Description and analysis of El Fuerte de Samaipata in the light of new research, and a proposal of the relative chronology of its main elements

Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

doi:10.37190/arc200210

Abstract

The study describes the results of the documentation of El Fuerte de Samaipata, a pre-Hispanic archaeological site in Bolivia on the UNESCO World Heritage List. It summarizes the most recent observations of the carvings on the rock of Samaipata, which, due to progressing erosion, are slowly vanishing. Thanks to modern documentation technologies, laser scanning, structural light scanning, digital photogrammetry, and advanced digital analysis of images and 3D models of the site, all.

Full article view is only available on bigger screens.

2020

2(62)

DOI: 10.37190/arc200210

Introduction

The importance, history, and past investigations of the

UNESCO World Heritage Site of El Fuerte de Samaipata

have already been studied by numerous scholars [1]–[9].

A more comprehensive current study of former research is

included in the same issue of this journal

. Here, we will

concentrate on describing particular sectors of the rock in

light of the results of our multidisciplinary documentation

project carried out in the years 2016–2019.

* ORCID: 0000-0003-0623-8071. Faculty of Architecture,

Wrocław University of Science and Technology, e-mail: jacek.kosciuk@

pwr.edu.pl

** Italian Centre for Pre-Columbian Archaeological Studies

and Research in Nazca (Peru).

*** ORCID: 0000-0003-4137-0799. Centre of Andean Studies

of the University of Warsaw in Cusco (Peru).

**** ORCID: 0000-0001-5442-3947. Faculty of Architecture,

Wrocław University of Science and Technology.

***** Italian Centre for Pre-Columbian Archaeological Studies

and Research in Nazca (Peru).

Cf. M. Ziółkowski, El Fuerte de Samaipata in the light of current

research, in the same issue of “Architectus”.

Jacek Kościuk*, Giuseppe Orefici**, Mariusz Ziółkowski***,

Anna Kubicka****, Rosario Muñóz Risolazo*****

Description and analysis of El Fuerte de Samaipata

in the light of new research,

and a proposal of the relative chronology

of its main elements

Opis i analiza El Fuerte de Samaipata

w świetle nowych badań

oraz propozycja względnej chronologii

jego głównych elementów

Applied techniques and methods

During this project, we used several methods to doc-

ument the site – 3D terrestrial laser scanning (TLS)

and

multisensory close-range digital photogrammetry

for the

whole site, as well as structured light scanning for the

most important petroglyphs

. The data from TLS and

digital photogrammetry were collected during the eld-

work season in June and July 2016, and from structured

light scanning in July 2017. All data were compiled in

the Laboratory of 3D Scanning and Modeling at the De-

partment of History of Architecture, Arts and Technol-

ogy at Wrocław University of Science and Technology.

Cf. J. Kościuk, B. Ćmielewski, M. Telesińska, A. Kubicka, 3D

terre strial laser scanning of El Fuerte de Samaipata, in the same issue

of “Architectus”.

Cf. B. Ćmielewski, I. Wilczyńska, C. Patrzałek, J. Kościuk, Digi-

tal close-range photogrammetry of El Fuerte de Samaipata, in the same

issue of “Architectus”.

Cf. J. Kościuk, M. Telesińska, M. Nisztuk, M. Pakowska, Docu-

mentation of the most important petroglyphs by structured light scanning

and analysis of the most damaged petroglyphs by vPTM and vRTI meth-

ods, in the same issue of “Architectus”.

80 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

The results of the data compilation were raster orthoim-

ages (monochromatic and in RGB) with a resolution of

1 mm for photogrammetry and 10 mm for TLS. For both,

dierent hill-shading algorithms were used to emphasise

characteristic features of the rock surface. This proved

extremely important in the next stage of work, when or-

thoimages were used to prepare vector plans for the entire

rock, including all its terraces, niches, reservoirs, canals,

and petroglyphs. Replacing raster references during on-

screen vectorisation facilitated the interpretation of details

invisible on colour imagery or, on the contrary, provided

additional information for the grayscale terrain model.

Vector maps obtained courtesy of Albert Meyers, the head

of the German research project at Samaipata carried out

by the University of Bonn from 1992 to 1997, were also

very helpful when interpreting the most eroded parts of

the rock [9, p. 28, Fig. 9]. At this stage of our project, we

also consulted all previous topographical plans. The vec-

tor plan resulting from this back-oce work was brought

to the site during the 2017 eldwork season to check for

possible omissions or misinterpretations. The nal plan,

together with raster references, is now available in open

geographic information system (GIS)

format

Since the part of Samaipata rock that has been surveyed

in detail measures roughly 110 × 240 m, it is technically

impossible to present it on one illustration in this paper.

Unless otherwise stated, all illustrations are by J. Kościuk. The

le gend for the graphics on the plans can be found on page 78.

I. Wilczyńska, P.B. Dąbek, B. Ćmielewski, J. Kościuk, Integra-

tion of project results on a GIS platform and its impact on conservation

strategies, in the same issue of “Architectus”.

Therefore, the site has been divided into ve areas. Each

area, according to principal morphological, and to some

degree, functional, features, was then split into several fur-

ther sectors. Each sector name starts with the letter den-

ing its rough location – “W” for the western part (Fig. 1),

“N” for the northern part (Fig. 2), “S” for the southern part

(Fig. 3), and “C” for the central and “E” for the eastern

parts (Fig. 4). Sector names are then completed with num-

bers. These numbers begin at “01” for each of the ve are-

as of the site and follow on sequentially from this.

Results and discussion

The description of the results will follow the naming

convention described above. It will start with the western

part of the site; then it will cover the northern and southern

slopes of the rock, continue with the central sectors, and

terminate with the easternmost end of the rock.

Western part of the rock

Sector W01

This sector lies in the westernmost part of the rock,

just below and behind the modern viewing platform

erected for visitors (Fig. 1). In this sector, earth has ac-

cumulated and covers the natural rock. Three circular

structures and an L-shaped fragment of a wall, all con-

temporary reconstructions made of stone and clay, are

extant there. Since the main aim of the project was to

document and analyse original, Incan or pre-Inca carv-

ings on the rock, this sector is out of the scope of this

paper and is mentioned only for the sake of completeness.

Fig. 1. The western group of sectors of the rock

Fig. 3. The southern group of sectors of the rock

Fig. 2. The northern group of sectors of the rock

Fig. 4. The central and eastern groups of sectors of the rock

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 81

Sector W02

Sector W02 lies directly northward from sector W01, and

it gently slopes downwards to the east (Fig. 5). The eastern

limit of sector W02 forms a well-dened edge with trac-

es of two triangular recesses in its southern part. Close to

the western border of the sector, two raised terraces can be

seen. The area further to the west is currently unexcavated,

and earth, as well as small bushes, are possibly covering the

next sequence of terraces. The slope of this sector provides

perfect conditions for spectators watching events and cel-

ebrations in the central part of the rock located to the east.

A line of circular holes (ca. 16 cm in diameter) drilled

into the rock marks the northern extent of sector W02.

These are easily distinguishable from natural cavities in

the sandstone, which are typically ellipsoid in shape. The

circular holes are irregularly spaced with distances from

ranging from 0.4 m to 1.2 m. According to our interpreta-

tion, they were used to place the wooden posts of a quin-

cha wall that was once standing there

Sector W03

This triangular sector lies to the north of the quincha

wall line (Fig. 5). The remaining borders are the natu-

ral(?), very eroded edge of the rock (north-eastern border)

Cf. J. Kogut, J. Kościuk, A. Kubicka, Interpretation of traces of

hypothetical quincha walls in Samaipata, in the same issue of “Archi-

tectus”.

Fig. 5. Sectors W01–W06

82 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

and the north-western extent of the excavated area where

the rock is still visible. The only anthropogenic feature

in this sector is a small natural cavity in the rock that

might have been used as an oering (libation) hole, as it

has traces of an intensely eroded channel running to it.

Sector W04

Sector W04 lies directly to the east of sector W01,

and it slopes downwards to the south (Fig. 5). The south-

western part is covered with accumulated earth, while the

face of the natural rock is exposed on the remaining part

of the sector. A line of natural cracks in the rock marks

the northern limit of the sector. A similar crack, this time

north–south orientated, divides the sector into two parts.

Aside from traces of what might once have been a small

rectangular shelf, there are no other identiable signs of

intentional shaping of the rock in the western part of the

sector. Many more traces, although extremely eroded, are

found in the eastern section. In the northern area of the

eastern section, a shallow terrace ca. 1.2 m wide cuts into

the slope of the rock. Its upper edge marks the northern

extent of this part of the sector. Another east–west orien-

tated terrace, this time only ca. 0.75 m wide, is located

further towards the south.

In the northern part, the eastern boundary of the sec-

tor corresponds with a step leading down to a wide ter-

race located further to the east, while in the southern part,

a row of four round holes drilled into the surface of the

rock mark the edge of the sector. The arrangement of

these holes strongly resembles those found in the northern

part of sector W02. Thus, we interpret them as traces of

another quincha wall. We do not exclude the possibility

that this hypothetical quincha wall continued further in

a north-eastern direction, and climbed the narrow terrace

described above.

Sector W05

The main feature of this sector is a spacious (ca. 8.5

× 6.0 m), roughly horizontal platform with well-dened

edges on its northern and eastern side (Fig. 5). In the cen-

tral part of this platform, a circular shape (roughly 2.25 m

in diameter) rises ca. 10 cm above the platform level. In

the centre of this shape, very weather-beaten traces of

a gural petroglyph are extant. Earlier research described

it as a representation of a felid – possibly a puma or jag-

uar. Today, it is dicult to see the gure, let alone assign it

to a particular subspecies. However, with some help from

virtual polynomial texture mapping (vPTM) and virtual

reectance transformation imaging (vRTI) techniques,

it was possible to recognise a shape suggesting the rep-

resentation of a puma. Nevertheless, this interpretation

must be treated with great caution since the surface of the

petroglyph is highly eroded and additionally disturbed

by mosses and lichens. Further doubts are raised when

comparing our reconstruction with a drawing by Alcide

d’Orbigny en Ponce from 1832, where the schematic rep-

resentation of the puma(?) is turned 180 degrees (Fig. 6).

On the other hand, in the textual description of the draw-

ing clearly representing animalistic gures, d’Orbigny

writes Dentro de esta hipótesis, sería lícito creer también

que las medias lunas FF de la parte baja representarían

los primeros y los últimos cuartos de la luna, andando de

este a oeste [1, p. 1465] so perhaps the gure in question

is rotated in order to t his hypothesis.

The rough rock surface, especially around the north-

western edge of the circle, may suggest that the gure has

been intentionally recut (and perhaps also erased) by later

occupants of the site. This observation may point to the

pre-Inca origin of this petroglyph.

One can get a similar impression by analysing the very

shallow carvings of two T-shaped reservoirs(?) located

south of the petroglyph. They are visible only on the hill-

shaded 3D model without texture, as texture obscures

the weak traces of these shapes. These carved reservoirs

show further evidence of a second phase of rock shaping,

as the upper edge of the northern terrace of sector W04

cuts into the eastern reservoir. Thus, both carved reser-

voirs can be attributed to the pre-Inca phase of the shaping

of the rock.

The western edge of the central platform of W05 has

two steps. The rst one forms a roughly 50 cm-wide bench

elevated ca. 12–25 cm above the main platform level. It

continues along the northern edge of the sector and runs

until it reaches a narrow channel connecting this sector

with sector W06. The position of this channel corresponds

with a natural crack in the rock, and probably this connec-

tion between both the sectors is also of late provenance.

The second step on the western edge of the central

platform consists of two recesses cut into the slope. The

Fig. 6. Fragment of drawing by Alcide d’Orbigny en Ponce from 1832

(reproduced from [9, p. 25, Fig. 5])

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 83

northern one is smaller (60 × 80 cm) and forms a small

seat, while the southern one is much more spacious and

has a less regular shape. Further to the west lies a at area

that extends until another step about 25 cm in height.

Sector W06

This sector lies directly north of sector W05 and is very

similar (Fig. 5). Its central part is also a large (6.5 × 10.5 m),

roughly horizontal platform with a gural petroglyph em-

bedded into a circular shape raised above the platform

level. However, in this case, the circle is surrounded by

a ca. 35 cm-wide and 15–25 cm-deep channel that directs

water to the north.

The gural representation is relatively well preserved,

and the silhouette of a sneaking felid (possibly a jaguar)

is easy to recognise. The petroglyph gives some evidence

that it was altered in later periods. The channel surround-

ing it might be a later addition or at least have been recut.

Further to the north-west in sector W06 lies the area

where, according to the earlier investigators [5], a rep-

resentation of a coiled snake was once clear. Today, no

traces of it can be seen. Using vPTM and vRTI tech-

niques, we were able to identify at most a vague outline of

a snake crawling towards the north-west. We did not nd

any archival photographs that could indisputably conrm

that there was a depiction of a coiled snake. Instead, we

found the opposite – d’Orbigny in his drawing from 1834

(Fig. 6) clearly shows, although schematically, a snake

crawling towards the north-west. However, this random

possible coincidence does not prove the correctness of

our reconstruction. The preserved traces are so indeci-

pherable that they also allow for several other interpreta-

tions. Despite this, the one proposed here seems to be the

most reliable.

In this area of the sector, there are traces of another

small channel running to the east and joining the other

channel at a position where yet another channel joins from

the west. All three channels disperse water down the slope

in a north-western direction, across the line of a hypothet-

ical quincha wall.

The main platform of sector W06 is well dened from

the south and the west. Its southern border is a 6 m-wide

edge separating sectors W06 and W05. To the west, there

is a bench (roughly 50 cm wide) elevated ca. 12–25 cm

above the main platform level – similar to that in sector

W05. Two small (roughly 30 × 60 cm) symmetrical seats

have been cut into the upper edge of the back of the bench.

Further to the west lies a 3 m-wide at area that continues

until another step about 50 cm high.

In the middle of the western margin of the platform,

roughly adjacent to the gural petroglyph, a spacious

(ca. 2.5 × 2.5 m) recess with three nearly vertical sides at

least 70 cm high cuts into the slope. At present, the bottom

of this recess lies ca. 0.5 m below the level of the central

platform, but this might be due to local erosion of the rock

or later modications. The original bottom surface of the

recess is still visible on the western side in the northern

and southern corners. It is ca. 30–40 cm higher than the

level of the central platform. One can only speculate about

the original function of this recess. Perhaps it was a kind of

ceremonial seat associated with rituals celebrated around

the two petroglyphs in the central part of this sector.

The eastern margin of sector W06 is not as sharply de-

ned as the western one. Only a very shallow step, not

more than 20 cm high, separates it from neighbouring

sector W09. Since this step cuts into a at area where

there was once a representation of a rhea (ñandú), the

main platform of sector W06 must have been arranged

(or recut) in later periods, with the rhea representation

pre-dating it. The same applies to the central platform of

sector W05. Its north-eastern corner also cuts into the area

of the rhea representation. This observation does not ex-

clude the possibility that the representations of the rhea,

puma or jagu ar are contemporary and that the two central

platforms of sectors W05 and W06 were recut when all

three petroglyphs already existed. However, due to the

dierences in the state of preservation of particular petro-

glyphs, we are inclined to suggest that the representa-

tion of the jagu ar(?) amended an earlier, probably pre-

Inca petroglyph.

Sector W07

This sector shares its western border with sector W04

(Fig. 7). Accumulated earth covers its southern, lowest

part. The original rock emerges further to the north in two

steps. The rst one leads to an irregularly shaped platform

raised roughly 1 m above the southern, lowest part of this

sector. It forms an east–west elongated trapezium. The

second step leads to a rectangular (ca. 2.4 × 17 m) plat-

form situated ca. 60 cm higher. The southern edge of this

platform is highly eroded due to natural cracks in the rock.

At the back of this platform, there is a sharp step more

than 1.5 m high separating this sector from sector W08

in the north. A stairway with at least ve steps located in

the centre of the platform probably facilitated movement

between both sectors.

In the east, another stairway marks the border of the sec-

tor. This one is much longer, and consists of 14 steps ena-

bling access to sector S02, located further down the slope.

Sector W08

Sector W08 can be divided into three distinct parts

(Fig. 7). The westernmost part that neighbours sector

W05 does not show any traces of intentional shaping. Its

incline towards the south is disturbed by several parallel

natural cracks. In the north of the central part, three small

(ca. 1 × 2.5 m) platforms can be found. These could be

a continuation of the stairway that reaches this point from

sector W07 below.

