Introduction

Although the first observations on Inka domestic and public architecture had already been published in the 1940s, a systematic and comparative examination regarding aspects of Inka stone masonry did not occur until the 1970s. By concentrating primarily on the basic structural form and characteristics of building shapes, door niches, floors and roofs, first Gasparini and Margolies, then Hemming, and eventually Kendall, have been able to clarify and define the fundamental compounds of the general nature of Inka architecture. From these studies it has emerged that the rectangular plan with a gabled roof was a dominating feature of Inka buildings, from humble rural houses to halls and sacred temples. Occasionally curved walls were also built, but this building style, according to Hyslop was used in response to particular irregular terrains. Examples of these latter types of building are found in Cuzco, with the Temple of the Sun (Qori Kancha), the Temple of Pisaq (Urubamba Valley) and the Torreon building in Machu Picchu. An unmistakable trait of Inka architecture is its wall apertures, which were generally trapezoidal in form. They had various sizes but were placed in the walls by following specific patterns. Doors, for instance, were not wider than 1.65 m, and windows were usually around 1.25 m above ground. An elegant characteristic of Inka architecture was the use of double or triple-jamb doorways, of which there are many examples in Cuzco and in the Lake Titikaka region.
 

Inka stone block ornamentation was limited. Walls were rarely adorned, and when they were, it was usually in the form of small animals or geometric figures engraved directly onto the stone blocks.

Two types of Inka architecture are generally found: royal and common architecture. The latter was used by common people to build their own houses or agricultural structures, whereas the former was employed for royal constructions such as temples, priests’ houses and palaces. This kind of architecture is also known as fine masonry, and it is indeed on this aspect of Inka stone constructions that my research is focused.

Most of the comments written on the origins of Inka architecture and in particular its stone masonry are unfortunately purely speculative. According to Rostworowski, Inka architecture did not have a common origin, and in spite of some similarities with the pre-Inka civilisations, the Andean stonework of the 15th century may indeed be a peculiarity of the Inka Empire. Undoubtedly, the characteristics of the Inka stone masonry could not have been generated from nothing; the influence of previous Andean cultures is evident. The problem arises when one has to decide where the sources were. Kendall, for instance, argues that the Inkas were strongly influenced by the Wari (or Huari) culture. In fact, her argument is founded on the similarities between the two kinds of architectures, especially in regard to rectangular enclosures and  use of a grid pattern.

Other characteristics of Inka stone masonry appear to have come from the Lake Titikaka region and in particular from Tiwanaku. The Tiwanaku culture (AD 300-1100) apparently developed an extraordinary level of stonework which shares several similarities with that of the Inkas. A number of speculations have been made regarding this matter, the most famous being that of the Spanish chronicler Cobo. He argues that when Pachakuti (the 9th Inka ruler) went to Tiwanaku, he saw those magnificent buildings and being especially impressed by the stonework, ordered his men to copy the construction techniques because he wanted the building project in Cuzco to be of the same standard of work. Referring to Cobo’s chronicle, the reason why there are so many similarities between Inka and Tiwanaku stone masonry is that a lot of mitimaes (tax labour force) from there were sent to Cuzco to help build the town. This theory is backed by the fact that the Inkas claimed to be originally from the Lake Titikaka area. Therefore, having their new heartland built using some traditional Tiwanaku techniques would have helped to reinforce their beliefs about their mythical origins. From this point arises another interesting  fact for speculation. Claiming that their origins lie in the Lake Titikaka region, if that were truly the case, the Inkas might simply have brought their stonework traditions with them when they moved to Cuzco. This kind of statement is difficult to test because even with archaeological evidence, it is hard to define whether or not a specific type of construction is an invention of a particular culture. For instance, the trapezoidal form of apertures was already in use before the Inka Empire, but with some further innovations and variations the Inkas made it the hallmark of their architecture. Another point of disagreement about Inka stone masonry is in relation to its chronological development.

The historical Inka Empire lasted less than a century and it was within that period that the Inkas developed and applied more widely different types of stone masonry. Obvious questions at this point are: were these diverse building techniques related to different periods? Or, was their diversity just a matter of function? Some scholars believe that variations in fine Inka masonry have no chronological implications (which coincidentally would fit with Inka concepts of time). Other researchers argue that diverse types of stone masonry developed in different periods. For instance, flat-surfaced rectangular-coursed masonry dates back to the time of Pachakuti (9th Inka), whereas the adoption of sunken joints and small nearly square stones was developed very late, possibly at the beginning of the 16th century. Anyway, whether or not  differences  in  Inka  masonry  were  determined  by  specific periods is not the main concern of this chapter. Instead, in the following sections, these various masonry styles will be discussed from a technical point of view, from stone extraction (quarrying) to the final construction of the Inka walls.

