additional details of the capitals, which can be useful factors in the dating of Qizqapan, are described below.
Central Large Palmette
Palmettes like that of Qizqapan can be seen in the art of the Neo-Assyrian and Achaemenid periods, but none of these examples are located between the volutes of Ionic or Quasi-Ionic capitals (Koch Reference Koch1992: figs. 41, 42). Herzfeld believed that a palmette like that of the Qizqapan capitals was suggested by similar designs on objects imported from Greece as early as 600–550 B.C. (Herzfeld Reference Herzfeld1941: 247). However, Hopkins and Porada believe this kind of palmette to be rarely older than the end of the sixth century B.C. (Hopkins Reference Hopkins1942: 406; Porada Reference Porada1965: 139).
The central palmettes of Qizqapan are distinctly different from the archaic examples, which include older ornaments termed ‘Cypriot palmettes’ by Frankfort (Frankfort Reference Frankfort1970: 323, fig. 385; Marquand Reference Marquand1909: figs. 94, 95). Yet they are very similar to the classical palmettes of Greece (fifth–fourth centuries B.C.), for example, in the Erechtheion (fifth century B.C.) and Minerva Polias. This strengthens the possibility of a later dating than the Median period for the central large palmette of Qizqapan.
Egg-and-Dart
Some details of egg-and-dart ornaments used on the abacus of the Qizqapan capitals, including the rounded ovals and rhombic darts, were quite common in the fifth–third century B.C., for example in the Oxus temple (Litvinsky and Pichikian Reference Litvinsky and Pichikian1998: 234, 247). But on later capitals, such as the Temple of Artemis in Magnesia of the early third century B.C (University of Michigan Library n.d.) and the Pergamum altar (197–159 B.C.), ovals were mainly pointed, darts were spear shaped and reliefs were deeply cut to create contrasts of light and shadow (Litvinsky and Pichikian Reference Litvinsky and Pichikian1998: 236, 239). The egg-and-dart ornament at Qizqapan includes no such features, which indicates that the Qizqapan columns are probably older than the early third century B.C. Therefore, a date of the fifth–third century B.C. is suggested for Qizqapan.
Angle Palmettes
On the capitals of the Hellenistic period (for example, the Temple of Athena at Magnesia and Temple of Dionysus on Teos), and in the Roman periods, angle palmettes were either large or very large (Litvinsky and Pichikian Reference Litvinsky and Pichikian1998: 236). The angle palmettes of Qizqapan, on the corners of the capitals, are not as large as the above examples and are comparable with classical Greek examples (Boardman Reference Boardman2000: 67).
Column Bases
The plinths of the Qizqapan columns are square. The closest examples can be found in the Achaemenid period (for example, Schmidt Reference Schmidt1953: fig. 72: I, J, K). Square single plinths have been identified at the gates of Pasargadae, Persepolis, and Susa, and double square plinths or two-stepped plinths can be seen in the Achaemenid royal tombs. Both types were also common in the Hellenistic era (for example, at Khurha and Ai Khanoum). Therefore, it seems that the single plinths of Qizqapan have an eastern identity which was very common during the Achaemenid period, and then, as a traditional and popular element, was preserved in the architecture of the Seleucid era.
Column Shafts
The column shafts in Qizqapan, similar to those of the Achaemenid royal tombs, are plain and do not have any vertical flutings. One of the principles of the Ionian order is the harmony between the components of the pillar, as the height of genuine Ionic columns is at the beginning eight, and never more than ten, diameters (Herzfeld Reference Herzfeld1941: 285). At Qizqapan this proportion is about 6 diameters to the column height, which is shorter and thicker than the standard Ionic style. According to Herzfeld, the lack of proportionality in the columns of the Hellenic monument of Khurha is the result of clumsiness and imitation (ibid.: 286). The same claim could be made about the Qizqapan columns. Generally, we can divide the Ionic capitals of the eastern regions, outside of Turkey, into two general categories based on the presence or absence of ornaments:
1. Ionic capitals which are plain and do not include elements such as echinus, egg-and-dart ornaments or palmettes. These Ionic capitals can be found in Iran (Khurha and Dā-ū-dukhtar), Central Asia (Ai Khanoum) and India.