In the eastern part of the sector, two terraces, nearly

10 m long, are extant. The lowest one, which is more than

3 m deep, can be directly accessed from the stairway on

the border between sectors W07 and W11. In front of the

last step of this stairway, an oval shape (ca. 2 × 1.4 m)

protrudes ca. 15 cm from the surface of the terrace. We

consider this to be an eroded, intentionally destroyed, or

unnished petroglyph – possibly gural. Since all of the

gural petroglyphs identied on Samaipata rock are asso-

ciated with traces of water used for rituals performed there,

our interpretation is further supported by the

presence of

84 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

shallow traces of what might be a small basin in front of

the petroglyph.

Behind the remains of the petroglyph, four steps lead

to the upper terrace, which is about 1 m high. On both

sides of the steps, there are narrow benches not more than

0.5 m deep.

Sector W09

Sector W09 is situated on the top of the western part

of the rock (Fig. 8). Its surface is slightly inclined to the

west (the dierence in height between the eastern and the

western part is ca. 1 m), and on its northern and southern

edges, rapidly slopes down. It borders sector W06 in the

west, sector W08 in the south, and sector W10 in the north.

The western side of a north–south oriented wall across the

top of the rock is the eastern border of sector W09. The

southern section of this wall runs through a ca. 35 cm-

deep reservoir that collects rainwater. Thus, the wall must

be of a later origin.

Although the wall is a modern reconstruction, sev-

eral lines engraved and still visible on the rock surface

well attest its original shape and dimensions. Three nich-

es are extant on each side of the wall. The niches look-

ing to the west have double recesses. This is typical in

Incan architecture, which clearly points to the origin of

this wall.

Since the modern reconstruction well follows original

dimensions, the Incan provenance of the wall can be fur-

ther conrmed using the cosine quantogram metrological

analysis of niches. When comparing dimensions of all the

recesses (Fig. 9), the results are inconclusive. However, if

we limit our investigation only to the inner small recesses,

the resulting quantogram matches units of measurement

typical for Incan buildings of high prestige, as proposed

by Anna Kubicka in her doctoral thesis [10].

Two other elements in this sector can however be at-

tributed to the pre-Inca period – the already mentioned

representation of the rhea in the western part and a cir-

cular depression in the eastern part of the sector. Of the

depression, only its eastern edge – a small, ca. 3 cm-high

step – can be identied, but this is enough to reconstruct

its diameter as ca. 4 m. No traces are left on the surface

of this depression that might point to its function. It is

too shallow to have been used only as a place to collect

rainwater, but perhaps it once housed another (gural?)

petroglyph in its centre. If so, this must have been entire-

ly erased by later occupants of the site – probably dur-

ing the same time that a T-shaped platform with a small

Fig. 7. Sectors W07 and W08

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 85

recess pointing to the east was carved. The outline of this

platform overlaps the reconstructed extent of the circular

depression. The petroglyph representing the rhea has also

probably been intentionally hammered out. Our attempts

to reconstruct it using vPTM and vRTI ended only with

a general suggestion about the size and orientation of the

gure. Generally, this suggestion is similar to the sche-

matic icon on d’Orbigny’s drawing (Fig. 6). There, the

gure of the rhea is shown in an entirely dierent manner

than all other petroglyphs, which, as we can see today, are

bas-relief. Can this dierence in the way of depicting the

rhea on d’Orbigny’s drawing mean that in this case, we

are dealing with a sunken relief? If so, this might be one of

the reasons that it has completely disappeared

The central part of sector W09 has been further re-

shaped during the last phases of occupation. A shallow,

Z-shaped platform cuts into the rock surface. In the middle

of the eastern, roughly square part of this platform, a rep-

resentation of another felid (possibly puma) is carved in

bas-relief. A at water channel surrounds it. This channel

continues further down the slope in a north-western direc-

tion (Fig. 8). A rounded hole (ca. 25 cm in diameter) drilled

into the rock in front of this petroglyph might have been

used for oerings. It seems that the petroglyph was created

together with the platform around it. The highest points of

the “Puma(?)” petroglyph are well below the reconstructed

Our attention was drawn to this possibility by Marta Pakowska.

line of the ridge of the rock from the time before the plat-

form was carved (Fig. 10). Thus, the wall with niches, the

Z-shaped platform, and the puma representation could all

be attributed to the period of Inca occupation of the site.

Another oering hole has been cut into the rock at the

western end of the Z-shaped platform, in front of the place

where the rhea is no longer extant. A small channel con-

nects it with a natural crack in the rock that is more than

35 cm deep.

Sector W10

Sector W10 forms a nearly horizontal platform (ca.

5 × 28 m) on the northern slope of the rock (Fig. 8). The

quincha wall already described marks the northern ex-

tent of the sector. Two steps starting on a nearly square

(ca. 2.5 × 3 m) platform in the south-western corner of the

sector probably facilitated movement between this sector

and sector W09 in the south, as there is nearly 1 m dier-

ence in height between these two sectors. The steps lead

directly to the area where the representation of the rhea

was once extant. This observation goes hand in hand with

our previous remark about the earlier provenience of this

petroglyph.

In front of the platform with steps, a water channel

ca. 6 m long and 0.5 m wide cuts into the surface of the

rock. It collects water running down from the puma petro-

glyph located in sector W09. The second branch of water

owing from sector W09 goes directly to the north and

splits into two streams.

Fig. 8. Sectors W09 and W10

86 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

The eastern part of sector W10 is organised dierently.

Although in the far east, there is a platform similar to that

found at the western end, the southern and northern bor-

ders dier. A ca. 50 cm-wide bench runs along the south-

ern edge. A small (ca. 1 × 1.5 m) platform or possibly seat

projects from the middle of this bench. In front of it, trac-

es of an extremely eroded petroglyph are extant. Again,

using vPTM and vRTI, we were able to identify this as

the representation of an animal that has caught a snake’s

head with its snout. Another snake is grabbing the tail of

the rst one, and yet another, emerging from a hole in the

rectangular centre of the whole scene, is catching the tail

of the second one. This complicated layout of four inter-

acting animals covers approximately 1 m

. It has no par-

allel on Samaipata rock. The representation of the main

animal is also very untypical and points rather to being

carved during the pre-Inca period. It does not resemble the

puma or jaguar petroglyphs on the rock

. An oering hole

Attempts to identify the species suggest that it is the white-nosed

coati (lat.: Nasua narica; local names: pizote, antoon, tejón), the only

similar animal from the area, known as a snake catcher. Cf. https://

en.m.wikipedia.org/wiki/White-nosed_coati [accessed: 15.03.2019].

(ca. 25 cm in diameter) accompanies this petroglyph in

the west.

The northern perimeter of sector W10 has further pe-

culiarities. Roughly adjacent to the snake-catching petro-

glyph described above, the quincha wall is split into two

parts running parallel for a distance of ca. 2 m. Between

them, traces of a channel directing water down the slope

are visible. The upper part of this channel is ca. 25 cm

wide and has well dened, straight edges. In the middle,

extremely eroded traces of yet another petroglyph repre-

senting a snake crawling up the slope were detected.

Sector W11

Sector W11 is located at the base of the opposite,

southern slope of the rock (Fig. 11). From the south, it

borders a at area covered with reconstructed remains of

Incan habitation [9].

In the west, the ight of steps already described for sec-

tor W07 separates these two sectors. Similar steps mark

the eastern extent of the sector. There, 18 narrow steps

(ca. 75 cm wide) climb to the north from the foot of the

rock. Traces of other, less regular steps are extant in the

middle of the sector. East of these, an oval shape protrudes

Fig. 10. North–south section across sector W09

Fig. 9. Cosine quantogram

analysis of the wall with niches.

A – The double system of Incan

units of measurement proposed

by A. Kubicka

[10, p. 171, ryc. 114]

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 87

10–15 cm from the surface of the rock. It is probably a rel-

ic of another petroglyph that is impossible to see today.

The presence of a small platform to the north of this shape

renders this interpretation even more likely.

Further to the north, irregularly arranged terraces and

platforms are probably the result of many modications

that the rock underwent during various phases of occu-

pation.

The southern area of sector W11 is divided into three

distinct units. Four double-recessed, square niches deco-

rate the face of the central part that protrudes to the south.

A narrow, 50 cm-wide bench continues to the west. The

back of the bench, which is ca. 75 cm above the bench level,

is decorated with eight small, square niches. To the east

of the double-recessed niches, a small platform (ca. 1 ×

2.5 m) cuts into the slope. A sequence of four, rectangular

terraces can be seen in the easternmost part of the area.

In front of the four double-recessed niches, there is

a 50 cm-wide and 3.50 m-long possible fragment of a wall.

Its eastern and western extents are not clear. It may be the

remains of the front wall of a roofed room that once exist-

ed in this place, but no conrmation of such a hypothesis

has been found in any of the archaeological reports.

Northern part of the rock

Sector N01

Sector N01 is located in the western part of the northern

slope of the rock, on the edge of the area covered with vege-

tation (Fig. 12). The sector comprises three dierent zones:

– A small plateau in the south-western part shaped as

a system of at-stepped terraces;

– A steep slope running latitudinally in the central part

of the sector at the foot of which traces of ve dierently

sized terraces are preserved;

– A at zone in the northern part that marks the north-

ern extent of currently identiable traces of anthropogenic

shaping of the rock.

The dierence in levels between the southern and

northern zones reaches 1.5 m in some places. In the steep,

Fig. 11. Sector W11

Fig. 12. Sector N01

88 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

eastern part of the slope, where the dierence in levels is

particularly signicant, six narrow, not more than 40 cm

wide, steps have been carved. They lead directly to the

foot of a quincha wall running along the upper edge of the

slope. Either they precede the construction of this wall,

or they were hewn in the rock when the wall ceased to

exist. They could also have been used only during the con-

struction of the quincha wall. The very negligent manner

of execution may speak in favour of the last interpretation.

In three places, at rainwater channels intersect sector

N01. They drain water from sectors W06, W10, and C02

located higher up the slope.

Sector N02

Sector N02 is a continuation of the lowest, at part of

sector N01 (Fig. 13). The quincha wall mentioned above

surrounds it from the north, west, and partly from the

south. The southern boundary is a steep, nearly 1.5 m-high,

cli that separates this sector from sector C02.

In the eastern, almost at part of sector N02, there

are relics of an indecipherable petroglyph in the form of

a round disk about 1.5 m in diameter. The disk was once

surrounded by a ca. 25 cm-wide water channel. The dif-

ference in elevation between the centre of the disk and the

surrounding channel does not exceed 10 cm. In the vicin-

ity, a few oering holes have been drilled into the rock.

The only other features of anthropogenic origin are a small

fault across the sector and the outline of a ca. 2.3 m-wide

platform at the foot of the cli.

Sectors N03 and N04

The topography of sector N03, which is located further

to the east, is similar to sector N02. Its northern border is

the continuation of the quincha wall traversing the slope

of the rock that descends towards the north (Fig. 14).

A terraced rock edge almost 2 m high marks the southern

border of this sector.

In the northern part of the sector, no traces of inten-

tional carving of the rock have been found. A sequence

of parallel natural cracks running east to west cuts this

sloping surface. In the higher, southern part of this sector,

three at terraces bear traces of intentional shaping. The

overall state of preservation of the rock in this sector does

not allow for a more conclusive interpretation of these

terraces. However, one can get the impression that the

layout of the terraces is a continuation of the layout of the

much more well preserved terraces in sector N04 located

to the east.

Sector N04 has a clearer layout (Fig. 14). The line of

circular holes that we interpret as traces of a quincha wall

again marks its northern border, while a more than 60 cm-

high, sharp step delimits its southern perimeter. Between

these two borders, four terraces are arranged. To the east,

they end at a narrow stairway ca. 70 cm wide. Seventeen

steps negotiate nearly a 3 m dierence in height between

the northernmost terrace of N04 and sector N05 located

further to the south.

The southernmost terrace in sector N04 is trapezoidal in

shape, but the remaining three are rectangular, and their per-

fectly parallel edges suggest one coherent concept of layout.

Sector N05

This sector (Fig. 15) has the shape of an irregular, elon-

gated, polygon (ca. 6 × 40 m). It is located on the northern

slope of the rock, just to the south of sectors above sectors

N04 and N06. The main elements of sector N05 are three

terraces, 2.5–3.5 m wide, falling gently towards the west.

The dierences in levels between terraces do not exceed

30 cm.

The most interesting element of this sector is the trace

of a rounded shape, located in the northern part of the sec-

tor, protruding 5–15 cm from the surface of the terraces.

Only one-quarter of the original shape has survived, but

it is enough to reconstruct the extent of the whole shape,

Fig. 13. Sector N02

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 89

which must have been over 5.5 m in diameter. The re-

maining three-quarters of the shape was destroyed when

the eastern and western terraces of sector N05 and the ter-

races of sector N06 in the north were carved. Undoubt-

edly, this is a relic of an earlier phase of carving the rock

surface – probably the pre-Inca one. Perhaps the eastern

part of the central terrace of sector N05 belongs to the

same chronological phase.

There are no traces on the surface of the circular

shape clear enough to attempt a reliable reconstruction

of the petroglyph that may have been carved here. Local

height dierences (no more than a few centimetres) on

the shape surface suggest that there may once have been

a gural representation. It would have been one of the

largest known petroglyphs in Samaipata had it survived

to this day, comparable in size only to the “Choir of the

Priests” carving situated slightly further south-east in

sector C05.

Sector N06

Sector N06 (Fig. 16) is similar to sector N04. A long

set of steps separate both the sectors. The northern and

southern borders of sector N06 are respectively also the

quincha wall and the sharp step of the rock. Between two

and four terraces step down the slope in a northern di-

rection. Since the slope of the rock also inclines to the

west, some terraces are split into horizontal sections sep-

arated by steps 10–30 cm high. The surface of the lowest,

northern most terrace is disturbed by a series of parallel,

natural cracks running east to west.

The eastern boundary of the sector is marked by the

place where the last traces of the quincha wall vanish. It

should be stated, however, that the easternmost section of

sector N06 shows continuation with neighbouring sector

N07. The division is therefore arbitrary in this case.

Sector N07

Four dierent areas constitute this sector (Fig. 17): an

area alongside the sector’s northern border, ve platforms

in the centre that step down to the west, two spacious plat-

forms located to the south, the southern edge with a se-

quence of narrow shelves and two water reservoirs(?), and

an irregularly shaped platform at the westernmost end of

the sector.

Fig. 14. Sectors N03 and N04

Fig. 15. Sector N05

90 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

Two natural cracks cut the whole area. The smaller one

runs alongside the northern edge of the sequence of ve

platforms. The much deeper, southern crack is a continu-

ation of a series of ssures running alongside the northern

slope of the rock, already known from sectors N04 and

N06.

The northern, irregular edge of the westernmost plat-

form has no parallel on the whole rock, as all other plat-

forms are trapezoidal or rectangular in shape with angles

nearing 90°. It might be that this platform did not get n-

ished, thus can be associated with the last stages of the

rock being carved. The question of whether this was dur-

ing the last pre-Inca phase or the last stage of Inca activity

remains open. A semi-circular recess in the eastern part of

the platform looks, however, like it has been carved inten-

tionally. Very unclear traces suggest that once this could

have been a circular (ca. 60 cm in diameter) protrusion

raised a few centimetres above the surrounding surface –

perhaps another petroglyph.

The levels of all platforms in sector N07 have been

shaped so that rainwater ows towards the northern,

lowest platform, and thus down the slope (Fig. 18). This

observation also applies to two recesses, one elliptical

and the other L-shaped, cut into the southern edge of sec-

tor N07. The inclination of the bottoms of these recesses

excludes the possibility that they could be reservoirs for

collecting rainwater – any water on the bottom of both

recesses would have immediately owed to the lower

platform. Such a situation leads to the supposition that

both recesses were supposed to remain dry, even if water

was in use here. It is not possible to point to more spe-

cic functions that both recesses could have served, but

they could have been related to rituals in the “Choir of the

Priests” located about 25 m to the west, or more likely,

with rituals in sector C06 adjacent from the south.

Since the southern edge of sector N07 cuts into an

earlier circular petroglyph, a fragment of which can be

found in sector C06, we can conclude that the platforms of

Fig. 16. Sector N06

Fig. 17. Sector N07

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 91

sector N07 as well as the recesses came from a relatively

late, probably Inca phase of carving the rock.

In the north-eastern corner of sector N07, an irregular,

roughly circular protrusion can be interpreted as traces of

another petroglyph whose original form cannot be recog-

nised today. Its diameter was probably up to 3 m.

Sector N08

Sector N08 (Fig. 18) is located on the edge of a steep

northern slope and in the northernmost part of the rock where

traces of intentional human activity are still preserved. It

is one of the most architecturally interesting sectors.

