Evaluating the Design Principles of Traditional Safranbolu Houses ^†

Abstract

Safranbolu, located in the Western Black Sea region of northern Turkey, consists of three historical settlements distinguished by monumental buildings and especially traditional houses. For this reason, the city was declared a ‘World Heritage Site’ by UNESCO in 1994. Traditional Safranbolu houses are examples of Turkish houses built with mixed techniques. In research on Turkish residences and traditional Safranbolu houses, the focus has been on the origin of dwelling construction culture, the spatial reading of the parts of the house, the sociocultural life of the owners, and the surveying of traditional construction techniques. In this study, the design principles affecting the spatial arrangement of Safranbolu houses were evaluated from a holistic perspective. Dwellings selected from Safranbolu sites were analyzed at the urban scale according to the environmental design data of the residential area. It can be seen that houses were designed in a certain order according to these factors. In the building scale analyses, the diversity in the domestic space configuration was surveyed by the drawings of immediate environments at the plans and sections of the houses. The study contributes by explaining the relationship between parcel characteristics and house design.

1. Introduction

Humanity’s building knowledge dates back to very ancient times. Shelters in various forms were built to provide basic needs such as eating, drinking, sleeping, and resting in a safe area. In the historical process, societies have carried out the act of building at different scales in line with their geographical possibilities. The Turkish house consists of spaces developed and interpreted by inheriting various aspects of nomadic life order, Anatolia’s housing culture and geographical data, and Islamic beliefs, which are reflected in the houses.

The spatial organization principles of the Turkish house originate from Central Asia [1,2,3]. Turks lived in the steppes in a system that required constant displacement. They established temporary settlements in the surroundings, which they limited with a covering system known as a tent or yurt. After embracing Islam, they came to Anatolia in the 11th century and synthesized their own building knowledge with the design dynamics that existed in this new geography. Accordingly, the Turkish house contains traces of the mudbrick houses in Central Anatolia, houses with courtyards and iwan in Mesopotamia, and the Byzantine house. From the 16th century onward, the arrangement of ‘house with hayat’ was introduced [4]. For this reason, the earliest examples of the classical Turkish house are dated to the 16th century, and the latest examples belong to the 19th century [5].

The most basic unit constituting the Turkish house is the ‘room’, which represents an independent and specialized space. The ‘hayat’ is a circulation space connecting the rooms and maintains social relations. The Turkish room is like a tent, where all needs are met under a cover with special and distinctive interior space elements [6]. The house form is established by arraying the tents, where family members live separately, around a common area. This diagram explains the relationship between rooms and hayats in the Turkish house [3]. Semi-open hayats were converted into closed spaces called ‘sofa’ in time [4]. In the sofas, which bring together various numbers of rooms, some areas are specialized and turned into ‘eyvan’ (iwan). The iwan placed between the rooms provides privacy. At the same time, the sofa, which is surrounded by rooms, receives daylight due to the windows in the iwans. Apart from this, spaces such as an animal shelter, kitchen, storage/cellar, shop, toilet, and hammam can be found in different parts of the house.

In the first research on Turkish houses, the upper floors where the living units are located were classified according to the room–sofa relationship. Plan types were diversified by arranging rooms of a standardized shape and size around the sofa (hayat) of variable form. Taking the number of rooms and sofas as a reference, Eldem categorized the houses as without sofas, open sofas, inner sofas, and central sofas [5]. This approach has influenced many studies. Kuban drew attention to the socio-cultural dynamics that constituted the home, showing that the concept of privacy and the woman’s role within the family shape the space [4].

Küçükerman considered the location and number of rooms in the residence. He stated that the sofas, which provide diversity in the plans, were shaped to adapt to climatic conditions. He mentions that the upper floor was built with the same logic in every region, and that the natural relations with the topography diversified the ground floors [3]. In Asatekin’s study, the impact of the lifestyles of communities, producing unique house examples in various cities of Anatolia, on the space was emphasized. The relationship between the owner and the house was discussed with the spaces shaped according to the concept of privacy [7].

Dalkılıç and Bekleyen stated that Diyarbakır traditional houses developed around the courtyard just like Middle Eastern houses. The spaces around the courtyard were arranged in the plan to adapt to the climate, and the inhabitants lived in various rooms on different floors of the house according to the seasons. The study highlighted the importance of the solarization factor in the spatial arrangement principle [8].

The first study to mathematically express the formation of the Turkish house was conducted by Çağdaş. In the study, which was based on sofa–room relations formed according to specific rules, the genealogy of the Turkish house was explained by the grammatical derivation method [9]. Similarly, Güzelci showed the plan grammar by defining the iwans and kiosks on the sofa [10]. Form analyses were based on Eldem’s classification approach, which only includes floor plans. The design factors affecting spatial organization were not questioned. Çolakoğlu, on the other hand, analyzed and interpreted the grammar of Sarajevo hayat houses. The hayat, which is the most important design element in the plan, was also the focus of new house designs. The mass composition of the house was created by the grammatical derivation method according to plan diagrams and third-dimensional relationships. It has also been pointed out that traditional houses contain interpretable spaces for modern housing designs [11].

Yürekli developed a matrix-based method to identify houses with similar stylistic characteristics without harming residential privacy. The ground floor was evaluated according to the entrance, and the house, garden, and street relations were described [12]. The study is noteworthy for mentioning the relationship of traditional dwellings with their immediate surroundings. However, a cause-and-effect relationship between interior house spaces and the environment has not been investigated. Instead, a method facilitating external observation has been developed.

Yürekli states that traditional houses should also be evaluated in a cross-section plane. The Turkish house has a fluid connection, visually and physically, with the street or garden. Integrated with the garden, there is passive balance with the climate, providing fresh air, functional arrangements such as animal care, vegetable processing, bread baking, etc. Its characteristics show that the Turkish house is very close to the universal pursuits of 20th century of modern architecture, and even a pioneer [13].

Petruccioli analyzed Mediterranean and Islamic cities at the urban and building scale. The number of spaces that make up the house and their relationship with the topography were evaluated together with plans, sections, and elevations. The housing typology in cities was analyzed with a morphological approach. In the study, it is mentioned that the Turkish house should be analyzed together with its immediate surroundings; it is not enough to evaluate only the upper floor plans to express the spatial richness and design practices of the houses [14].

In light of these studies, it has been determined that the perspective of research should be extended, and in-depth analyses of space should be conducted. The aim was to examine the Turkish house at the urban and building scale together with its immediate surroundings. For this purpose, Safranbolu houses, which are included on the World Heritage List, were analyzed within the scope of the study. The effects of environmental elements on the design of the house were emphasized. Traditional h ouses were analyzed in terms of user profiles and urban use type according to the sites they are located in. Subsequently, parcels, the most fundamental design element in urban planning, were evaluated. The impact of parcel data on the place where the building will be constructed at the parcel, and the effect of the orientation of the parcel entrance street on the layout of the domestic spaces of the house are highlighted. With this approach, the spatial usage logic of the Turkish house has been identified.