2. Ionic capitals that include the above-mentioned elements and are more similar to the Ionic capitals of Greece and Turkey than the first group. The column capitals at Qizqapan, Dokhāharān, and the Temple of Oxus are of this group.
The worship scene
Two human figures can be identified in the worship scene of Qizqapan; each holds a bow in his left hand and stands on either side of a fire-altar. This is a common scene in Achaemenid religious iconography (Briant Reference Briant1996: fig. 28b).
The Worshippers’ Costume
Most analyses of the worshipper figures at Qizqapan, in terms of dating and identification, are made regarding their costume type. According to many researchers, both men are wearing Median dress (Diakonoff Reference Diakonoff and Keshavarz1998: 375; Haerinck Reference Haerinck and Curtis1997: 34; Herrmann Reference Herrmann1977: 40; Herzfeld Reference Herzfeld1941: 205; Porada Reference Porada1965: 138). In the Persepolis reliefs, the Medes and some of their neighbours wear a tight belted tunic that reaches down to their knees. Their trousers are tight and convenient for horsemen, and they have strapped shoes or boots (Koch Reference Koch1992: fig. 50). The costume of the right-hand figure of Qizqapan is similar, although due to the erosion of the reliefs, it cannot be stated for sure whether or not he is wearing a belt. The comparison of the Qizqapan figures with other figures on the Persepolis reliefs (e.g., Roaf Reference Roaf1983: fig. 111) indicates that the left-hand figure of Qizqapan wears the same costume but with a mantle over it. In addition, the head-covering of the worshipers of Qizqapan, despite some differences in details, is the same type as the one worn by the Mede delegation at Persepolis; it covers their chin, head, and neck. It has been proposed that this kind of head-covering, also known as bashlyq, would protect the person against the dust (Walser Reference Walser1966) or, as seen in Qizqapan, precluded the “polluting human breath reaching consecrated objects” (Boyce Reference Boyce1982: 20), in this case, the fire on the altar.
The left-hand figure wears a mantle with empty and long hanging sleeves. It is believed that this attire is the kandys (κάνδυς), which Greek historians such as Xenophon have mentioned (Thompson Reference Thompson1965: 121). Such mantles were worn by dignitaries on the Persepolis reliefs, Achaemenid seals and coins, and some Achaemenid reliefs in Turkey (Culican Reference Culican1965: fig. 52; Curtis and Tallis Reference Curtis and Tallis2005: fig. 57; Herzfeld Reference Herzfeld1941: fig. 314). Therefore, despite the absence of sufficient evidence from the Median period, we see that attire similar to that of the Qizqapan figures, including the head-covering, shirt, trousers, shoes, and mantle, have been worn by Medes and their neighbors, such as Armenians, Sagartians, Cappadocians, and Parthians, as depicted in the Achaemenid reliefs. Moreover, on some Frataraka coins, the king wears a similar attire to the left-hand figure of Qizqapan (Haerinck and Overlaet Reference Haerinck and Overlaet2008: pl. 1). Also, according to Strabo (15.3.15; trans. Jones Reference Jones1961), some parts of this attire, such as the head-covering or bashlyq, were still common among Cappadocian Magi during the Parthian period. However, the carving style of the Qizqapan figures is different from that of Parthian reliefs, important examples of which can be seen at Bisotun and Sarpol-i Zahab. Moreover, according to von Gall, some of the technical features of the costume of the Qizqapan worshipers, including the shirt folds at the elbow, were derived from Greek art and did not previously exist in art from the East (von Gall Reference Von Gall1988: 572). Therefore, the most closely comparable examples for the costume of the Qizqapan worshipers is found among Achaemenid and early Hellenistic images. However, although the author, like the majority of researchers, believes that the costume of the Qizqapan figures is of Median type, Median costume alone is not a good reason to attribute the tomb to the Median period. Nor are the many comparanda of the Achaemenid and post-Achaemenid periods the only acceptable reason to attribute Qizqapan to later periods.