Three distinctly separate units can be distinguished

here. The most important is that of what is known as the

“Temple of Five Niches”. It was once a roofed building

measuring ca. 5 × 13.3 m. From the northern side, its front

wall is still preserved. The lower part of it survived only

as a 25–35 cm protrusion from the rock surface. The upper

part is a ca. 60 cm-thick wall constructed from unshaped

stones with local soil as a binder. Although partially

reconstructed, the stone wall is still visible in the western-

most part of the temple front.

A narrow, ca. 50 cm-wide passage in the centre of the

front wall leads to the single room (ca. 2.5 × 11.5 m) of

this building. Its oor is a carefully levelled rock surface.

Additionally, the back wall of the room has been carved in

the natural rock. A well levelled bench, ca. 30–40 cm high,

extends for the whole length of this wall. In the upper part

of the wall, ve niches (0.99–1.07 × 1.70 m) are regularly

distributed. The back wall of the niches is semi-circular,

so their depth varies from 45 to 75 cm.

Metrological analysis

of the width of the niches and

the distances between them indicates that in the process

Again, the cosine quantogram method was used for metrological

analysis.

of their planning, a basic unit of measurement was used

whose metric value corresponds to 54.5 cm. This obser-

vation ts with the results of the metrological research on

the Machu Picchu sanctuary, indicating the use of such

a measurement in buildings of high prestige [10], and

conrms assumptions about the Incan provenance of the

“Temple of Five Niches”.

The natural rock on the rear side of the room extends

30–50 cm above the niches, so the total height of the back

wall of the room reaches 2.6 m in some places. On the

top of this back wall is a 60 cm-wide shelf with shallow,

irregular elliptical depressions – apparently troughs for

stone blocks of the upper-most part of the back wall of the

room. However, no single block has survived, and we do

not have any indication of the height of this extension of

the wall. We can only estimate that it was at least 40–50 cm

– just enough to put canopy beams over the entire room,

since there is no doubt that the building was once roofed.

In his study, Rolando Marulanda [11, p. 47, Fig. 25]

came to similar conclusions as ours given above. How-

ever, his hypothetical reconstruction assumes the exist-

ence of a gable roof. Due to the lack of a typical channel

draining rainwater from the southern roof slope, such a re-

construction seems doubtful. A single-pitched roof seems

more likely – similar to the one proposed by Boero Rojo

[12, p. 90, Fig. 45]. However, the latter reconstruction does

not take into account the existence of a stone front wall

of the building, and Rojo erroneously interprets its traces.

Irregular elliptical depressions (troughs for stone blocks)

are visible on the top of the side walls of the room. Due

to the nature of the slope of the rock, the tops of the lower

parts of the side walls that are carved in the natural rock

are also sloping. As a result, the troughs for the stones in

the side walls are similar to roughly cut steps. However,

they should not be confused with them. The real steps are

on the east side of the building and run parallel to the line

of troughs.

Fig. 18. Sector N08

92 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

In front of the “Temple of Five Niches”, a spacious,

roughly horizontal, platform (5.5 × 14 m) has been cut into

the rock. In the centre, in front of the entrance to the room

with ve niches, traces of another circular petroglyph

have been identied, but the poor state of preservation

prevents any attempts to identify the nature of this petro-

glyph. From the northern side, the platform is limited by

a reconstructed stone wall standing on the edge of the cli.

The eastern border is another protrusion that is 30–40 cm

higher than the surface of the rock and separates this sec-

tor from neighbouring sector N09.

To the west, the platform borders a group of four ter-

races separated by a Z-shaped protrusion. The three north-

ernmost terraces were probably open to the sky, while the

last one most likely functioned as a separate room, cov-

ered by a roof. The only entrance was from the east, at

the corner of the room – a feature not typical for Incan ar-

chitecture. It should be noted, however, that the traces al-

lowing for the hypothetical reconstruction of the southern

and western walls of this room are not as apparent as in

the “ Temple of Five Niches”, so the existence of a roofed

room here cannot be taken for granted.

The most interesting element in this part of sector N08

is the southern wall of the highest terrace, which, accord-

ing to our hypothesis, was roofed. The wall consists of two

parts separated by a rectangular projection. The smaller,

western part is devoid of decorations. In the eastern part,

however, four small niches (ca. 25 × 28 cm) with charac-

teristic double recesses have been arranged symmetrically

(Fig. 19). Here, too, metrological analysis indicates the

use of an Incan unit of measurement typical for high-pres-

tige buildings – this time, due to the smaller dimensions

(13.3 cm) of the niches, this is yuku [13] – a fractional

part of the previously identied unit of measurement [10].

This observation can be considered as further conrma-

tion that the relics in sector N08 relate to the Incan phase

of occupation of El Fuerte de Samaipata.

Sector N09

Sector N09 (Fig. 20) consists of three separate areas.

Remnants in the northernmost area form a T-shaped pro-

trusion elevated 20–40 cm above the rock surface. This

separates the sector from sector N08 in the west and at

the same time divides the lowest, northernmost unit of

this sector into two parts. Several small niches or seats

(ca. 30 × 45 cm) are preserved in this protrusion. Six of

them face a bigger, rectangular area (ca. 2.5 × 6.0 m) and

two others point to the north, where much smaller and less

regu

lar space has been arranged. The form and the way

this T-shaped protrusion has been carved into the rock are

very similar to the bottom part of the front wall of the

“Temple of Five Niches”. Perhaps, here too, it served as

a foundation for a stone wall that stood on it. However,

clear traces allowing for a convincing reconstruction of

the shape of the rooms that could have been separated by

such a wall are missing. Besides, it is highly possible that

either the original intention was abandoned or the original

form of this protrusion was disturbed by later work in the

central part of the area.

There, two distinct groups of rock carving can be dis-

tinguished. The bigger, more westerly located carving

con sists of three terraces (ca. 1.0 × 6.5 m) cut into the

northern slope of the rock. The middle terrace has a more

complicated form than the two others (Fig. 21). Two plat-

forms (ca. 52 × 206 cm), each roughly 45 cm high, are set

symmetrically against the back of the terrace. In front of

them, ca. 27 cm-high, rectangular steps (51 × 82 cm) are

placed. The surface of the rock is well nished there and

the edges of the platforms and steps are sharp and well

dened. All this gives the impression that skilled stone-

cutters worked on this area. Metrological analysis of main

dimensions reveals again a unit of measurement typical

for Incan architecture of high prestige

The lowest of the terraces described above continues

to the east. Approximately 10–15 cm above the surface

of this terrace, ve semi-circular seats (each ca. 60 cm

in diameter) have been carved in its rear wall (Fig. 22).

The semi-circular shape is somewhat surprising, as is

their upper part, which gives the impression of having

never been nished. It is, therefore, possible that work in

this area was abandoned. Perhaps the seats were intend-

ed to be rectangular, and their present shape is a kind of

semi-nished product – the rst stage of carving them.

Since metrological analysis again indicates the use of

units of measurement typical of Incan high-prestige ar-

chitecture, one can risk the hypothesis that the work was

interrupted in the last phase of Inca presence at El Fuerte

de Samaipata.

On the slope above the terraces to the south, traces of in-

tentional carving are limited to only two rectangular water

This time, the results obtained were more unambiguous than in

the “Temple of Five Niches”. Measurements of 13.8 [13.5], 26.0 [27.0],

and 50.3 [54.0] cm were detected. Incan units of measurement observed

by Kubicka [10] in the central part of Machu Picchu are given in square

brackets.

Fig. 19. Sector N08. The northern façade of the “Temple of Five Niches”. Ortho-image from TLS

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 93

reservoirs (ca. 1.1 × 3.1 m and 1.25 × 3.35 m). Their depth

does not exceed 7–10 cm. The outlet of the rst of them

leads towards the steps separating sectors N08 and N09.

The second reservoir discharges water to the slope above

the terraces located in the central part of sector N09.

Sector N10

Sector N10 is located east of sector N09 (Fig. 23).

10 m-

long series of 26 steps, each less than 80 cm wide,

separates both the sectors. The eastern border of sector

N10 is an L-shaped wall decorated on both faces with

double-recessed niches.

The main feature of this sector is a group of three water

reservoirs located in the highest, north-western part. The

surface of the largest L-shaped reservoir exceeds 12 m

but the slope of its bottom and the elevation of the out-

let mean that it could accumulate at most about a quarter

of a cubic metre of water. Water overowing from this

reservoir ran to a much smaller one (0.9 × 1.3 m) located

on the eastern side, and then to another slightly larger one

(ca. 2.0 × 3.0 m). A small intermediate reservoir could act

as a sediment trap where all impurities owing down with

the water could accumulate.

The last of the reservoirs is also very shallow, and

taking into account the fact that its outlet is only 3–4 cm

above the bottom, it could not store more than about

150 litres of water. It supplied water to a carefully carved

channel on its eastern side. The preserved traces of this

channel allow its course to be traced to over a length of

ca. 4 m, but it was probably much longer. The bottom of

the channel has the form of a meandering snake. On both

sides of the channel are shelves about 25 cm wide on

Fig. 20. Sector N09

Fig. 21. Terraces in the central part of sector N09 as recorded with TLS

Fig. 22. Semi-circular niches in sector N09 as recorded with TLS

94 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

which a similar meandering pattern has been carved, but

this time as a bas-relief.

From the west, the lowest of the reservoirs is accom-

panied by a small platform (ca. 1.65 × 3.00 m) raised

around 0.5 m above the bottom of the reservoir. Poorly

decipherable traces preserved on the surface of this plat-

form may suggest that another zoomorphic petroglyph

once existed at this place [11, p. 38]. Due to the poor state

of preservation, attempts to reconstruct the form of the

petroglyph should be treated with extreme caution. It can-

not be excluded that the traces are the result of natural

weathering and erosion. This reservation also applies to

two round openings interpreted as oering holes – their

anthropogenic origin is not indisputably conrmed.

Marulanda also puts forward a hypothetical reconstruc-

tion of a building (a temple, as he calls it) erected on the

slope above the lowest reservoir and the accompanying

platform [11, Fig. 28, p. 50]. Similarly to the “Temple

of the Five Niches”, placing the walls of such a building

on a slope would require preparation of a proper base on

the ground (troughs for the lowest row of stones), so that

the foundation part of the walls would not slip down the

slope. No such traces can be found today, which makes

the whole hypothesis disputable.

Near to the reservoirs, almost entirely blurred edges of

subsequent recesses and small platforms or niches can

be observed. To the east of the reservoirs, there are also

traces of numerous oering holes and three circular petro-

glyphs. Their poor state of preservation prevents any in-

terpretation.

Traces of articially shaped surface are also extant in the

natural rock crack along the southern border of the sector.

Perhaps it was used as another channel to drain rainwater.

Sector N11

A few traces of pre-Hispanic activity can be found in

the area bordered from the west and south by an L-shaped

stone wall decorated on both faces with double-recessed

niches (Fig. 23). These include traces of two round petro-

glyphs in the western part that are hard to interpret, and

very eroded edges of two terraces in the eastern part. The

remains of a rectangular building (ca. 6–12 m) standing

in the central part of the sector were built during the His-

panic phase.

Fig. 23. Sectors N10 and N11

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 95

Central part of the rock

Sectors C01 and C02

Sector C01 (Fig. 24A, B) is located on the ridge of the

rock, east of a transverse wall decorated with double-re-

cessed niches. Three water reservoirs are located in the

westernmost, lowest part of this sector. Most probably, the

L-shaped northern one belongs to the later Incan phase of

carving Samaipata rock. It links with the sequence of shal-

low terraces in sector C02, from where it is also supplied

with water. In turn, the south-western reservoir, across

which the transverse wall decorated with double-recessed

Fig. 24. Sectors C01 and C02: A – the western part; B – the eastern part

96 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

niches has been built, is evidently of earlier origin and

might be associated with the remains of the circular petro-

glyph from sector W09. The third, south-eastern reservoir

has a semi-oblong form and is ca. 1 m deep. It could have

held nearly 10,000 litres of water if only the outlet in its

south-western corner was closed.

The last reservoir is part of a more signicant arrange-

ment called the “Big Snake” or “Rattlesnake”. The most

characteristic feature of this arrangement is three stripes,

each almost 30 m long, covered with rhomboid decora-

tions imitating the pattern on the back of a rattlesnake. Be-

tween the stripes, two channels have been cut into the rock

– both slightly shorter (ca. 26.8 m long), 36–30 cm wide,

and only 10–12 cm deep. The whole design is recessed

1–2 cm below the surrounding surface of the rock. This

creates additional shadows along the longer edges, em-

phasising the entire composition.

The rhomboid pattern is essentially a system of shallow

channels, 1–2 cm deep and ca. 2 cm wide. Water from the

reservoir (with a capacity of ca. 2,000 litres) located in the

highest, eastern part of the sector (Fig. 24B) could ow

into them. There are well-preserved traces of a shallow

channel facilitating supplying the middle stripe with wa-

ter. Slightly less well preserved is a small channel that

directed water to the southern stripe. Although the east-

ern end of all three patterns is well preserved, no traces

have been discovered of a channel leading to the northern

stripe.

The western ends of the stripes are in a much worse

condition. The northernmost is partially erased by shallow

terraces (and possibly steps) in sector C02, so there is no

evidence of where water running through it could have

been directed. The most extended stripe is the middle one.

It ends with a short channel directing water to the already

mentioned semi-oblong reservoir, as does the southern-

most one.

It is worth noticing that the geometry of each of the

rhomboid patterns diers, resulting in a dierent number

of rhombuses in each. Supposing that the western extent

of all stripes ended at more or less the same position, there

would have been 85 rhombuses in the northern stripe,

101 in the middle, and 81 in the southern one. These dif-

ferences could have had a signicant role if one assumed

that the whole system could have been used for magical

divination purposes. By knowing the number of shapes

in each of the patterns, it was possible to predict in which

of them water would appear rst at the western end when

simultaneously poured into their eastern ends.

Sector C02 (Fig. 24A, B) runs parallel to the northern

border of sector C01 described above. It consists of two,

and in some places three, levels of terraces located on the

southern slope of the rock. They descend towards the east.

Fig. 25. Sectors C03 and C04

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 97

The largest of them, located in the eastern part at the base

of the wall with double niches (Fig. 24A), has an elon-

gated, complicated shape and several interesting features.

Along its northern edge runs a channel draining water to

the next, lower level of terraces. A similar channel runs

along the southern border of the sector, but this time it

discharges water to the already mentioned L-shaped res-

ervoir. There are faint traces between both channels that

can be interpreted as relics of an earlier petroglyph that

existed here before this terrace was cut into the surface

of the rock.

Sectors C03 and C04

Both sectors are located east of sectors C01 and C02.

The western part of sector C03 (Fig. 25) might be con-

sidered as a continuation of sector S12, but its character

is dierent. There are still two terraces descending south-

wards, but there are none of the double-recessed niches

that are characteristic for sector S12. This is also where

the petroglyph in the form of a meandering snake ends

after running for more than 20 m alongside the middle ter-

race of sector S12. Further to the east, the arrangement of

terraces is less regular, and the original layout is dicult

to identify due to the highly eroded surface of the rock and

many natural cracks that pass through this area. It is, how-

ever, evident that there, on a bigger platform located in the

eastern part of this sector, a long channel begins that runs

to the west alongside the northern border of sector S11.

Sector C04 occupies the northern part of the ridge of

the rock and descends in a northern direction in two ter-

raced steps. The most characteristic features are four tri-

angular recesses (or possibly seats) facing the north and

an oval reservoir located to the west of them. The recesses

have the shape of equilateral triangles with sides about

1.5 m long. The inside and outside corners are chamfered.

Although such recesses can be found in other sectors (for

example, W05, S10, S13, S14, S29, E01, and E02), their

character diers from the forms typical for Incan architec-

ture. Therefore, they might belong to an earlier phase of

the carving of Samaipata rock.

The oval reservoir has a diameter of about 2 m, and its

depth does not exceed 30 cm. The shape of the rock surface

in the immediate surroundings excludes the idea that this

place could have acted as a reservoir for rainwater ow-

ing there. Its bottom part is also slightly inclined towards

the north, where a channel leading to the lower terrace is

located. It could have served as a water reservoir only if

this channel was blocked o and water was poured there

manually.

Sector C05

Sector C05 (Fig. 26) is situated on the ridge of the rock,

almost at the highest part of it. It is here that the most fa-

mous, symbolic petroglyph is located – the “Choir of the

Priests”. The main element of the whole composition is

a circular channel with an outer diameter exceeding 7 m

and a width of just over 1 m. It surrounds a central, circu-

lar section (diameter of about 4.75 m) which at the highest

point is raised about 75 cm above the bottom of the chan-

nel. The outer edge of the circular section is decorated

with nine square niches or seats (roughly 65 × 65 cm)

cut nearly 40 cm into the rock. On the vertical surface

between each niche, traces of small (15 × 15 cm) square,

double-recessed niches are extant.