2. Material: Safranbolu, Historical, Physical Context, and Climate

Safranbolu, Karabük Province, is in northwestern Turkey (Figure 1). This historical town, remarkable for its natural sites, archeological sites, monumental buildings, and traditional houses, is the first place where the cultural heritage conservation research in Turkey began. Safranbolu houses are an ideal reflection of the traditional architectural characteristics and cultural richness of Turkish houses.

According to the maps in Charles Texier’s book ‘Asia Minor’ (1862), Safranbolu and its surroundings are called Paphlagonia. The existence of a city called Zağfiranboli, famous for the saffron trade, was mentioned in the region. However, it has been reported that the architectural monuments here have been destroyed [15]. Safranbolu and its surroundings were used as a residential area during the Romans, Byzantines, Seljuks, various Beyliks, and the Ottoman Empire [16]. The construction culture in the region was inherited from these societies and has been passed on to the present day. In 1196, until the Turkish conquest of Safranbolu, local people lived in the citadel; after, Turks built a new city around this stronghold [17].

Safranbolu is located on the caravan route connecting Central Anatolia to the Western Black Sea region. With the public improvement works carried out according to these commercial routes, Safranbolu developed between the 16th and 19th centuries and gradually attained its present appearance. Today, the historical city of Safranbolu allows one to experience traditional life, and consists of three settlements: Şehir (Old Bazaar), Kıranköy, and Bağlar, which were identified as areas whose architectural character should be preserved in the 1968 Master Plan [18]. In the 1990 Master Plan, these areas were taken under preservation as urban sites [19], and the natural beauty of Safranbolu and the tumulus on the hills were also declared as natural and archeological sites (Figure 2). With the 2010 Master Plan, a buffer zone was defined around these sites [20], thus aiming to preserve the authenticity of Safranbolu.

The urban sites are located between the lowest point at 300 m and the Sarı Çiçek Hill at 1750 m (Figure 3). Summers are hot, winters are cold, and springs are mild and cool in the region. Spring and autumn are quite long. The prevailing wind direction is primarily west, southwest, southeast, and northwest [21].

The place where the caravan route reached the city, where commercial activities are concentrated, and where the Citadel is located, is the Şehir urban site known as ‘Old Bazaar’. This is the oldest settlement, and the place where non-Muslims lived after the conquest is known as the Kıranköy urban site. Located in nearby elevations, Kıranköy and Şehir are naturally separated from each other by the canyons formed by the Gümüş River. With increasing prosperity during the Ottoman period, Bağlar became a secondary residential area. Bağlar is in the north, on the south-facing slopes of the hill, and Muslims had vineyard houses (bağ evi) in this neighborhood. Urban sites are structurally quite rich. Traditional Safranbolu houses were built of different types such as Şehir houses, Greek (Rum) houses, vineyard houses, and mansions with pool (Figure 4, Figure 5 and Figure 6). There were also outbuildings such as pool pavilions and pool rooms in the gardens of the houses. The housing typology emerged with dominant rules in the Şehir urban site, and diversified in other sites.

At the Şehir site, utilizing the variable terrain slope, climatic comfort, and security have been ensured. Here, traditional houses are located around a commercial center consisting of a mosque, han, arasta, hammam, fountain, and craft-based workshops. The commercial center is within the triangle where the Akçasu and Gümüş Rivers meet. Houses are arranged in this area and along the valleys formed by these two rivers.

The harsh winters in Safranbolu made it necessary to live side by side in protected areas in the valley between the canyons. Therefore, Şehir houses were built in sloping parcels on the outskirts of the valley. The layout scale was organized so that each house could benefit from the daylight and see the view without blocking each other. Families of Şehir lived in the vineyard house, which was cooler in summer, and in Şehir house, which was warmer in winter. At this site, houses with or without small gardens were dense, and there were also families living permanently in houses with large gardens. There were very few residences with commercial units on the ground floor [21].

Kıranköy emerged simultaneously with Şehir in terms of historical development. With the establishment of Turkish rule in the region, it was settled on higher and flatter ground outside the citadel. The non-Muslims of Safranbolu always resided in Kıranköy and commuted to Şehir every day for trade, and there were usually commercial sales units on the ground floor of the houses [21].

Bağlar has been a summer resort for its inhabitants since the 19th century. It is quite simple and integrated with nature. Due to the transformation of the terrain structure into flat and wide parcels, the housing texture is sparse. The large south-facing gardens were irrigated with water from the mountains. High quality agricultural products were grown here. In addition to vineyard houses, there were also houses where crowded families lived permanently [21].

3. Traditional Safranbolu Houses

Turkish residence plan typologies are divided into seven, according to geographical regions [23]. Safranbolu houses, the subject of the study, are examples of the Black Sea coast and hinterland group. These traditional houses express the cultural level and economic prosperity of the public in space and have attracted attention since the 1970s due to their spatial characteristics. The events of the European Architectural Heritage Year in Turkey (1975) were held here, and efforts to document the historical houses with drawings were initiated. Although their owners and functions have changed, traditional houses have preserved their existence until today (Figure 7).

The inhabitants of Safranbolu engaged in agriculture and animal husbandry and produced food and clothing at home. Crowded family members lived together respectfully, observing privacy. The inside of the house and female owner of the house could not be seen from the street. Depending on the family’s economic status and the size of the house, there were arrangements for the separate use of entrances, floors, or parts of the house for men and women [21]. In some houses, the house was entered through different doors on the same street, called the haremlik entrance for women, and the selamlık entrance for men. These entrances were designed to be completely independent of each other by being positioned on different streets according to the condition of the parcel and on different floors according to the slope usage. The houses, reflecting the lifestyle of the inhabitants, were built in a detached layout, with one or two stories above the ground (Figure 8), and spaces with different functions on each floor (

Table 1. Parts of a traditional Safranbolu house.

Space Name	Spatial Characteristic	Spatial Role
Hayat	This is covered with soil where daily work is conducted.	Circulation unit
Taşlık	This is covered with stone where daily work is conducted.	Circulation unit
Ahır (animal shelter)	This is the area where the animals feed.	Service unit
Sofa (hall)	It is a common area that connects the rooms.	Circulation unit
Eyvan (iwan)	This is elevated with steps or surrounded by railings, where the sofa receives light and where the household sits.	Circulation unit
Oda (room)	This is a special place designed for the basic needs of family members such as eating, drinking, and resting.	Private space
Mutfak (kitchen)	This is the cooking area.	Service unit
Hela	This is the toilet unit.	Wet area
Abdestlik	This is the washbasin unit.	Wet area
Gusülhane (ablution place)	This is the bathing cubicle.	Wet area
Dış çardak	This is an outdoor space for drying vegetables and fruits, accessed from the kitchen or the sofa.	Service unit
Depo (storage)	This is an area where the crops are stored.	Service unit
Mahzen (cellar)	This is an area where the wine is stored.	Service unit
Dükkân (shop)	This is the commercial unit connected to the street.	Service unit

).