The worshippers’ bows
The profile view of the bow held by the Qizqapan worshippers is similar to the letter ‘B’, and thus this kind of bow is commonly called the B-Shaped bow, although it is also referred to as the Double-Convex bow (Zutterman Reference Zutterman2003: 142), Scythian bow, Scythian–Median bow, and Cupid bow (Ghirshman Reference Ghirshman1963: 319). So far, no intact example of such a bow has been found, but scattered fragments attest that it was compositeFootnote 1 (McLeod Reference McLeod1965: 2). Scythians are one of the ethnic groups to whom the invention of the B-Shaped bow has been attributed (Zutterman Reference Zutterman2003: 141). They arrived in the Near East during the first half of the seventh century B.C. Although archaeological evidence confirms that the Scythians used the B-Shaped bow, all of the available evidence recording Scythians’ use of this kind of bow belongs to the fifth–fourth centuries B.C., which is either contemporary with or younger than the Achaemenid period (Hinds Reference Hinds2010: 38; Zutterman Reference Zutterman2003: fig. 7). Some scholars believe that the Scythian arrowheads found at Tepe Nūsh-i Jān indicate the probable use of the B-Shaped bow by Medians in the Median period and reflect their relationships with the Scythians (Mollazadeh Reference Mollazadeh2014: 352). However, this conclusion is indirect and not entirely dependable, because it is not certain that such arrowheads could only be shot by B-Shaped bows.
According to Herodotus (1.73; trans. Godley Reference Godley1975), Cyaxares entrusted Median boys to the Scythians to be taught their language and the craft of archery. But according to Zutterman (Reference Zutterman2003: 141), we cannot be certain if the Medes used the Scythian Double-Convex composite bow or adopted it and modified it to their own needs. Ghirshman believes that, due to the smaller size of these bows, Median horsemen preferred them to other kinds of bows (Ghirshman Reference Ghirshman1963: 319). So far, no B-Shaped bows or reliefs have been found that can be reliably attributed to the Median period. On a seal from Tepe Nūsh-i Jān, as well as on some artifacts from Lurestan, dating to the eighth–seventh centuries B.C., contemporary with the Median period, a motif of a single convex bow can be seen (Curtis Reference Curtis1995: fig. 8b; Ghirshman Reference Ghirshman1963: fig. 388). This evidence may suggest that the B-Shaped bow appeared only after the Median period. Although discussing the usage of B-Shaped bows by the Medes in the Median period is fairly difficult, discussing its usage by the Medes in the Achaemenid period is much easier, thanks to relevant contemporary archaeological evidence. Some objects have a reliable date, for example, the coin of Datames, satrap of Cappadocia in the time of Artaxerxes II (404–359 B.C.), that shows Datames sitting on a chair, wearing a costume similar to that of the left-hand figure of Qizqapan and holding a B-Shaped bow (Briant Reference Briant1996: fig. 45a). Study of the Achaemenid motifs shows that people who wear Persian attire usually use the single convex bow which often has a duck's head tip (Garrison Reference Garrison, Curtis and Simpson2010: fig. 32.1), while those wearing the Median costume use a B-Shaped bow (Head Reference Head1992: fig. 13c). After the Achaemenid period, usage of the B-Shaped bow continued; for example, this bow can be seen on the coins of Frataraka, Parthian and Scythian kings (Cernenko et al. Reference Cernenko, McBride and Goreli1983: 39; Haerinck and Overlaet Reference Haerinck and Overlaet2008: pl. 2; Herzfeld Reference Herzfeld1941: fig. 388).