Outside the circular channel, there is an alternating pat-

tern of nine square (also ca. 65 × 65 cm) and nine triangular

seats (or possibly recesses) that could have had a decorative

function. The latter are in the form of equilateral triangles

with sides of ca. 75 cm and they are embedded less deep-

ly into the rock surface than the square seats. The whole

arrangement, when looked at from above, resembles a so-

lar symbol, although solar symbology is not particularly

Fig. 26. Sector C05

98 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

common in this region of South America. Much more typ-

ical for the area would be a representation of a feline. Very

weak trace of what once might have been such a symbol

preceding the sculpturing of the “Choir of the Priests” can

be seen in the central part of the sector (Fig. 27).

The bottom of the circular channel is inclined slightly

northwards so that the water that accumulates there ows

into a channel drilled below the surface of the rock. The un-

derground part of the channel is over 4 m long, and it does

not follow a straight line. This observation may explain the

function of three oval openings spaced evenly along this

channel – each of them with a diameter of at least 75 cm.

They were to facilitate drilling short, ca. 50 cm long, sec-

tions of the underground channel between the openings.

They may have also provided some ritual functions in the

celebrations performed here.

The channel leads to a T-shaped reservoir located in

the north-east of the sector. Its shape is connected to Pan-

American ancestral cosmological symbolism and is par-

ticularly common in the Amazonian area

. It is possible

that this reservoir that collects the water coming from the

“Choir of the Priests” was modied, and it erased some

older gures in bas-relief, of which we can still identify

One of the most characteristic examples can be found in Pantia-

colla in Peru.

some traces in its deepest parts (on its western and eastern

sides). A shallow meandering channel has been cut in the

bottom of this reservoir to direct the ow of water to its

north-east corner. Several round holes (10–30 cm in diam-

eter) are extant east of the “Choir of the Priests”. Further

to the east, very light traces of a cross-shaped water res-

ervoir are still preserved. Additionally, close to the border

with sector C06, a very low oval protrusion suggests the

remains of yet another petroglyph.

Sector C06

Two large water reservoirs are the main feature of this

sector (Fig. 28). From these, water drains to either the

southern or northern slope of the rock. On the bottom of

the northern, L-shaped reservoir, highly eroded traces of

two (or possibly only one) petroglyph are visible.

Fig. 27. Section A-A across the “Choir of the Priests”

Fig. 29. Sector C07

Fig. 28. Sector C06

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 99

The most interesting feature of this sector is evidence

of a large sculpture of a snake that separates both the res-

ervoirs. It has been modied, probably during the Inca

occupation, into a sequence of small niches or seats to

hide the original ophidian form. Three of the seats are

facing towards the north, and another three southward.

Once, the snakelike egy had its head directed to the

east, but it has been intentionally(?) erased after its shape,

as a symbol related to the cult of water, was deformed by

the abovementioned seats. It is possible that the faint trac-

es of the snake on the opposite, western side of the gure

are also its erased parts.

Remains of the one of the biggest round petroglyphs

(ca. 3.75 m in diameter) occupy the northern part of this

sector. It was split in half when platforms of neighbouring

sector N07 were sculptured.

Sector C07

Sector C07 (Fig. 29) lies where the top of the rock be-

gins to drop slightly eastwards. Only very faint traces of

one circular petroglyph on the border with sector N07,

and four, small rectangular platforms in the centre part,

remain. All of them, judging from their state of preserva-

tion and interference with evidently latter carvings, can be

attributed to the earlier, possibly pre-Inca phases of carv-

ing Samaipata rock.

On the southern margin of this sector, on the southern

slope of the rock, traces of six slightly better preserved

niches are extant. They seem to form two separate groups.

The three niches on the western side are roughly of the

same dimensions. The other three are narrower, but the

distances between them are wider. Both groups were

probably carved during two dierent phases, but there is

neither evidence of the sequence of their execution, nor

any hints to attribute them to the Inca or pre-Inca period.

Southern part of the rock

Sectors S01–S05

Sectors S01–S05 are located on the at area at the foot

of the southern slope of the rock. Four housing complex-

es can be identied there. They were examined in detail

during work by the German Samaipata Archeological

Research Project (Proyecto Arqueológico en Samaipata,

PIAS) [9]. According to the results of these investigations,

all of the housing complexes, except the central, U-shaped

group of buildings that is dated to the post-Conquest pe-

riod, were built during the Inca phase of occupation. All

these sectors are outside of the scope of this study, which

is limited to Samaipata rock itself.

Sector S06

The architectural structures that constitute sector S06

are carved in the lithic wall at the base of the southern

slope of Samaipata rock (Fig. 30). They consist of ve

units that evidently were once roofed (Fig. 31). The roof,

Fig. 31. Section A-A across sector S06

Fig. 30. Sector S06

100 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

today convincingly reconstructed, rests on one side on

a wall built on the slope of the rock. The front side of the

roof is supported by single row of wooden posts.

The three central units, measuring roughly 4 × 2 m,

have a very similar layout. From the west and south, the

units are enclosed by an L-shaped wall. A more than 2 m

wide entrance provides access to each room. On the back

wall, opposite these entrances, rectangular niches (75–80

× 160–167 cm) have been carved. These are accompanied

by a pair of smaller (roughly 50 × 50 cm), almost square

niches or blind windows, under which platforms protrud-

ing ca. 25 cm from the oor level are located. It seems that

originally the oors of these units as well as the platforms

were sculptured directly into the rock. In the same way,

the L-shaped walls separating individual units were exe-

cuted, or at least their lower parts. The upper parts might

have been built as stone rubble walls.

The easternmost and westernmost units show dierent

layouts. The westernmost is smaller (only ca. 2 × 3 m),

and it lacks a front wall. On the centre of its back wall is

one rectangular niche without a platform in front of it. The

easternmost unit has undergone several modications.

From the west and south, it is enclosed by an L-shaped

wall analogous to the kind already described. However,

there is no access from this side. It has been blocked by

a later stone rubble wall and a new entrance has been

arranged from the east by pulling down a stone wall that

once existed there. The negative of this wall is still vis-

ible on the back wall of this room. The presence of the

platform in the north-western corner of the room suggests

that the back wall might have been arranged in a similar

way as in the three central units. Unfortunately, it has been

badly damaged by treasure hunters, so its original layout

cannot be unambiguously reconstructed. Originally, there

was clearly no large rectangular niche here, as can be seen

today. Preserved traces suggest one, ca. 2 m long and

50 cm high, niche placed ca. 80 cm above the mentioned

stone platform. This niche might have had extensions in

its lower part, so its shape resembles the letter “T”. The

lower part of the back wall is so badly damaged that such

reconstruction must remain only hypothetical.

Rectangular-shaped niches, and in some cases with

a kind of enthasis, are not typical for Incan architecture.

This kind of form is more related to Tiwanacota tradition

than to Inca. This whole sector is often referred as the

“Sacristy”.

Sector S07

The main feature of this sector is a ca. 80 m-long ramp

enabling movement between western and eastern sectors

of the rock (Fig. 32). It runs just above sectors S06, S08,

and S09, and connects two ights of steps – the western

Fig. 32. Sector S07

Fig. 33. Sector S08

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 101

one located between sector W011 and S07, and the eastern

steps on the border between sectors S09 and S23. This

ramp, also played an important role in draining rainwa-

ter. Along its entire course runs a channel collecting water

from the terraces located above.

Sector S08

Sector S08 is located at the foot of the southern slope

of the rock (Fig. 33), east of sector S06. It seems that the

sculpting of this part of the rock began after the comple-

tion of sector S06, but this is only a technological phase;

it does not necessarily reect subsequent cultural chang-

es. From a formal point of view, sector S08 presents an

almost vertical wall, over 3 m high and almost 25 m long,

with 10 roughly rectangular niches carved at equal inter-

vals (Fig. 34).

The niches are slightly bigger than those of sector S06.

Their width ranges from 105 to 112 cm while their height

is between 180 and 200 cm. Their shape is however not

uniform. Some are nearly rectangular, in some cases we

can see a kind of enthasis in the middle, and one of them

is trapezoidal in shape. This variety of shapes does not

point to any specic period, but the overall impression is

similar to that of sector S06. Above this set of niches, the

long ramp leading from sector S07 continues in an eastern

direction

Sector S09

Sector S09 (Fig. 35) can be divided into two zones:

the lower one located at the foot of the southern slope of

the rock and the upper one stretching up to the long ramp

constituting sector S07.

In the lower zone, which is an almost vertical wall

2.3–3.5 m high, three separate groups of niches can be

distinguished (Fig. 36). The rst group adjoins the almost

11 m-long series of steps connecting the foot of the rock

Fig. 34. Front view of sector S08

Fig. 35. Sector S09

Fig. 36. Front view of sector S09

102 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

with the eastern end of sector S07. They are four small

(68–72 cm wide and 77–81 cm high) double-recessed

niches of a trapezoidal shape, which is characteristic of

Incan architecture. The front wall where these niches are

located is almost 6 m wide and ca. 2.3 m high, and n-

ishes with a nearly horizontal ca. 1.25 m-wide platform

from which the next slightly inclined wall extends by

ca. 1.6 m. Five smaller (49–54 × 54–58 cm) niches are

placed symmetrically on this wall, conversely to the lower

row of niches where only four are extant. This asymmetry

might be the result of a later modication in this sector.

In the central part of the 2.5–3.5 m-high wall, the front

wall is retracted by more than 1 m and three large (1.56–

1.59 × 2.35–2.45 m) double-recessed niches have been

carved here. They are rectangular rather than trapezoidal

and all of them have been altered in some way by robbers.

Nevertheless, the niches have the look of being never ful-

ly completed.

Further to the west, the front wall of sector S09 is again

retraced by ca. 0.5 m. Another two double-recessed niches

are placed there. Their width corresponds with the niches

from the central part of the wall, but they have a slightly

smaller height of between 2.29 and 2.34 m. Their shape is

even less regular, and the inner recesses also look unn-

ished. Perhaps the carving of the front part of this sector

was interrupted by the collapse of Incan control related to

the arrival of the Spaniards.

In front of all these three groups of niches are three at

platforms carved in the solid rock. From the south, they

are bordered by a channel draining rainwater. All three

platforms are located approximately 2 m lower than the

area directly to the west (sector S08). The dierence in

levels is overcome by narrow (ca. 60 cm wide) steps con-

necting both sectors.

The upper zone of sector S09 consists of a sequence

of irregular terraces that climb up a further 5 m until they

reach the eastern end of the ramp in sector S07. At least

ve platforms are furnished with niches or seats that vary

in size. Four of these platforms are in the eastern part of

this zone, which is separated from its western counterpart

by a long, ca. 4.5 m, sequence of steps that also leads to

the ramp in sector S07. It seems that these steps may be

from an earlier time than the three groups of niches in the

lower zone, which are clearly from the Incan period. The

inclination of these steps may correspond to the original,

virgin slope of the rock from the period before rst terrac-

es and platforms were cut in.

It looks like in this area, the base of the rock underwent

decisive changes, most probably in the Inca epoch. How-

ever, the possibility that some of these modications were

made during the Chané occupation of the site cannot be

excluded.

Sector S10

Another unnished part of Samaipata rock can be

found in sector S10 (Fig. 37), directly to the east of the

steps separating this sector from sector S09.

There might have been similar steps on the eastern bor-

der of sector S10, but the highly eroded rock surface in

this area render a convincing interpretation impossible.

A room more than 13 m long and 2 m wide is located in

the lowest, southern part of the sector. The room was cut in

into the slope of the rock, so its roughly horizontal oor is

solid rock. The lower part of its southern wall, which is more

than 1 m wide, was sculptured into the solid rock. There are

no indications as to whether the upper part of this wall was

ever built, but evidently the intention was to build it, like

in other cases, from unshaped stones set on clay mortar.

The entrance, less than 70 cm wide, that leads to this

room is decorated with double-recessed jambs that point

to an Incan origin and the importance of this room. Two

shallow steps are placed in the entrance passage.

Fig. 37. Sector S10

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 103

On the back wall of the room, which is nearly 3 m

high, several vertical grooves (6–7 cm wide) are placed

in nearly equal intervals ranging from 95 to 105 cm. The

most plausible interpretation is that the grooves mark the

edges of niches that were planned to be carved there but

were never nished. A small bench (ca. 30 cm wide and

ca. 40 cm high) runs alongside the base of this wall. On

the vertical face of the bench, a sequence of small square

(ca. 17 × 17 cm) niches has been carved – nearly perfectly

in between the ends of the grooves above (Fig. 38). Mis-

alignment does not exceed 1.5 cm.

When applying the cosine quantogram method for ana-

lysing distances between grooves marking the edges of fu-

ture big niches and the distances between the centres of the

small square niches below, we nd a unit of measurement

of 50.4 cm, which might have been used to lay out their

plan (Fig. 39). According to Kubicka’s studies on Machu

Picchu metrology [10, p. 171–175], units of measurement

between 50 and 55 cm are typical at Machu Picchu for

high-prestige buildings. This observation ts well with the

entrance adorned with double-recessed door jambs.

On the back wall of the room, which is nearly 3 m high,

several vertical grooves (6–7 cm wide) are placed in near-

ly equal intervals ranging from 95 to 105 cm (Fig. 38).

The most plausible interpretation is that the grooves mark

the edges of niches that were planned to be.

We can presume that this room was planned to serve

some important functions. As such, it was undoubtedly

intended to be roofed – probably with a one-pitched roof

sloping south. The shelf, almost 1 m wide, above the back

wall of the room could have been successfully used to

support rafters whose other ends would have rested on the

front wall. This room has several features that suggest it

was built during the Inca occupation of the site. The fact

that it was never completed suggests that construction

may have been interrupted due to the fall of Inca rule in

these areas.

Two terraces, nearly 15 m long, located in the upper

part of sector S10 are decorated with rows of triangular

recesses. Although triangular recesses are not foreign

to Incan architecture [14], these decorative patterns are

probably of earlier, pre-Incan origin. The lower terrace is

decorated with 17 small (ca. 50 cm wide and 30 cm deep)

triangular niches. A total of 18 much bigger and deeper

(ca. 80 × 70 cm) niches decorate the upper terrace.

Sectors S11 and S12

Sectors S11 and S12 lie in the central part of the south-

ern slope of the rock, to the north of sector S07 (Fig. 40).

Together with the latter, they form three long, east-rising

ramps connecting sectors located in the western and east-

ern parts of Samaipata rock.

Fig. 39. Cosine quantogram

analysis of niches in the back

wall of the room in the lowest

part of sector S10.

Green areas mark the Incan units

of measurement found by

A. Kubicka [10, pp. 171–175]

as typical for high-prestige

buildings at Machu Picchu

Fig. 38. The back wall of the room in the lowest part of sector S10

104 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

The steeply rising, over 38 m long, sector S11 is di-

vided by a series of terraces that help overcome a height

dierence of over 5 m between the sector’s eastern and

western ends. Along the rear wall of this sector, which is

over 2 m high, there is a channel draining rainwater from

sectors C03 and C04. Next to the series of steps consti-

tuting the western border of sector S11, this channel con-

nects to its counterpart in sector S07.

Sector S12 (Fig. 40) lies ca. 1.8 m higher than sector 11.

It has a similar length to its southern neighbour, but it is

much wider. In the middle, its width exceeds 4.5 m. A series

of ve narrow (1–1.5 m wide) irregular platforms rising east

have been cut into the southern edge of the sector, except

in the western end. Another platform (0.5–1.3 m wide) runs

alongside the sector’s northern border, at the foot of a near-

ly vertical wall ca. 2 m high that separates this sector from

the ridge of the rock (sector C01). This platform rises not

more than 15–20 cm above the central part of the sector.

Along the southern edge of this platform runs a narrow, at

least 21 m long, meandering channel resembling a crawl-

ing snake. Originally, it was probably much longer, but its

western part has vanished due to rock erosion.

There are 34 square, double-recessed niches evenly dis-

tributed on the vertical surface of the already mentioned

back wall. They are all severely eroded, so their dimen-

sions can be only roughly estimated as around 30 × 30 cm.

The average distance between the niches is 63.7 cm, which

with a good approximation corresponds to three values

of the Incan yuku unit of measurement

reconstructed as

12–14 cm [13]. This observation, along with the distinc-

tive, double-recessed niches, may again indicate Incan

origin. At the western end of sector S12, alongside its

southern edge, are four round openings (ca. 13.5 cm in di-

ameter) very reminiscent of the holes for the posts of the

quincha wall located on the northern slope of the rock. Per-

haps a similar structure was also erected here.

