The Yenikapi 12 Shipwreck A 9th Century

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 34

bs_bs_banner

The International Journal of Nautical Archaeology (2018) 47.2: 357–390


doi: 10.1111/1095-9270.12325

The Yenikapı 12 Shipwreck, a 9th-Century Merchantman


from the Theodosian Harbour in Istanbul, Turkey:
construction and reconstruction
Işıl Özsait-Kocabaş
Istanbul University, Department of Conservation of Marine Archaeological Objects, Ordu Cad. Laleli, Fatih, İstanbul,
Turkey, isilozsait@gmail.com

Shipwreck Yenikapı 12 was discovered in Yenikapı, Istanbul, Turkey in 2007 during rescue excavations carried out by Istanbul
Archaeological Museums. The majority of the bottom of the shipwreck, found within the sediment-filled Theodosian Harbour, is
intact and part of the cargo was found in situ. According to the results of reconstruction studies, YK12 was a small merchantman
working coastal waters, approximately 9.24m in length and 2.64m in breadth. The shipwreck is dated to the 9th century AD and
was built with a mixed construction using the shell-based method for the lower hull.
© 2018 The Author

Key words: Yenikapı shipwrecks, Theodosian Harbour, mixed construction, ship reconstruction, plank-edge dowels.

A
rchaeological excavations in Yenikapı were digitizer (FaroArm) drawings, documentation and
conducted within the scope of the Marmaray cataloguing of the hull members were completed, and
and Metro projects and were carried out non- construction techniques and reconstruction studied
stop by Istanbul Archaeological Museums (IAM) from (Özsait-Kocabaş, 2013: 45–50). In this article, the basic
2004 to 2013. A 12m-thick deposit was excavated construction features, design, and building philosophy
with stratigraphic layers dating from the end of the of YK12 are discussed, along with its structural
Ottoman Empire to the Neolithic Age (Kızıltan, 2010; similarities to other Yenikapı vessels. In order to
Kocabaş, 2015b). The Theodosian Harbour, which understand the original form and dimensions of vessel,
was discovered during the excavations, produced many reconstruction studies are briefly outlined.
archaeological finds that have enriched the cultural
history of Istanbul, most notably the remains of 37
vessels that sank at different times from the 5th to
11th centuries AD.1 The excavation of the Theodosian Finds and dating
Harbour and the recording methods used have been YK12 was one of four shipwrecks that contained
previously published (Kocabaş, 2008; Kızıltan, 2010; cargo when found. As well as the amphoras and a
Kocabaş, 2010; Kocabaş, 2015a; Kocabaş, 2015b, Pulak large number of amphora fragments spread around
et al., 2015). the hold, cooking utensils were found in the stern
Yenikapı 12 shipwreck (YK 12) with a cargo of area (Fig. 2), including a stove-like brazier and its
amphoras was discovered by IAM archaeologists in lid, a cooking pot, two cups, a trefoil juglet, glass
the east part of the harbour in February 2007. The goblet sherds, and two amphoras of a different type
shipwreck was located in the fifth sediment layer, in (Denker et al., 2013: 204, 209). Furthermore, gaming
grid squares F-G/13-14 at the Yenikapı site (Fig. 1) pieces made of ivory, a wooden comb, a bronze needle,
(Perinçek, 2010: 210). After the amphoras had been and a thimble found inside the shipwreck inform us
removed, an Istanbul University (IU) team started about daily life aboard. Also, two toggles were found
to study the wreck. The shipwreck was documented as rigging equipment (Denker et al., 2013: 198–203).
in situ with 3D measurements and drawings, full-scale Cherry seeds that were found in a basket, and olive
hand drawings, photography, and photo-mosaics. and peach seeds provide information about the crew’s
Then the hull members were dismantled and placed food supplies. The amphoras and the wreck itself
in wooden chests. In July 2007 they were moved to were dated to the 9th century AD (Denker et al.,
storage tanks in the IU Yenikapı Shipwrecks Research 2013). Additionally, a 9th-century AD coin was found
Centre directed by Ufuk Kocabaş. At the Centre, 3D in the shipwreck. The shipwreck has been dated to

© 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society.
Published by John Wiley & Sons Ltd. 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.
NAUTICAL ARCHAEOLOGY, 47.2

Figure 1. Site plan of the Yenikapı excavation, showing the locations of YK 12 and the other shipwrecks studied by Istanbul
University (Istanbul University Yenikapı Shipwrecks Project Archive).

Figure 2. a) The amphoras as found on YK 12; b) the personal belongings found in the storage compartment (IUYSP Archive).

358 © 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society.
I. ÖZSAIT-KOCABAŞ: YK 12: A 9TH-CENTURY MERCHANTMAN FROM YENIKAPI

Table 1. Radiocarbon analysis results for shipwreck YK12. All samples analysed by Oxford Radiocarbon Accelerator Unit,
calibrated using OxCal V4 1.7

Sample ID Timber type 14C age years BP Calibrated Probability Note

OxA-23836 Floor-timber (Fr 11) 1246 ± 25 BP cal AD 681–870 95.4% Quercus sp


OxA-23837 Plank (PS7-2) 1233 ± 24 BP cal AD 690–876 95.4% Castanea sativa
cal AD 762–876 60.1%
OxA-23838 Central keel timber 1258 ± 25 BP cal AD 672–859 95.4% Fagus orientalis L.
(keel 3) cal AD 672–818 93.4%

Figure 3. In situ plan of YK 12 and cross-sections at frames Fr4, Fr12, Fr18, and Fr25 (I. Özsait-Kocabaş, processing the digital
version/PDV M. Sağır).

AD 672–870 by radiocarbon analysis of its timbers two amphoras in the storage compartment resembles
(Kocabaş, 2015a: 13) (Table 1).2 an amphora type from the Black Sea region found on
There are different opinions about the typology the Bozburun ship (Hocker et al., 1998: 4, 5; Denker
of the amphoras. Some researchers state that the et al., 2013: 204). YK 12 sailed during a period in which
ship’s cargo was transported in Crimean amphoras trade from Constantinople was directed from the south
(Trakadas, 2014: 84; Polat, 2016: 382). Also, one of the to the north, to the Aegean and Black Sea (Bortoli

© 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society. 359
NAUTICAL ARCHAEOLOGY, 47.2

Figure 4. Keel timbers (C. Ciner, PDV Ç. Şanlıgençler).

Table 2. Keel timber dimensions (see Fig. 4)

Moulded (mm) Sided (mm)

Keel Timber Length (m) Forward Middle Aft Forward Middle Aft Notes

Keel 1 (Stem) 0.21 – – – – – – Broken


Keel 2 (Curved keel timber) 1.54 100 - 95 67 – 90 Intact
Keel 3 (Central keel timber) 3.63 100 110 115 90 100 90 Intact
Keel 4 (Curved keel timber) 1.78 110 – 135 90 – 85 Intact
Keel 5 (Sternpost) 0.10 – – – – – – Broken

and Kazanski, 2002: 662; van Doorninck, 2002: 902;


Asal, 2010: 156). Based on the cargo, YK 12 may
have been involved in regional trade along the southern
coasts of the Black Sea, or it may have been a local
trading vessel involved in transshipping cargo carried in
Crimean amphoras from a seagoing vessel to a nearby
port.
Other researchers have suggested the amphoras
could be prototypes of a Günsenin type 1 Ganos
amphora, dated to the 10th–11th century, from a
workshop on the Marmara Sea coast (Günsenin, 2016:
401; Vroom, 2016: 163). If these are earlier Ganos
amphoras, the vessel could have been used by a local
trader in Marmaran and Thracian foodstuffs (Laiou,
2002: 725; Vroom, 2016: 163–164). The divergent views Figure 5. Keyed hook scarf joining the central keel timber
(keel 3) and curved keel timber (keel 4) (IUYSP Archive).
of the amphoras’ provenance and uncertainty about the
vessel’s route and home port, may be resolved when the
analysis of the amphoras and goods the vessel carried 25 frames, the mast-step timber and 43 ceiling planks
is concluded. (Özsait-Kocabaş and Kocabaş, 2008) (Fig. 3).
Keel and posts
Five timbers made up the backbone of the vessel,
The hull remains numbered keel 1 to keel 5 from bow to stern. They
The vessel might have sunk as a result of a violent included a central, rockered keel timber (keel 3), with
storm from the south as it was found covered with sand curved timbers scarfed to each of its ends (keel 2 and
likely raised from the seabed during a storm (Perinçek, keel 4), all of which are well preserved. The posts (keel
2010: 211). Since YK 12 sank with a slight incline 1 and keel 5) are largely missing, except for small pieces
to starboard, sitting on its bottom, and was quickly attached at the scarfs at the ends of keel 2 and keel 4
buried, most of the timbers under the amphora cargo (Fig. 4, Table 2). The scarfs between the five keel and
were preserved in their original places. The hull remains post pieces are intact and all are keyed hook scarfs
have a length of 7m and a width of 2.3m, and consist of (Fig. 5). Rabbets were cut on both sides along the
five scarfed keel timbers, 17 planking strakes, two wales central keel and curved keel timbers. At the stem
on the starboard side, a piece of the starboard caprail, and sternpost the rabbets end in a smooth transition.

360 © 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society.
I. ÖZSAIT-KOCABAŞ: YK 12: A 9TH-CENTURY MERCHANTMAN FROM YENIKAPI

Figure 6. Plan of planking of YK 12.

Table 3. Planks dimensions. Widths measured at the widest Table 4. The different thicknesses on the inboard and
points of the planks; * indicates intact planks outboard edges of some original planks

Plank Width (mm) Length (m) Thickness

PS1 122 5.72* Starboard Port edge


PS2 210 PS2-1: 5.18* / Plank Measured location edge (mm) (mm)
PS2-2: 2.08*
PS3 215 PS3-1: 4.63 / PS1 Fr2-Fr3 16 24
PS3-2: 1.33* Fr13-Fr14 20 26
PS4 203 PS4-1: 0.96* / Fr24-Fr25 26 23
PS4-2: 2.75* PS2 Fr1-Fr2 20 21
PS5 160 PS5-1: 2.25 / Fr10-Fr11 15 17
PS5-2: 4.76* Fr24-Fr25 21 30
PS6 181 PS6-1: 3.75 / PS6 Fr4 16 14
PS6-2: 2.68 Fr11-Fr12 19 16
PS7 158 PS7-1: 1.48 / Fr22-Fr23 21 20
PS7-2: 4.60 PP1 Fr3-Fr4 19 14
PS8 150 PS8-1: 1.38* / Fr13-Fr14 23 16
PS8-2:1.90 Fr23-Fr24 26 25
PS9 140 PS9-1: 3.70 PP2 Fr11 21 12
PP1 124 5.72* Fr22-Fr23 17 14
PP2 212 PP2-1: 5.02 / Fr24-Fr25 27 18
PP2-2: 2.10 PP5 Fr4-Fr5 19 16
PP3 210 PP3-1: 4.52 / Fr8-Fr9 10 19
PP3-2: 1.74* Fr26-Fr27 19 14
PP4 186 3.08
PP5 148 PP5-1: 1.77 /
PP5-2: 4.80*
PP6 176 PP6-1: 3.65 / Planking
PP6-2: 2.59 The bottom of the hull was found in a very robust state
PP7 165 PP7-2: 4.20 with nine surviving strakes to starboard and eight to
PP8 90 (broken) PP8-1: 1.10* port (Fig. 6). The garboards (PP1, PS1) were made from
PP8-2: 0.69 single planks with tapered ends and are 5.72m long
(Tables 3 and 4). At their widest points PS1 measures
122mm and PP1 is 124mm.
Rabbets average 10mm in depth and were cut 12mm They were fastened in the keel rabbet with treenails,
below the inboard surface of the keel. Approximately driven at intervals of 0.22–0.45m, from the outboard
in the middle of the central keel timber and to the stern towards the upper surface of the keel, where they were
of keel 4, there are transverse holes cut through the port cut flush (Fig. 7). Resin and luting material were found
and starboard faces of the timber. in the join.

© 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society. 361
NAUTICAL ARCHAEOLOGY, 47.2

Figure 7. Characteristics of the joint between the garboard and the keel (I. Özsait-Kocabaş, PDV Ç. Şanlıgençler).

Figure 8. FaroArm drawing of planking strakes PS1 and PS2 (C. Ciner, PDV Ç. Şanlıgençler).

The other strakes are generally made up of two determined between wale 1and PS9; caulking material
planks, combining one long and one short plank. These was probably used between the planks of the upper hull.
planks also taper towards the ends and range 0.14– The planking strakes are organized so that the long
0.215m at their widest points. Some planks are very planks, and thus their scarf directions, alternate from
worn and thickness ranges 0.01–0.03m (Fig. 8). bow to stern (Fig. 6). The long planks of the second
The planks forming a strake are joined together and third strakes start at the bow, while the fourth and
with S-shaped scarfs, varying in length from 0.20 to fifth strakes have the short plank at the bow (except for
0.45m. The alignment and locations of these scarfs PP4). On the sixth strake, the direction changes once
is approximately symmetrical to port and starboard. again and the long planks were placed at the bow. Then
The planks are fastened together, edge to edge, with the long planks of the seventh and the eighth strakes
regularly spaced dowels (coaks) from the garboard up start at the stern.
to the first wale (Fig. 9). Space between dowels ranges The planks were scarfed in the same style throughout
from 0.20m to 0.36m. The dowels are polygonal and the hull. The scarf end of the first, long plank was
11mm diameter on average. Depending on the width of attached to the plank below with a dowel, and the
the planks the dowels are about 0.07–0.10m in length. scarfed end of the short piece was joined to the long
Dowels 0.09–0.13m long fasten the planking scarfs. plank with a dowel, also (Fig. 10).
Only two dowels were found between the eighth strake
(PS8-1, PS8-2) and the first wale, near frames Fr22 and Wales, caprail, and through-beams
Fr23 (Fig. 10). There is no evidence of dowels above the A 4.69m length of the first wale (wale 1) was found
first wale. Luting was laid between the plank seams and in situ on the starboard of the hull, between the 7th/8th
appears around many of the dowels. Also, caulking was and 9th strakes (Fig. 11). The wale has a curved outer

362 © 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society.
I. ÖZSAIT-KOCABAŞ: YK 12: A 9TH-CENTURY MERCHANTMAN FROM YENIKAPI

Figure 9. Edge-dowel from the PP5–PP4 plank seam (IUYSP Archive).