Another debated issue is the position or carrying method of the Qizqapan bow. According to Porada, the position suggests a tradition of what may have been a ceremonial military pose from the early first millennium B.C. (Porada Reference Porada1965: 138). Probably the oldest comparable position can be seen on the Golden Cup of Hasanlu, which dates to the late second or early first millennium B.C. (Porada Reference Porada1965: fig. 63). However, the bow on the Hasanlu cup has the older triangular shape. The closest comparable scene to Qizqapan, both in the position of the archers and the general scene of worship, can be seen on the tomb of Darius I and his successors. There, in the same exact position, the Achaemenid kings carry a Persian single convex bow (von Gall Reference Von Gall2009: fig. 4). This method of carrying the bow can also be seen on the coin of Vadfradad I (first half of the second century B.C.), the ruler of Frataraka (Haerinck and Overlaet Reference Haerinck and Overlaet2008: pl. 2). Therefore, the best available evidence regarding the usage of the B-Shaped bow and its carrying position can be found in the art of the Achaemenid period and then in the art of post-Achaemenid dynasties such as the Frataraka rulers.
The Fire-Altar
Older examples of fire-altars have been identified at Karmir Blur in Armenia (eighth century B.C.), at Khorsabad (second half of the eighth century B.C.) and Nineveh (first half of the seventh century B.C.) in Assyria (Houtkamp Reference Houtkamp and Kellens1991: 33), and on two rock reliefs from the Neo-Elamite period (late eighth century B.C.), in the Kul-e Farah of Izeh (Sarraf Reference Sarraf2008: designs 1 and 5). However, the stepped fire-altar of Qizqapan bears little resemblance to these cases. The similarity of the Qizqapan fire-altar to the only known Median fire-altar found at Tepe Nūsh-i Jān (Roaf and Stronach Reference Roaf and Stronach1973: fig. 6) is also questionable, since there are ambiguities regarding the function of the latter (ibid.: 135–136). On the other hand, this kind of stepped fire-altar is seen frequently on Achaemenid rock reliefs (in the royal tombs, for example; Schmidt Reference Schmidt1970: fig. 63) and on other Achaemenid artifacts (Razmjou Reference Razmjou2004: fig. 5), as well as in the form of a free fire-altar, for example the Achaemenian Fire-altars of Pasargadae and Dahān-i Ghulāmān. stepped Fire altars also appear in Parthian art (such as the fire-altar of Kūh-i Khwājah) and as motifs on Sassanian artifacts (for example on the reverse of coins). However, the shafts of the later fire-altars (Frataraka and Parthian cases) are usually narrower than those at Qizqapan and in the Achaemenid period.
The Symbolic Motifs
The four-winged figure, human bust in a circle, and eleven-rayed star at Qizqapan form a trinity of symbols, along with the figures of worshippers, and make up an important part of the worshipping scene, as well as having a role in the dating of the tomb.
The Four-Winged Figure
Herzfeld believed that the symbol of the winged figure in Qizqapan was an archaic shape (Herzfeld Reference Herzfeld1941: 205), but he did not describe any reason for its supposed older origin. He also used the term “archaic” for other elements of Qizqapan, such as the capitals and palmettes, in order to emphasize the Median identity of this monument. However, no four-winged figure has been found yet dating to the Median period. On the other hand, several examples of this rare motif can be seen in Achaemenid art, for example, on a Pasargadae seal (Moorey Reference Moorey1978: fig. 6), a seal now in Freiburg (Keel and Uehlinger Reference Keel and Uehlinger1990: pl. IV), a cylinder seal in the British Museum possibly from Babylon (British Museum n.d.b), a disc from the Oxus Treasure (British Museum n.d.c), and a gold piece from the Sardis graves (Koch Reference Koch1992: fig. 158). In the art of the post-Achaemenid period, the four-winged figure similar to that of Qizqapan has not been identified, but a double-winged figure can be seen on some coins from the Frataraka and other dynasties. Therefore, the suggestion of an Achaemenid dating for the four-winged figure of Qizqapan seems to be more acceptable than that of other periods. In a contribution to this debate, using analyses of the components of the four-winged figure, the author argues here that the symbol is neither from the Median period nor the early Achaemenid period. In other words, this motif is more recent than the reign of Darius and Xerxes.