Sector S13

Sector S13 (Fig. 41) is located to the north of the cen-

tral part of sector S07, and consists of several roughly

horizontal platforms elongated in an east–west direction.

The lowest platform, which is not more than 1 m wide,

runs alongside the southern border of the sector and is

elevated roughly 2 m above the surface of sector S07.

Above the western end of the lowest platform, three seats

(ca. 90 × 60 cm) are overlooking the plaza in front of the

“Sacristy” (sector S06). The seats are separated by pro-

trusions ca. 1.15 m wide projecting from the back wall of

the platform.

The uppermost platform borders sector S14 and consist

of two sections. The most interesting is the western one.

Its distinguishing feature is a set of ve triangular recesses

(or possibly seats) attached to each other. Small (30 × 60 cm)

steps are in front of each seat. This part of sector S13 is

Fig. 40. The western (A) part and eastern (B) part of sectors S11 and S12

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 105

likely to be of an earlier, pre-Incan origin but triangular

niches are also known in Incan architecture [14].

The eastern section of the northernmost terraces con-

sists of small platforms with two sets of small, triple seats

– one in the upper, north-eastern corner and the other in

the lower, south-western corner of the platform. Between

the uppermost and the lowest platforms, there is an irreg-

ular arrangement of single and double seats – all looking

to the south.

Sector S14

Sector S14 is located up the rock, directly above sector

S13 (Fig. 41). A channel that diagonally crosses the slope

from the north-eastern corner of this sector and terminates

with a small cascade at the westernmost end divides the

sector into two parts. At the western end of the channel,

water runs down from a height of almost 2.5 m into a shal-

low pool at the foot of the rear wall of sector S07.

The most spectacular feature of the upper part of this

sector is the well-known representation of two felines (pu-

mas?) facing each other. A small (45 × 70 cm) rectangular

depression above the heads of the felines resembles the

face of a creature (possibly a feline head as seen from the

front) from whose mouth originates a channel leading be-

tween the heads of both pumas

Two at gutters 65 cm wide border the “Double

Puma(?)” petroglyph on its western side. Both of them ter-

minate with two elliptical pools along the channel run-

ning below the petroglyph. Between the gutters, a ca. 35 cm-

wide ight of steps has been cut into the rock surface.

Similar steps might have once existed further to the west,

but the degree of rock erosion does not allow for unambig-

Cf. J. Kościuk, M. Ziółkowski, M. Pakowska, Formal and ico-

nographic analysis and interpretation of the most damaged petroglyphs,

in the same issue of “Architectus”.

uous interpretation there. West of these gutters and steps,

alongside the channel, a series of irregular platforms and

seats are extant.

It is dicult to unequivocally interpret whether both of

the elements described above (“Double Puma(?)” petro-

glyph and two gutters with accompanying steps) were

carved at the same time, or were created and functioned in-

dependently. Nevertheless, both of the elements had a very

great signicance in the liturgical life of Samaipata, as can

be attested by their predominant position on the southern

slope of the rock, in front of the terraced plaza located below.

The south-eastern section of sector S14 (below the

diagonal channel) consists of three rows of platforms.

The most spacious (ca. 1.20 m deep) is the southern one.

A nearly 1 m-wide protrusion divides it into two equal

parts. Three narrow (ca. 50 cm wide) platforms extend

west of it. The highest platform is only partially preserved.

The diagonal channel cuts its western part.

The middle platform has the form of a set of ve trian-

gular recesses (or possibly seats). They are slightly smaller

than those in the neighbouring sector S13. It is possible that

this sequence of triangular recesses once extended further

to the west, but later on, one of the platforms neighbouring

from the west (that split by the channel) was cut into it.

The last, northernmost platform consists of four rectan-

gular seats separated by narrow (ca. 20 cm wide) protru-

sions. The area gives the impression of being of an earlier

origin. At the western end of this platform, there are traces

(just the corner) of the next seat, but the diagonal channel

has disturbed it.

To summarise, one can speculate that the oldest part of

this sector is the set of triangular recesses and perhaps the

platform above them with the series of rectangular seats.

The next chronological group would be the set of three

narrow platforms, the northernmost one cutting into the

triangular recesses. The diagonal channel that disturbs

Fig. 41. Sectors S13 and S14

106 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

both the set of rectangular recesses in the uppermost

platform and the three narrow platforms on the western

side probably constitutes the third chronological phase.

The chronological relationship between the “Double

Puma(?)” petroglyph (together with the two gutters and

steps to the west) and the diagonal channel is not clear. Since

the cult of water and venerating felines go well together in

Andean cultures, both the parts are ideologically and func-

tionally combined. However, there is no physical evidence

conrming that both were actually executed together.

Sectors S15 and S16

Sectors S15 and S16 (Fig. 42) border sectors S13, S14,

and sector C03 in the west. The eastern edge of both sec-

tors is a 12 m-long series of steps leading from the top of

the rock down the slope. However, these steps reach nei-

ther the base of the slope nor even the ramp of sector S07.

They could have served to enable movement between the

signicant petroglyph the “Choir of the Priests” located

in sector C05 on the top of the rock and sectors S15–S18

below, thus they would have played an essential role in the

rituals performed there.

Sector S15 is a complex system of platforms, seats, and

recesses. It does not give the impression of being a homo-

geneous creation, but looks rather like the result of many in-

dependent activities not necessarily far apart in time. How-

ever, among this rich diversity of forms, some elements

are particularly striking due to their thoughtful layout.

Two symmetrically placed rectangular seats separated

by a protrusion nearly 1 m wide are extant in the lowest

part of the sector. Above these is another set of ve trian-

gular recesses (seats?) with a broad platform in front of

them. West of this platform is an even broader platform

with four nearly uniform rectangular seats placed above it.

The last interesting feature is a small platform located di-

rectly to the west of the two symmetrical, rectangular seats

in the lower part of this sector, just on the edge of a 1 m-high,

nearly vertical rock fault. Two recesses were added to the

back of this platform, giving the platform an unusual T-like

shape. The uppermost platforms bordering sector S16 no

longer give the impression of a thoughtful layout, although

they are also equipped with one or two rectangular seats.

Sector S16 (Fig. 42), located further up the rock, bor-

ders sector C05 in the north and it does not have a consist-

ent layout. Only in its south-eastern corner is one deliber-

ately planned feature. Three rectangular seats are placed

30 cm above two narrow, only ca. 35 cm-deep platforms,

which look like they may have been wide steps.

Sectors S17–S20

Sectors S17 and S18 (Fig. 43) are on the eastern side of

the long steps leading down from sector C05. In the east,

they border two dierent sets of steps separating them

from neighbouring sectors S19 and S20.

Sector S17, located further in the south than sector S18,

consists of several platforms of dierent sizes and shapes.

However, it can be divided into three groups showing some

common features and traces of a more orderly layout.

Two small trapezoidal platforms separated by a cen-

trally placed 1 m-wide projection are located on the steep

Fig. 43. Sectors S17 and S18Fig. 42. Sectors S15 and S16

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 107

edge to the north of the long ramp of sector S07. West

of these, the same design is repeated twice, although on

a much smaller scale.

Each of the two terraces lying above the trapezoidal

platforms is divided into unequal parts by very eroded

ledges 60 cm wide – perhaps relics of two steps used to

negotiate the relatively large dierence in levels between

the terraces. A similar ledge separates two spacious plat-

forms located above these terraces. To the west, four trian-

gular seats with two descending platforms in front of them

is the next important feature of sector S17. Particularly

interesting is, however, the eastern platform. Three deep,

rectangular recesses are set into the wall behind the plat-

form at regular intervals. At the same time, the southern

edge of the platform is raised ca. 15–20 cm.

Using analogies with some constructions at the foot of

the southern slope of the rock (sectors S06, S27, and prob-

ably W11) and with the “Temple of Five Niches” (sector

N08 on the northern slope), one can conclude that this

platform was once walled up and probably also roofed.

In this case, the traces on the eastern and western sides

of this platform should be interpreted not as steps, but as

foundations for the upper part of the structure built of rou-

ble stones.

Located to the north of sector S17, sector S18 (Fig. 43)

is an example of a homogeneous complex with a clear

layout. Its main feature is the set of three equal rectangular

seats raised above the platform in front of them. Two deep

recesses on both sides complete this symmetrical design.

An irregular hole cut into the rock surface in front of the

middle seat seems to be a later addition.

Sector S19, situated east of sector S17 and to the north

of the long ramp of sector S07 (Fig. 44) has a very uni-

form plan. In its lower part, a narrow strip running from

north to south and raised above the neighbouring plat-

forms separates three pairs of terraces. Located above

this, another pair of platforms is devoid of such a parti-

tion. It appears again in the highest part of the sector, but it

is much broader here and both terraces dier signicantly

in width. Except for a small recess in the south-western

corner of the lowest terrace, no other special features like

seats or benches can be found in this sector. Most likely,

the whole of sector S19 was created from one concept,

and it probably comes from the same period as the adja-

cent sectors of S17 and S21.

Sector S20 (Fig. 44) shows a much more appealing

layout. A big seat measuring ca 1.90 × 2.60 m can be

found in the centre of the back wall of its only platform.

The sides and back of this seat are raised, so it resembles

a large stone armchair. Behind it, a deep rectangular re-

cess has been cut into the rock. Two pairs of rectangular

recesses (ca. 70 × 90 cm each) have been cut into the back

wall symmetrically on both sides of this undoubtedly cer-

emonial seat. This symmetry is additionally emphasised

by steps anking the platform on both sides. They en-

abled movement to sector C05, where one of the most

important elements of the rock is located – the “Choir of

the Priests”.

In the middle of the eastern part of this sector, there is

a small (ca. 1.2–1.5 m) space. It is adjacent to small steps

separating this sector from neighbouring sector S22. The

front and western edges of this area are raised, and the

only passage is in the south-western corner. In the back

wall of this “room”, a single recess is centrally placed.

Its size is similar to the recesses adorning the ceremonial

seat. Probably, this “room” was associated with some cer-

emonial rites performed on the platform.

Sector S20 is one of the best sectors for allowing us to

understand the principles of composition that the builders

were guided by, and their attempts to construct possible

functional relationships with neighbouring elements. It is

also preserved in almost complete shape. A small recess

in the south-western corner of the sector may have been

added later. The south-eastern corner was also slightly

modied by a large platform in sector S22 that was anoth-

er later addition.

Sector S21

Sector S21 (Fig. 45) occupies the east–west extended

area at the eastern end of the long ramp of sector S07. It

is accessible by the steps leading down from sector S22

located further to the north. This sector does not have any

regular layout. It consists of sequences of many irregular

terraces, platforms, and recesses. Its present state is rather

the result of many transformations that have altered the

southern slope of the rock.

However, two parts of this sector are particularly inter-

esting. The rst is a trapezoidal platform located across

from the beginning of the steps leading up from the foot

of the rock. Three rectangular recesses have been sym-

metrically cut into its rear wall. This part of the whole

complex distinguishes itself as it seems to have a regular

Fig. 44. Sectors S19 and S20

108 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

plan. Chronologically, it was created during the most re-

cent – probably Incan – phase of rock transformation.

It was cut into an earlier platform located in the west and

disturbs one of its recesses. Possibly, it also intruded into

a narrow rock ridge leading to the north, located just be-

hind the central recess of the platform in question. This

ridge might be the last remains of earlier steps that once

led down from sectors C06 and C07.

The other interesting part is a ca. 8.5 m-long trapezoidal

platform in the north-eastern corner of the sector. Ten nich-

es of a semi-elliptical shape, untypical for Incan architec-

ture, decorate its back. Close to the western end, eroded re-

mains of what might have been once a small (45 × 80 cm)

petroglyph projects from the surface of the platform.

Sector S22

Sector S22 (Fig. 46) lies to the north of sector S21, and

from the west it borders sector S20. Its layout is dominat-

ed by a spacious (4.5 × 7.5 m) triple-recessed platform. Its

shape bears no similarity to anything else on Samaipata

rock and may resemble half of the twelve-cornered form

commonly known as the Chakana or Andean Cross.

This motif was frequently chiselled in stone or embed-

ded in ceramics and textiles. It can be found in many An-

dean civilisations, including Marcavalle and Pukara cul-

tures, the Tiawanako and Wari civilisations, and nally,

Inca culture, so alone cannot be considered as providing

a conclusive date of construction. How ever, the fact that

this platform cuts into an earlier, well planned and homo-

geneous sector S20 points to a later origin – most likely

that of Inca occupation of the site. An interesting observa-

tion that may also inuence our understanding of the rela-

tive chronology of the site is the fact that the principal axis

of this platform is parallel to the steps that separate sectors

S19 and S21. The orientation of some terraces adjacent to

these steps also corresponds with this axis. Although not

decisive alone, it may suggest a close temporal relation-

ship for all these elements.

Further to the east of Chakana-like platform lies another

small unit with a typical layout. Its plan resembles a small

room rather than any other platform known from Samai-

pata rock. A central passage leads to the C-shaped unit,

which is surrounded by raised sides upon which rouble

stone walls might have once been erected. However, no

traces of such are in evidence to support this hypothesis.

Four terraces of an earlier origin are located on two

parallel platforms (each about 1 m deep) above the Chaka-

na-like platform and C-shaped room.

Fig. 45. Sector S21

Fig. 46. Sector S22

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 109

Sector S23

Sector S23 (Fig. 47) extends to the east of the steps that

separate it from sectors S07 and S09. It consists of several

smaller units – each of them of a dierent layout, so as

a whole, it cannot be treated as one coherent complex. The

single name of “sector S23” was given to all these units

only for ease of description.

A long terrace divided into ve platforms extends

along the southern border of this sector. A small channel

collects water at its base. On the back of this terrace, one

can distinguish four dierent units. The rst from the west

is a set of six semi-circular seats (roughly 30 × 50 cm in

size) embedded into the rear wall of the rst two plat-

forms. Their axial distances range between 79 and 81 cm,

and might t well with the Incan unit of measurement

sikya [13], [15]. Further to the east, two benches over 3 m

long and ca. 45 cm deep are placed on top of each other.

They are framed on both sides with roughly 1 m-long sec-

tions of rear wall without any decoration. The next unit

consists of two pairs of short benches (ca. 90 cm wide and

30 cm deep), again on top of one another. This sequence

of dierently arranged seats ends with an elliptical recess

framed with two buttresses projecting from the rear wall

of the terrace. This recess, which looks more like the low-

er part of a niche than a regular seat, is orientated in a dif-

ferent direction to the seats already described. Its line of

sight orientation points in the same direction as the seats

of sectors S15–S20 located further to the west.

The back wall of the remaining, eastern part of the low-

est terrace of sector S23 is curved and follows the natural

shape of the original rock surface at this part. As a con-

sequence, the line of sight axes of four roughly square

(ca. 60 × 60 cm) seats cut into the back wall intersect just

at the lower, southern border of the sector. We do not

know whether this had any signicance or if it was only

the result of natural conditions, but it bears a distant sim-

ilarity with the circular “Orchestra” from the western part

of sector S25 (Fig. 49).

Behind these four seats lies another curved platform.

Its backside is equipped with three curved benches with

the same orientation as the lower seats. The easternmost

bench disturbs the earlier, now completely indeciphera-

ble, remains of an oval petroglyph in bas-relief. Back to

the west, three rectangular seats are cut into the back of

the platform. On the western side, they are anked by yet

another bench. Between them and the two pairs of bench-

es below, there is long and narrow protrusion from the

surface of the platform. Its surface, as well as the surface

of the western end of this platform, is rough and diers

considerably from the relatively smooth and even face at

the eastern part. One of the possible interpretations as-

sumes that the work of shaping the platform was never

completed there. This, in turn, may imply that this part

of the sector dates back to either the arrival of the Inca or

the end of their occupation of the site. The orientation of

all the seats in this sector, consistent with the orientation

of the seats in sector S21, and in particular with the ori-

entation of the Chakana-like platform (sector S22) may

indicate the second of these two possibilities more likely.

Further to the west lie two platforms and a spacious

(ca. 2 × 6 m) terrace above them. At the opposite north-east-

ern corner of this sector, this terrace has its counterpart,

although of a much more irregular shape. A narrow water

channel that starts in the northern part of sector S26 passes

through it diagonally following a natural crack in the rock.

Sector S24

Little can be said about sector S24 (Fig. 48), which is

squeezed between the top of the rock (sectors C07 and

N10) and sectors S21 and S23 on the southern slope of

the rock. It consists of a series of irregular platforms and

terraces with no consistent layout; however, they all seem

oriented in the same direction – towards the same part of

the plaza located to the south of the rock.