The most important element affecting the mass expression of houses is the street [21]. Compared to the courtyard house type, it has been observed that the houses follow a street-garden-house layout. Public space and semi-private space are separated by garden walls. Ground floors consisting of semi-open spaces sheltered from the outside are associated with the garden. The private space is on the upper floors, where the living units are located. The ground floors consist of service units that connect the private and semi-private spaces. As in Turkish houses, the visual relationship between the private space and the public space is provided on the upper floors by projections with frequent windows overlooking the street.

Another design factor affecting the construction of traditional houses is the parcels. The ground floor or garden walls of the house follow parcel and street boundaries. When evaluating the parcels in Şehir, Kıranköy, and Bağlar, it was observed that the ground floor of the houses in sloping, small, and non-uniform parcels followed the street boundary. In this case, the middle floor prepares the upper floor like a transition story. On the upper floor, a steep-sided layout is established with projections, and the rooms become larger. Flat and large parcels enabled the ground and upper floors to develop in the same arrangement [21].

It is noteworthy that environmental design data were considered in the design of the dwellings, thus ensuring spatial comfort. In the houses, the principles of the right orientation on the winter (middle) floor, keeping the floor height low, and reducing the window sizes on the facades provided heat gain and reduced the heating load. On the summer (upper) floor, natural ventilation was achieved and the spaces cooled by increasing the ceiling height and placing windows according to the predominant wind direction [25,26].

Safranbolu houses are examples of outer sofas, inner sofas, and central sofa types. Houses with outer (corner) sofas are common, and iwans provide diversity in this type [27]. Row-room plans lined up along one side of the sofa, appealing to small families, were not preferred. Plan types suitable for living in crowded families, with rooms facing each other and generally featuring symmetry between the rooms, were developed [28]. The maximum number of rooms in the floor plans was four, and three- and four-room plans were common. The number of iwans varied between one and four. Plan schemes required the usage of iwans as toilets, storage, and stairs [29].

Bayazıt divided the floor plans of traditional houses in Bağlar, Kıranköy, and Şehir into types according to the number of rooms, sofas, and iwan usage types and schematized the spatial relations. In the schematic drawings, balconies (dış çardak) were not considered, and stairs and toilets were shown as solid. In the study, the spatial relations of all floor plans were expressed graphically. However, the morphological structure was analyzed according to the upper floors [24].

It is possible to classify houses according to the presence of a haremlik-selamlik entrance, the number of floors, the sofa–room relationship, and the number of rooms. This study did not aim to create any classification-typology. Traditional houses were analyzed with their parcels and their immediate surroundings, and the design factors impacting the plan scheme of the house plans were expressed. It has been observed that despite being constructed in different sites, the houses had recurring planning decisions.

4. Method

The methodology of this study was determined by considering Petruccioli’s observation that the immediate environment of the Turkish house had not been investigated in a multidimensional way, Yürekli’s more holistic approach to traditional housing within the framework of form, and Bayazıt’s research of Safranbolu houses with a focus on the number of sofas and rooms [12,14,24]. The focus of the research framework was the parcels that shaped the dwelling.

In this study, 52 traditional houses built in historical Safranbolu settlements were analyzed in depth. Unlike other studies on the Turkish house, Safranbolu residences have been systematically examined step by step from the urban scale to the building scale, therefore, this study tried to provide a scientific contribution by evaluating the multifaceted design logic of Turkish house culture with a holistic perspective.

In the first phase of the study, Safranbolu’s 2010 Master Plan and the site plans of all houses were schematized, compared, and evaluated at the settlement (urban) scale. Then, all houses were analyzed at the building scale with schematic floor plans and cross-sectional drawings indicating the immediate surroundings. The spatial information of the houses was explained in tables according to the design principles defined under headings, and numerical data were obtained. Thus, all plan schemes were compared and the reasons for the differences in interior space solutions were investigated. In the discussion part of the study, plans and sections of twenty-three typical traditional houses, five from Bağlar, three from Kıranköy, and fifteen from Şehir, are presented.

The stages of the analysis that examined the relationships between traditional Safranbolu houses and their immediate surroundings are summarized as follows:

-

Site scale analysis (information on the parcel)

• Parcel usage: parcel form, size, and accordingly determination of the sunlight period direction of the parcel;
• Parcel street relationship: defining the parcel type according to the number and width of the streets around the parcel;
• Determining the orientation of the entrance street of the parcel: main street west, east, south, or north;
• Parcel landscape relationship: defining the dominant view direction of the parcel.

-

Building scale analysis (information on the structure)

• Building location: determining the type of building positioning on the parcel according to the condition of the street;
• Parcel entrance and number of entrances: determination of entrance characteristics according to the type of building positioning;
• Defining the number of floors of the building;
• Slope usage: determining the use of elevation in the basement or ground floor according to the slope of the parcel;
• Location and orientation of projections: defining open and closed projections according to the dominant view and entrance street direction and sunlight period factor.

-

Analysis according to the orientation of the entrance street (information on spatial arrangement)

• Hayat/Taşlık;
• Animal shelters;
• Sofas;
• Rooms;
• Kitchens;
• Wet areas (toilets and washbasins);
• Balconies (dış çardaklar).

As part of this research, the settlement plans of traditional houses with close surrounding expressions were analyzed. In the first stage, in order to compare the design data of the parcel on which the house was built, traditional parcel types in Safranbolu were examined according to the headings of form, size, orientation (sunlight period), number, and width of streets around the parcel, dominant view direction, orientation of the street providing access to the parcel/house, and elevation difference within the parcel. Accordingly, the comparison criteria within each category varied. Traditional houses were analyzed according to the location at the parcel and entrance of the parcel/building, the number of entrances, the slope usage, and the location of projections. Thus, the impact of the parcel–street relationship on the location of the building and interior units (hayat, animal shelter, room, sofa, wet areas) and the diversity of spaces in different parcels was evaluated.

5. Results

Environmental design data were effectively incorporated into the design of Safranbolu houses, resulting in logical solutions [30]. According to the evidence of the study, the planning approach of traditional houses, each of which appears unique, was similar. The environmental design data of the parcel on which the house was built significantly influenced the plan, section, and façade design of the house as well as the major design decisions emerging from the site assessments. The quality and diversity of the interior space changed according to the spatial needs of the owners and functional usage differences. The similarities and differences in the design principles were identified by comparing the site where the house is located, the natural characteristics of the parcel where the house was built (environmental design data), and the orientation of the street that provides access to the parcel or the house.

5.1. Site Scale Analysis (Information on the Parcel)

The most basic unit that constitutes the traditional urban structure of Safranbolu is the building parcel. The general settlement logic of the residences can be explained by comparing the design factors in the parcels.