Hat of the Four-Winged Figure
The oldest Achaemenid winged figure in the Relief of Bisotun has a cylindrical hat adorned with a horn and an eight-pointed star. In the Assyro-Babylonian iconographic tradition, deities invariably wear horned crowns of varying form, as does the hovering human winged-disk at Bisotun (c. 520 B.C.). But thereafter at Persepolis the human winged-disk and the encircled bust wear the royal tiara, and only ‘bull-men’, as guardians and on column capitals, and winged sphinxes retain their traditional horned crowns (Moorey Reference Moorey1978: 147). The hat of the four-winged figure at Qizqapan does not resemble the older tall horned hat of Bisotun; it is a rather cylindrical and relatively short hat with vertical flutes that is similar to type C of Persian hats in Thompson's categorization. In the Persepolis reliefs such hats are worn by the king and his nobles without any apparent distinction (Thompson Reference Thompson1965: 125). This suggests that the hat of the four-winged figure in Qizqapan is more recent than the Bisotun relief.
Beard of the Four-Winged Figure
The winged figure of Bisotun has a long, squared Assyrian style beard. In the later Achaemenid reliefs, such as the reliefs of Artaxerxes I in Persepolis, the beard of the winged figure is still long, but it has a rounded end rather than a rectangle. The beard of the four-winged figure of Qizqapan is the same as the recent type, which can be counted amongst motifs newer than those of the early Achaemenid period.
Wings of the Four-Winged Figure
The oldest examples of the Achaemenid winged figures, including Bisotun and the Tomb of Darius I, include quadrangular or box-like wings; however, the wings of the four-winged figure at Qizqapan are round and have curved ends. Some researchers believe that the rounded wings are a later type (Roaf Reference Roaf1983: 138), and curved wings have often been taken as indicative of a later iconographic type, occurring only irregularly in the time of Darius (Garrison and Dion Reference Garrison and Dion1999: 9). Moreover, in the older motifs of Bisotun and Darius's tomb, the feathers in the wings of the winged figure are arranged in horizontal rows, but in the newer reliefs of Persepolis, the wings of the winged figure are composed of vertical feathers. The wings of the Qizqapan symbol are different from that of the older motifs and similar to the newer version of the winged figures in Persepolis.
Disc of the Four-Winged Figure
In the older Achaemenid reliefs of Bisotun, as well as the ones from the tombs of Darius I and Xerxes, the disc of the winged figure has a flat surface and adorned margin. Subsequently, in the tombs of Artaxerxes I and the later Achaemenid kings, the disc does not include the decorative margin. The ring of the winged figure at Qizqapan differs from the older Achaemenid examples and does not include a decorative margin either. Also, its surface is not flat, but in the form of a half cylinder, similar to the ring of the winged figure in the tombs of Artaxerxes II (404–359 B.C.) (Calmeyer Reference Calmeyer and Bagherzadeh1975: fig. 8a) and Artaxerxes III (359–338 B.C.). However, it should be noted that in the Achaemenid royal tombs, a general similarity can be seen in the plan and the pattern of the reliefs, yet in the later tombs some details of the reliefs, such as the form of the disc, are affected by the technical evolutions of their time.
Yoke and Base of the Four-Winged Figure
Old Achaemenid winged figures from Bisotun and the Tomb of Darius have tendrils on each side near the base, ending in three points. Alternative tendrils are spiral-ended, which according to Roaf seem to be a later type (Roaf Reference Roaf1983: 138). The four-winged figure at Qizqapan has tendrils that coil at the ends, which seems to be a more recent form than the motifs of Bisotun and the Tomb of Darius. Another aspect of the dating debate is the yoke. According to Roaf, yokes are found in the reigns of Darius and Xerxes but not later (Roaf Reference Roaf1983: 138). The winged figure of Qizqapan does not have a yoke and therefore should be younger than the reign of Xerxes.