Figure 10. a) Two dowels between wale 1 and PS8; b) fastening detail of the scarf ends observed on worn planks (PS4, PS5-1
and PS5-2 (IUYSP Archive); c) locations of dowels fastening planking scarfs (I. Özsait-Kocabaş, PDV Ç. Şanlıgençler).

surface, and flat inner surface. Both of its ends are waterline above plank PS9: a wedged treenail identified
broken. It ranges 92–115mm in width and 55–70mm in at the top end of floor-timber Fr14 supports this
thickness. hypothesis, as do carpenter’s marks on the on the outer
Based on the proposed reconstruction (see below), surfaces of the long arms of floor-timbers Fr12 and
a second wale (wale 2) was positioned at the loaded Fr14 and on starboard futtock Fr15 (see below).

© 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society. 363
NAUTICAL ARCHAEOLOGY, 47.2

Figure 11. Starboard planking and wale 1 (IUYSP Archive).

Figure 12. FaroArm drawing of the caprail and wale 3 (C. Ciner and A. Çetiner, PDV Ç. Şanlıgençler).

A third wale (wale 3) was found under the vessel On wales 1 and 3 the inner surface has raised sections
towards the stern on the starboard side, disarticulted in line with and the same width as some of the frames.
from the main structure, but still fastened to a caprail These areas were left 2–4mm higher than the rest of the
(Fig. 12). It has a similar rounded section to wale 1 wale surface. Similar protrusions were also found under
with flat edges about 0.03m wide to match the planking some frames on plank PS9 (Fig. 13).
below and the caprail above. Both of its ends were The presence of through-beams fitted on to the first
broken. No dowels were recorded on the lower edge starboard wale were identified ahead of Fr18 (through-
of wale 3. This piece of wale 3 is 2.63m long, ranges beam 1) and abaft Fr20 (through-beam 2). Through-
90–110mm in width, and is c.60mm thick. The caprail beam 1 was fastened to wale 1 with a treenail, and a
fragment is 2.18m long, about 100mm wide and 32mm notch 0.07m wide and 0.025m deep was cut for it on
thick. It was attached to wale 3 with treenails, which the bottom edge of PS9. Only a small part of through-
were driven into angled drilled holes. Caulking was beam 1, 0.06m wide and 0.022m thick and 0.064m long,
recorded between the caprail and wale 3. Notches were was preserved (Fig. 13).
recorded cut in the inner side of the caprail that seem to Through-beam 2 was not found; its location was
be related to the sail rigging (Fig. 13). determined by a treenail and a pressure mark recorded

364 © 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society.
I. ÖZSAIT-KOCABAŞ: YK 12: A 9TH-CENTURY MERCHANTMAN FROM YENIKAPI

on the upper surface of wale 1, and a notch in the edge


of PS9.

Framing
Remains of 25 floor-timbers (Fr2-Fr26) and 15 futtocks
(five to port and ten to starboard) survived (Fig. 14).
Most of the floor-timbers were found intact (Fig. 15).
The futtocks are well preserved at the scarfs with the
floor-timbers but the upper ends are broken at the
height of strake 10. The floor-timbers are moulded
0.082–0.096m and sided 0.05–0.07m at the keel. The
futtocks are moulded 0.07–0.09m and sided 0.05–
0.07m. They have rectangular cross-sections with one
entirely flat side, while the other has a 20mm chamfer
on one top corner. The chamfered sides of the frames
in the array change direction at a certain point: the
chamfered side of Fr2–Fr7 face the bow, the mast-step
timber starts on Fr8 and thereafter the chamfer is
on the stern side of each floor-timber. On the inner
surface and side surfaces some floor-timbers have bark
layers; therefore, irregular shapes on these surfaces were
recorded. The floor-timbers have triangular limber-
holes, one on each side of the keel with a vertical face
aligned with the outer edge of the keel.
The framing system consists of L-shaped floor-
timbers, and short and long futtocks. The floor-timbers
are asymmetrical, one end finishing at the turn-of-the-
bilge and the other extending to strake 9 or wale 2 (Fig.
16). They are placed with the long arms alternating
to port and starboard. Corresponding short and long
futtocks are scarfed into the ends of the floor-timbers
but not fastened (Fig. 14). The long futtocks meet the
floor-timbers at the turn-of-the-bilge at the fifth or the
sixth strake. The short futtocks start at the ninth strake
Figure 13. a) Caprail and wale 3; b) protrusion on the inner or at the second wale (Fig. 16).
surface of plank PS9; c) through-beam fragment in mortise Fr12, which has the narrowest angle of deadrise (4°)
in PS9 (IUYSP Archive). and is the broadest floor-timber, has been determined

Figure 14. Framing plan (I. Özsait-Kocabaş, PDV Ç. Şanlıgençler).

© 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society. 365
NAUTICAL ARCHAEOLOGY, 47.2

Figure 15. FaroArm drawing of frames a) Fr4; b) Fr12; c) Fr19, and d) Fr24 (C. Ciner and E. Ertuğrul, PDV Ç. Şanlıgençler).

as the midships frame. The floor-timbers were placed planking with one or two treenails per strake. They go
on the keel with room and space ranging 0.23–0.27m right through the frame and planking. These treenails
with an average of 0.25m. Eight of the floor-timbers were driven with spaces ranging 0.04–0.15m along
(Fr4, Fr8, Fr10, Fr12, Fr14, Fr16, Fr22, and Fr24) the inner surface of each frame. Iron nails were also
were attached to the keel with iron nails, while 11 of used as frame-plank fasteners; nails were driven from
them (Fr3, Fr6, Fr9, Fr11, Fr13, Fr15, Fr17, Fr19, the outboard and only go into the thickness of the
Fr21, Fr23, and Fr25) are fastened with treenails. frame. These nails were found less often than treenails
Seven floor-timbers (Fr1, Fr2, Fr5, Fr7, Fr18, Fr20, and with an irregular distribution throughout the
and Fr26) were not fastened to the keel at all. A hull.
pattern was identified: the floor-timbers with long
arms to starboard were attached with iron nails, while Mast-step timber
those with long arms to port were fastened to the The mast-step timber was placed on the floor-timbers
keel with treenails. The frames were attached to the between Fr8 and Fr16, close to the centre of the vessel,

366 © 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society.
I. ÖZSAIT-KOCABAŞ: YK 12: A 9TH-CENTURY MERCHANTMAN FROM YENIKAPI

Figure 16. a) Alternating pattern of floor-timbers with long arms to port and starboard, L-shaped scarf connection between
floor-timber and futtock, and triangular limber-holes (IUYSP Archive); b) schema showing arrangement of short and long
futtocks in a bend (I. Özsait-Kocabaş, PDV Ç. Şanlıgençler).

Figure 17. Ceiling plan (IUYSP Archive).

and was found still set firmly in place and intact


(Figs 17–19). It has a rectangular section with
chamfered edges. It is 2.05m long with a maximum cross
section 0.145m wide and 0.10m deep at the step. It has
notches cut 0.03–0.04m deep on the underside that fit
tightly over the floor-timbers. In addition, notches were
cut into floor-timbers Fr8, Fr12, and Fr16 to fit the
mast-step timber. It is attached to Fr13 with an iron
nail.
There are two mortises on the upper surface of the
mast-step timber. One is the step for the mast, the
other a small rectangular mortise foreward of the step.
The small mortise is about 0.096m long, 0.04m wide
and 0.046m deep. The mast-step itself is 0.206m long,
Figure 18. Detail of fitting the mast-step timber over the 0.056m wide and 0.058m deep. Along the long edges
floor-timbers (IUYSP Archive). of this mortise there are thin wooden rods nailed to

© 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society. 367
NAUTICAL ARCHAEOLOGY, 47.2

Figure 19. FaroArm drawing of the mast-step timber (C. Ciner and A. Çetiner, PDV Ç. Şanlıgençler).

Figure 20. Wood species used in YK 12 (Identification Ü. Akkemik, PDV Ç. Şanlıgençler).

the mast-step timber that are approximately 0.44m long appear to be exceptions—they are 1.16–1.36m long,
and 0.04m wide. 0.105–0.20m wide, and 1–15mm thick.
Ceiling planks C22 and C24 were placed on each
Ceiling and stringer side of the mast-step timber. C22 is 3.08m long, 120–
A total of 43 ceiling planks have survived (Fig. 20), 140mm wide, and 15–30mm thick. C24 is 2.10m long,
that were systematically placed and secured to the floor- and 32mm wide on average, and 20–33mm thick. These
timbers with 1–3 iron nails. The ceiling was placed by two ceiling planks are thought to be repair timbers and
superimposing one plank edge 20mm over the next, are notable for their larger dimensions, differences in
starting from the keel working up the sides to starboard colour and texture, and mismatch with the rest of the
and port (Fig. 17). Five groups of ceiling planks were ceiling system. These may have been added because of
placed in this way, starting from the bow and continuing wear, or to provide additional support to the mast-step
towards the stern, with the ends of the planks in a timber over time.
group overlapping the previous set. The ends of the The ceiling stops at Fr5 to the bow, and there is a gap
ceiling planks are finished in a variety of shapes, such in the ceiling between Fr18 and Fr20 that, according
as triangles, curves, and straight cut. Based on ceiling to the finds, was the space used as stowage by the
planks with intact ends—excluding C22 and C24 that crew (Fig. 17). Ceiling was found in situ up to PS9

368 © 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society.
I. ÖZSAIT-KOCABAŞ: YK 12: A 9TH-CENTURY MERCHANTMAN FROM YENIKAPI

and empty nail holes on the innerface of some frames futtock Fr20-P1 also had a groove. There is a broken
suggest the ceiling continued up to wale 2 (Fig. 17, see stanchion approximately 0.34m high, 0.14m wide, and
Fig. 39). 0.054m thick, attached to the inner surface of Fr18 with
No stringers were found: however, slight depressions, two dowels. The stanchion has grooves cut along two of
only about 1–2mm deep, observed on the upper ends of its sides. Two dowels and a pressure mark placed side-
a few long futtocks along strake 9 indicate the existence by-side on floor-timber Fr20 suggest there was a second
of a stringer (Fig. 21). stanchion. The panels that would have been fitted in the
grooves and the upper parts of the bulkhead structure
Bulkheads were not found.
Towards the stern, bulkheads were constructed on
frames Fr18 and Fr20. The starboard futtocks (Fr18-
S1 and Fr20-S1) were attached with two treenails to Repairs
the inner surface of the long arms of the floor-timbers The YK 12 vessel had spent a long time travelling
(Fig. 22). A groove was cut in the top surface of futtock on the seas as evidenced by the three layers of pitch
Fr18-P1, and it can be assumed that the missing port found on its hull, each separated by mussel remains.
Moreover, the inner surfaces of the planking and ceiling
were in some places worn extremely thin. Despite
this, few repairs have been identified. Indications of
repairs following accidents were recorded on the port
and starboard sides of the planking, possibly as a
result of different events, based on differences in the
workmanship and materials used. On the port side
of the vessel towards the stern, there is a roughly
shaped chestnut repair timber (PP4-2), about 0.54m
long, between frames Fr16, Fr17, and Fr18, fitted in
the third to fifth strakes (PP4 and PP5)(Fig. 23a). A
cleanly cut and skilfully fitted repair plank about 0.57m
Figure 21. Carpenter’s mark on Fr11 indicating the position long was identified on the starboard side of the vessel,
of a stringer (IUYSP Archive). towards the stern of the sixth strake (PS6-3), under

Figure 22. a) The futtocks with grooves in the strorage compartment (IUYSP Archive); b) FaroArm drawing of frame Fr18
and bulkheads. (C. Ciner and A. Çetiner, PDV Ç. Şanlıgençler)

Figure 23. a) Repair plank PP4; b) repair plank PS6-3; c) grey resin between the repaired part of arm of floor-timber Fr21
(IUYSP Archive).