5.1.1. Parcel Usage

When the parcels where the Safranbolu houses were built were analyzed, it is noteworthy that there were quite large and flat parcels in the Bağlar site sloping toward the south. While the parcels on the eastern and southern slopes of Kıranköy were quite steep, there were smooth and slightly inclined parcels in the northern and southwestern parts. Small and sloping parcels were predominant in the Şehir site. The general rule in the location of the building floor area in the parcels is that the dwelling was close to the street (Figure 9). According to the slope of the parcel land, dwellings were placed on the elevated side of the parcel concerning the street at the highest elevation. Thus, it aimed to dominate the view and benefit from daylight for as long as possible. For example, the houses in Bağlar were located on the north side of the parcel. With this placement method, it is noticeable that the amount of garden facing south was kept large.

5.1.2. Parcel–Street Relationship

It is not clear which concept emerged first in the parcel–street–house relationship in traditional Turkish cities. It is accepted that parcels arose spontaneously over time, according to the prevalent street usage. When the urban structure of Safranbolu was analyzed, it was determined that the number of streets had a significant impact on the parcel usage and house design. Accordingly, even the smallest parcel had an efficient space usage. The tendency to establish direct spatial relations with the urban transportation artery was common in houses, therefore, as the number of streets increased, spatial usage diversity was achieved by providing independent entrances from streets at different heights.

Safranbolu urban sites have two types of streets: main streets and side streets. Main streets, which are considered as urban transportation arteries, are generally over 3 m wide. The width of the side streets (sometimes dead-end streets) is less than 3 m. When the streets around the parcel were analyzed according to their width, number, and function in urban transportation, four types of parcels could be identified (Figure 10). The parcels surrounded by one main street are examples of parcel attached to one street (S1). In parcels attached to two streets (S2), and attached to opposite streets (KS2), one of the streets may be a main street and the other an alley. In another case, both streets are main streets. In parcels attached to three streets (S3), all streets can be used as main streets or at least one street is an alley.

When the houses were evaluated according to parcel type (

Table 2. Types of parcels on urban sites.

Urban Site	Parcel Attached to One Street	Parcel Attached to Two Streets	Parcel Attached to Opposite Streets	Parcel Attached to Three Streets	Total Number of Houses
Bağlar	4	5	3	1	13
Kıranköy	5	1	0	2	8
Şehir	8	11	5	7	31
Total	17 (33%)	17 (33%)	8 (15%)	10 (19%)	52

), it can be seen that parcels attached to one and two streets were common. Parcels attached to one and two streets were more common at the Bağlar and Şehir sites. In Kıranköy, while parcels attached to one street were predominant, during the scope of the study, no examples of parcels attached to opposite streets were encountered.

5.1.3. Orientation of Parcel Entrance Street

The direction of the main street associated with the parcel determined the location of the house’s entrance façade and the entrance door. Accordingly, it affected the layout of units in the interior space arrangement. The streets provided entrances from the west, east, south, and north (

Table 3. Orientation of entrance street at the parcel.

Total Number of Parcels	West	East	South	North
52	19 (37%)	16 (31%)	9 (17%)	8 (15%)

). Accordingly, the entrance facades of traditional Safranbolu houses were generally oriented to the west and east (Figure 11).

5.1.4. Parcel–Landscape Relationship

Traditional Safranbolu houses have the principle of seeing the view [21,31]. According to the investigations, the dominant view direction varied in the parcels. Depending on the view’s direction, open and closed projections were located accordingly. Projections in various forms carry the quality of interior space to a different point. The parcels in the Bağlar had a view of lower Kıranköy to the south and the forest-covered hill to the north. The houses on the eastern and southern slopes of Kıranköy faced toward Şehir. In Şehir, the canyons surrounding the bazaar and the hills they form are the scenery. Some parcels overlook the valleys, while some overlook the water. According to this approach, it can be concluded that the view direction was included in the design of traditional houses.

5.2. Building Scale Analyses (Information on the Structure)

After analyzing the parcels where Safranbolu houses were built, the decisions determining the location of the building spots on the parcel were examined. According to the defined criteria, the situations impacting buildings were analyzed.

5.2.1. Building Location

In Safranbolu houses, integration with the street and establishing visual and spatial relationships were constructed in various ways. The number of streets around the parcel and the width of the streets affected the positioning of the building on the parcel. Accordingly, traditional buildings were placed on the parcel in four different ways: building attached to one street, building attached to two streets, building attached to three streets and building positioned on the parcel.

In each parcel type, the building locations differed (Figure 12). In parcels attached to one street, it was common for buildings to be built attached to the street (88%). In parcels attached to two streets, buildings tended to be attached to one street and two streets (47%). In parcels attached to opposite streets, the building was either attached to the street or positioned on the parcel (50%). In parcels attached to three streets, buildings attached to two streets predominated (70%).

Building location types are presented according to urban sites (

Table 4. Types of building locations on urban sites.

Urban Site	Building Attached to One Street	Building Attached to Two Streets	Building Attached to Three Streets	Building Positioned at the Parcel	Total Number of Houses
Bağlar	10	0	0	3	13
Kıranköy	5	2	1	0	8
Şehir	13	13	0	5	31
Total	28 (54%)	15 (29%)	1 (2%)	8 (15%)	52

). One building in Kıranköy was the only example attached to three streets. Buildings attached to one street were common. Other factors that determined the type of building location were the size of the parcel, the presence of neighboring buildings, or neighboring parcels. In some large parcels in Bağlar and Şehir, it was observed that the buildings were freely located. The presence of neighboring buildings or neighboring parcels restricted the positioning of the building. In each building location type, unique house examples were built according to the street direction and slope.

5.2.2. Parcel Entrance and Number of Entrances

When the parcels were analyzed according to the number of entrances, it was seen that one to three entrances were given. There were twenty-eight parcels with one entrance, twenty with two entrances, and three with three entrances. Only one house had five entrances to the parcel (

Table 5. Number of parcel entrances and entrance characteristics according to building location type.

Number of Parcel Entrances	Entrance Location	Building Attached to One Street	Building Attached to Two Streets	Building Attached to Three Streets	Building Positioned at the Parcel
Single Entrance Parcel (28) (54%)	Entrance through the building	10	7	1	-
	Courtyard entrance	2	-	-	7
	Courtyard Entrance (building with haremlik and selamlık entrance)	-	-	-	1
Double Entrance Parcel (20) (38%)	Entrance through the building and shop entrance	1	-	-	-
	Courtyard entrance and entrance through the building	12	-	-	-
	Haremlik entrance, courtyard entrance, and selamlık entrance	2	-	-	-
	Selamlık entrance, courtyard entrance, and haremlik entrance	1	1	-	-
	Building with haremlik and selamlık entrance	-	3	-	-
Three-Entrance Parcel (3) (6%)	Entrance through the building, shop entrance, and courtyard entrance	-	1	-	-
	Haremlik entrance, Selamlık entrance, and courtyard entrance	-	2	-	-
Multi- Entrance Parcel (1) (2%)	Selamlık entrance, courtyard entrance, haremlik entrance, and shop entrances	-	1	-	-

). In single-entrance parcels, it was common to enter the building from the ground floor wall attached to the street. In parcels with double entrances, it was common to enter through the building and from the courtyard. Diversity was provided with building entrances in houses (Figure 13). In houses with a commercial unit on the ground floor, it was observed that there was another entrance door for the shop independent from the house entrance. In some houses, the entrances were divided into haremlik (women’s entrance) and selamlık (men’s entrance).