Tail of the Four-Winged Figure
At Bisotun, the tail of the winged figure consists of two rows of vertical feathers; however, on the tomb of Darius I, the tail of the winged figure includes three rows of vertical feathers. The three-part tail can be seen in the subsequent royal tombs and other Achaemenid rock reliefs, as well as on the four-winged figure of Qizqapan. Moreover, at Bisotun and the Achaemenid royal tombs, the division of the tail into the double and triple sections is done with horizontal stripes, but in the later reliefs (for example, in the Council Hall), horizontal bands are not used, and the division is achieved by the use of perspective, which seems to be more modern and professional in terms of carving technique. The tail of the four-winged figure in Qizqapan must be considered to be from the later type.
The Encircled Bust
So far, no examples of the encircled bust similar to the middle symbol of Qizqapan have been found in either the Median or the post-Achaemenid periods. But this symbol can be seen on more than twenty Achaemenid artifacts, especially on the seals known as the Achaemenid court style (Finn Reference Finn2011: no. 14; Kantor Reference Kantor1957: fig. 10a–c; Keel and Uehlinger Reference Keel and Uehlinger1990: pl. IV; Krückman Reference Krückman1933: no. LXXVIII; Maras Reference Maras2009: fig. 5.4; Moorey Reference Moorey1978: fig. 7; Zettler Reference Zettler1979: fig. 10). According to Moorey, the symbol of the encircled male bust in royal costume is one of the primary symbols in the Achaemenid court style, appearing on seals and jewelry of this period (Moorey Reference Moorey1978: 146).
The Eleven-Rayed Star and the Trinity of Symbols
The Eight-Pointed Star can be seen in the art of the Median, Achaemenid, and post-Achaemenid periods, but so far, an Eleven-Pointed Star, similar to that of the right symbol at Qizqapan, either alone or as part of a trinity, has yet to be identified from these periods. Generally, the closest and the most comparable examples for the Qizqapan symbols can be found in Achaemenid art. A very important note is that amongst the various motifs from the Median, Achaemenid, Seleucid and Parthian periods, there are only a few rare examples of the trinity of symbols, all of which solely belong to the Achaemenid period (Kuhrt Reference Kuhrt2007: fig. 5.4; Perrot and Chipiez Reference Perrot and Chipiez1890: fig. 504). One such case which illustrates the same trinity pattern as the symbols of Qizqapan is a clay bulla from Persepolis (Calmeyer Reference Calmeyer and Yarshater1986: fig. 42 bottom).
The carving style at Qizqapan
According to Herzfeld, the style of the Qizqapan reliefs is far from the perfection of the sculptures of Pasargadae, but it is in conformity with that of the small gold plates of the Oxus treasure, the single figure under the Dukkan-i Daud, and the sculpture over the small tomb of Sakawand, the first two of which, in his opinion, belong to the pre-Achaemenid period (Herzfeld Reference Herzfeld1941: 205). However, Diakonoff believed that the rock reliefs attributed to the Median period, including Qizqapan, are not significantly different from the famous Achaemenid reliefs (Diakonoff Reference Diakonoff and Keshavarz1998: 376). The author also believes that the Qizqapan reliefs, in terms of style and features such as monotony, dignity and having a profile view, are fully similar to the Achaemenid rock reliefs. During my visit to Qizqapan in 2014, traces of the usage of a toothed chisel were found in different parts of this tomb. It is believed that in Greece this type of tool was known somewhat earlier, perhaps around 570 B.C. (Nylander Reference Nylander1970: 54). Study of the tool-marks confirms the very limited uses of the toothed chisel in Iran before 530 B.C. (Nylander Reference Nylander1966: 376; Stronach Reference Stronach1969: 157). However, the tomb of Darius I is the first known Achaemenid rock tomb which includes the very limited use of the toothed chisel (for example, on coffins) (Nylander Reference Nylander1970). The further use of this tool in Qizqapan probably indicates that this tomb was built in a time after the Tomb of Darius.