© 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society. 369
NAUTICAL ARCHAEOLOGY, 47.2

Table 5. Summary of woods used according to the place of use in the hull

Keel The lower parts of the hull The upper parts of the hull

Central keel Posts and curved Futtocks and Wale1 and


timber keel timbers Planks Floor-timbers Ceiling bulkheads wale 3

Oriental Hornbeam Chestnut Oak or white oak, Oak or Ash Oak


beech walnut, white oak
hornbeam, ash

Fr20, Fr21, and Fr22. This plank (PS6-3) was made of


oak (Fig 23b).
The arms of floor-timbers Fr21 and Fr22 were also
broken and then stuck back together: grey resin was
identified between the repaired parts of these frames
(Fig. 23c). These frames are above the repair plank and
were likely damaged in the same incident.
Of C22 and C24, which are thought to be repairs to
the ceiling as they differ from the other ceiling planking,
C24 was made of plane wood.
Splits and cracks that occurred in the planking, such
as PS2-1, were repaired by being filled with pitch,
caulking material, or grey resin. Figure 24. A pair of floor-timbers Fr11 and Fr12 cut from
the same timber (IUYSP Archive).

Building technology the posts (keel 1 and keel 5). These species, which have
Many features, such as symmetry in the planking high-friction and wear resistances, are the preferred
to starboard and port, very regular frame scantlings, construction material for keels of ships built in the
carefully worked chamfers on the frames and mast- Black Sea region of Anatolia (Nutku et al., 1963: 4;
step timber, and well-cut floor-timber to futtock scarfs, Dinçel et al., 1977: 82–85). For the futtocks in the
and other standardized joints demonstrate that YK 12 upper part of the framing system, with less contact
was built with great care. In addition, as evidenced by with water, ash was preferred (Akkemik, 2015: 188–
the presence of unprocessed bark left on some floor- 189). The bulkhead timbers, and the mast-step timber
timbers, the builder was careful to use the materials were also processed from smooth, fibrous, and resistant
economically and, whenever possible, two floor-timbers ash. A plane tree (Platanus L.) timber was used for
where cut from a single trunk (see below). Considering repair (C24) following the construction of the vessel. It
the properties of the wood used, it can be seen that the is noted that the ceiling is a mixture of woods, as are
builder also consciously selected different species for the floor-timbers, while the planking is consistent. This
specific uses on the vessel. These choices are indicative could be an indication of the builder’s focus on the shell
of balancing expenditure incurred during construction in the lower hull. A similar consistency is seen in the
with both reducing the risks encountered on voyages futtocks which are cut from ash (Table 5).
and costs of repairs (Unger, 1980: 27). The grain structure is parallel to the length of the
planks, which indicates that naturally curved trees were
Wood use and processing selected and used in the construction. This method,
Seven tree species were used in the construction of which prevents deformations and splits along the seams
the vessel: oak (Quercus L.), ash (Fraxinus L.), walnut occasioned by internal stresses in the wood, is known
(Juglans L.), chestnut (Castanea Miller), hornbeam from the traditional boats of several regions and is the
(Carpinus L.), beech (Fagus orientalis L.) and plane preferred method used in the traditional Black Sea-built
(Platanus L.) (Akkemik, 2015: 43–47). This variety vessels, known as saçula (Pla Y Rave, 1996; Kaygın and
arises because species were selected according to the Aytekin, 2005: 22).
requirements of their positions in the vessel (Fig. 20). During timber studies, it was determined that two
Chestnut, the strength of which increases as it gets wet, floor-timbers, of similar size, were extracted from a
was used on all the planking (Akkemik, 2015: 188–189). single trunk with one branch. For example, Fr12 was
Oak, which is water-resistant and hardwearing, was derived from the same trunk as Fr11; they have almost
used for the wales (wale 1 and wale 3) and also preferred identical measurements and the grain structure and
for floor-timbers and ceiling planks (Akkemik, 2015: knots match (Fig. 24).
188–189). For the central keel 3, beech was used, while According to the cross-sections taken at different
hornbeam was chosen for the curved keel 2 and 4 and places along the beech keel 3 it was determined that

370 © 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society.
I. ÖZSAIT-KOCABAŞ: YK 12: A 9TH-CENTURY MERCHANTMAN FROM YENIKAPI

Figure 25. a) Saw marks on the plank (IUYSP Archive); b) overlapping saw marks on side face of floor-timber Fr20 (C. Ciner,
PDV Ç. Şanlıgençler, IUYSP Archive).

the heartwood, which is the strongest and most stable frequent intervals. These marks come from different
part of the log, was aligned before the keel was shaped. directions and overlap in certain places. According to
In the same way, the mast-step timber, which would be these marks, the carpenter started cutting from one
subject to a variety of stresses, was obtained from the end, the saw was removed before the turn and cutting
heartwood of an ash log. recommenced from the other end (Fig. 25). On the
Chestnut, ash and eastern beech used in the round side of the floor-timbers, adze or axe marks
construction of YK 12 are abundant in the Black can be seen, but no saw marks. The bark encountered
Sea region and were often used in ship construction on the surfaces was dubbed with an adze/axe, which
(Nutku and Küçük, 1963). These tree species, which shows that most parts were hewn first, rather than
vary according to the part of the vessel, have been cut with a saw. Like the floor-timbers, one side of the
recorded in the traditional vessels of this region between futtocks was cut with a saw, while the other side was
the 16th century and the last quarter of the 20th century trimmed with an adze or axe. The floor-timber and
(Turna and Pirim, 2015: 119). The current natural range futtock joints are carefully worked L-shaped scarfs.
of these trees includes the Black Sea, Marmara, and the Tool marks show that floor-timbers and futtocks were
Aegean Regions. Bithynia and Paphlagonia woodlands trimmed with an adze after the two elements were
were very suitable for shipbuilding in the Byzantine era joined together (Fig. 26a). Adze marks are also very
(Morrisson, 2013). Akkemik, who carried out wood clear in the limber-holes of the floor-timbers (Fig. 26b).
analysis of all the Yenikapı shipwrecks studied by IU, The adze/axe marks are also observed on the chamfers
stated that in the vessels dated after the 7th century, of the side and bottom edges of the mast-step. Some
timbers from the Black Sea-Marmara vicinity were used planks have had traces of sawing erased by dubbing that
(Akkemik, 2015: 197). has left them very smooth and with few marks. An adze
was used on the surfaces of the planking and marks
cross over the seams; thus, they were dubbed after
Tool marks and shaping the planking was assembled. Different sizes of adze
It is possible to ascertain which tools were used and used can be determined from blade marks of different
how the wood was shaped from the different tool marks widths.
recorded, the shapes of the finished timbers, and the Saws were mostly used for major conversion of
orientation of the wood grain. Also, similar marks timber into planks and frames. Additionally, smaller
related to certain tools are reported in other shipwrecks hand saws were used for smoothing the transition
(Katzev, 1982; Hocker, 2004b). From the tool marks on from the curved keel timbers (keel 2, keel 4) with
the frames and planking of YK 12, it was determined rabbets to the posts (keel 1 and keel 5) without
that several different saw types were used. Planks were rabbets, fitting the planking scarfs and shaping the
cut longitudinally using a saw as ascertained by the plank ends. Clear adze/axe marks are visible especially
wide spacing of marks and the way they cross over each on the curved side of the floor-timbers and futtocks
other at regular intervals (Fig. 25). Moreover, short saw and the floor-timber chamfers, the top and bottom
marks detected at the ends of the planks, made in the surfaces of the keel members (Fig. 26c) and at the
opposite direction to the wide saw marks, indicate the keel scarfs, and on the top, bottom and side surfaces
use of smaller saws to adapt them to their individual of the mast-step timber. This indicates the use of
positions. adzes/axes for fine finishing timbers once joined, as
Generally, the frames have one sawn side, and one well as shaping the main forms. These marks show
left round. The teeth marks left by the saw used are that the adze and axe were the favoured tools for
all at the same angle, very regular, with even and construction.

© 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society. 371
NAUTICAL ARCHAEOLOGY, 47.2

Figure 26. a) Adze/axe marks on the outer surface of the frame; b) adze marks on the limber-hole; c) adze marks on the inner
surface of the keel; d) saw and chisel marks on the scarf of the keel (IUYSP Archive).

Tool marks show that chisels were used to shape Fasteners


treenails, for cutting treenails to length, cutting and The elements of the hull were fastened with treenails
trimming scarfs and mortises (Fig. 26d), surface and iron nails. Treenails were driven at regular intervals
matching, and finishing. Chisels of different widths, and to connect all the frames to the planking, providing a
both straight and curved blades have been recognized solid connection by traversing the combined thickness
from the tool marks (Katzev, 1982: 246–248). of the frame and plank. Iron nails did not follow a
A sharp tool, such as an awl or knife, was used for regular pattern and were hammered in from the outside
marking the position of the floor-timbers on the inner of the planking and end in the middle of the thickness
face of planks, the locations of the plank-edge dowels, of the frames. Similarly, iron nails fastening the frames
the end points of scarfs on the edges of the planks, to the keel only penetrated to the middle of the keel. The
and areas of timbers requiring additional shaping, as frame–keel joints, ceiling–frame joints, and mast-step
well as a series of special builder’s marks of unknown timber–floor-timber joints are other examples of the
function. use of iron nails in the construction. Treenails are used
Drill marks were identified in the treenail holes in higher numbers and more regularly to attach the
that connect the YK 12 hull elements. Some of these frames to the planking. In addition to this, all frames
holes are believed to have been made with bow drills, attached to the first and third wales were fastened with
examples of which were found in different sizes in treenails. Based on this evidence, the main fastening
the Yenikapı excavation area (Gökçay, 2010: 145). device in the construction of YK 12 is treenails. Iron
Holes 110/120, 150, and 170mm diameters were found nails were used for temporary fastening, reinforcing,
indicating that drill bits of different sizes were used. and strengthening. Moreover, it is concluded that
Two wooden mallets were found in the stern of some of the iron nails were added for repairs and
the vessel; these carpentry tools may have been for reinforcement later in the vessel’s working life, rather
repair and maintenance during voyages, being similar than during the construction. For example, nails
to those used in construction (Denker et al., 2013: are found grouped on repair planks and the planking
200). around them or coinciding with the top of treenails that

372 © 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society.
I. ÖZSAIT-KOCABAŞ: YK 12: A 9TH-CENTURY MERCHANTMAN FROM YENIKAPI

Table 6. Locations, cross-section shapes and dimensions of nails and dowels

Average length
Fasteners Cross-section (mm) Average diameter/width (mm)

Iron nails
Floor-timber and keel Square 155 12 × 12 (on the inner surface of the frame)
Frame-planking Square 60 6 × 6 (on the outer surface of the frame)
Ceiling and frame Square 40 4 × 4 (on the inner surface ceiling)
Frame and upper member Square ? 3.5 × 3.5/4 × 4 (on the inner surface of the
frame)
Planking and keel Square 30–40 4 × 4 (on the inner surface of ceiling)
Mast-step, floor-timber and keel Square 155 12 × 12 (on the inner surface of the
mast-step)
Mast-step and mast base Square ? 6 × 6 (on the inner surface of mast base)
Treenails
Floor-timber and keel Polygonal 170–210 Ø 16/17
Frame and planking (with small Ø Polygonal 80–140 Ø 11/12
hole)
Frame and planking (with large Ø Polygonal 80–140 Ø 14/15
hole)
Frame and first wale Polygonal 190 Ø 11/12–14/15 (Ø dependant on frame)
First wale and through-beam Polygonal 50 Ø 14
Garboard and keel (slopped nail) Polygonal 50–70 Ø 11
Unidentified holes in the planking Rounded – Ø 12 (diameter of hole)
Dowels
In plank-edge seams Polygonal 80 Ø11
Between scarf of planks (edge-joint) Polygonal 90–130 Ø11
Garboard and keel (fitting dowel) Polygonal ? Ø11
Floor-timber and bulkhead stanchion Polygonal 75 Ø14
Floor-timber and bulkhead futtock Polygonal 75 Ø14

Figure 27. a) An iron nail hole to the right of a treenail hole on the underside of frame Fr17 (IUYSP Archive); iron nail and
hole b) before and c) after application of EDTA solution (IUYSP Archive).

fasten the frames to the planking throughout the hull 12mm; 6 × 6mm; and 4 × 4mm, varying according
(Table 6). to their position and use in the hull. The lengths of
these nails as determined by reconstruction drawings
Iron nails are 155mm, 60mm, and 40mm respectively. The nails
Deposits adhering to the hull indicate nail holes covered have a square-section shank and taper towards the
with corrosion and rust stains (Fig. 27a). The sizes of point. Further, during conservation of the timbers and
nails used were determined from measurements taken cleaning the iron stains with EDTA solution, 12mm-
from corrosion shapes or the nail holes. Three different diameter holes were discerned (Fig. 27b). These holes
upper shank sizes were detected: approximately 12 × are very hard to see with the naked eye, being filled with

© 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society. 373
NAUTICAL ARCHAEOLOGY, 47.2

a hard layer of wadding, with iron nails driven through pegs, were placed between two members, and in the
it. Nail-head marks, about 30mm in diameter, were plank-edge seams (Fig. 28), for fitting the garboards to
distinguished by corrosion and pressure marks around the keel, between the starboard plank PS8 and the wale,
some of the holes on the outer surface of the planking. and between the bulkhead components. Treenails which
This revealed two different forms of nailing: iron nails are usually polygonal wooden nails, are driven from one
were driven through pilot holes previously opened and side of two connected members. Treenails were detected
packed with wadding in the planking and into the joining floor-timbers to the keel, frames to the planking
frames; or, nails were hammered directly through the and wales, wales to the through-beam, and wale 3 to the
planking into the frames with no evidence of pilot caprail. Additional holes were also detected irregularly
holes. In both actions the same sized nails were used, distributed in the planking, unrelated to fastening the
and they have an irregular pattern: however, those frames, in some of which treenails were identified.
driven through pilot holes were commonly found in the Pressure marks were not identified around these 12mm-
repaired areas of the hull. diameter holes (Fig. 29).
Three different diameters of treenails were
Treenails and dowels determined, with diameters of 11/12, 14/15 and
Wooden fasteners were used in two different forms: 16/17mm. The treenails are slightly conical tapering
treenails and dowels. Dowels, which are polygonal small by about 2mm. For the floor-timber–keel joints both
wedged or unwedged treenails were noted. The treenails
were driven into holes drilled through the frame and the
keel, and then the protruding parts were trimmed. For
the frame–planking joints Ø11/12mm and Ø14/15mm
treenails were used. Moreover, treenails were placed
in two distinct locations on the upper face of frames:
in the centre of the frame, and close to the side edge
of the frame. Similarly, some of the futtocks were also

Figure 28. a) Drawing of planking edge-dowel; b) drawing


of planking scarf edge-dowel (G. Turan and M. Sağır, PDV Figure 29. Treenail hole in the planking of unknown
Ç. Şanlıgençler). function (IUYSP Archive).