5.2.3. Number of Stories

It was observed that the use of floors in the houses was in the form of a basement, ground, middle, and upper floors. The floor on which the hayat/taşlık space is located is characterized as the ground floor. Of the fifty-two houses analyzed, six of them had two floors, one floor above ground. Forty-five houses had three floors (87%) including ground, middle, and upper floors. Only one house had four floors including a basement, ground, and upper floors. Accordingly, traditional Safranbolu houses were predominantly built with three floors, as the Safranbolu family structure and lifestyle are effective with this preference.

5.2.4. Slope Usage

The main space on the ground floors of the houses is the hayat/taşlık. Floor heights varied between 3 and 4.5 m on average. The slope usage in the building was evaluated according to the relationship of the basement and ground floors with the natural ground. On flat parcels, the building was completely above ground. If the elevation difference within the parcel varied by up to 3 m, one wall of the building could be partially buried in the ground (Figure 14). On parcels of 3 m and above, one or two walls of the building were designed to be completely buried in the ground. Thus, the house could be accessed from different streets with independent entrances on different floors.

Houses built on flat plots in the vineyards did not use slopes. No basements were found, and the ground floors were built above ground. In some houses in Kıranköy, partial basements under the ground were found to be used as wine cellars. The cultural diversity of the residents diversified the spatial use. In sixteen houses in Kıranköy and Şehir, it was observed that the slope difference within the parcel was 3 m or more. Therefore, the basement floor of five houses and the ground floor of eleven houses (69%) were buried in the ground. Only in one house in Kıranköy was the basement completely under the ground, even though the in-parcel elevation difference was less than 3 m. In two traditional houses, one in the Şehir and the other in Bağlar, it was observed that although the building was attached to one street, the entrance to the building was not through the ground floor wall. From the lower elevation of the parcel, one entered the courtyard through the garden gate and reached the building entrance. On the rising side of the parcel, the building facade was partially buried in the ground. This is remarkable in terms of obtaining a smooth ground floor by arranging the entrance to the parcel according to the slope (Figure 15).

5.2.5. Location and Orientation of Projections

Traditional houses have two types of projections, open and closed. Open projections, called ‘dış çardak’, were designed for the daylight needed in traditional production methods. These are located at one end of the sofa and are also oriented toward the view. Closed projections establish a visual relationship with the street and the landscape. In a house, closed projections can be in the sofa or room. With these geometric extensions, it is possible to see the street or the view from different perspectives.

5.3. Analyses According to the Orientation of the Entrance Street (Information on Spatial Arrangement)

In addition to the site and building scale analyses, the third stage focused on the principles that defined the interior design. The most important of these decisions was to plan the house by choosing the most favorable direction from the streets around the parcel. The basic units that emerged in the Safranbolu house were the hayat/taşlık and animal shelter on the ground floors, and the sofa, room, wet areas, and balconies on the middle and upper floors. The kitchen and storage could be located on each floor according to need. The location of all spaces with regard to the entrance street is explained in the tables.

5.3.1. Hayat/Taşlık

The ground floor of the houses was entered through the circulation and service unit named the ‘hayat’ or ‘taşlık’ (Figure 16). This area, where various household chores were carried out, was named as a hayat if the floor was covered with earth, and a taşlık if it was covered with stone [21]. In this study, it was determined that the orientation of the entrance street determined the location of the hayat/taşlık in the plan. Animal shelters were positioned according to the orientation of the hayat/taşlık.

When forty-nine houses were evaluated, in the parcels with predominantly western and eastern entrances, and in the parcels with southern entrances, the hayat/taşlık were entirely in the south (

Table 6. Hayatlar/taşlıklar according to the orientation of the entrance street.

Orientation of Entrance Street	Hayat/Taşlık (Circulation)				House without Hayat/Taşlık	Total
	South	North	West	East
West (17)	11 (65%)	5	1	0	2	19
East (15)	13 (87%)	2	0	0	1	16
South (9)	9 (100%)	0	0	0	0	9
North (8)	1 (12%)	7	0	0	0	8
Total	34 (69%)	14	1	0	3	52

). In the hayat/taşlık, large window openings (gliste) were designed to receive direct southern light. In the parcels with northern entrances, the hayat/taşlık necessarily faced north in order to provide access to the house.

5.3.2. Animal Shelters

The residents of Safranbolu built animal shelters on the ground floors for the animals they utilized. Each animal (horse, cow, donkey) was kept in a different stable. For this reason, the number of stables in the houses varied between one and three. Entrances to the stables were either through a door inside the house or through a door opening to the garden/street (Figure 16).

When forty-five houses with animal shelters were evaluated, it was determined that there were a number of animal shelters located in the north (

Table 7. Animal shelters according to the orientation of the entrance street.

Orientation of Entrance Street	Animal Shelter				House without Animal Shelter	Total
	North	South	West	East
West (15)	11 (73%)	4	0	0	4	19
East (13)	10 (77%)	3	0	0	3	16
South (9)	9 (100%)	0	0	0	0	9
North (8)	2 (25%)	5	0	1	0	8
Total	32 (71%)	12	0	1	7	52

). According to this placement method, the unit with the longest sunlight period was the hayat/taşlık. It is noteworthy that animal shelters were placed on the cold (north) side of the house (heating with animals), and blind facades were preferred to prevent heat loss.

5.3.3. Sofas

The floor plans of the houses within the scope of the study were schematized according to the evaluation logic of Küçükerman, who explained the Turkish house sofa–room relationship, and Bayazıt, who expressed the Safranbolu house sofa–eyvan–room relationship (Figure 17) [3,24]. On the ground floor, the room is an animal shelter and the circulation area is the hayat or taşlık. The middle and upper floors have outer, central, and inner sofas according to Eldem’s plan types [5]. Bayazıt’s method, based on the occupied-unoccupied status of the iwans placed between the rooms, includes classifications according to the number of rooms and the iwan usage. The most important factor (orientation) in the settlement scale analysis of this study was the sunlight period. Depending on the dominant view direction, the presence of neighboring parcels and neighboring buildings around the parcel, the location and function of the iwan changed. In the planning, it was seen that the sofa–rooms–iwan and other units were placed within a framework that would reveal the most effective design.

In order to understand this multifaceted design logic of Safranbolu houses, it was investigated whether the principle of orientation in the ground floor units was present in the living floors. The arrangements made in the plan for the sofas, which are called ‘çardak’ by the residents of Safranbolu, to see the street or the view were identified. In the plans of the houses that appealed to small families, it was seen that the rooms were gathered together and the sofas were used as a room. Thirty-seven houses with outer sofas and corner sofas serving two, three, or four living units were analyzed (Figure 18). Accordingly, in twenty-four houses with a rate of 65%, the sofa was oriented to the south (

Table 8. Outer and corner sofa (hall) according to the orientation of the entrance street.