The five most important quarries used by the Inka in the Sacred Valley were: Saqsaywaman, Huaccoto, Yucay, Kachiqhata and Rumiqolqa.

From  these  five  quarries  the  Inkas  extracted  the  material  they  needed   according to the different types of constructions. For example, the blocks of calcite and diorite were extracted from Saqsaywaman Hill, as well as the Huaccoto Quarry. Sandstone blocks were obtained from Yucay Quarry, and andesite at Rumiqolqa and Huaccoto. Finally, the huge red rhyolite blocks used at Ollantaytambo were derived from the Kachiqhata Quarry.

Because of the enormous availability of labour, the Inkas did not particularly care about the distance which separated the quarry from the building site. In fact, apart from Machu Picchu and Saqsaywaman whose construction materials were available directly from the sites themselves, most of the other sources of stone blocks were some considerable distance away from their intended construction sites. For instance, most of the andesite used to build the finest buildings in Cuzco, such as Acllahuasi (the convent of the Virgins of the Sun), and Qori Kancha (the Temple of the Sun), was obtained from the Rumiqolqa Quarry situated about 35 km south of Cuzco.

The gigantic rhyolite blocks (some of them weighing up to 75 tons) of Ollantaytambo, 70 km north of Cuzco, used to build the Temple of the Sun, were extracted from the Kachiqhata quarry, lying some 5 km across the Urubamba River. However, the Temple of the Sun was never finished perhaps owing to the civil war and the Spanish invasion.

Of  the  above  mentioned  quarries,  Rumiqolqa  has always  been  the  most studied, not only because of its long extraction history, as it was already in active use in the Killke period (AD 1000-1200), but also because of the clear evidence of Inka stone block extraction and dressing especially found in the Llama Pit. In fact, some parts of this quarry are still in use.

The basic method of stone extraction was apparently very simple. Once a coarse piece of rock was fractured from the natural bedrock, the process of dressing could start. The first part to be done was the top surface,  then  the  two long  lateral sides, and finally the two other small sides. When the blocks had assumed a parallelepiped form, they were finely dressed and in some cases polished (see the following pictures).
 
 

Various scholars, such as Protzen, Heffernan, Agurto Calvo and Kendall, agree with the conclusion that the most common tool used by the Inkas to dress stone blocks was the stone hammer.

Protzen in his researches at Rumiqolqa Quarry found about 68 hammerstones scattered around the area. These stone tools are easy to recognise. They are rounded, have a smooth surface with pitted small ends and are normally of an exotic material. Most of these stones are river cobbles weighing from 200 g to 8 kg. They are made of quartzite, rhyolite and  hematite, a  non-magnetic iron  ore.

The hardness of both quartzite and hematite is between 5 and 7 on the Mohs scale, which corresponds to about the same hardness as that of some of the materials (e.g. basalt, rhyolite, diorite) on which hammerstones were used.

What makes quartzite and hematite ideal materials for hammerstones is that both are often very tough, and sometimes tougher than the rock to be worked on. Toughness, as opposed to brittleness, is defined by the resistance of the material to the propagation of incipient fractures. This level of fragmentation depends mostly on the differential cooling which occurred when the rock was originally formed, especially  as  far  as igneous rocks are concerned.

Recent studies have brought up the possibility that some copper and bronze tools might have been used to dress the stone blocks. According to Heffernan, a use-wear study of several bronze implements based on micro-structural deformation, combined with inferences based on tool form, has revealed the possibility of bronze tool usage for chipping stones. As briefly mentioned in the last section, Protzen agrees with the idea that bronze tools were used; however, he argues it was only as wedges in order to separate stone blocks or detach them from bedrock. Evidence of this kind of bronze tool usage is in fact visible in two of the most used quarries situated in the Sacred Valley, the quarries of Kachiqhata and Rumiqolqa. Recent excavations at Qori Kancha Temple have also unearthed some small copper chisels, possibly employed to polish the stone blocks in the circular structure of this Temple of the Sun.

There has been much speculation about how the Inkas dressed their stone blocks. The most accepted technique is the use of hammerstones, as mentioned earlier. In order to understand exactly the technique used by the Inka masons to dress the stone blocks, I will follow Protzen’s experiment carried out in the 1980s.  Protzen began his experiment by choosing three different groups of hammerstones. The first weighed between 8 and 10 kg, the second 2 to 5 kg and the last less than 1 kg. As a second step, he selected a block of andesite measuring about 25 by 25 by 30 cm. After knocking off the largest protrusions to give the stone block the shape of a parallelepiped, he started pounding the block with a 4 kg hammerstone. If the hammerstone is dropped from a height of about 30 cm, it will rebound by 15-25 cm.  It can be allowed therefore to fall again without much effort. The efficiency of the strike can also be increased by giving the hammerstones a twist with the wrists just before it drops onto the surface of the stone block. By employing this sudden movement, the angle of impact is increased by up to 30-40 degrees, and consequently, also the efficiency of cutting is augmented. Once the block’s six faces have been smoothed, a small hammerstone (half  kilogram) must then be used to graze the edges. The business of changing the hammerstones (e.g. from big to small) should be repeated on each face of the stone block.