Table 7. Distributions of treenails with small and large diameters at the floor-timbers and futtocks (all of the futtocks are ash)

Floor-timbers with large Floor-timbers with large Floor-timbers with small Floor-timbers with small
treenail and their futtocks with treenail and their futtocks with treenail and their futtocks treenail and their futtocks
large treenail small treenail with small treenail with large treenail

Frame Wood species Frame Wood species Frame Tree species Frame Tree species

Fr2 Hornbeam Fr3 Oak Fr11 Oak Fr10 Oak


Fr6 Walnut Fr5 Walnut Fr15 Oak Fr22 Oak
Fr8 Walnut Fr7 Walnut Fr17 Oak Fr24 Oak
Fr12 Oak Fr9 Oak Fr19 Oak – –
Fr14 Walnut Fr13 Walnut Fr21 Oak – –
Fr16 Hornbeam Fr23 Oak – – – –
Fr26 Walnut Fr25 Walnut – – – –

374 © 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society.
I. ÖZSAIT-KOCABAŞ: YK 12: A 9TH-CENTURY MERCHANTMAN FROM YENIKAPI

Figure 30. a) The treenail cut at the scarf joint in the short arm of the floor-timber Fr6 (photo) (IUYSP Archive); b) schema
showing the treenail at the L-shaped end of the floor-timber Fr11 (I. Özsait-Kocabaş, PDV Ç. Şanlıgençler).

futtocks with small-diameter treenails placed close to


the edge are on the starboard side; the futtocks with
large-diameter treenails placed centrally are on the
port side (Table 7).
An additional detail concerns the L-shaped short
arm ends of the floor-timbers. For Fr5, Fr6, Fr8, Fr11,
Fr12, Fr14, Fr15, Fr17, Fr18, Fr19, Fr20, Fr22, Fr24,
Fr25, a treenail joining the end of the short arm of
the floor-timber to the planking had been cut through
leaving part of the treenail in the floor-timber and
the other part in the plank, and a gap between them
(Fig. 30). This suggests that the floor-timbers were cut
long and fastened to the planking before the futtock
scarfs were cut and the futtocks fitted. The cut edges of
the L-shaped scarfs of the floor-timbers are parallel to
the planking seams beneath them, which also supports
this argument.
Two types of treenails were used in the frame–wale
joints: wedged and unwedged. Wedged treenails were
found at the first starboard wale (wale 1) joints with the
long arms of floor-timbers only (Fig. 31).

Bedding
In addition to the nails, a grey resin was noted in
locations requiring extra strength, such as the gaps
in the keel scarfs, between the garboard and the keel
(Fig. 32a), between the mast-step timber notches and
Figure 31. Wedged treenail connecting wale 1 and floor- the frames, between some ceiling planking and the
timber Fr16: a) photo (IUYSP Archive); b) drawing (G.Turan innerface of frames (Fig. 32b), and between frames, keel
and M. Sağır, PDV Ç. Şanlıgençler). and planking, and the inner surfaces of the transverse
holes in the keel (Fig. 32c). Resin was also found in
deep cracks and broken parts of two frames (Fr21 and
attached with treenails placed at the centre, and some Fr22) (Fig. 32c). Resin was used for bedding (sealing)
at the edges of the upper face. Of the surviving futtocks and perhaps to strengthen the joints. It was applied as
and those evidenced by holes in the planking, 10 were a fluid as evidenced by it having penetrated the wood
affixed centrally with large-diameter treenails (Fr2, and flowed into the joints. The resin, which becomes
Fr6, Fr8, Fr10, Fr12, Fr14, Fr16, Fr22, Fr24, Fr26); 12 very hard on drying, maintained its gluing qualities and
were held in place by small-diameter treenails driven it was difficult to separate some of the joints during
close to the edges of the upper face (Fr3, Fr5, Fr7, Fr9, our investigations of the timbers in the lab. The resin
Fr11, Fr13, Fr15, Fr17, Fr19, Fr21, Fr23, Fr25). The was analysed by Armaggan Culture Tourism Trade and

© 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society. 375
NAUTICAL ARCHAEOLOGY, 47.2

Figure 32. Locations of grey resin: a) at the the keel scarf; b) in the mortise for the mast-step on floor-timber Fr12; c) in the
transverse hole (IUYSP Archive).

Industry Incorporated Company’s DATU Laboratory,


which defined it as a mastic resin.

Waterproofing
The waterproofing of the vessel was provided by
luting, caulking, and a layer of pitch. The luting, the
presence of which was reported in Port Vendres 1 (Liou,
1974: 422; Santamaria, 1995: 149–50) and Dramont E
(Bockius, 2006: 118, 120) and has been identified in
many Yenikapı shipwrecks (Pulak, 2007: 133; Kocabaş,
2015a: 11), is present on all the surviving planking of
YK 12 (Fig. 33a). The dowel holes in the plank edges
were drilled longer than the dowels and luting was also
stuffed into these holes.
Larch resin/turpentine was found in analyses carried
out in the DATU Laboratory of Armaggan Culture
Tourism Trade and Industry Corporation. The inner
and outer surfaces of the planking of the YK 12 hull
and the upper surfaces of the ceiling planks are covered
with a yellow resinous layer (Fig. 33b), a mixture of
larch resin/turpentine and horsehair.
The shipbuilder improved watertightness by luting
the plank seams, and by coating with a thick layer
Figure 33. a) Luting on the outer surface of plank; b) yellow of pitch the entire inner and outer surfaces of the
pitch layer on the inner surface of planking. Saw marks and planking. Also, caulking material was probably used
grey resin pitch on the inner surface of planks below the floor- between the planks of the upper hull. At the same time,
timber E14 (IUYSP Archive). the pitch coating gave the hull timbers a protective

376 © 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society.
I. ÖZSAIT-KOCABAŞ: YK 12: A 9TH-CENTURY MERCHANTMAN FROM YENIKAPI

Figure 34. Two scribe marks on the outboard surface of plank PP2 indicating the beginning of a scarf (IUYSP Archive).

future placement of frames cross over plank seams. In


addition, a particular mark of the shipbuilder, three-
four parallel, stepped lines, was found on the inner
surface of the planking, on keel 1–keel 4 (Fig. 35a), on
the first wale, on the mast-step timber, and on a few
frames. A series of six or seven lines marked on the inner
surface of the keel may also have had a special purpose
(Fig. 43b). Scribe marks on the frames are fewer than
on the planks. Furthermore, there is no evidence of any
mark indicating the position of planks on the outer
surface of the frames; however, on the outer surfaces
of Fr12 and Fr14’s long arms, and on starboard futtock
Fr15 lines have been identified that are thought to mark
the position of starboard wale 2’s inboard edge.

Design, construction, and assembly


Shell strength
When we look at YK 12, it is clear that elements of
two traditions, namely shell-based and skeleton-based,
are visible in its construction and that the vessel was
built using a mixed method. Much care and effort was
applied to building the shell structure and maximizing
its strength: for example, the five keel/post timbers are
joined with keyed hook scarfs and any gaps filled with
grey resin and caulking material. Having joined the
keel timbers and posts, the construction continued with
fastening the garboards to the keel. The garboards, each
Figure 35. a) The builder’s sloped 3–4 scribe marks on made from a single plank, were fastened with treenails
the inner surface of the keel; b) the builder’s side-by-side and placed on a bed of grey resin and luting applied to
perpendicular six-seven scribed lines on the inner surface of the keel rabbet. Indeed, the fastening of the keel and
the keel (IUYSP Archive). garboard, a feature that played a major role in hulls
built in the shell concept, continued to have importance
layer against wear and tear, splitting, and seawater in mixed construction where a shell-based philosophy
damage. still had a strong influence (Pomey, 2004: 27; Pomey
et al., 2012: 297).
Carpenter’s marks The plank-edge dowels exhibit a homogeneous
A large number of scribed carpenter’s marks were distribution, positioned between 0.20m and 0.36m
identified on the inner surfaces of the planks: some apart throughout the hull up to the first wale (Fig. 36).
were used to align the holes for the dowels; some mark As they are numerous and placed at short intervals
structural details. For example, two small scribed marks they contribute to the strength of the hull in addition
were made on the outboard surface of the lower plank to aligning the planking and also provide resistance
PP2 at the start of scarf of PP3-2 (Fig. 34). Scribed to the planking opening during a voyage. Their task,
marks on the inner face of the planking indicating the contrary to the dowels with an uneven distribution

© 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society. 377
NAUTICAL ARCHAEOLOGY, 47.2

Figure 36. Locations of plank-edge dowels on planking plan. Grey lines mark frames (I. Özsait-Kocabaş, PDV Ç. Şanlıgençler).

found on the Bozburun vessel (Harpster, 2005a: 91), can be seen in many other ancient Mediterranean
was similar to the mortise-and-tenon joints found on shipwrecks including Yassıada I, which has a mixed
the bottom of Yassıada I (van Doorninck, 1982: 59), construction using the shell-based method for the
Pantano Longarini (Kampbell, 2007: 53, 57) and YK 11 bottom of the hull (Steffy, 1982: 73).
(Pulak et al., 2015: 49; Ingram, 2018: 110–113), ships Additionally, the through-beams are fastened to
dated to the early 7th century AD. the first wale, rather than the frames, as seen on
The use of S-scarfs, which are one of the earlier shipwrecks with shell-based construction, such
characteristic features of Greco-Roman hulls, would as Laurons 2 (Pomey et al., 2012: 242, 249), the County
have maximized the strength of the hull by increasing Hall ship (Marsden, 1974: 59), and with partly shell-
the contact surface of the scarf. S-scarfs were also based construction such as Yassıada 1 (van Doorninck,
documented in shipwrecks such as Ma'agan Michael 1982: 53). This evidence suggests that a shell philosophy
dated about 400 BC (Steffy, 1994: 41; Kahanov, 1998: had a strong influence in the construction of YK 12.
155), County Hall dated late 3rd century AD (Steffy,
1994: 72; Marsden, 1974: 62), Pantano Longarini Framing pattern
dated early 7th century AD (Throckmorton and Analysis has shown that the builder of YK 12 placed the
Throckmorton, 1973: 262; Kampbell, 2007: 78), frames before completing the planking (see below) and
Bozburun dated AD 874 (Harpster, 2005b: 255) and a skeleton-based construction philosophy is apparent
several Yenikapı shipwrecks dated to the 5th–10th in the framing pattern and the lack of plank-edge
centuries AD. The butt-joints observed on the vessels dowels above the first wale. It is suggested that the
thought to have been built according to the skeleton transition occured just below or just above the first wale.
philosophy in Yenikapı, such as YK 17 and YK 27 Some floor-timbers were attached to the assembled
emphasize the difference between the two main lower planking shell. From this level up to wale 3, the
techniques (Türkmenoğlu, 2012; Kocabaş, 2015a: 19, planks and wales were attached to these floor-timbers.
31). Similarly, the planking of the Dor/Tantura lagoon The alternating arrangement of the floor-timbers, the
wrecks that have been interpreted as skeleton-based fastening of most of them to the keel, and the precision
vessels, such as Dor 2001/1, Tantura A, Tantura B, of floor-timber–futtock scarfs reflect the importance
Tantura E, and Tantura F, also featured butt-joints given to the frame system. A floor-timber dominated
(Kahanov et al., 2004: 116, fig. 8.5; Barkai and framing pattern was reported in the County Hall ship
Kahanov, 2007: 29). dating to AD 300 (Marsden, 1974: 60), Dramont F
The planking pattern noted on YK 12 is similar dating to the second half of the 4th century AD in
to the alternating planking arrangement reported in which pegged mortise-and-tenon joints were used in the
the 4th-century-BC Kyrenia shipwreck (Steffy, 1985: planking seams (Joncheray, 1977: 6), and also Parco
91–92; Steffy, 1994: 48), which is an archetype of the di Teodorico dating to the 5th century AD, which has
shell-based construction method. This pattern was used unpegged mortise-and-tenon joints for the planking
for the same purpose in the hull of YK 12, that is (Medas, 2003: 46). A similar pattern can be seen on
spreading the weakness introduced by the joints around the Bataiguier (Jézégou et al., 1997: 35), Bozburun
the hull to improve the strength of the shell. This (Harpster, 2005a: 93) and Serçe Limanı vessels (Steffy,
alternation of plank placement and scarf orientation 2004: 158), which are said to have been built with a