Orientation of Entrance Street	Outer and Corner Sofa (Hall) (Circulation)							Total
	South	North	West	East	Southeast	Southwest	Northeast
West	1	1	1	2	3	4	0	12
East	1	1	2	0	1	6	0	11
South	0	0	1	0	3	1	1	6
North	0	0	2	0	2	2	2	8
Total	2	2	6	2	9	13	3	37

).

The plans of large houses inhabited by crowded families were of the central and inner room types. With the increase in the number of rooms, symmetrical plan schemes providing access to the rooms from the central circulation area were preferred. Instead of a dark hall in the center or inside, sofas, receiving daylight via iwans located between the rooms, were constructed. In fifteen houses with this characteristic, the iwans could see the street or view (Figure 19). Thus, sofas with a sunny and visual quality enabled the residents to establish relations with the environment. In front of some iwans seeing the view, there is a balcony (dış çardak). When there were neighboring parcels or neighboring buildings around the parcel, stairs or wet areas were located in these iwans. In houses with central and inner sofas, the iwans faced the courtyard or the street providing access to the house (

Table 9. Inner and central sofa (hall) according to the orientation of the entrance street.

Orientation of Entrance Street	Inner and Central Sofa (Hall) (Circulation)		Eyvan (Iwan)
			South	North	West	East
West	7	Seeing the Street	1	0	4	0
		Seeing the View	2	1	0	5
East	5	Seeing the Street	1	0	0	4
		Seeing the View	3	0	3	1
South	3	Seeing the Street	2	0	2	0
		Seeing the View	0	0	0	1
Total	12	Total	10 (67%)

).

5.3.4. Rooms

The most special unit of the Turkish house is the room. This is designed with interior spatial equipment to meet all the needs of a family in the same place. The use of rooms in Safranbolu houses is expressed with different names such as the chief room, bride’s room, winter room, and room without stove. The rooms specified as kitchens, storage, and cellars were kept separate from the daily rooms analyzed under this heading. The main goal in the middle and upper floors that make up Safranbolu houses is for the rooms to face the street as much as possible and receive plenty of daylight. Rooms that did not see the street had a view. Rooms received daylight mainly from two walls. If there were neighboring parcels or neighboring buildings around the parcel and there was no appropriate parcel distance between them, the wall of the room on this facade was left without windows. For this reason, some rooms received light from only one wall. In houses in small parcels, some upper-floor rooms were enlarged with closed projections by crossing the street boundaries on the ground floor. Thus, the room received more daylight and different street/landscape perspectives were obtained.

The rooms in the houses were analyzed (

Table 10. Rooms according to the orientation of the entrance street.

Orientation of Entrance Street	Room								Total
	South	North	West	East	SouthE	SouthW	NorthE	NorthW
West (19)	3	3	8	12	21	20	7	12	86
East (16)	5	4	8	13	17	8	17	13	85
South (9)	3	2	7	7	8	6	5	7	45
North (8)	4	7	5	2	0	8	11	3	40
Total	15	16	28	34	46	42	40	35	256
	5.90%	6.25%	10.90%	13.30%	18.00%	16.40%	15.60%	13.60%

). In general, there was a tendency to orient toward the southeast, southwest, northeast, and northwest, respectively. If intermediate directions were not possible, there was a tendency to locate to the east, west, north, and south, respectively. In the rooms of Safranbolu houses, southeast, southwest, and northeast light was preferred (Figure 20). Thus, effective and controlled daylight was provided in the rooms at different times of the day and year. The number of rooms facing northwest was close to the number of rooms facing east and west. The low number of rooms facing north and south was quite remarkable. Accordingly, in traditional houses, they avoided building rooms that did not receive any sun or were exposed to unwanted sunlight. The rooms were also located in a manner in which to see the street.

In houses where the entrance street was at the west, there were orientations to the southwest (23.25%) and northwest (13.95%) to see the street, and to the southeast (24.4%) and east (13.95%) to benefit from daylight (Figure 21). In houses where the entrance street was at the east, there were orientations to the southeast (20%) and northeast (20%) to see the street, and to the east (15.3%) to benefit from daylight. Rooms that needed less light faced northwest (15.3%).

In houses where the entrance street was at the south, there were orientations to the southeast (17.8%) and southwest (13.4%) to both see the street and benefit from daylight, and to the east (15.6%) and west (15.6%) to benefit from daylight only (Figure 22). Rooms that needed less light faced northwest (15.6%). In houses where the entrance street was at the north, there were orientations to the northeast (27.5%) and north (17.5%) to see the street, and to the southwest (20%) and west (12.5%) to benefit from daylight.

As a result of the evaluations specific to the entrance streets, it was observed that the rooms were placed by considering the sunlight period of the parcel and the orientation of the main street that provided access to the parcel/building. It is noteworthy that the ideal orientation criteria were adhered to in parcels where the distance to the neighboring building or neighboring parcel was short and in compelling parcels with a small parcel area. To ensure privacy, there were no windows in the rooms that had a view of the neighboring parcel and the neighboring building. In the parcels attached to two or more streets, the privacy of the building adjacent to the narrow street had also been taken into consideration. No windows were opened on these facades. In some houses, although there was no neighborliness, the north-facing cold walls were built completely without windowless (Figure 23). Cultural and climatic knowledge were also included in the design of the rooms, the most important unit of the traditional dwelling.

5.3.5. Kitchens

In traditional houses, kitchens could be located on any floor. Of the 52 houses whose plans were analyzed, 41 had kitchens. It was observed that the kitchens were placed to the north in the parcels where the entrance street was at the west, east, and south (

Table 11. Kitchens according to the orientation of the entrance street.

Orientation of Entrance Street	Kitchen				House without Kitchen	Total
	North	South	West	East
West (13)	7 (54%)	5	0	1	6	19
East (13)	9 (69%)	2	2	0	3	16
South (8)	7 (88%)	1	0	0	1	9
North (7)	1 (14%)	6	0	0	1	8
Total	24 (59%)	14	2	1	11	52

). Accordingly, kitchens, which need less daylight, were located on the cold façade with the aim of sun lighting the rooms (Figure 24).

5.3.6. Wet Areas

The units characterized as wet areas in the plans were hammams, ablution places (gusülhane), toilets (hela), and washbasins (abdestlik). One of the two houses with hammams was on the upper floor of a mansion in Bağlar. The other was on the ground floor of a house in Kıranköy. There were various numbers of ablution places in each house. Ablution places, which were placed on top of each other in the floor plans, were usually located in rooms with a stove. It is noteworthy that the ablution places were located side by side or close to other wet areas in the plan.

Toilets and washbasins (hand washing units) were considered wet areas (Figure 25). All 52 Safranbolu houses analyzed in the study had toilet-washbasins ranging in number from one to three. If there was a toilet-washbasin in the house, it was located on the upper floor. If there was a toilet-washbasin on the middle and upper floors or all floors, it showed vertical continuity. In cases where the street providing access to the parcel/home was located on the west, east, and south, it was observed that the wet areas were placed especially on the north façade (

Table 12. Wet areas according to the orientation of the entrance street.