The only noteworthy observation in this sector con-

cerns the relative chronology. Two terraces at the eastern

Fig. 47. Sector S23

110 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

end of this sector seem to be of a later origin since they cut

into small platforms located to the west.

Sector S25

Sector S25 (Fig. 49) lies at the foot of the southern

slope of the rock. From the west, it borders an extreme-

ly eroded ramp climbing up the rock. Perhaps this is the

remains of earlier steps leading to sector S23, but at the

present state of preservation, no traces resembling steps

could be identied. From the north, sector S25 borders

sector S28.

The western end of the southern border of this sector

reaches the foot of the rock and ends with a nearly vertical

rock fault more than 1 m high. Below, but more to the

east, a small (ca. 2.35 × 0.35 m) terrace with two rectan-

gular, unequal recesses is set. Their line of sight orienta-

tion strongly resembles that of platforms and seats from

sector S26 located higher up the rock. Still further to the

east, the western part of neighbouring sector S27 cuts into

the at, hori zontal platform constituting the central part

of sector S25.

An interesting feature of the western part of sector S25

is a semi-circular “Orchestra” (roughly 2 m in diameter)

accompanied by three benches at the back. No other ar-

rangements of this kind have been found on Samaipata

rock. Small rectangular recesses cut into the front part of

the “Orchestra” do not t the design and might be a later

alteration.

In front of the “Orchestra” are highly eroded traces of

a shallow, double-T-shaped depression. It abruptly ends at

the southern part of the rock. It seems the vertical face of

the rock was cut down there when the U-shaped building

from the colonial period was erected in the eastern part of

sector S04.

Fig. 48. Sector S24

Fig. 49. Sector S25

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 111

Fig. 50. Sectors S26 and S28

The western part of the central platform of sector S25 is

separated from the rest by a narrow water channel that leads

from the north down the rock. The part of it that lies behind

the “Orchestra” borders a steep rock scarp ca. 1.2 m high

in the north. Four double-recessed niches are equidistant-

ly placed on its western section. This decoration seems to

be an Incan addition to this evidently earlier sector.

Another channel that follows the natural crack in the

rock crosses the platform roughly at the centre. It brings

water from as far as the northern part of sector S28 and

terminates in front of the semi-circular “Orchestra”.

The remaining, eastern part of the central platform in

sector 25 does not bear any features except for a small

L-shaped terrace and very at, unclear traces of what might

once have been a small recess and a bench separating the

slightly raised eastern end of the platform.

Sectors S26 and 28

These two adjacent sectors (Fig. 50) are probably the

most interesting and intriguing parts of Samaipata rock.

Both lie to the north of sectors S25 and S27 and extend up

the rock until the big at area that constitutes sector S29.

The more westerly located sector S26 shows a very

consistent arrangement. Its main part is composed of three

successive platforms climbing to the north. Except for

small modications in the uppermost one, all of them have

an identical layout.

Their main part is a centrally located rectangular

(roughly 50 × 70 cm) seat anked by two raised buttresses

ca. 60 cm wide. Small benches are placed symmetrically

on both sides of these possibly ceremonial seats, except

on the third platform. The symmetrical layout of these

platforms is additionally emphasised by the coaxial steps

leading from the south.

Further to the north, another platform, this time of

a double-T shape, has been cut into the surface of the rock.

A rectangular, centrally located seat is placed at the back

of the platform. Faint traces of smaller seats accompany-

ing this seat on both sides are still noticeable. Directly to

the east, there might once have been a similar platform.

However, its rear side is arranged in a slightly dierent

manner, and the front side is very eroded. The rock sur-

face behind these two platforms is very irregular. It might

be that once a kind of east–west elongated petroglyph ex-

isted there. It must have been of an earlier origin since

both the platforms cut into it.

Behind this elongated petroglyph, there is a corridor,

0.45 m wide, ca. 1 m deep, and more than 5 m long, that

leads from sector S23 to yet another platform situated in

the north-eastern corner of the sector. Three rectangular

seats, the central one bigger and slightly raised, are set

against the back wall of this platform. Both the corridor

and the platform were apparently executed at the same

time as the upper terrace in sector S23.

Nearly ten separate units make up sector S28 (Fig. 50),

located adjacent from the east from sector S26. The rst

unit in the west consists of two terraces towering above

the western part of sector S25. The lowest one is divided

112 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

into three platforms, with the side ones being slightly el-

evated. An interesting feature is the rhomboid shape of

platforms, not seen in other sectors. The back wall of this

terrace is 0.7 m high. Even higher (ca. 1 m) is the back

wall of the next terrace adjacent from the north. Such sig-

nicant dierences in levels must have resulted in an im-

pressive cascade of water owing down the channel cut

along the side walls of both platforms.

The next unit of this sector has a very irregular arrange-

ment. Several small, dierently shaped platforms have

been cut to the east and west of the water channel running

from the north. Above the uppermost platform of this unit,

this water channel ows through a small (ca. 1 × 1.5 m)

rectangular reservoir at the bottom of which there are trac-

es of three, no longer identiable, petroglyphs. The whole

layout of this unit seems to be subordinated to the cult of

water, and this is further emphasised by the two rectangu-

lar ceremonial seats adjacent from the east of this unit. In

front of them, in the centre of a small (roughly 50 × 50 cm)

rectangular depression, traces of a rounded (ca. 23 cm in

diameter) oering hole are preserved. A small channel

branching from the area of the three petroglyphs might

have “fed” this oering hole. Below this, traces of some

other terraces, probably created earlier based on the state

of their preservation, are extant.

Two rectangular platforms located further to the east

are much better preserved. In the east, they border a much

bigger unit laid out over three terraces climbing the slope.

The lower one consists of a spacious (ca. 1.7 × 7.3 m) plat-

form with raised western and eastern borders isolating it

from neighbouring units. Four rectangular seats are placed

against the rear wall. Three buttresses, nearly 1.5 m wide,

separate them. Poorly preserved traces of much smaller

seats (ca. 50 × 60 cm) are visible on the top of each of the

buttresses. The front (southern) edge of this platform is

slightly raised so that the rainwater accumulates in a at

rectangular (ca. 60 × 80 cm) reservoir with indecipherable

traces of an oval petroglyph on its bottom. From there,

water passes a small meandering channel and runs down

from the edge of a nearly 2 m-high slope into the reservoir

located below. The level of the next, located further to the

north, terrace is raised for about 1 m. Three steps placed in

the centre of its back wall lead to another terrace that com

pletes the whole unit. There, in the centre of its 1.2 m-high

rear wall, two benches are set on top of one another. They

probably played the role of a ceremonial podium towering

over the terraces below.

Located above, the northern part of sector S28 narrows

gradually, squeezing between the neighbouring sectors

S26 and S31. Three further units constituting independ-

ent entities can be distinguished there. The westernmost

one starts with a terrace over 6 m long. The already men-

tioned reservoir with the three unclear petroglyphs on

the bottom attaches at the south-western corner of the

terrace. The rear wall of this terrace is nearly 1 m high

and ve triangular recesses decorate the top of it. Behind,

dicult to interpret traces of two narrow depressions ac-

companied by traces of rectangular recesses (or seats)

are extant. Possibly, these were sculpted earlier than the

triangular decorations. East of this terrace lies a unit of

a much consistent layout. A trapezoidal (ca. 3.0 × 3.7 m)

platform there opens to the south and is surrounded on

three sides by low benches. Two narrow buttresses in

the centre of the northern bench separate a more prom-

inent individual seat. Traces of rectangular recesses (or

possibly seats) to the east might be of an earlier origin,

so the whole unit was probably cut into an earlier (pre-

Incan?) phase.

The uppermost part of sector S28 is occupied by a at,

spacious (more than 5 × 10 m) terrace. Two small plat-

forms in its north-eastern corner facilitate movement up

to sector S29 located ca. 1 m above. Traces of at least

three at depressions in the western part of the terrace are

signs of an earlier decoration that was erased when this

terrace was arranged.

Other vestiges of earlier phases can be found on the cen-

tral part of this terrace. Fifteen round at holes (ca. 10 cm

in diameter), visible only on the hill-shaded mo nochroma-

tic 3D model, form a rectangular enclosure of 2.4 × 4.6 m

(Fig. 51). One possible interpretation assumes the exist-

ence of a small hut here with quincha type walls. Round

marks would then be remnants of the holes for the wood-

en posts of the wall. The hut, if it really existed in this

place, would have to have been from an earlier period

than the present form of this platform. Quincha wall post

holes are usually over 20 cm deep – in this case, levelling

the platform surface has led to almost completely eras-

ing them.

Sector S27

Sector S27 (Fig. 52) lies in the most southerly part of

Samaipata rock. It consists of two functionally distinct

parts that are connected by a special passage.

On the western side of sector S27 is a spacious court

separated from the south by a wall built of stone rubble.

A narrow entrance leads to it directly from the plaza at the

foot of the rock. The rear side of the court is bordered by

a steep edge of the rock and decorated with four niches.

The average width of niches and distances between them

is 105.2 cm, and again, this is a good approximation of

the unit of measurement used according to Kubicka [10]

Fig. 51. Round marks visible on the hill-shaded 3D model

of the central part of the upper terrace of sector S28

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 113

for buildings with the most prestigious function at Ma-

chu Picchu.

Spacious benches surround the court from the west,

north, and east. This courtyard was initially roofed, as is

clear from the presence of the front wall with the well

dened entrance and the side walls rising 2.25 m above

the stone oor inside.

The eastern part of sector S27 had a more utilitarian

function. A set of three deep reservoirs has been cut into

the rock there. The southernmost reservoir, 2 m deep,

collects water overowing from the other two reservoirs lo-

cated up the rock and has an outlet at a level close to its bot-

tom. Interestingly, a special Z-shaped entranceway connects

this reservoir with the court (building?) on the western side.

The side walls of the southernmost reservoir show ma-

ny steps and recesses perhaps facilitating climbing down

Fig. 52. Sector S27

Fig. 53. Andean Cross partially

preserved in a reservoir

in sector S27

(photo by G. Orefici, 25.06.2016)

for inspection and cleaning of the bottom part. On the

northern wall, Giuseppe Oreci found remains of what

might have once been an Andean Cross (Fig. 53). Be-

sides the already mentioned Chakana-like platform, it

is the only example of this symbol found on Samaipata

rock, and probably manifests the inuence of the Tiwan-

aku civilisation or was created during the period of Wari

culture expansion into the area of Bolivia, as can be at-

tested by similar examples found in the areas of Bolivia

and Brazil.

It is then evident that at least the upper part of the res-

ervoir in question had an earlier origin – however, dur-

ing Inca occupation of the site it was probably deepened.

The outlet on its bottom might also be of Incan origin and

probably provided water for the Incan building located di-

rectly to the south.

114 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

Sector S29

Sector S29 is located at the top of the rock (Fig. 54). Its

southern margin borders sectors S26 and S28. The north-

ern extent is limited by the eastern wing of an L-shaped

wall with double-recessed niches.

The central part of sector S29 was radically modied

in the Inca era, at the time when the L-shaped wall was

built. The archaeology of this wall was already examined

by the German mission led by Albert Meyers [12]. Ismael

Montero and Joanna Broda also studied it in 2006, determin-

ing that it functioned as an astronomical observatory [16].

On the southern edge of this sector, traces of a present-

ly indecipherable petroglyph can be seen. This petroglyph

is surrounded by narrow water channels and eight small,

irregularly distributed oering holes.

At the eastern border of sector S29, remains of what might

have once been a representation of a meandering snake exe-

cuted in bas-relief are extant. Its overall length of more than

11 m makes it the largest snakelike sculpture on the entire

Samaipata rock. The origin of this image, closely related to

the cult of water and fertility, might go back to the period

when the site was occupied by Chané-Arawak cultures

(ca. between 800 and 1450 AD). During Inca dominance, or

even earlier, the snake gure was severely modied with tri-

angular seats cut into both of its sides. The head of the snake

also completely disappeared, although this was partially

due to intentional erasing and partially to natural erosion.

Some modications were also introduced in the narrow

area between the snake and the eastern end of the L-shaped

wall. Two rectangular seats were cut in with two narrow

(ca. 45 cm wide) steps between them facilitating access

to the area above where several round oering holes are

irregularly distributed. In the present state of preservation,

only a few centimetres are left between the L-shaped wall

and the northernmost seat, making the latter completely

afunctional. Since the rectangular seats seem to have been

sculptured at the same time as the triangular ones, this

suggests that both alterations were executed before the

L-shaped wall was erected.

Sector S30

Sector S30 occupies the south-eastern edge of the rock

and borders sector S27 from the west and sectors S31 and

S32 from the north. A steep slope forms its southern border.

Two elongated terraces have been cut into this slope. Steps

leading from the north provide a secure entrance to the up-

permost terrace. However, there are no traces of any access

to the lower terrace located down the slope. It only had the

technical function of collecting rainwater owing down the

slope and channelling it o the rock. None of the terraces

show traces of any seats or decoration, which may be due

to the very poor state of preservation of this part of the rock.

Sectors S31 and S32

Sector S31 extends between sector S30 in the south and

S29 on the top of the rock (Fig. 55). Two long ights of

steps set ca. 12 m apart delimit its western and eastern

borders. Along the western edge of the sector runs a long

channel that drains rainwater from the top of the rock and

the terraces located below.

Fig. 54. Sector S29

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 115

Six consecutive terraces extend across the entire width

of the sector in its lower part. Apart from the southern-

most one, where two rectangular recesses of unequal size

tower above the steep slope, the layout of the terraces

is simple.

The upper part of the sector is divided into two units

separated by a narrow, very eroded strip that might once

have been a ight of steps. Now, no traces of such are

visible. The slightly broader western unit consists of a se-

quence of four towering terraces. Again, their layout is

very simple and except for the rst platform (the lowest

one), no any additional features such as seats or recesses

are present. The eastern unit is composed of four simple

trape zoidal platforms of which the uppermost one is the

narrowest.

Located to the east, sector S32 shows a much more

complicated arrangement (Fig. 55). Two units constitute

its lowest, southern part. The western one consists of two

platforms. The front edge of the lowest one is decorated

with three unequal recesses. A more symmetrical layout

is shown on its rear side, where three recesses have also

been carved, but in a more orderly equidistant manner.

Two of them of an equal size (ca. 0.7 × 1.0 m), while the

last one is much smaller so as to avoid cutting into the

semi-circular unit located to the east. The upper terrace

of this unit is much deeper and extends for more than 3 m

in a north–south direction. A single trapezoidal bench

(or possibly seat) is asymmetrically placed against its

0.8 m-high rear wall.

The second, more easterly unit consists of a semi-cir-

cular platform nearly 6 m in diameter. At its apex, a cer-

emonial seat with raised shorter edges is attached to the

0.8 m high rear wall. There is a small (45

65 cm) recess

cut into the upper edge of this wall to the west of the seat.

A similar one probably also existed on the eastern, now

completely eroded side. The front part of this platform has

also already collapsed. Only at its western end are three

rhomboid recesses preserved. They form a diamond-like

pattern that is unusual for Samaipata.

Above these two units is a sequence of several terraces

that extend for the whole width of this sector. The rst two

show an orientation similar to the lower, westerly unit,

although they are probably of an earlier origin since the

last cuts into the central recess of the lowest of these two

terraces. Their much more eroded surface further conrms

this relative dating. The uppermost terraces of sector S32

are dierently orientated and are probably even older than

the middle ones. The surface of the rock is very weath-

er-beaten there, and it is dicult to determine the plan-

ning concept at this part of the rock.

Fig. 55. Sectors S31 and S32

116 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

Eastern part of the rock

In this part of Samaipata rock, three separate sectors

can be distinguished: E01, E02 and E03 (Fig. 56). All of

them show traces of very advanced erosion; therefore, lit-

tle can be determined about their original layout. What

can still be seen today is a sequence of terraces along the

arched, eastern slope of the rock.

The only features that may attract more attention are

two sets of double, triangular recesses in sectors E01 and

E02. In sector E02, they ank a big trapezoidal terrace,

to which a smaller, also trapezoidal, platform attaches

Fig. 56. Sectors E01, E02, and E03

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 117

from the north. Similar triangular recesses are however

missing on the eastern side of the big trapezoidal terrace.

The other two triangular recesses are in the central part

of sector E01. They have been cut into the rear side of

the eastern end of the middle terrace, but a further contin-

uation of this kind of layout cannot be traced. The surface

of sector E03, as well as the northern part of sector E01,

is very fractured and battered due to harsh climatic con-

ditions at this part of the rock. The less compact structure

of the sandstone rock also contributes to this quickly pro-

gressing erosion.