Burial Chamber
Frame of the Main Doorway
According to von Gall, the doorway frame in Qizqapan is apparently derived from the local wooden architecture (von Gall Reference Von Gall1988: 560). It seems that most rock-cut tombs with a similarly outlined doorway are situated in regions west and north of Qizqapan, such as the Levant, Asia Minor, and eastern Europe. Many of these examples belong to the Hellenistic period, including the tomb of Kaş, Gemlik (ancient Kios), and Tomb 70 at Myra, all located in Asia Minor, the tomb of Palatitza and Tomb 3 at Vergina in Macedonia, Tombs 1 and 4 at Basse-Selce in Albania (Fedak Reference Fedak1990: fig. 148), the entrance to the Petra Treasury in Jordan, and the tombs of Hegra in the northwest of Saudi Arabia (Anderson Reference Anderson2002: fig. 33).
Burial Type
There are two main theories regarding the burial method in Qizqapan. First is the hypothesis held by some scholars who accept the Median dating and believe that in Qizqapan and other Median tombs the full body, as a primary burial, was buried in a real tomb, probably with a rich inventory of tomb furniture and small objects (Herzfeld Reference Herzfeld1941: 202). Second is the hypothesis of secondary burial, mainly suggested by scholars who support the new dating (Achaemenid and post-Achaemenid), such as von Gall (Reference Von Gall1988: 562), Haerinck (Reference Haerinck and Curtis1997: 34), and Basirov (Reference Basirov, Curtis and Simpson2010: 78), based on the small size of the burial pits or cists. They propose that Qizqapan cists do not seem to be big enough to contain the outstretched body of a grown person and thus each cist should instead be an astōdān or ossuary, in which the bones of the dead were gathered after exposure of the corpse (Haerinck Reference Haerinck and Curtis1997: 33). This kind of burial corresponds to the Zoroastrian rules mentioned in Vendidad (Basirov Reference Basirov, Curtis and Simpson2010: 77). However, these scholars usually ignore the possibility of corpse burial in a foetal or flexed position, a method quite common during the prehistoric and historic periods.
Discussions regarding the burial methods used in Qizqapan still continue. The main problem is that we cannot determine precisely which methods emerged in what historical era or geographical zone and in what period they became obsolete. Indeed, burial method has an irreconcilable link with the ideology and religion of the tomb owner, and our information regarding this is still quite insufficient. The royal tombs of Darius I and his successors include burial cists which are invariably too large to justify their use as ossuaries; for example, the cists in Darius’ tomb are 2.10 m long and 1.05 m wide (Basirov Reference Basirov, Curtis and Simpson2010: 77). On the other hand, the earliest attestation of the term astōdān is in the early fourth century B.C. Aramaic epitaph of Artima, son of Erəzifiya, a Persian official in Limyra (Lycia) (Shahbazi Reference Shahbazi1987: 851). In this rock-cut tomb, a pit 130 cm long, 68 cm wide and 80 cm deep is cut into the floor of the funerary chamber (ibid.). These dimensions are close to the size of the Qizqapan burial pits: 115-135 cm long, 60-75 cm wide and 75 cm deep. Thus, despite whether Qizqapan is an astōdān or not, its similarity to Limyra in the dimension of cists may attest a later date than the early Achaemenid period for Qizqapan.
The possibility of using new scientific techniques for the dating of Qizqapan
The invention of new scientific techniques during the recent decades means that today scholars are able to date ancient rock paintings by using small amounts of their organic matter (Rowe Reference Rowe, Chippendale, Whitley and Loendorf2005: 294–319; Rowe Reference Rowe and Reddy2007: 218–231). In Qizqapan, traces of red-colored paint have remained on the upper part of the eleven-pointed star and the crown of the encircled bust. Using this method, we might be able to determine a more exact date for Qizqapan in the future.