Orientation of Entrance Street	Wet Areas (Toilet and Washbasin)				Total
	North	South	West	East
West	12 (63%)	4	1	2	19
East	7 (44%)	4	5	0	16
South	6 (67%)	1	1	1	9
North	1 (12%)	4	1	2	8
Total	26 (50%)	13 (25%)	8 (15%)	5 (10%)	52

).

If there was a neighboring parcel or neighboring building around the parcel, the plan arrangement changed (

Table 13. Wet areas according to the parcel dynamics.

Wet Areas (Toilet and Washbasin)	Total	Neighboring Parcel	Neighboring Building	Courtyard	Street	Unusable Edge of the Parcel
North	26	16(62%)	6(23%)	4(15%)	0	0
South	13	5	1	5	1	1
West	8	3	0	5	0	0
East	5	1	1	3	0	0
Total	52	25 (48%)	8 (15%)	17 (33%)	1 (2%)	1 (2%)

). If the distance to the neighboring parcel or neighboring building was insufficient, wet rooms were placed on the façade facing this direction to make the most effective use of all facades.

5.3.7. Balconies (Dış Çardaklar)

Dış çardaklar (balconies), which serve traditional production methods, are often part of the sofa (Çardak), which open to the outside (Figure 26). There were twenty-three balconies, nine in Bağlar, six in Kıranköy, and seven in Şehir. In houses where the streets providing access to the parcel were to the west, east, and north, the balconies tended to face south (

Table 14. Balconies according to the orientation of the entrance street.

Orientation of Entrance Street	Balcony (Dış Çardak)				House without Balcony (Dış Çardak)	Total
	South	North	West	East
West (12)	5 (42%)	0	0	7	7	19
East (7)	5 (71%)	0	2	0	9	16
South (1)	0	0	1	0	8	9
North (3)	3 (100%)	0	0	0	5	8
Total	13 (57%)	0	3 (13%)	7 (30%)	29	52

).

6. Discussion

Cities developed in an organic form until the 18th century, when technological and industrial breakthroughs took place. Since then, social living conditions have changed on a global scale. The problems of increasingly crowded cities have been attempted to be solved with modern planning principles. Urban spaces created by modern planning was the subject of much scientific research in the 20th century. In this sense, in addition to critical approaches, studies have also been conducted that examined traditional cities and pointed to a rethinking of the built environment created by cultural accumulations. Cities have been evaluated according to the space created by cultural life differences and the cognitive relationship between people and the environment [32,33]. Some researchers have identified typologies by analyzing cities with a morphological approach and described them as design tools [34]. However, it is also possible to mathematically analyze spatial patterns in continuity and syntax [35]. Alexander (1977) defined urban space with forms. From the urban scale to the building and construction scale, all forms relate to each other and create the whole [36]. This approach is visible, especially in traditional cities.

Within the scope of the study, the factors that constituted the traditional Safranbolu urban fabric were emphasized. In studies analyzing the Turkish house, the impact of social life in traditional houses on spatial utilization has mainly been explained. The special names of the spaces in the region and their conceptual origins are emphasized. It is stated that the region where the house is located affects the mass composition and building materials, and climatic conditions diversify the way in which open, semi-open, and closed spaces come together in the house. However, the lack of holistic studies on the principles of spatial organization is remarkable. For this purpose, in-depth research was conducted on the traditional Safranbolu houses, which are considered to be an impressive regional example of the Turkish house.

When the traditional Safranbolu houses were evaluated according to the urban site they were located in, the obtained findings were as follows:

It was observed that the houses in Bağlar were built on large and less-sloping parcels. Therefore, slope usage and basement floors were not observed. The buildings were placed on the north side of the parcel, resulting in large gardens facing south. There were examples of all types of traditional parcels. The most common building location type was the building attached to one street. Buildings positioned at the parcel were rare.

Kıranköy had both flat and highly sloping parcels. Due to the elevation difference within the parcel, slope usage was preferred. It was noticeable that basement floors were built completely under the ground, even on flat parcels. There were examples of parcels attached to one, two, or three streets, but no examples of parcels attached to opposite streets were found. Buildings attached to one street type were widespread. The only building attached to three streets was located on this site. Commercial units located on the ground floor of the houses changed the spatial arrangement by establishing a direct relationship with the street.

In Şehir, the parcel sizes and the elevation difference within the parcel were quite variable, and the parcel types and building location types were diverse, containing examples of all traditional parcel types. It was observed that parcels attached to two streets were common. The number of parcels attached to one and three streets were equal in number. The number of parcels attached to opposite streets was low. It was observed that as the number of streets around the parcel increased, the building was affected. The buildings attached to one and two streets were common while there were few buildings positioned at the parcels.

Fifty-two buildings were constructed in different urban sites, and twenty-three were compared within the scope of the study. The houses were numbered and explained to better understand the planning principles. The entrance street of twenty-three typical houses was in the west in five houses (Figure 27), in the east in seven houses (Figure 28), in the south in five houses (Figure 29), and in the north in six houses (Figure 30). Since the direction of the parcel entrance street affected the internal organization of the spaces, the houses were shown in groups.

If there was only one street around the parcel, the house established a connection with this street. The location of the building was determined according to neighboring parcels and neighboring buildings around the parcel. It was observed that buildings attached to one street where the entrance street was in the west, east, and north were positioned very close to the neighboring parcel (Houses 1, 2, 3, 8, 9, 10, 19, and 21). If there was a neighboring building around the parcel, the houses were placed close to this side of the parcel and adjacent to the street (Houses 6 and 18). There were no windows on the facades of the houses facing the neighboring parcel and the neighboring building, which was to ensure privacy. To utilize these facades, wet areas were especially placed here.

If there was more than one street around the parcel, the characteristics of the streets determined the building locations and interior solutions. If one of the streets was narrow, the facade of the house was left without windows, just like the facades facing the neighboring parcel and the neighboring building (Houses 4, 5, and 22). Units such as wet areas, storage, and stairs were placed on this unusable edge of the parcel. If all streets were wide, the direction of the slope within the parcel determined the location of the building. For the houses to receive more sunlight, ventilation, and view, it was aimed to be located on the higher side of the parcel. Thus, buildings attached to the two streets at the highest elevation were built (Houses 12 and 16). With the use of slope, independent entrances were provided for women and men from the streets to different floors of the house.

There were parcels with different characteristics in urban sites. The settlement principles provide for the production of unique interior solutions in each parcel. For example, the logic of the interior space settlement changed in parcels rising toward the south. It was observed that the settlement method according to the duration of sunshine in houses lost its importance, and the principle of seeing the view and the street was prioritized (Houses 17 and 23). For this reason, units such as storage and wet areas were placed on the south facade.

Within the scope of the research, we tried to explain the spatial usage logic of Safranbolu houses from a holistic perspective by examining 52 houses in terms of the settlement scale, building scale, and interior space scale. When the settlement decisions of traditional Safranbolu houses were evaluated at the urban scale, the obtained findings were as follows:

Parcel usage: Directions were determined at the parcel. The aim was to place the house on the higher part of the parcel, close to the main street and on the north side. Thus, the house had a better view and the sunlight period of the parcel increased.