The only anthropogenic traces preserved in this sector

are two terraces in its eastern part. They are both orien-

tated directly to the east. The more spacious (ca. 5.5 by

more than 7.0? m), lower one has a centrally placed re-

cess in its oblong front edge. The rear side also shows

a symmetrical layout – three small, rectangular benches

of roughly similar size (ca. 0.5

1.0 m) have been carved

there at equal intervals. Behind these, a much smaller

(ca. 2.6

3 m) platform completes the layout. It seems that

this unit could be from an earlier period than the neigh-

bouring trapezoidal platform from sector E02. Perhaps

two similarly oriented terraces with the same kind of lay-

out once existed in the northernmost part of sector E01.

However, only very faint traces of them have survived,

visible only under the right angle of oblique lighting

Summary and conclusions

Generally, chronological data for rock art are very lim-

ited, and few studies attempt to determine a chronological

sequence for this form of expression. Due to this, it is dif-

cult to obtain reliable data. However, in the last 30 years,

considerable progress has been made in dating some sites.

Studies about archaeological sites in the Cochabamba

Department [17], [18] and from the area of de la Serranía

San José in the Santa Cruz Department [19], particularly

that of Santa María Chica Cave, have shown that for some

kinds of rock engravings and paintings, absolute dating

can be provided.

The relative or absolute chronology of this class of

rock art can be, at least theoretically, established with the

following techniques and methods:

1. Superposition analysis that makes it possible in sev-

eral cases to establish which of the engravings is earlier or

later than others.

2. Stylistic and formal comparative analysis that al-

lows some of the representations to be associated with

ones from other, better dated sites.

3. Analysis of the technique of execution, taking into

account not only the type of rock art (e.g., rock painting,

relief) but also the technologies used to create them, for

example, in the case of reliefs, concave, convex, incised, or

dotted relief. Such analyses may allow particular images to

be linked with specic cultural circles of a known dating.

4. Some techniques for the absolute dating of engrav-

ings such as lichenometry or the analysis of the micro-ero-

sion of the petroglyph surface.

5. Dating of the archaeological contexts associated

with the rock. However, this method only points since

when human presence existed in the investigated area; it

does not determine when the engravings began to be made.

We will start with this last method. Ceramics from the

Formative Period (at least the end of the rst millenni-

um BC) have been found in the excavations carried out

near Samaipata rock [12, p. 79], particularly in the res-

idential area and in the lling of the terraces [20], [21,

p. 121]. However, for the moment, it is not possible to

attribute such an ancient dating to any of the petroglyphs

of the rock.

In terms of absolute dating (point 4 above), there was

some hope that analysing the micro-erosion of petroglyph

surfaces would be useful [22], [23]. Unfortunately, this

method is limited only to crystalline rocks, so at the pres-

ent stage of its development, it cannot be used for the sed-

imentary rock at Samaipata. Lichenometry cannot be used

for various reasons, either. The main one is that the rock

surface has been modied several times by dierent an-

thropic factors, making it practically impossible to deter-

mine places in which lichens began to grow immediately

after the execution of any particular petroglyph. Addition-

ally, the climatic conditions of Samaipata result in a dif-

ferent growth process of lichens compared to, for exam-

ple, North America or Easter Island, where lichenometry

was relatively successful. The third and nal reason is that

there is no calibration curve based on growth rate data for

local lichens at Samaipata [24].

Going back to the rst method, superposition analysis,

chronological dierentiation can be based on observa-

tions of rock shaping phases, marks of intentional recut-

ting, and characteristic formal features. Analysis of those

features makes it possible to create a tentative “relative

chronology” map of the engravings (Figs. 57 and 58). The

main distinction achieved in this map is between “earlier”

and “later” phases (A and B on Figs. 57 and 58). Only in

some cases was it possible to suggest particular features

as “possibly Incaic” (C on Figs. 57 and 58). It is, however,

important to mention that category “B” should not always

be considered as before the Incaic period. Sometimes the

carvings in this category may be modications made in

Inca times but without the typical stylistic features of this

period. From this point of view, one cannot add much more

to the ndings of previous researchers, who dieren tiated

two main stages in the cultural activities at the rock:

– pre-Incaic period, probably divided into dierent

cultural phases,

– Incaic period, of which, apart from the buildings on

top of the rock, the most outstanding feature is the cere-

monial “reorientation” towards the great plaza south of

the rock (see below).

Stylistic and formal comparative analysis, as well as

analysis of the technique of execution of particular en-

gravings (methods 2 and 3 above), bring other problems,

particularly with respect to the pre-Incaic phases of the

shaping of the rock. The diculty in absolute or even

relative dating of carvings at Samaipata rock also arises

from the universality of the occurrence of specic icono-

graphic motifs. Several rock art symbols are common

in the territory of South America, and they are related

to ancestral themes that characterise the Amazonian and

118 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

Fig. 57. Eastern part of the rock – relative chronology:

A – earlier;

B – later;

C – possibly Incaic

Fig. 58. Western part of the rock – relative chronology:

A – earlier;

B – later;

C – possibly Incaic

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 119

Andean area. For example, representations of serpents

and felines are typical over all the South American terri-

tory, and these iconographic motifs are the most common

expressions of the main divinities of the religious world.

Thus, for example, the meandering snake is found on

the site of the large El Buey rock shelter. Roy Querejazu

Lewis [17] identies the El Buey site as belonging to dif-

ferent moments in the development of several agro-pot-

tery periods and compares them to Yampará ceramics

(1100–1470 AD). However, interpreting the iconography

of the reliefs sculptured on Samaipata rock as belonging

exclusively to the Late Intermediate Period or Late Re-

gional Development (800/1000–1400 AD) seems to be

a bit extreme. There are several examples of reliefs with

serpentine elements, for example on stelae at the Khonkho

Wankane site in the Lake Titicaca Basin [25], which can

be dated back to the Late Formative Period (approximate-

ly 200 BC–500 AD), if not earlier, to the Middle Forma-

tive Period (ca. 800–200 BC).

In Eastern Bolivia, generally, examples of sculpted

rocks related to rituals of water worship and fertility are

most frequent in the vicinity of rivers and at the conu-

ence of them. They are often associated with round holes

serving as oering pits to be lled with rainwater or spe-

cial liquids. This strong association between petroglyphs

and oering holes is also present at Samaipata rock.

As already stated, the gural and geometrical petro-

glyphs found on Samaipata rock do not belong to a sin-

gle chronological moment. We can rather assume that

the sculpturing of these images had an intense dynamic

and that even though they all exist in the same place, the

importance of dierent symbols has varied throughout

time. Therefore, there were probably many more gural

representations sculptured on the rock surface from since

the time that it became a spiritual centre for the local

population. These were added, modied, or erased ac-

cording to the evolution of local beliefs and changes in

associated rites.

Among dierent zoomorphic petroglyphs, those of

snakes are particularly frequent and important. We found

them on the top of the rock as well as in the lower parts.

They are usually in the form of a meandering snake. The

only exception might have been the representation of

a coiled serpent, which, according to historical sources

[5], was located in the western part of the rock, just north

of the two depictions of felids. Today it is dicult to even

identify this spot exactly, let alone recognise the image.

It seems, however, that this petroglyph might have been

rather a representation of a meandering snake

, as repre-

sented by d’Orbigny on his drawing (Fig. 6).

Both the representations of felids and the now lost

coiled(?) serpent are coeval and can be possibly dated to

the Late Intermediate Period (ca. 1000–1400 AD). How-

ever, the neighbouring, nowadays completely indecipher-

able, image of a rhea can be of even earlier provenance

since as an exception, it was probably executed in sunk-re-

lief

. Known examples representing rheas might be petro-

glyph no. 2 from Tarija, Orozas (Fig. 59), or a pe troglyph

from Poopó, Oruro (Fig. 60). Generally, repre sentations

of rheas can be traced back to as early as the second mil-

lennium BC [27], [28].

On the other hand, all these petroglyphs belong to the

feline-snake-bird triad, which is the most representative

Cf. J. Kościuk, M. Ziółkowski, M. Pakowska, Formal and ico-

nographic analysis and interpretation of the most damaged petroglyphs,

in the same issue of “Architectus”.

Cf. J. Kościuk, M. Ziółkowski, M. Pakowska, Formal and ico-

nographic analysis and interpretation of the most damaged petroglyphs,

in the same issue of “Architectus”.

Fig. 60. Representation of a rhea (ñandú)

on a petroglyph from Poopó/Oruro

(photo by M. Strecker, March 2012, source: [26])

Fig. 59. Representation of a rhea (ñandú) on a petroglyph no. 2

from Tarija/Orozas (source: extract from YouTube movie,

https://www.youtube.com/watch?v=jXvhVrpvDB0)

120 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

for the religious world of the South American area and

for the eastern Bolivian jungle; therefore, despite possible

dierences in sculpturing technique, all of them may have

the same cultural provenance that however cannot be pre-

cisely dated. This type of iconography already existed at

least from the second millennium before our era, and can

be seen in petroglyphs from Argentina, Peru, and Bolivia

[29], [30].

All the gural petroglyphs found on Samaipata rock

are connected to the cult of water. The two felids in the

western part of the rock and the so-called Altar (“Choir

the Priests”) are even surrounded by circular water chan-

nels. For other petroglyphs, water channels pass beside or

through the petroglyphs (for example, the “Double Puma(?)”

or “Triple Snake”, both in the eastern part of the rock), or

rounded pits for liquid oerings are located nearby (the

“Big Snake”, “Snake-catching Animal”, or the lost rhea

petroglyphs). The eastern part of the rock is particularly

abundant with this kind of pit. They can be distinguished

from natural openings in the rock by the shape and charac-

ter of the inner surface, and in some cases by the presence

of channels that guide liquid into the holes. Possibly, they

formed part of rituals related to sunrise or water. Similar

circular holes, which were elements of water worship, can

also be seen in the Roca de Mairana [31]. Samaipata rock

also has another type of rounded hole meant for oering

deposits. These are found on the entire surface of the rock

and in particular on the southern and northern sides.

They are also many traces of now indecipherable cir-

cular petroglyphs. Typically, these are between 1.5 and

3 m big, but some of them have a diameter of more than

5 m. They are mostly located on the top of the rock and its

northern slope. As they were erased or even cut through

by later activities (sculpturing terraces, benches, or seats),

they are evidently of an earlier, pre-Incaic origin. The fact

that they bear not only traces of intentional erasing, but

they are generally much more eroded than the other exam-

ples, also suggests a much earlier origin.

The association of circular petroglyphs with earlier

phases of the sculpturing of Samaipata rock may lead to

a conclusion about preferences for circular shapes in ear-

lier periods. However, since there are other, non-circular,

petroglyphs attributed to earlier periods (for example, two

big meandering snakes into which zig-zag seats were cut

during later times), a circular shape cannot be considered

as the only decisive factor in chronological analysis.

The most emblematic gure on Samaipata rock is the

“Big Snake” – the long geometric petroglyph more than

30 m long that is interpreted as a representation of a rattle-

snake. Since generally, the edges of the rhomboid patterns

as well as of the two accompanying gutters are relatively

sharp, this petroglyph has probably not been exposed for

a long time to atmospheric and bio-erosion factors. There-

fore, the most probable dating is not very much earlier

than the Late Intermediate Period. During the Inca period,

this petroglyph lost or changed its symbolic and religious

signicance. A north–south oriented wall erected across

the top of the rock made the petroglyph inaccessible for

viewing from the only place from where it might have

been admired – from the slope of the hill rising to the west.

The earliest anthropogenic traces on Samaipata rock

are to be found on its eastern side. From there, access to

the top of the rock was also the most convenient. Possibly,

this was the oldest main route to the summit of the rock,

from where one could proceed further to the west or down

the southern slope. Several steps, often separating dier-

ent groups of terraces and platforms, facilitated move-

ment down the steep southern slope. Some of these were

modied over time due to the erosion aecting the struc-

tural integrity of the rock and changes in the function of

particular sectors. Modications were also made to other

features sculptured on the rock – oering pits, water res-

ervoirs, seats and benches, as well as many petroglyphs.

The triangular seats(?) that frequently appear on the

rock, often in long rows, are also worth attention. Al-

though this shape is untypical for Incan architecture,

some examples of triangular niches are known [14, p. 256,

Fig. 242]. However, generally, they were more likely to

have been made during earlier periods of Samaipata rock

shaping. This is also true for the triangular seats in the

“Choir of the Priests”.

One of the typical architectural forms on Samaipata

rock is niches. Some of these are double-recessed, trap-

ezoidal in shape, or show metrological similarities with

Incan architecture. None of these features, however, can

be considered as clearly indicating Incan origin. Such ar -

chitectural elements are also typical for the Tiwanaku cul-

ture; therefore, they are considered by us only as “possi-

bly Incan”.

The possible inuence of Tiwanaku culture on Samai-

pata carvings presents another problem. This inuence

may be seen in the architectural elements of the niches

as well as the traces of the Andean Cross found in Sec-

tor S27 and the Chakana-like platforms. On the one hand,

we have clear evidence of the inuence of Tiwanaku cul-

ture in the Cochabamba valley [32] – so approximately

halfway between Samaipata and Lake Titicaca. However,

there is generally a view that in Bolivia, there is no evi-

dence of the inuence of Tiwanaku culture on rock art.

In a UNESCO report about rock art in Bolivia, Matthias

Strecker even states […] An absence of rock art belonging

to the Tiwanaku culture has been noted by several inves-

tigators […] It is a curious fact that the only clear case of

Tiwanaku style rock art has been found outside Bolivia, in

northern Chile […] [33, p. 142].

The restructuring of the rock in Inca times

Summarising, we can state that over time, all the sculp-

tured forms, including petroglyphs, platforms, seats, nich-

es, terraces, and the steps connecting them, underwent

several modications due to the changing religious and

ceremonial needs. In most cases, it is not possible to deter-

mine the exact sequence of these changes and the specic,

pre-Incaic, periods in which they were made. There is no

doubt that generally the Inca expansion to new areas often

eliminated previous cultural elements, particularly when

they were deeply rooted in the social, political, and espe-

cially religious structure of the populations that had previ-

ously used the site. One other place of the many where this

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 121

occurred is the Intihuatana in Pisaq, which was erected

in a place where much older symbols had been engraved

in the rock. Everything was removed from the previous

sacred element and replaced with the Intihuatana [34].

We must not forget a very specic factor: the Incas

erected a large ceremonial complex, with a great plaza and

kallanka, immediately south of the rock. As the main In-

can ceremonial activities were carried out in large squares,

usually located in the centre of settlements, presumably

this was also the case at Samaipata. An interesting analo-

gy to this particular relationship between the rock sculpted

with niches and benches and the adjacent plaza is found in

the Incan ceremonial centre of Coyoctor in the province of

Cañar in Ecuador [35].

The areas of worship and sacrices with ceremonial

seats or thrones, often of considerable dimensions and of-

ten several times remodelled, are generally oriented toward

the south. This southern edge of the rock, especially in its

eastern and central parts, was modied very drastically dur-

ing the Inca occupation of the site. The edge of the rock

was cut almost vertically, and several sets of niches were

placed there. Subsequent carving of platforms and terraces,

as well as the addition of new functions in some sectors,

caused remarkable transformations to be made to this part

of the rock. It seems that except for a long row of ten niches

in the central part of the southern edge of the rock, all other

niches might have been roofed. A similar arrangement

was also introduced at the northern edge of the rock – at

the “Temple of Five Niches”. It is, however, a possibility

that some of these modications were executed during the

Middle

Period or the Chané occupation of the site.

Since our knowledge about rites performed at the rock

is less than limited, we cannot reconstruct the “choreog-

raphy” of any ceremonies. Theoretically, two options are

possible. Certain rituals were performed on the rock, and

spectators gathered on the southern side, at the foot of the

rock. On the other hand, the presence of long benches that

could accommodate many people on the rock points to

another possibility. Seats and benches were indeed meant

for celebrants and their acolytes, but in addition to per-

forming certain rites, they watched processions and cere-

monies that took place at the foot of the rock. Most like-

ly, depending on the particular occasion, both scenarios

were possible.

Visibility analysis (analysis of the line of sight) can

shed some light on this. Benches, seats, and terraces, es-

pecially on the southern slope of the rock, mostly follow

the natural shape of the rock. However, drawing sightlines

perpendicular to the backside of the seat or bench leads to

some interesting observations (Fig. 61). Along the south-

ern side of the rock, within a range of 20–30 m, there are

four zones suitable for spectators where sightlines tend

to concentrate (Figs. 61A–D). However, the results of

analysis of the line of sight do not determine whether

the designated zones were intended for viewers observ-

ing the rites on the rock, or they were the route (or plac-

es) of the celebrations that were observed by privileged

spectators gathering on the benches and seats on the rock.