378 © 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society.
I. ÖZSAIT-KOCABAŞ: YK 12: A 9TH-CENTURY MERCHANTMAN FROM YENIKAPI

Figure 37. Frame (Fr14) planking treenail running into


edge-dowel (between PP4-PP3) (IUYSP Archive). Figure 38. Schema showing fastening of a plank end (I.
Özsait-Kocabaş, PDV Ç. Şanlıgençler).

skeleton-based construction technique. However, since


the alternating floor-timbers on these vessels are used to the plank ends to the plank below using two dowels
determine the shape of the hull and are fastened to the spaced 0.25m apart. Between these, a single long dowel
keel before the planks (Harpster, 2009: 301; Harpster, joins the plank above, the plank end, and the plank
2010: 45), they differ substantially from YK 12. Even below (Fig. 38).
if the frames undoubtedly provide the majority of To identify the assembly sequence more precisely
the transverse strength and contribute to the upper the plank-edge dowels positioned at frame stations, the
shape of the hull, YK 12’s floor-timbers support rather intervals between dowels, and carpenter’s marks on the
than shape the hull’s bottom because the construction inner surface planks at the corresponding frame station
started with the planks. were analysed in more detail. Analysis of the plank-
It is certain that the frames strengthen the structure edge dowels under the floor-timbers showed that no
of the hull and carry a significant part of the structural floor-timber could have been placed prior to the fifth
load; however, the fact that the planks exhibit an strake being fitted. Moreover, Fr12, the midship frame,
alternating pattern, like the frames, indicates that was placed after at least the seventh planking strake.
planks were also considered important for the strength The floor-timbers that could have been inserted after
of the hull. Put another way, the fact that the planks as the seventh planking strake are Fr15, Fr21, and Fr22,
well as the frames in YK 12 have an alternating pattern, and the ones likely placed just before wale 1 are Fr4,
much as the planking documented on Kyrenia (Steffy, Fr7, Fr9, Fr10, Fr11, Fr13, Fr14, Fr16, Fr17, Fr18, E19,
1985: 92) and Yassıada 1 (Steffy, 1982: 73), indicates Fr20, Fr23, and Fr26. A particular capenter’s mark on
that the frames are not the only structures giving the the inner surface of the wale 1 corresponding to the
hull its strength. The builder used both the planking lower section of Fr8 was detected. This detail suggests
and the frame system together in order to strengthen that Fr8 was inserted after wale 1.
the hull structure and spread stresses. One of the three important characteristics used to
determine construction sequence according to Sean
Assembly sequence McGrail is how the ship was made watertight (McGrail,
There are a number of indications that the bottom of 1981: 43): for YK 12 the watertightness of the bottom of
the hull was built before the frames were inserted. Many the vessel was ensured by the planking. The long planks
of the treenails identified along the upper surface of the start at the bow of the vessel and cover the load-bearing
keel that fasten the garboards were positioned under the areas of the hull, with the planking joints concentrated
floor-timbers; however, no holes or pressure marks on in the curved stern.
the underside of the corresponding floor-timbers were Another indication that supports that the floor-
found, proving that none of the floor-timbers were in timbers were added after the fifth strake is the
place when the garboard treenails were driven and cut. differences in shape seen on each side of the hull. The
The plank-edge dowels fastening the planking scarfs angle of deadrise, as indicated by the floor-timbers is
follow a standard system regardless of the presence not consistent to port and starboard, for example the
of frames behind them, reinforcing the notion that deadrise angle of the starboard arm of Fr4 is wider
the planking was assembled before the framing. The than that to port, while the starboard arm of Fr22 is
presence of frames would have made driving many of narrower than that to port. These differences in the
the edge dowels awkward where they are placed directly symmetry of both sides disappear after the turn-of-the-
behind the dowel. In addition, many of the treenails that bilge, coinciding with the fifth strake and the floor-
fasten the frames to the planks pass through dowels in timber–long-futtock scarf. Above this level the sides of
the plank seams (Fig. 37). the hull are a closer match. Based on this detail, the hull
Another significant detail is that none of the stealer bottom appears to have been formed before the floor-
plank ends, in the third and fourth strakes, were timbers were placed. Following a mixed construction
fastened to the frames. Instead, the shipbuilder attached concept, the bottom of the floor-timbers would have

© 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society. 379
NAUTICAL ARCHAEOLOGY, 47.2

been shaped to fit the planking, while above the turn-of- forms of wooden elements, such as the floor-timber–
the-bilge the framing determined the shape of the hull futtock joints (Harpster, 2005b: 115); however, neither
(Basch, 1972: 16). the construction sequence nor the design of YK 12
The futtocks are not fastened to the floor-timbers, resemble the 9th-century Bozburun vessel in which
so, despite the close-fitting L-shaped scarfs, they require some duplicated frames based on the midships frame
the planking to support them. This situation could have were placed before the planking (Harpster, 2009: 301).
been resolved, as Beltrame and Bondioli have pointed The plank-edge dowels used to align the planks have
out (2006: 92), by the pre-installation of side planking, an irregular distribution throughout the Bozburun hull
battens, or the top wale. All three solutions emphasize (Harpster, 2005a: 92–93). For YK 12, in constrast, the
the importance of the longitudinal shaping of the upper hull planking was erected without the frames at least up
hull. to the fifth planking strakes. The frames are therefore
It is possible to broadly describe the process as ‘mixed not pre-designed, but rather formed to fit the shape of
design and mixed construction’: however, this does not the lower part of hull. This constitutes the basic design
adequatley describe YK 12. The vessel is evidence of a and construction difference between Bozburun and
technological transition that is thought to have begun YK 12. Although they were built at similar dates, the
with increasingly loose and widely spaced mortise-and- Bozburun ship tends towards a skeleton-based mixed
tenon joints from the middle of the 2nd century AD construction technique, whereas YK12 is representative
and continued with unpegged mortise-and-tenon joints of a period when shell-based principles and design were
from the end of the 4th century and the beginning used in combination with mixed construction methods.
of the 5th century AD (Pomey et al., 2012: 297). The (Özsait-Kocabaş, 2012: 118–119).
design, construction, and assembly sequence of YK 12 In terms of some structural features and its
reflects the 9th-century projection of a ‘longitudinal construction philosophy, YK 12 was observed to
perspective’ defined by Steffy for ancient hulls (1995: be closer to the 5th-century wreck of Parco di
419); it is an example of one of the last stages of Teodorico than its contemporary vessels. For example,
mixed construction techniques prior to the adoption of its mast-step/keelson was placed directly on to the
skeleton-first principles (Hocker, 2004a; Pomey, 2004; floor-timbers and a full ceiling was fastened to
Pomey et al., 2012). the frames with iron nails (Medas, 2003: 43). In
terms of construction philosophy, YK 12 is linked
Comparanda to the early 7th-century shipwrecks: Yassıada 1
Mediterranean shipwrecks dated to the 8th–10th (van Doorninck, 1982); and Pantano Longarini
centuries are few outside of Yenikapı. Tantura E, (Throckmorton and Throckmorton, 1973; Kampbell,
Tantura F, and Tantura B, all dated to the 7th–9th 2007), which were much larger; and YK 11 (Pulak
century and found in the Dor/Tantura Lagoon, Israel; et al., 2015: 47, 49; Ingram, 2018). These vessels have
the Agay A, and the Bataiguier wrecks are dated to the mixed construction features, while their shapes were
10th century and found off the southern coast of France determined substantially by their planking.
(Kahanov, 2000; Mor and Kahanov, 2006; Barkai and
Kahanov, 2007; Israeli and Kahanov, 2014; Pomey
et al., 2012: 269–272, 275–277), all have skeleton- Yenikapı merchantmen and YK 12
based features, such as planking without edge fasteners, Of the 37 shipwrecks discovered in the Theodosian
the garboards not joined to the keel, and planking Harbour, six have been classified as galleys and 31
butt-joints. In addition, the 9th-century Bozburun as merchant ships (Kocabaş, 2015a; Kocabaş, 2015b:
shipwreck, discovered off the south-western coast of 48–53; Pulak et al., 2015). Post excavation analysis
Turkey, which has widely spaced edge dowels, has is still in progress for most of them; however, in situ
been presented as an example of mixed construction recording, photographs, and preliminary publications
with skeleton principles (Harpster, 2005a). Of the 31 provide an initial data set. Here, the focus is on the hull
merchantmen recovered at the Yenikapı site, most have construction of the Yenikapı merchant ships and their
plank-edge dowels and are in this date range, making similarities to YK 12.
a considerable contribution to our knowledge of ship Yenikapı merchantmen have been classed in three
archaeology for this period (Pulak, 2007; Kocabaş and groups based on the hull planking joints: vessels with
Özsait-Kocabaş, 2013; Kocabaş, 2015b; Pulak et al., pegged or unpegged mortise-and-tenon joinery; vessels
2015). with edge fasteners in the form of dowels, and vessels
Even if the hull form of YK 12, with its flat without edge fasteners. Alternatively, the ships can be
floors and round turn-of-the-bilge, has similarities to classified by the geometry of the hull cross section
Tantura B, Tantura F and Agay A at first glance, amidships as flat-floored or wine-glass shaped; or by
analysis of the construction features suggests it belongs size as small, medium, and large vessels.
to a completely different shipbuilding tradition. The The merchantmen without edge-joined planking
Bozburun shipwreck, also built in the 9th century and have also been considered here to avoid missing any
found on the Turkish coast, shares common features clue. Briefly, most have rabbeted keels. Their planks
with YK 12, with plank-edge dowels and similar are joined with diagonal, butt, and three-planed scarfs

380 © 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society.
I. ÖZSAIT-KOCABAŞ: YK 12: A 9TH-CENTURY MERCHANTMAN FROM YENIKAPI

(Z scarfs). The framing pattern has alternating floor- also be divided in two subgroups according to their
timbers and half-frames. Joints are mostly fastened with hull shapes. The first is represented by the box-shaped
iron nails. YK 31 differs from these with S-scarfs, and YK 7 and YK 8 with very flat floors. The limber-holes
the planking–frame fasterners include both treenails of YK 7 and YK 8 are rounded. YK 7 has bilge keels
and iron nails. The vessels without edge-joined planking along the hull’s bottom (Özsait and Kocabaş, 2008: 138,
have been dated between the mid 7th century AD and 150). YK5, studied by the INA fits in this group (Pulak
the end of the 9th century AD (Türkmenoğlu, 2012: et al., 2015: 56, 58). The second group includes vessels
121–125; Kocabaş, 2015a: 29–30). that have a very shallow deadrise angle: these are YK 6,
In considering YK 12 we should also mention YK 9, and YK 20. For YK 1, it is reported that its keel
the earliest group of Yenikapı shipwrecks, which was rockered and lacking rabbets (Pulak et al., 2013:
have mortise-and-tenon joints. Of these, only one has 33, Pulak et al., 2015: 61). Bilge keels or runners are
pegged mortise-and-tenons (YK 34). They are wine- also reported in YK 24 (Pulak et al., 2015: 58). YK 1
glass shaped at their midships cross-sections. They and YK 24 are classified in one of these two subgroups
have thick planks, and diagonal and S-shaped plank according to their structural characteristics.
scarfs. Their thick and closely spaced frames are The vessels with plank-edge dowels and rabbeted
generally alternating floor-timbers and half-frames. keels are YK 3, YK 12, YK 15, YK 18, YK 21, YK 30,
Stringers were nailed on to the frames. Mortise-and- and YK 32.3 These are discussed in two subgroups
tenon fastened vessels at Yenikapı have been dated according to the framing pattern. The first subgroup
between the 5th and the beginning of 7th centuries with shallow (YK 12 and YK 21) or a curved (YK 18)
AD (Kocabaş, 2015a: 8, 30; Pulak et al., 2015: 47–50; garboards, and a round turn-of-the-bilge, has a framing
Ingram, 2018). pattern of floor-timbers with long arms alternating to
Here focus is on the vessels with plank-edge dowels. port and starboard side. YK 14, studied by the INA
The majority have planks joined with plank-edge fits in this group (Pulak et al., 2015: 54–55; Jones, 2017:
dowels in the lower hulls up to the waterline; diagonal 259). YK 20 has curved keel timbers with rabbets. The
and S-scarfs are predominantly used to join planks in possibility that the central keel with no rabbet is a
a strake; and they commonly have flat floor-timbers replacement piece is suspected; if so, YK 20 can be
amidships. The framing pattern is floor-timbers with considered in this subgroup.
long arms alternating to port and starboard. The The other subgroup is represented by YK 3, which
futtocks are placed at the ends of floor-timbers in L- has hollow garboards or a wine-glass-shaped hull. YK 3
shaped scarfs or placed to the fore or after sides of was built with a framing pattern of alternating floor-
floor-timbers. Floor-timbers are fastened to the keel timbers and half-frames. Futtocks were placed on the
either with iron nails or by a combination of treenails L-shaped ends of the half-frames or to the fore or
and iron nails. In some vessels, not all floor-timbers after side of the floor-timbers (Çetiner, 2013: 61). With
are fastened to the keel. The floor-timbers have mainly hollow garboards and a round turn-of-the-bilge or
triangular or rounded limber-holes on either side of wine-glass-shaped hull, YK 32 and YK 23 are also
the keel. In some vessels, the right-angled side of the likely to be in this subgroup (Pulak et al., 2015: 52).
triangles are towards the keel, while in others they This subgroup, with its hull shape and framing pattern,
face the opposite direction. The frames are generally resembles ancient Mediterranean ships (Pomey, 2004:
fastened to the planking with treenails and iron nails. Pomey et al., 2012: 298).
The mast-step timbers, even if their dimension and YK 12 has the some of the characteristics of the
locations are not standard, were placed directly on the Yenikapı shipwrecks with plank-edge dowels; however,
floor-timbers. The stringers are narrow, flat timbers there are some differences in terms of construction
placed at or above the waterline. There are transverse characteristics and the variety of use of wood species
holes through the keels amidships and towards the used. With its rabbeted keel, YK 12 resembles YK 18
stern, presumably used for hauling the vessels ashore. and YK 21. The stern of YK 14 is narrower than the
It is known that in the shipbuilding tradition of the bow, as in YK 12 (Pulak et al., 2015: 54; Jones, 2017:
Black Sea region of Anatolia, a hole is cut at the centre 254), which is one of the rare Yenikapı vessels that
or towards the stern of the keel for hauling the vessels closely resembles YK 12 in its construction features.
smaller than 10m ashore; such holes were blocked when However, certain YK 12 features find similarities in
not used (H. Bakan, a shipbuilder from Amasra, pers. other Yenikapı shipwrecks. The slightly rockered keel of
comm., October 2009). These vessels, which constitute YK 12 resembles YK 1 (Pulak et al., 2013: 33), YK 6,
the majority of wrecks recovered from the Yenikapı and YK 7. S-shaped scarfs were predominanatly used
excavation site, date between the 7th and the late in YK 1, YK 3, YK 9, YK 14, YK 15, YK18, YK 20,
10th centuries AD (Kocabaş and Özsait-Kocabaş, 2013: YK 21, YK 32 and YK 31, though without plank-edge
43–44; Pulak et al, 2013: 29; Kocabaş, 2015a: 8, 29; fasteners; and YK11, YK22 and YK35 with mortise-
Pulak et al., 2015: 50–62; Jones, 2017). and-tenon (Kocabaş, 2015a: 8; Pulak et al., 2015: 61;
The vessels with plank-edge dowels, studied by the Jones, 2017: 262; Ingram, 2018: 114).
IU team are divided according to whether the keels The garboards of YK 14, similarly to YK 12, were
have rabbets or not. The ships without keel rabbets can fastened to the keel with treenails that were driven