Conclusion
All possible options for the date of the rock-cut tomb of Qizqapan, including the Median, Achaemenid, Seleucid, and Parthian periods, were studied here in detail. The main difficulty in suggesting a Median date for Qizqapan first is the lack of comparable Median artifacts with certain dating. Second, the details of some motifs, such as the symbol of the winged figure, suggest an evolution that does not coincide with older periods such as the Median era. Also, the use of the toothed chisel at Qazqapan is not compatible with the theory of a Median date, since such instruments are inventions which were only used in the major eastern centers such as Pasargadae and Persepolis after the Median period. Therefore, calling Qizqapan a Median tomb or attributing it to Cyaxares, the king of the Medians, lacks sufficient evidence. However, based on evidence such as the costume, the bow type, and the geographical location of this monument, the author believes that Qizqapan does have a Median identity (but not a Median period date).
The investigation of the details and components of the architecture, figures, and symbols at Qizqapan, and their comparison with monuments of different periods, clearly show that there is a great similarity between Qizqapan and the art of the Achaemenid period. For example, the style of carvings in Qizqapan is very similar to that of the Achaemenid style at Persepolis, and some important motifs at Qizqapan, such as the symbol of the encircled bust, can only be found amongst Achaemenid art or artifacts. Also, the general scene in which the worshipper(s) hold a bow alongside a fire-altar was a very popular scene in Achaemenid art, especially on the Achaemenid rock-cut tombs of Persepolis and Naqsh-i Rustam. Therefore, all these pieces of evidence tempt the author into suggesting an Achaemenid dating for Qizqapan. However, there is some evidence that points to the late Achaemenid era, especially the fourth century B.C., which is more acceptable than the early Achaemenid period. This conclusion mainly comes from the comparison between some details of the symbolic motifs, including the type of hat, headdress and beard, the type of ring or disc, shape of the tail and tendrils, and so forth, to which previous researchers paid little attention.
On the other hand, some elements and motifs at Qizqapan, such as the costume of the left-hand worshipper, the B-shaped bow, and the three stepped fire-altar, can also be seen in the art of the Seleucid period or their contemporary local dynasties, such as the Frataraka. Undoubtedly, one of the most important elements of Qizqapan, which is considered to belong to the Classical or Hellenistic periods, is the Ionic column. Following the establishment of the Seleucid dynasty, the Ionic order expanded into a vast region from the Levant to Bactria. However, there is no evidence to invalidate the hypothesis that the Ionic style may have entered northern Mesopotamia during the late Achaemenid period, i.e., before the forcible entry of the Macedonians into the East, merely due to an increase of cultural connections. Moreover, some elements and motifs at Qizqapan, such as the Ionic capitals, the attire of the left-hand worshipper and the B-shaped bow, were common during the Parthian period. Yet there are no similarities between the reliefs of Qizqapan and Parthian reliefs in terms of style or carving techniques. In fact, what we see at Qizqapan is a combination of Achaemenid carving style, with western Ionic column style and traces of indigenous traditions. Therefore, by taking all factors into account it has been concluded that the fourth century B.C., i.e., the last decades of the Achaemenid period and the early decades of the Seleucid and Macedonian rulers, is the most probable dating for the construction of the rock-cut tomb of Qizqapan. In this era, an intersection between Achaemenid and Ionic carving traditions is an expected possibility, the realization of which is seen at Qizqapan.
Acknowledgments
I would like to express my gratitude to Professor Michael Roaf as well as Dr. Shahrokh Razmjou for their guidance, reviewing this paper and proposing valuable suggestions. I thank Dr. Sajjad Alibaigi, Sirwan Manhubi and Tahsin Osmani for their help and Mr. Naser Aminikhah, who prepared a new drawing of the Qizqapan motifs. I also thank Mr. Kamal Rashid Rahim, head of the Sulaymaniyah Directorate of Antiquities, Mr. Hashim Hama Abdullah, Director of the Sulaymaniyah Museum, and Mr. Akam Omar Ahmad Qaradaghi for their help and paving the way for me to visit Qizqapan. My final thanks are to Mr. Hüseyin Vural, Director of the Sinop Archaeological Museum, and Mr. Sayed Ali Hosseini, for sending me images of Turkish rock-cut tombs.