The entire process, from squaring the stone block to drafting five edges and finishing three sides, took Protzen no more than 90 minutes. What the experiment revealed was that, contrary to modern day thinking, the technique of stone dressing is very efficient and a skilled Inka mason, after years of experience, could have dressed a similar-sized stone block in less than one hour. Protzen repeated the same kind of experiment at the Kachiqhata Quarry where the huge Ollantaytambo stone blocks were quarried. Using the same procedures, Protzen worked out that a block of stone measuring 4.5 by 3.2 by 1.7 m would have taken 20 quarry workers about 15 days to complete. Consequently, if we assume that at Kachiqhata Quarry the labour force was say c. 300  workers, the total number of 150 rhyolite blocks found next to the quarry would represent no more than eight months of work.

Physical evidence on Inka stone blocks for usage of the hammerstone pounding technique is abundant. On almost all stones of Inka walls, regardless of the type of rock, one can find scars resembling those left by the pounding on Protzen’s experimental stone block. Unfortunately, this evidence does not prove with certainty that the Inkas always used this technique, but it does at least confirm the usage of  hammerstones as mentioned in some 16th and 17th century chronicles.

One of the most debated aspects of Inka stone masonry is the transport of stone blocks from quarry to building site. According to a number of scholars, stone blocks were transported with the help of ropes and wooden rollers. The usage of animals (llamas in particular) in stone block transport has also been well discussed. It has been argued that llamas were employed to transport small stone blocks weighing up to 60 kg.

While Protzen is uncertain as to how the finely dressed stone blocks were transported from Rumiqolqa, he firmly denies any employment of wooden rollers for moving the gigantic blocks at Kachiqhata Quarry. Bengtsson holds the same opinion, even though she argues that the archaeological evidence for wooden implements could have easily been destroyed by the area’s seasonally humid  conditions. However, an examination of some of the stone blocks from Kachiqhata Quarry reveals clues which fully sustain Protzen’s hypothesis. In fact, at least one side of the huge stone parallelepipeds has a smooth, yet uneven, polish traversed by fine, more or less parallel striations.

The presence of drag marks does not exclude the possibility that the stone blocks were moved in some other way, especially if one considers the large number of people needed to drag these enormous blocks, and the fact that labour was the standard and only means of taxation. The force (K) required to drag any stone block depends on the coefficient of friction (f) between the stone and the material of the ramp, the weight of the block (P) and the slope of the ramp (a):

K= f . P . cos  a  +/-  P . sin  a

The + is used to compute the force needed to pull uphill and the - to drag the stone downhill. After some experiments on ballast of broken rock and compact dirt, a coefficient of friction of 0.75 was obtained for the former and 0.7 for the latter. Knowing that one of the largest stone blocks weighed about 110,000 kg and assuming that a person could pull 50 kg, some 2,000 people would be required to drag such a rock up an 8% gradient.

Quite often scholars try to find answers that are logical within the balance between labour, time and environment, but as archaeologists we make poor palaeopsychologists. I mean, it needs to be remembered that the thinking of ancient people would often have been different to that of today. A similar waste of energy to transport stone blocks also happened within the Tiwanaku culture long before the Inkas. Maybe, in spite of the long distance between the quarry and the building yard, the transport matter was easier than what we can imagine today, or owing to certain cosmological beliefs difficult stone block transport had other cultural implications.

As briefly mentioned in the introduction to this chapter, contrary to what was believed a couple of decades ago, recent studies have revealed that there are more than two main types of stone wall construction. There are five distinct types of Inka stone wall masonry: rustic, cellular, encased, sedimentary and cyclopean. These classifications can be further subdivided according to stone block shape, wall cross-section, wall block lay-out, profile and texture. The adoption of these subdivisions depends on the type of wall. In fact, one wall may have them all, while another may have just one type.

Cross-sections: There are eight diverse types of cross-sections: piled-up, wedged, denticulate flat-fitted, encased, braced, dowelled and stapled. In some cases these cross-sections can only be used with specific wall styles, while in some others a single style can have two or more cross-sections.

Stone block shapes: The Inkas did not have a specific standardised block shape for their stone wall constructions. According to the type and lay-out of  the wall, the Inka masons adopted diverse block shapes which have been classified into five different forms: polygonal, pentagonal, natural, tetrangular and rectangular. The first three types (polygonal, pentagonal and natural) can be used in various wall constructions from rustic to sedimentary, whereas the last two, tetrangular and rectangular, can only be employed in the encased type of wall.