© 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society. 381
NAUTICAL ARCHAEOLOGY, 47.2

diagonally from the outboard surface of the planks, repair carried out in the north of Anatolia rather than
and additionally with iron nails (Jones, 2017: 262). being characteristic of its construction. In contrast,
Garboards fastened with treenails in keel rabbets almost half of the planks and a part of the ceiling of
were observed on the curved keel timber of YK 20. YK 31, which has no edge-joints, are chestnut.
YK 7, YK 8, and YK 9 have keels without rabbets It is noted that YK 12 most resembles the ships
but with treenails fastening the garboards to the that have keels without rabbets (YK 6, YK 7, YK 8,
keel. YK 9, and YK 20), in terms of planking wood species.
The other shipwrecks, in addition to YK 12, where a However, the specific use of ash for the futtocks was not
combination of treenails, iron nails, and no fastneners detected in any other vessel.
are identified in the floor-timber–keel joints are YK 6, Vessels with plank-edge dowels differ from the
YK 7 and YK 24 (Özsait and Kocabaş, 2008: 104, 132– other Yenikapı vessels in their dimensions and
134; Pulak et al., 2015: 58). construction features such as framing patterns and
The triangular limber-holes with the vertical side wood species/types. Predominant wood types of the
facing the keel as in YK 12, are also present in vessels with mortise-and-tenons, dated before the
YK 5, YK 6, YK 9, YK 18, YK 21, in YK 34 with 7th century, are Mediterranean cypress and red oak
mortise-and-tenon, and in YK 31 and YK 27 without (Akkemik, 2015: 188). For the vessels without edge-
edge-joints (Özsait and Kocabaş, 2008: 110, 131, 145; joints, dated to the 7th–9th century, stone pine and oak
Pulak et al., 2015: 56;). species were found to be the predominant wood used in
Bulkhead components were found in other vessels. YK 17, YK 27, and YK 28, and Mediterranean cypress
The ones that have similar characteristics to YK 12 and Turkish pine in YK 10 and YK 29. The trees used
are YK 20 and YK 14. These vessels have grooved for these two groups originate from the Aegean and
futtocks and floor-timbers without grooves. In contrast Mediterranean regions (Akkemik, 2015: 188).
to YK 12 and YK 20, YK 14 has bulkheads located The hull of YK 14, which resembles YK 12 in terms
forward of midships (Pulak et al., 2015: 55; Güler, of construction features such as fastening of the plank
2017). Another system with grooves cut in the inner ends, keel-garboard fastening system, and bulkhead
surface of the floor-timber is seen on YK 3 and YK 21, timbers, is primarily built of Turkey oak, a type of
while YK 29 andYK 31 (without edge-joints) and red oak (Jones, 2017: 255). Turkey oak grows almost
YK 11 (with mortise-and-tenons) have bulkhead bases everywhere in Anatolia (Akkemik, 2015: 199).
on the floor-timbers (Çetiner, 2013: 62; Kocabaş, 2015a: YK12 can be placed with the small-sized vessels with
17, figs 18–20; Jones, 2017: 265; Ingram, 2018: 127). flat floors and plank-edge dowels, built with chestnut
Through-beams, or their positions, have so far been planking (YK 7, YK 8, YK 6, YK 9, and YK 20).
detected in only a few of the Yenikapı ships. The It is suggested that this group belongs to a regional
through-beams of YK 20, YK 11, and YK 12 were type from the Marmara and Black Sea area (Akkemik,
positioned towards the stern, next to the bulkheads 2015: 188). This situation corresponds to the routes
(Güler, 2017; Ingram, 2018: 129). In YK 5 and YK 14 used for trade activities from the Theodosian Harbour
the through-beams were fitted slightly forward of (Asal, 2010: 156). But when technology is considered,
midships (Pulak et al., 2015: 54, 57). YK 12 lies between those with hull types similar to the
The wood genera/species used in YK 12 were also ancient Mediterranean ships and those that have no
preferred for some of the other merchantmen with keels rabbets.
plank-edge dowels found at the Yenikapı site. The keel
timbers of YK 6 were made of hornbeam, while those of
YK 8 and YK 20 are made of beech, resembling the keel Reconstruction
of YK 12. One of the wales of YK 20 is oak, as in YK 12. Considering the time elapsed since deposition, the
Ash was used in a few frames of YK 3 and YK 29, bottom of YK 12 was well preserved: minimal changes
which has no edge-joints; walnut was used in a few to the shapes and dimensions of timber elements
frames of YK 6, YK 8, and YK 35, which has planking were observed. According to reconstruction studies,
with mortise-and-tenon joints. White oak is the wood to starboard only one wale and two planking strakes
type preferred for frames in most Yenikapı ships. In the were missing between the existing 9th strake and the
samples from YK 12 identified at species level, some uppermost wale 3. To port, planks above the 8th strake
floor-timbers were white oak (Akkemik, 2015: Jones, were lost. The ends of the planks following the second
2017: 255). strake in the bow and the fifth strake in the stern were
The predominant wood genera of the shipwrecks that broken towards the ends and were lost above the curved
have the keels without rabbets and plank-edge dowels keel timbers (keel 2 and 4). In contrast to the hull
are chestnut for the planks and white oak for the frames. bottom, which maintained its integrity, the majority of
Of the shipwrecks that have rabetted keels and plank- the futtocks, deck and upper elements, mast, rigging,
edge dowels, the use of chestnut has only been identified and steering components were not found (Fig. 39).
in the planks of YK 32 and YK 18, other than YK 12. Information derived from other Yenikapı shipwrecks
It might be concluded that a few chestnut timbers of have contributed to the reconstruction of YK 12 to
YK 29, which has no edge fasteners, are indication of a complete its missing parts. Likewise, the image of a

382 © 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society.
I. ÖZSAIT-KOCABAŞ: YK 12: A 9TH-CENTURY MERCHANTMAN FROM YENIKAPI

Figure 39. Preserved hull remains and the reconstructed lines (I. Özsait-Kocabaş, PDV H. Güler and Ç. Şanlıgençler).

ship with a lateen sail on a 9th-century amphora For the purposes of the reconstruction it was
found in the excavation site (Günsenin and Rieth, assumed that the both sides of the ship’s hull were
2012) (Fig. 40) is invaluable considering the few symmetrical in the longitudinal axis. This approach, as
iconographic sources dated to this century. Ship images a starting point, was appropriate since the asymmetrical
on 9th-century manuscripts, such as the Homilies of St bottom of the hull survived intact, while the less
Gregory of Nazianzus (Galavaris, 1969; Zafiropoulou, complete sides appeared symmetrical above the turn-of-
1997), and archaeological and ethnographic examples bilge. The position of missing futtocks was determined
of vessels from the Marmara and Black Sea region following the existing floor-timbers. The starboard
with similar dimensions as YK 12, have also helped to strakes were mirrored to port and the missing floor-
develop hypotheses regarding the missing parts of the timbers completed.
vessel. At this stage, a physical 3D research model at a 1/10
A reconstruction based on attempting to maximize scale was constructed. This model yielded clues about
the use of the tangible data with precision and accuracy the shape and dimensions of the missing floor-timbers
was initiated. First, a plan was made by placing the and the course of the missing planking ends. The keel
disarticulated pieces found around the hull in place. and frames (actual and estimated) were assembled, then
The hull was found listing 12° to starboard, thus, the scaled copies of the planks were cut from balsa and
plans were digitally righted. Afterwards, missing parts attached to the frames by matching up nail holes. The
of existing members of the hull were estimated. Planks strakes were extended to the posts, enabling the angle
PP2, PS2, PP3, and PS3 at the bow; PS5, PS5, and PS6 of the planks and the height at which they met the
at the stern are intact or with only small pieces broken posts to be determined. The same method was used
from the ends. The curve of the posts at the bow and for wale 1. Wale 1, which is broken near the bow and
stern are calculated according to the shapes of these stern, helped to accurately determine the hull form at
planks, enabling the widest and the narrowest possible the waterline. Similarly, possible forms for the posts
angles of the posts to be determined. at the waterline were estimated by extending the ends

© 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society. 383
NAUTICAL ARCHAEOLOGY, 47.2

Figure 41. A pressure mark, nail, and a carpenter’s mark on


the inner surface of Fr6 (IUYSP Archive).

addition, two toggles found inside the hull are evidence


of rigging equipment. A similar toggle was found in the
Tantura B ship, which had a lateen sail and is dated
to the 9th century AD (Polzer, 2008: 227). Byzantine
ship depictions portray lateen sail and pairs of quarter
rudders (Albani, 1997: 82, 84; Evgenidou, 1997: 37–
38). Also, Yassıada 1 (Steffy, 1982: 85; Castro, 2008:
351), Bozburun (Harpster, 2009: 297–298) and Serçe
Limanı (Matthews, 2004) reconstructions have used
Figure 40. The amphora from the Yenikapı site with a archaeological evidence from these ships to suggest
depiction of a lateen-rigged ship (Istanbul Archaeological lateen sails and a pair of quarter rudders in each
Museums Archive).
case.
YK 12 was likely propelled by a single-mast, lateen-
sail rig and steered with a pair of quarter rudders, as was
of wale 1. The shape of the posts above this were characteristic of the period (Casson, 1994: 19; Pryor,
estimated by trial and error using the physical 3D 2004: 43). These parts are proposed as hypotheses in the
model and referring to descriptions in the Homilies of reconstruction of YK1 2, based on the archaeological
St Gregory of Nazianzus dated c.880 (Omont, 1929: evidence, iconographic and ethnographic images, and
plate CXVII.16, LII). Wale 3 and the caprail assisted according to the dimensions and structural features of
in increasing the accuracy of the proposed sheer line. the vessel. The hypothesized support mechanism and
The builder’s precision and the regularity of patterns size of the quarter rudders were tested during the first
in the construction also helped us to apprehend the sailing trials of a full-scale reconstruction of the ship
system used. Missing parts of the YK 12 vessel were launched in 2017.
estimated from clues such as the framing and planking The amphoras were found in the middle of the hull.
patterns, pressure marks and nail holes found on The ceiling plank C23 stopped on frame Fr6 at the
the planking and the frames related to the upper bow. A pressure mark on this frame corresponds to
members. For example, the marks on Fr6 are possibly a possible stanchion for a partial deck on which the
for a stanchion supporting a partial deck at the bow anchors could have been placed, perhaps with ropes
(Fig. 41) and dowels on the inner surface of Fr20 and other equipment stowed beneath. Examples of such
are evidence of the location of a missing bulkhead equipment were found in the bow of YK 1 (Pulak
stanchion. The bulkhead structure and through-beam et al., 2015: 59). There was no ceiling between Fr18
mortises indicate that the vessel had a compartment and Fr20. The bulkhead structure at these frames may
towards the stern. have delimited the hold. No knees or other evidence,
The mast-step timber is the principal indicator of such as mortises, nail holes, or pressure marks on the
the rigging system. The small size of the vessel and ceiling planks, were found that would suggest a full
lack of any other mortise on the inner surface of the deck, so it is suggested that the hold was open, which
floor-timbers suggests that YK12 had a single mast. accords with the ship’s small size. Also, the narrow
There is a small mortise for a stanchion in front of the freeboard of the vessel would make a complete deck
mast-step. A mast-partner through-beam might have impractical. A partial deck at the stern would have
been located above this mortise fastened to wale 2, provided space for handling the quarter rudders. No
since there is no evidence for it on PS9 or wale 1. In finds other than a few amphoras were found in this

384 © 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society.
I. ÖZSAIT-KOCABAŞ: YK 12: A 9TH-CENTURY MERCHANTMAN FROM YENIKAPI

Figure 42. Reconstruction drawing showing sources of evidence used (I.Özsait-Kocabaş, PDV H. Güler and Ç. Şanlıgençler).

area of YK 12. Stowage beneath the stern may have beams or decking; however, ethnographic and other
been used for cargo or more valuable goods, or used archaeological examples provide some information.
to shelter the crew in stormy weather, or to sleep. As The Parco di Teodorico boat had an open hold and
the upper parts of the vessel were not found, there is no short decks at the bow and stern (Medas, 2003: 45). It
firm evidence for the location of the decks, such as deck was assumed that YK 14 also had fore and aft decks

© 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society. 385
NAUTICAL ARCHAEOLOGY, 47.2

Figure 43. Reconstructed linesplan of YK 12 (I. Özsait-Kocabaş, PDV M.O. Avcı).