The following analyses ignore the vast area further south,

where a much larger audience could gather. Due to the

Fig. 61. Analysis of sightlines for benches and seats of earlier (1), later (2), possibly Incaic periods (3) and of unknown provenance (4). A–G – areas convenient for spectators (explained in the text)

122 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

References /Bibliografia

[1] d’Orbigny A., Viaje a la América Meridional, T. 4, Institut Français

d’Études Andines, Lima 2002, doi: 10.4000/books.ifea.6812, http://

books.openedition.org/ifea/6812 [accessed: 14.04.2019].

[2] Nordenskiöld E., Meine Reise in Bolivien 1908–1909, “Globus”

1910, Bd. 97, 213–219.

[3] Nordenskiöld E., The Guarani Invasion of the Inca Empire in the

Sixteenth Century: An Historical Indian Migration, “Geographical

Review” 1917, Vol. 4, No. 2, 103–121, https://www.jstor.org/stable/

pdf/207290.pdf [accessed: 18.02.2019].

[4] Nordenskiöld E., Indianer und Weisse in Nordost Bolivien, Strecker

und Schröder Verlag, Stuttgart 1922, https://www.yumpu.com/de/

document/read/20650859/indianer-und-weisse-in-nordostbolivien

[accessed: 24.05.2019].

[5] Pucher L., Ensayo sobre el arte prehistórico de Samaypata, “Revista

de la Universidad de San Francisco Xavier”, Universidad Mayor de

San Francisco Xavier Sucre, Sucre 1945, 17–61.

[6] Ponce Sanginés C., Tadeo Haenke y su viaje a Samaipata en 1795,

Centro de Investigaciones Arqueológicas, INAR, La Paz 1974.

considerable distance (75–150 m), anyone gathered there

could

probably only participate in ceremonies as a passive

viewer.

Area A has a visible preference for benches and seats of

a later period and gives a good view for at least three big

ceremonial seats of sector S32 (arrows on Fig. 61A). Area

B shows the concentration of most sightlines coming from

benches and seats of probably Incaic origin. In this case,

viewers gathered in front of the roofed(?) building located

in sector S27. Above, sets of seats and platforms in sectors

S26 and S28 towered over the southern slope of the hill

(arrows on Fig. 61B).

The sightlines of two areas located further to the west

(C and D) do not show any consistent chronological pref-

erences. Area C – where a U-shaped building was erected

in colonial times, gives direct access to the terraces on

the southern slope of the rock. There, on the border be-

tween sectors S09 and S10, starts a long stairway leading

up the hill (the middle arrow on Fig. 61C). A spacious

building with unnished niches on its back wall anked

its eastern side (the right arrow on Fig. 61C). To the east,

a set of platforms with several niches (some of them dou-

ble-recessed) adorned the rock face (the left arrow on

Fig. 61C).

Nearly all petroglyphs on Samaipata rock were sculp-

tured on at parts of the hill, either on its top or on arti-

cially created horizontal terraces and platforms; therefore,

they were visible only to those who were gathering di-

rectly on the rock. Exceptions, besides the already men-

tioned “Double Puma”, are the felines on the western

edge of the rock, and the petroglyph most emblematic for

Samaipata – the over 30 m-long representation of a rat-

tlesnake (“Big Snake”). All these petroglyphs were practi-

cally invisible to viewers assembling on the southern side

of the rock. The only place from which a bigger group of

spectators could view them was area E on the slope of the

hill rising to the west (Fig. 61E). This area has not been

excavated yet, but the terrain suggests that there could

have been several terraces falling to the east. This pos-

sibility, however, is severely limited by the construction

of the aforementioned transverse wall across the hill. It

practically wholly obscured the view to the “Big Snake”,

so only three petroglyphs depicting felines (three arrows

on Fig. 61E) remained visible to a bigger audience. These

alterations introduced apparently during the Inca occupa-

tion of the site may reect changes in ritual symbology

and associated rites.

In this context, one must also mention two specic

areas – F and G on Figure 61. In these areas, the arrange-

ment of seats and benches focuses on the area in close

vicinity. Both areas, rather poorly visible from the area

south of the hill, were probably meant for special, col-

lective rituals.

* * *

In summary of this attempt to address the problem of

the chronology of the rock art on Samaipata rock, we want

to emphasise that the objective of the entire project was

mainly technical – that is, to prepare high-precision 3D

documentation of this exceptional site.

Our primary goal was not to carry out a detailed com-

parative study, in a Pan-American context, on the symbol-

ic and chronological interpretation of the rock art. Look-

ing for possible analogies, we have tried to limit ourselves

to these territorially and culturally closest to Samaipata:

mostly, according to terminology from the Bolivian Rock

Art Research Society [36], from the Valles or Subandina

regions [31], [37]. The only exceptions were, in this case,

comparisons with similar monuments of Incan origin scat-

tered throughout the Tahuantinsuyu [37], [38].

More comprehensive analysis of chronological phases

of Samaipata rock shaping would require profound com-

parative studies on beliefs, symbology, and rituals of sev-

eral South American cultures that left their traces in this

region.

We must, however, remember that Samaipata, which

is the largest and most important site of its kind in South

America, due to its conformation, type of rock, and geo-

graphical position, was considered by its ancient inhabit-

ants as a very important sanctuary from the earliest times.

Its central geographical position with regards to commu-

nication ows between the jungle and the valleys deter-

mined its importance not only during the Middle Horizon

and Early Regional Development. Most probably the site

might have been considered as a cult place at least from

the Formative Period.

As mentioned, the detailed chronology of the site as

well as its cultural interpretation are outside of the scope

of this particular project. Its main aim was to document

the vanishing traces of this important site of rock art. The

observations we have presented above should be consid-

ered only as working hypotheses or directions for future

studies, but in no way as rmly established evidence. We

hope that our work will serve experts in more advanced

studies on the subject.

ElFuertedeSamaipatainthelightofnewresearch / ElFuertedeSamaipatawświetlenowychbadań 123

[7] Trimborn H., Der skulptierte Berg von Samaipata, “Archäologis-

che Studien in den Kordilleren Boliviens”, Bd. 3, “Baessler Archiv,

Beiträge zur Völkerkunde” 1967, NF 5, 130–169.

[8] Tadeo Haenke. Su obra en los Andes y la selva boliviana, G. Ovan-

do-Sanz (ed.), Los Amigos del Libro, Cochabamba 1974.

[9] El Fuerte de Samaipata: Estudios arqueológicos, A. Meyers,

I. Com bès (comp.), Biblioteca del Museo de Historia Universidad

Autó noma Gabriel René Moreno, Santa Cruz de la Sierra 2015.

[10] Kubicka A., Zastosowanie metody cosine quantogram do analizy

metrologicznej kompleksu Machu Picchu/Application of a cosine

quantogram method in metrological analysis of the Machu Pic-

chu settlement, Doctoral thesis submitted to the Faculty of Archi-

tecture, Wrocław University of Science and Technology, Wrocław

2019.

[11] Marulanda R., La roca esculpida del fuerte de Samaipata: Elemen-

tos de discusión para un enfoque interpretativo, [in:] A. Meyers,

I. Combès (comp.), El Fuerte de Samaipata. Estudios arqueológi-

cos, Biblioteca del Museo de Historia Universidad Autónoma Ga-

briel René Moreno, Santa Cruz de la Sierra 2015, 34–51.

[12] Meyers A., Los trabajos arqueológicos en “El Fuerte de Samaipa-

ta”, 1992–1996, [in:] A. Meyers, I. Combès (comp.), El Fuerte de

Samaipata. Estudios arqueológicos, Biblioteca del Museo de His-

toria Universidad Autónoma Gabriel René Moreno, Santa Cruz de

la Sierra 2015, 52–115.

[13] Rowe J.H., Inca culture at the time of the Spanish conquest, [in:]

J. Steward (ed.), Handbook of South American Indians, Vol. 2,

Smithsonian Institution, Washington, DC 1946.

[14] Gasparini G., Margolies L., Inca Architecture, Indiana University

Press, Bloomington–London 1984.

[15] Farrington I., Cusco. Urbanism an Archaeology in the Inka World,

University Press of Florida, Gainesville–Tallahasseee–Tampa–Boca

Raton–Pensacola–Orlando–Miami–Jacksonville–Ft.

Myers–Sara-

sota 2013.

[16] Alcázar De La Fuente R., El Sitio Ceremonial y Administrativo de

El Fuerte de Samaipata, “Boletín Tesape Arandu” 2015, Año 4,

Nº 21, 2–9, https://www.academia.edu/12331379/El_Sitio_Cere-

monial_y_Administrativo_de_El_Fuerte_de_samaipata [accessed:

10.07.2019].

[17] Querejazu Lewis R., El arte rupestre de “El Buey”, Departamento

de Cochabamba, Bolivia, “Boletín SIARB” 1989, Nº 3, 30–35.

[18] Querejazu Lewis R., Arte Rupestre en la Cuenca del Río Mizque,

[in:] M. Strecker, F. Huaranca (eds.), Arte Rupestre en los Andes de

Bolivia, SIARB, Sociedad de Investigación del Arte Rupestre de

Bolivia, La Paz 1986, 17–19.

[19] Kaier C., Tres sitios de pinturas rupestres en la parte occidental

de la Serranía San José, Departamento de Santa Cruz, Bolivia,

“Boletín SIARB” 1993, Nº 7, 59–95.

[20] Meyers A., Toward a Reconceptualisation of the Late Horizon

and the Inka Period: Perspectives from Cochasquí, Ecuador, and

Samaipata, Bolivia, [in:] R.L. Burger, C. Morris, R. Matos Men-

dieta (eds.), Variations in the Expression of Inka Power. A Sym-

posium at Dumbarton Oaks 18 and 19 October 1997, Dumbarton

Oaks Research Library and Collection, Washington, DC 2007,

223–254.

[21] Szykulski J., Investigaciones en los Sectores 1 y 2, [in:] A. Meyers,

I. Combès (comp.), El Fuerte de Samaipata. Estudios arqueológi-

cos, Biblioteca del Museo de Historia Universidad Autónoma Ga-

briel René Moreno, Santa Cruz de la Sierra 2015, 116–132.

[22] Bednarik R., Age Estimates for the Petroglyph Sequence of Inca

Huasi, Mizque, Bolivia, “Andean Past” 2000, Nº 6, 279–291.

[23] Bednarik R., A new method to date petroglyphs, “Archaeometry”

2007, Vol. 34, Iss. 2, 279–291.

[24] Benedict J.B., A Review of Lichenometric Dating and Its Applica-

tions to Archaeology, “American Antiquity” 2009, Vol. 74, No. 1,

143–172.

[25] Janusek J.W., Of Monoliths and Men: Human-Lithic Encounters

and the Production of an Animistic Ecology at Khonkho Wankane,

[in:] T.L. Bray (ed.), The Archaeology of Wak’as: Explorations of

the Sacred in the Pre-Columbian Andes, University Press of Colo-

rado, Boulder (Colorado) 2015, 335–365.

[26] Strecker M., Methfesse L., Rivera C., Taboada F., Lima P., Cami-

nos destruyen sitios de arte rupestre en Bolivia, “Boletín SIARB”

2012, Nº 26, 33–40.

[27] Recalde A., Colqui E., Las representaciones rupestres zoomorfas en

el centro de Argentina y la construcción de identidades (ca. 1500–

450 AP), “Boletín del Museo Chileno de Arte Precolombinom”

2019, Vol. 24, No. 1, https://scielo.conicyt.cl/scielo.php?script=s-

ci_arttext&pid=S0718-68942019000100083&lng=es&nrm=iso

[accessed: 21.05.2019].

[28] Zárate Bernardi M.S., Arte rupestre del Cerro Tunduqueral:

deniciónestilísticayperspectivascronoculturales, [in:] I Jorna-

das de Ciencia, Técnica y Posgrado de la Facultad de Filosofía

y Let ras, Universidad Nacional de Cuyo, Argentina, 2017, https://

www.academia.edu/36552429/ARTE_RUPESTRE_DEL_CER-

RO_TUNDUQUERAL_DEFINICI%C3%93N_ESTIL%C3%8D-

STICA_Y_PERSPECTIVAS_CRONOCULTURALES [accessed:

20.04.2019].

[29] Hostnig R., Los petroglifos de Boca Chaquimayo: reliquia arque-

ológica de la Amazonía puneña, Perú, 2009, http://www.rupes-

treweb.info/chaquimayo.html [accessed: 6.07.2019].

[30] Guroy J., El Arte Rupestre del Antiguo Perú, Institut Français

d’Études Andines, Lima 1999.

[31] Avilés Loaiza S.V., Piedras sagradas: Samaipata, Achira y Maira-

na: ¿Casos aislados o una tradición por descubrir en tierras bajas de

Bolivia?, 2015, http://www.rupestreweb.info/piedrassagradas.html

[accessed: 12.08.2019].

[32] Anderson K., TiwanakuInuenceon the CentralValley of Coch-

abamba, [in:] A. Vranich, Ch. Stanisch (eds.), Visions of Tiwan-

aku, The Cotsen Institute of Archaeology Press, [s. l.] 2013, 87–

112.

[33] Strecker M., Zone 4: West-north-west South America (north of the

Andes, Peru, Bolivia, Paraguay), 138–150, https://www.icomos.

org/studies/rockart-latinamerica/zone4.pdf [accessed: 16.07.2019].

[34] Angles Vargas V., Písac and the sacred valley of the Incas, Industrial

Gráca, Cusco 2001, 16–29.

[35] Siguencia Pinos R., Coyoctor, Casa de la Cultura Núcleo del Cañar,

Quito 2007.

[36] http://siarb-bolivia.org/regiones/ [accessed: 24.08.2019].

[37] Querejazu Lewis R., Arte Rupestre en el límite andino-amazónico,

“Boletín Ocial de la Asociación Peruana de Arte Rupestre” 2012,

Vol. 4, N

13–14, 349–360, https://issuu.com/apar/docs/boletina-

par4_13_14 [accessed: 3.08.2019].

[38] Van de Guchte M.J.D., Carving the World: Inca Monumental

Sculpture and Landscape, PhD dissertation, University of Illinois,

Urbana-Champaign 1990, https://www.scribd.com/document/

269527866/Van-de-Guchte-1990-Thesis-1 [accessed: 17.06.2019].

Acknowledgements /Podziękowania

The presented work is a part of the research sponsored by the grant giv-

entotheWrocławUniversityofScienceandTechnologybythePolish

National Science Centre (grant No. 2014/15/B/HS2/01108). Addition-

ally, the municipality of Samaipata, represented by Mayor Falvio Ló-

pes Escalera, contributed to this research by providing the accommo-

dationduring the eldworkin June andJuly 2016, aswell as inJuly

2017. The Ministry of Culture and Tourism of Bolivia kindly granted

all necessary permits (UDAM No. 014/2016; UDAM No. 060/2017).

The research was conducted in close cooperation with the Centre for

Pre-Columbian Studies of the University of Warsaw in Cusco. Spe-

cialists from many other universities and research centres also joined

the project.

124 Jacek Kościuk, Giuseppe Orefici, Mariusz Ziółkowski, Anna Kubicka, Rosario Muñóz Risolazo

Abstract

The study describes the results of the documentation of El Fuerte de Samaipata, a pre-Hispanic archaeological site in Bolivia on the UNESCO World

Heritage List. It summarizes the most recent observations of the carvings on the rock of Samaipata, which, due to progressing erosion, are slowly

vanishing. Thanks to modern documentation technologies, laser scanning, structural light scanning, digital photogrammetry, and advanced digital

analysis of images and 3D models of the site, all still visible traces of petroglyphs as well as “architectural” elements like terraces, platform, niches

and seats have been documented and described.

Key words: Samaipata, Bolivia, rock art, Andean archaeology, UNESCO World Heritage List

Streszczenie

W pracy opisano wyniki projektu dokumentującego El Fuerte de Samaipata, prehiszpańskie stanowisko archeologiczne w Boliwii wpisane na Listę

Światowego Dziedzictwa UNESCO. Artykuł podsumowuje najnowsze obserwacje dotyczące rytów naskalnych w Samaipata, które powoli zanikają

z powodu postępującej erozji. Dzięki nowoczesnym technologiom dokumentacji, skanowaniu laserowemu, skanowaniu światłem strukturalnym,

cyfrowej fotogrametrii oraz zaawansowanej cyfrowej analizie obrazów i modeli 3D terenu wszystkie nadal widoczne ślady petroglifów, a także

elementów „architektury” takich jak tarasy, platformy, nisze i siedziska zostały udokumentowane i opisane.

Słowa kluczowe: Samaipata, Boliwia, sztuka naskalna, archeologia andyjska, Lista Światowego Dziedzictwa UNESCO