Stone block joints: Apart from the rustic type of wall, the precise joints of the stone blocks are an extremely important characteristic of Inka stone wall  masonry. The finest type of joint is the carved or sunken form which gives the whole wall structure a “plastic” appearance. There are three other styles of joints: natural, rustic and refined. Especially in royal buildings where refined joints are used, the fitting technique is so perfect that even a razor blade cannot be stuck in the joint of two blocks.

Profiles: The last elements of stone wall construction subdivisions which need to be taken into account are the profiles, which like the joints and block shape vary according to the masonry style used. Six different types of profiles have been recognised: natural, rough-hewed, cushioned, convex, bevelled edges and flat. These profiles also have three different kinds of textures: coarse, rough and smooth.

The usage of diverse profiles and textures follows specific patterns. For instance, with an encased rectangular-blocked wall, it was common to employ a flat profile with a smooth texture, whereas sedimentary and cyclopean stone wall structures required convex or cushioned profiles with a fairly rough texture.

In some cases, especially with the sedimentary type of wall, distinctive protuberances on the stone blocks can be noticed. The function of these projections is not completely clear. Some scholars maintain that they may have served as points to which ropes could be attached, or as places where the force of a lever could be applied. These protuberances were apparently cut at the building site and in some cases left as decorative features rather than being cut off.

The rustic style of stone wall masonry consists of a superimposition of stone blocks which have not previously been worked. The material used in this type of structure varies from place to place.  In the Cuzco Valley, for example, calcite and grey diorite were used. Most of the time they were randomly collected from natural stone outcrops because the stone blocks were not worked, and because  their size was never standardised. In a rustic wall either small or big stones can be found, depending on the wall structure and the shape of the stone blocks themselves. The rustic style was employed for the construction of retaining walls and in particular agricultural terraces, even though it was sometimes used in so-called royal architecture if the situation required it. As far as the cross-section of this stone masonry style is concerned, it can be wedged or piled-up with a natural-coarse texture profile. Obviously, the stone block  shape is  natural with  natural-rustic joints, which also in this case were very seldom mortared.
 

The stone elements of the cellular style are fitted in a such a way that they come to resemble various organic cell tissues. The most common block shapes are the polygonal and pentagonal, and they are accompanied by rustic, sometimes  carved joints. The profile of the wall is mainly convex, and according to the use of materials (generally calcite, andesite and sandstone) it can have either a rough or smooth texture. The cross-section is denticulate, but in some structures the encased type can also be found. The cellular wall structure is mainly used for canalisations, retaining walls and agricultural terraces, and only very rarely in royal buildings.

With the encased style of wall construction the stone blocks are fitted together like the pieces of a jigsaw puzzle, giving the structure a high level of stability. Because of the relevant size of the stone blocks (between 1 and 2 m2 on the side), this   style requires very hard materials such as andesite, diorite and basalt. Stone blocks in an encased wall normally have various sides, not only from a two-dimensional, but also from a three-dimensional perspective. The cross-section can therefore be difficult to determine. It can be encased, braced, dowelled and sometimes even stapled. The stone blocks do not have a standard shape, that is, it varies according to their size and the structure of the wall. The facing surfaces can be either convex or raised from carved sunken joints. This type of Inka wall masonry was used in royal  buildings around the whole sacred valley of Cuzco. One of the best examples is the virtually world famous twelve-sided stone block found in Hatunrumiyoc Street in Cuzco. This form also occurs in the Temple of the Three Windows in Machu Picchu.
 

The sedimentary style has acquired its name because the lay-out of the stone blocks resembles geological strata. The materials used in this stone masonry style are calcite, andesite and only rarely basalt. As far as the block shape is   concerned, it varies from tetrangular to rectangular and sometimes even cubical, but in all cases the blocks have fine carved or refined joints. The sedimentary type of wall construction has various cross-sections, but the most common ones are encased and denticulate. Concerning the profile, it can be  either cushioned, convex or with bevelled edges, but in most of the cases it has a rough texture. The sedimentary type of wall can be seen throughout the Cuzco region, but the best example is the circular structure, also called the “Solar Drum”, in the Qori Kancha  Temple at Cuzco itself.

Because of the enormous size of the stone blocks, this style has been named cyclopean, or “made by giants”. Some stone blocks are more than 7 metres high and apart from a few modifications, their shape was kept as natural as possible. The accentuated carved joints give the whole wall structure a “plastic” appearance, even though the convex surface is left coarse. This kind of stone wall style was employed for the construction of fortresses to protect sacred ceremonial places. A famous example is the huge zigzag wall of the Saqsaywaman ceremonial centre, just out of Cuzco.

Top