(Jones, 2017: 274–275). Beside these, side decks can be reconstruction drawings were drawn covering all details
proposed to facilitate movement along the vessel as the of construction based on working systems.
cargo would have completely filled the central part of The results suggest a length overall (LOA) of a
YK12. Side decks are seen on small vessels in the Black minimum 9.24m, a beam (Bm) of 2.64m, and depth
Sea and the Sea of Marmara. (D) of 1.10m. The reconstructed length-to-beam ratio
YK12’s hypothetical reconstruction drawings were of the vessel is 3.5:1. The hold, if full to the beam,
realized using both hand and digital drawings (Fig. 42). was approximately 3.20m long, and carried an average
Since archaeological ship forms may be less even than 210 or maximum of 250 amphoras. Displacement
modern ship drawing programs allow for, the linesplan is calculated as c.5.2 tonnes and the estimated
for YK12 was drawn by hand. The drawings also took cargo capacity of the vessel is approximately 3.26
account of the 3D model studies. Then, the vessel’s tonnes at the loaded waterline. The loaded waterline
lines were created digitally based on numeric values (WL3 on the reconstruction) was located at wale 2
from offset tables using NAPA software, as were the (Figs 34, 43).
hydrostatic properties (Fig. 42). Following completion of the reconstructed plan, the
The AutoCAD drawings made from 3D sections and the linesplans, preparations for a full-scale
measurements taken using a Total Station of the sailing reconstruction of YK 12 were started. It was
timbers in situ and the full-scale individual timber built according to the original hull shape, dimensions,
drawings realized with a FaroArm digitizer and and tree species. Construction features, such as the
Rhinoceros 4 software at the Yenikapı Shipwrecks framing and planking pattern, keel-timber scarfs, and
Research Centre of Istanbul University were plank scarfs were built following the original vessel.
superimposed by reducing them to the same scale The aim was to create a vessel with the form and
digitally to create a detailed 3D digital model of dimensions of a 9th-century ship as a physical and
the hull in situ. The completed drawings of missing visual resource. The reconstruction was completed in
parts following the results of the reconstruction collaboration with the RMK Marine shipyard team,
studies were combined with this 3D digital hull, and the IU Yenikapı Shipwrecks Project team in May
using 3D Studio max software. This hypothetical 2016. It was financed within the ‘LIMEN: Cultural
digital model was used in the initial calculations of Ports from the Aegean to the Black Sea’ project within
the shape of the hull. In addition, it was used to the framework of the European Union, Cross-Border
create several digital illustrations of the hull. As a Cooperation Programme in the ENPI Black Sea Basin.
full-scale sailing reconstruction of YK12 was planned, (Fig. 44).

386 © 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society.
I. ÖZSAIT-KOCABAŞ: YK 12: A 9TH-CENTURY MERCHANTMAN FROM YENIKAPI

Figure 44. Full-scale sailing reconstruction of Yenikapı 12 (IUYSP Archive).

Conclusions curved stem and sternpost. Besides fair-weather sailing,


YK 12’s wide bow would have increased its ability to
Yenikapı 12 is a small merchant vessel that sank
weather light storms, while its low-draught and flat-
in the 9th century in the Theodosian Harbour
bottom would have allowed it easy access to shallow
of Constantinople, the capital of the Byzantine
harbours and bays, beaching, and even navigating
Empire, where maritime trade was intense. This
rivers. Its dimensions are estimated as: length overall
date corresponds to the period when economic and
9.24m, beam 2.64m, and depth 1.10m.
political troubles such as Bulgar and Arab sieges,
Merchant ships have always been constructed
civil rebellions, and iconoclastic movements, began to
with their economic function in mind (Unger, 1980:
subside (Haldon, 2007: 102–103; Norwich, 2011: 45);
26). Function must have influenced the design and
the routes changed from the eastern Mediterranean to
construction of YK 12. In the area of the hold, the
the Aegean islands and the Black Sea (Asal, 2010: 156);
angle of deadrise is as small as possible—at Fr 12
new practices in trade were indroduced such as the
it is 4 degrees and at Fr11 it is 6 degrees—and the
State promoting private initiatives instead of holding a
turn-of-the-bilge was slightly curved. The widest part
quasi-monopoly (Dagron, 2002: 451, Magdalino, 2002:
of the vessel was designed a little forward of centre.
532); and vessels were built and may have been used by
Even if this feature reduced speed, it expanded the
the independent merchants who owned them.
cargo area located towards the bow. The hull was
According to reconstruction results, YK 12 was a
designed so as to increase the cargo capacity despite its
small merchantman working coastal waters. It was
small size for coastal trade on short, safe routes.
likely propelled by a single-mast, lateen-sail rig and
We don’t know where YK 12’s captain would have
steered by two quarter rudders, which is characteristic
set sail to from the Theodosian Harbour had it not
of the period (Evgenidou, 1997: 38; Pryor, 2004: 43).
sunk, but the vessel then started on a journey that
It had a storage compartment near the stern, and
has transported knowledge of 9th-century shipbuilding
bow, stern, and side decks. As for the shape of the
technology to the 21st century, a destination that its
hull, its plan was wide in the bow and amidships,
builder could never have imagined.
tapering towards the stern. It had a rockered keel, and a

Acknowledgements
I would like to extend my sincere thanks to the Istanbul Archaeological Museums, to Zeynep Kızıltan and Rahmi Asal; to
the Istanbul University Yenikapı Shipwrecks Project team, particularly to the project chairman, Professor Ufuk Kocabaş; to
Professor Abdi Kükner, Professor Ünal Akkemik, Dr Evren Türkmenoğlu, Taner Güler, Dr Namık Kılıç, Dr Gökçe Kılıç Can
Ciner, Hakan Kahraman, Ayşegül Çetiner-Dinçer, Erkan Baloğlu, Çağlar Çakır, Gökçe Turan Teker, Ebru Ertuğrul-Kocaçınar,
Gökçe Eğin, İlker Kızılçay, Deniz Öztekin-Eke and Mehmet Sağır, Hilal Güler, Çisil Şanlıgençler, Serhat Keskin; to RMK Marine

© 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society. 387
NAUTICAL ARCHAEOLOGY, 47.2

Shipyard team Özgür Numan, Orhan Can, M. Oğuz Avcı, Tayfun Bilgili, Mercan Tara Şişman; to Dr Orkun Köyağasıoğlu; to
lecturer Taner Karakoç, Dr Bilge Özsait-Selçuk, Seçil Kocabaş. The Yenikapı Shipwrecks Project was supported by the Istanbul
University Scientific Research Projects Unit (Project no: 2294, 3907, 7381, 12765).

Notes
1. Istanbul University team 27 shipwrecks, INA team 8 shipwrecks and IAM team 2 shipwrecks.
2. OxA-13836 1246 ± 25 BP (cal AD 681–870 at 95.4% probability), OxA-23837 1233 ± 24 BP (cal AD 690–751), OxA-23838
1258 ± 25 BP (cal AD 672–818).
3. The framing pattern of YK 15 andYK 30 could not be determined because their frames were lost.

References
Akkemik, Ü., 2015, Yenikapı Shipwrecks Volume II: Woods of Yenikapı Shipwrecks/ Yenikapı Batıkları Cilt II: Yenikapı
Batıklarının Ahşapları. Istanbul.
Albani, J., 1997, Sea Journey, Journeys on the Seas of Byzantium, D. Zafiropoulou (ed.), 82–87. Athens.
Asal, R., 2010, Theodosian Harbor and Sea Trade in Byzantine Istanbul, in U. Kocabaş (ed.), Istanbul Archaeological Museums,
Proceedings of the 1st Symposium on Marmaray-Metro Salvage Excavations 5th–6th May 2008, 153–160. Istanbul.
Ashburner, W. (ed.), 1909, Nόµος Pοδιῶν Nαυτ ικός : The Rhodian Sea-Law. Edited from the Manuscripts. Oxford.
Barkai, O. and Kahanov, Y., 2007, The Tantura F Shipwreck, Israel. IJNA 36.1, 21–31.
Basch, L., 1972, Ancient Shipwrecks and the Archaeology of Ships. IJNA 1, 1–58.
Beltrame, C. and Bondioli, M., 2006, A Hypothesis on the Development of Mediterranean Ship Construction from Antiquity to
the Late Middle Ages, in L. Blue, F. Hocker and A. Englert (eds), Connected by the Sea, Proceedings of the tenth International
Symposium on Boat and Ship Archaeology, Roskilde 2003, 89–94. Oxford.
Bockius, R., 2006, Coating, Sheathing, Caulking and Luting in the ancient shipbuilding, in L. Blue, F. Hocker and A. Englert
(eds), Connected by the Sea, Proceedings of the tenth International Symposium on Boat and Ship Archaeology, Roskilde 2003,
117–122. Oxford.
Bortoli, A. and Kazanski, M., 2002, Kherson and Its Region, in A. E. Laiou (ed.), The Economic History of Byzantium: From
the Seventh through the Fifteenth Century, 659–65. Washington, DC.
Casson, L., 1994, Ships and Seafaring in Ancient Times. London.
Castro, F., Fonseca, N., Vacas, T. and Ciciliot, F., 2008, A Quantitative Look at Mediterranean Lateen-and Square- Rigged Ships
(Part 1). IJNA 37.2, 347–359.
Çetiner, A., 2013, Yenikapı 3: A Merchantman Carring the Past, in Z. Kızıltan (ed.), Stories from the Hidden Harbor: The
Shipwrecks of Yenikapı, 56–63. Istanbul.
Dagron, G., 2002, The Urban Economy, Seventh-Twelfth Centuries, in A. Laiou (ed.), The Economic History of Byzantium: From
the Seventh through the Fifteenth Century, 393–461. Washington DC.
Denker, A., Demirkök, F., Kiraz, M., and Akbaytogan, T., 2013, YK 12, in Z. Kızıltan (ed.), Stories from the Hidden Harbor:
The Shipwrecks of Yenikapı, 197–209. Istanbul.
Dinçel, H., Çelebi, N., and Şanıvar, N., 1977, Ağaç Teknolojisi. Istanbul.
Evgenidou, D., 1997, Byzantine Shipbuilding, Journeys on the Seas of Byzantium, D. Zafiropoulou (ed.), 32–39. Athens.
Galavaris, G., 1969, The Illustrations of the Liturgical Homilies of Gregory Nazianzenus, Studies in Manuscript Illumination 6.
Princeton.
Gökçay, M., 2010, Selected Wooden Finds from Yenikapı, in U. Kocabaş (ed.), Istanbul Archaeological Museums, Proceedings
of the 1st Symposium on Marmaray-Metro Salvage Excavations 5th–6th May 2008, 135–152. Istanbul.
Güler, T., 2017, Construction Technique of theYenikapı 20 Shipwreck, Found in the Harbour of Theodosius (İstanbul, Turkey),
in J. Gawronski, A. van Holk and J. Schokkenbroek (eds), Ships And Maritime Landscapes: Proceeding of the Thirteenth
International Symposium On Boat and Ship Archaeology Amsterdam 2012, 280–282. Amsterdam.
Günsenin, G., 2016, Ganos Limanı’ndan Portus Theodosiacus’a, in P. Magdalino and N. Necipoğlu (eds), Trade in Byzantium:
Papers from the Third International Sevgi Gönül Byzantine Studies Symposium, İstanbul 24–27 June 2013, 399–402. Istanbul.
Günsenin, N. and Rieth, E., 2012, Un Graffito de Bateau à Voile Latine Sur Une Amphore (IXe s. ap. J.-C.) du Portus Theodosiacus
(Yenikapı), Anatolia Antiqua 20.1, 157–164.
Haldon, J., 2007, Bizans Tarih Atlası, A. Özdamar (transl.). Istanbul.
Harpster, M., 2005a, Dowels as a Means of Edge-to-Edge Joinery in the 9th-Century AD Vessel from Bozburun, Turkey. IJNA
34.1, 88–94.
Harpster, M., 2005b, A Re-assembly and Reconstructıon of The 9th-Century AD Vessel Wrecked off The Coast of Bozburun, Turkey,
PhD dissertation, Texas A&M University, Department of Anthropology. College Station, TX.
Harpster, M., 2009, Designing the 9th-Century-AD Vessel from Bozburun, Turkey. IJNA 38.2, 297–313.
Harpster, M., 2010, Designing the 11th-Century AD Vessel from Serçe Limanı, Turkey. IJNA 39.1, 44–55.
Hocker, F.M., Bozburun Byzantine Shipwreck Excavation: The Final Campain 1998. The INA Quarterly 25.4, 3–13.
Hocker, F.M., 2004a, Shipbuilding: Philosophy, Practice, and Research, in F. M. Hocker and C. A. Ward (eds), The Philosophy
of Shipbuilding: Conceptual Approaches to the Study of Wooden Ships, 1–11. College Station, TX.
Hocker, F.M., 2004b, Tools, in G.F. Bass, S.D. Matthews, J.R. Steffy and F.H. van Doorninck Jr. (eds), Serçe Limanı, An
Eleventh-Century Shipwreck: The Ship and Its Anchorage, Crew and Passengers, Vol. 1, 297–326, College Station, TX.

388 © 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society.
I. ÖZSAIT-KOCABAŞ: YK 12: A 9TH-CENTURY MERCHANTMAN FROM YENIKAPI

Israeli, E. and Kahanov, Y., 2014, The 7th–9th Century Tantura E Shipwreck, Israel: Construction and Reconstruction. IJNA
43.2, 369–388.
Ingram, R., 2018, The Hull of Yenikapı Shipwreck YK 11: a 7th-century merchant vessel from Constantinople’s Theodosian
Harbour. IJNA 47.1, 103–139.
Jézégou, M.P., Joncheray, A. and Joncheray, J. P., 1997, Les Épaves Sarrasines d’Agay et de Cannes. Archéologia 337, 32–9.
Joncheray, J.-P., 1977, Mediterranean Hull Types Compared 2. Wreck F from Cape Dramont (Var), France. IJNA 6.1, 3–7.
Jones, M., 2017, The Hull Construction of Yenikapı 14 (YK 14), a Middle Byzantine Shipwreck from Constantinople’s
Theodosian Harbour, İstanbul, Turkey. IJNA 46.2, 253–283.
Kahanov, Y., 1998, The Ma'agan Mikhael ship (Israël), A Comparative Study of its Hull Construction. Archaeonautica 14.1,
155–160.
Kahanov, Y., 2000, Tantura B Shipwreck, A Preliminary Report on its Hull Construction, in J. Litwin (ed.), Down the River to
the Sea: Proceedings of the eighth International Symposium on Boat and Ship Archaeology, Gdansk 1997, 151–154.
Kahanov, Y., Royal, J. and Hall, J., 2004, The Tantura Wrecks and Ancient Mediterranean Shipbuilding, in F. M. Hocker and C.
A. Ward (eds), The Philosophy of Shipbuilding: Conceptual Approaches to the Study of Wooden Ships, 113–127. College Station,
TX.
Kampbell, S.M., 2007, The Pantano Longarini Shipwreck: A Reanalysis, unpublished MA thesis, Texas A&M University, College
Station, TX.
Katzev, M.L., 1982, Iron Objects, in G.F. Bass, and F.H. van Doorninck Jr., Yassı Ada, Vol. I: A Seventh-century Byzantine
Shipwreck, 231–265. College Station, TX.
Kaygın, B. and Aytekin, A., 2005, Ahşap Tekne Konstrüksiyonu, ZKÜ Bartın Orman Fakültesi Dergisi 7.7, 14–23.
Kızıltan, Z., 2010, Excavations at Yenikapı, Sirkeci and Üsküdar within Marmaray and Metro Projects, in U. Kocabaş (ed.),
Istanbul Archaeological Museums, Proceedings of the 1st Symposium on Marmaray-Metro Salvage Excavations 5th–6th May
2008, 1–16. Istanbul.
Kocabaş, U. (ed.), 2008, The ‘Old Ships’ of the ‘New Gate’ 1, Yenikapı Shipwrecks, Vol. I. Istanbul.
Kocabaş, U. (ed.), 2010, Istanbul Archaeological Museums, Proceedings of the 1st Symposium on Marmaray-Metro Salvage
Excavations 5th–6th May 2008. Istanbul.
Kocabaş, U., 2015a, The Yenikapı Byzantine-Era Shipwrecks, Istanbul, Turkey: A Preliminary Report and Inventory of the 27
Wrecks Studied by Istanbul University. IJNA 44.1, 1–35.
Kocabaş, U., 2015b, Geçmişe Açılan Kapı, Yenikapı Batıkları. İstanbul.
Kocabaş, U. and Özsait-Kocabaş, I., 2013, A New Milestone in Ship Archaeology: The Yenikapı Shipwrecks Project, in Z.
Kızıltan (ed.), Stories From The Hidden Harbor Shipwrecks of Yenikapı, 35–46. İstanbul.
Laiou, A. E., 2002, Exchange and Trade, Seventh-Twelfth Centuries, in A. E. Laiou (ed.), The Economic History of Byzantium:
From the Seventh through the Fifteenth Century, 697–770. Washington, DC.
Liou, B., 1974, L’épave Romaine de L’anse Gerbal (Port-Vendres), Comptes rendus des séances de l’Académie des Inscriptions et
Belles-Lettres, Vol. 118.3, 414–433.
Magdalino, P., 2002, Medieval Constantinople: Built Environment and Urban Development, in A. Laiou (ed.), The Economic
History of Byzantium: From the Seventh through the Fifteenth Century, 529–537. Washington DC.
Marsden, P., 1974, The County Hall Ship, London. IJNA 3.1, 55–65.
Matthews, S. D., 2004, Evidence for the Rig of the Serçe Limanı Ship, in G. F. Bass and F. H. van Doorninck Jr., Yassı Ada, Vol.
I: A Seventh-century Byzantine Shipwreck, 32–64. College Station, TX.
Medas, S., 2003, The Late-Roman ‘Parco di Teodorico’ Wreck, Ravenna, Italy: Preliminary Remarks on the Hull and the
Shipbuilding, in C. Beltrame (ed.), Boats, Ships and Shipyards: Proceeding of the Ninth International Symposium on Boat and
Ship Archaeology Venice 2000, 42–48. Oxford.
Mor, H. and Kahanov, Y., 2006, The Dor 2001/1 Shipwreck, Israel. IJNA 35.2, 274–89.
Morrisson, C., 2013, Trading in Wood in Byzantium: Exchange and Regulations, in P. Magdalino and N. Necipoğlu (eds), Trade
in Byzantium, 105–127. İstanbul.
Norwich, J.J., 2011, Byzantium, The Apogee. New York.
Nutku A. and Küçük, F., 1963, Türk Kıyı Tekneleri I ‘Çektirme’ ‘Gulet’, İTÜ Gemi Enstitüsü Bülteni 8, 1–36.
Omont, H.A., 1929, Miniatures des Plus Ancients Manuscrits Grecs de la Bibliotheque Nationale. Paris.
Özsait-Kocabaş, I., 2013, The Centuries Long Voyage of Ship Yenikapı 12, in Z. Kızıltan (ed.), Stories From The Hidden Harbor
Shipwrecks of Yenikapı, 47–55. İstanbul.
Özsait-Kocabaş, I., 2012, Hull Characteristics of the Yenikapı 12 Shipwreck, in N. Günsenin (ed.), Between Continents:
Proceeding of the Twelfth Symposium on Boat and Ship Archaeology, Istanbul 2009, 115–120. Istanbul.
Özsait-Kocabaş, I. and Kocabaş, U., 2008, V. Technological and Constructional Features of Yenikapı Shipwrecks: A Preliminary
Evaluation, in U. Kocabaş (ed.) The ‘Old Ships’ of the ‘New Gate’ 1, Yenikapı Shipwrecks, Vol. I, 97–186. Istanbul.
Perinçek, D., 2010, Geoarchaeology of the Excavation Site for the Last 8000 Years and Traces of Natural Catastrophes in the
Geological Profile, in U. Kocabaş (ed.), Istanbul Archaeological Museums, Proceedings of the 1st Symposium on Marmaray-
Metro Salvage Excavations 5th–6th May 2008, 191–217. Istanbul.
Pla Y Rave, E., 1996, Construcción Naval y Madera: Extracto del Libro Editado en Madrid el año 1880 ‘Tratado de Maderas de
Construcción Civil y Naval’. Madrid.
Polat, M.A., 2016, Yenikapı’nın Yükleriyle Batmış Gemileri, in P. Magdalino and N. Necipoğlu (eds), Trade in Byzantium: Papers
from the Third International Sevgi Gönül Byzantine Studies Symposium, İstanbul 24–27 June 2013, 379–398. Istanbul.
Polzer, M.E., 2008, Toggles and Sails in the Ancient World: Riggigng Elements Recovered from the Tantura B Shipwreck, Israel.
IJNA 37.2, 225–252.

© 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society. 389
NAUTICAL ARCHAEOLOGY, 47.2

Pomey, P., 2004, Principles and Methods of Construction in Ancient Naval Architecture, in F.M. Hocker and C.A. Ward (eds),
The Philosophy of Shipbuilding: Conceptual Approaches to the Study of Wooden Ships, 25–36. College Station, TX.
Pomey, P., Kahanov, Y. and Rieth, E., 2012, Transition from Shell to Skeleton in Ancient Mediterranean Ship-Construction:
Analysis, Problems, and Future Research. IJNA 41.2, 235–314.
Pryor, J.H., 2004, Akdeniz’de Coğrafya, Teknoloji ve Savaş, Araplar, Bizanslılar, Batılılar ve Türkler, F. Tayanç and T. Tayanç
(trans). İstanbul.
Pulak, C., 2007, Yenikapı Batıkları: Fırtınanın Armağanı. ArkeoAtlas 6, 128–141.
Pulak, C., Ingram, R., Jones, M., Matthews, S., 2013, The Shipwrecks of Yenikapı and Their Contribution to the Study of Ship
Construction, in Z. Kızıltan (ed.), Stories From The Hidden Harbor Shipwrecks of Yenikapı, 22–34. İstanbul.
Pulak, C., Ingram, R., Jones, M., 2015, Eight Byzantine Shipwrecks from the Theodosian Harbour Excavations at Yenikapı in
Istanbul, Turkey: an introduction. IJNA 44.1, 39–73.
Santamaria, C., 1995, L’épave Dramont ‘E’ À Saint-Raphaël (Ve siecle ap. J.-C.). Archaeonautica 13. Paris.
Steffy, J.R., 1982, Reconstructing the Hull, in G.F. Bass and F.H. van Doorninck Jr (eds), Yassı Ada Vol. I: Seventh-Century
Byzantine Shipwreck, 65–86. College Station, TX.
Steffy, J.R., 1985, The Kyrenia Ship. An Interim Report on its Hull Construction. American Journal of Archaeology 89.1, 71–101.
Steffy, J.R., 1994, Wooden Ship Building and The Interpretation of Shipwrecks. College Station, TX.
Steffy, J.R., 1995, Ancient Scantlings: the Projection and Control of Mediterranean Hull Shapes, in H. Tzalas (ed.), Tropis III:
3rd International Symposium on Ship Construction in Antiquity, Athens 1989, 417–428. Athens.
Steffy, J.R., 2004, Construction and Analysis of the Vessel, in G.F. Bass, S.D. Matthews, J.R. Steffy and F.H. van Doorninck Jr,
Serçe Limanı. An Eleventh-Century Shipwreck. Vol. I, The Ship and Anchorage, Crew and Passengers, 153–170. College Station,
TX.
Throckmorton, P. and Throckmorton J., 1973, The Roman Wreck at Pantano Longarini. IJNA 2.2, 243–66.
Trakadas, A., 2014, A Sea in Transition: Ships of the Mediterranean, in S.M. Sindbæk and A. Trakadas (eds), The World in the
Viking Age, 82–85. Roskilde.
Turna, İ. and Pirim, A.E., 2015, A Study About The Brief History of the Vessel Named ‘Çektirme’ and its Role on the Commercial
Activities of Its Age. Sosyal Bilimler Araştırmaları Dergisi, 12, 119–135.
Türkmenoğlu, E., 2012, Preliminary Report on the Yenikapı 17 Shipwreck, in N. Günsenin (ed.) Between Continents: Proceedings
of the Twelfth Symposium on Boat and Ship Archaeology, Istanbul 2009, 121–125. Istanbul.
Unger, R.W., 1980, The Ship in the Medieval Economy 600–1600. London.
van Doorninck, F.H., Jr, 1982, The Hull Remains, in G.F. Bass, and F.H. van Doorninck Jr, Yassı Ada, Vol. I: A Seventh-century
Byzantine Shipwreck, 32–64. College Station, TX.
van Doorninck, F. H., Jr, 2002, Byzantine Shipwrecks, in A.E. Laiou (ed.), The Economic History of Byzantium: From the
Seventh through the Fifteenth Century, 899–905. Washington, DC.
Vroom, J., 2016, Byzantine Sea Trade in Ceramics: Some Case Studies in the Eastern Mediterranean, in P. Magdalino and N.
Necipoğlu (eds), Trade in Byzantium: Papers from the Third International Sevgi Gönül Byzantine Studies Symposium, İstanbul
24–27 June 2013, 157–177. Istanbul.
Zafiropoulou, D. (ed.), 1997, Journeys on the Seas of Byzantium. Athens.

390 © 2018 The Authors. International Journal of Nautical Archaeology © 2018 The Nautical Archaeology Society.

You might also like