Parcel–street relationship: Depending on the number and width of the streets around the parcel, there were parcels attached to one, two, or three streets and parcels attached to opposite streets.

Orientation of parcel entrance street: According to the direction of the main street associated with the parcel, the direction of the entrance facade and the entrance door of the house were determined.

Parcel–landscape relationship: Projections were placed according to the dominant view direction of the parcel.

When traditional Safranbolu houses were evaluated at the building scale, the obtained findings were as follows:

Building location: According to the number and width of the streets around the parcel, buildings were placed as attached to one, two, or three streets and as positioned on the parcel.

Parcel entrance and number of entrances: There were parcels with one, two, or three entrances according to the type of building location.

Number of Stories: According to the common preference of the users, the buildings were constructed with three floors as ground, middle, and upper floors.

Slope usage: If the elevation difference within the parcel was 3 m or more, the basement or ground floor was buried in the ground. If less than 3 m, one wall of the building was partially buried in the ground.

Location and orientation of projections: Open projections in the sofas provided sunlight and allowed for a view. Closed projections in sofas or rooms observed the street or the view.

When the interior units of traditional Safranbolu houses were evaluated according to the orientation of the entrance street, the obtained findings were as follows:

Hayat/taşlık: This was in the south.

Animal shelters: These were in the north.

Sofas: In houses with outer and corner sofas, these were to the south. In houses with central and inner sofas, the iwans observed the street or the view.

Rooms: These were placed according to the sunlight period of the parcel and the orientation of the entrance street. If the entrance street was to the west, they were oriented to the southeast, southwest, and northwest; if it was to the east, they were oriented to the southeast, northeast, and northwest. If the street was to the south, the orientation was to the southeast, northwest, and southwest; if to the north, the orientation was to the northeast, southwest, and north.

Kitchens: These were in the north.

Wet Areas: These were placed to the north or on the neighboring parcel side.

Balconies: These were in the south.

According to today’s architectural understanding, buildings are designed with the same spatial arrangements all over the world, ignoring their relationship with the context in which they are located. This construction approach of modern planning, which has gained validity on a global scale, contradicts the flexible design methods of traditional cities, which endured until the last two to three centuries, and had established a relationship with their surroundings in every way. The strong relations that each form, which constitutes only parts of the whole as emphasized by Alexander, establishes with the preceding and following forms are capable of forming an infinite number of combinations [36]. From this, it can be concluded that architecture and urban design and planning principles should be considered together to eliminate many of the spatial problems seen in modern cities.

In this sense, each analysis element in the study method expresses the holism of spatial relations in Turkish cities. Some of the prominent characteristics of historic Safranbolu houses are similar to the formal rules identified by Alexander. Principles such as settling to the north according to the parcel usage principle and keeping the garden in the south at large, building frequent and large windows on the two walls of the room to benefit more from daylight in the layout of the rooms, and leaving the cold north-facing walls of the room without windows are examples of these similarities. The study revealed that, just like other traditional houses, Safranbolu houses have an adaptable design system according to the dynamics of the parcel on which they are located. According to this design system, the location of the buildings on the parcel and interior solutions can be reorganized with specific rules. In this way, a different plan is created each time, with loosely interconnected elements, and urban diversity is achieved. In many other historical cities like Safranbolu, the environment created by the traditional adaptive method has been adopted by human beings for centuries, which has been sustained and transferred to the present day.

7. Conclusions

In this study, in which traditional residential architecture was examined from various perspectives, detailed analyses of the relationships between Safranbolu houses and their immediate surroundings were made to understand the site-specific design approach of the Turkish house. Studies were conducted on the design elements according to which traditional buildings were constructed. The impact of the environmental design data of building plots on spatial design was explained by a comparative method that evaluated a group of houses according to certain criteria. The houses in the study area are located within three different urban sites. Similar and different aspects of the spatial utilization practices of societies with different lives were observed in the examined houses. The most striking factor among the similar principles was the observance of privacy between houses at each site. The urban space was enriched with different types of houses at each site. For example, the use of basements for wine cellars and the presence of commercial units in Kıranköy houses are reflections of the socio-economic life of the householders. It is possible to say that Şehir houses reflect the type of housing required by the general family structure while Bağlar houses represent small and functional housing spaces where agricultural production activities were important. It can be seen that the Safranbolu residents commonly lived in 3-story houses, and this use did not provide diversity.

Houses in urban sites vary with the effects of topographic data as well as the differences created by social life. The land structure has affected the settlement pattern in the sites. There is a flat structure in Bağlar, both flat and sloping in Kıranköy and a highly variable sloping structure in Şehir. This has affected the parcel geometries. Another factor affecting the parcels is the streets. In the studies examining Safranbolu houses, the importance of the street can be explained from various aspects [37]. In the sites analyzed in this study, four types were identified according to the number of streets around the parcel, and the houses established direct relations with the street.

Another natural design feature affecting the parcel was the orientation according to the sun. At urban sites, buildings were positioned according to the elevation difference within the parcel, the number of streets, and the sunlight period. In each parcel type, it was important to settle on the higher and sunny side of the parcel. The presence of streets on the rising side of the parcel enriched the interior perspective of the houses. However, in parcels attached to one street, the buildings were positioned close to the neighboring building or the neighboring parcel. In sloping parcels attached to two streets, different floors of the house were reached from these streets. Thus, guests coming to the house used different parts of the house according to privacy. Similarly, in sloping parcels attached to three streets, the buildings were located attached to the two highest streets. This way, the house could be reached through independent entrances.

As the number of streets around the parcel increased, the building location, the number and location of parcel entrances, and the spatial arrangements inside the house changed. In addition to the natural design data, there are certain rules in the coming together of the units that make up the house. In this study, these rules were explained according to the orientation of the units such as the hayat/taşlık, animal shelter, sofa, room, kitchen, wet areas, and a balcony, which make up the floor plans. It was seen that these units were placed in certain directions depending on the direction of the street that provided the entrance to the house. On the ground floors, the hayat related to the street was ensured to face south. In houses where the street was in the west, east, and south, the hayat was in the south. The daily work of the house was ensured to be carried out in a space that received heat and light. In such houses, the animal shelter was in the north, which provided heating of the upper floors. If the entrance street was in the north, hayats were necessarily placed in the north, and the animal shelters were positioned.

The most important element on the middle and upper floors was the room. The aim was for the rooms to see the entrance street and receive sunlight. In houses with fewer rooms, sofas were used as a room and observed the street or the view. As the number of rooms increased, sofas became smaller inward, and turned into a hall. In these examples, iwans opened to the street or the view and enlightened the sofa. Kitchens were located to the north and generally observed the view. Wet areas were located to the north of the parcel and the neighboring parcel or neighboring building side. Thus, the unusable side of the parcel was utilized. Balconies were oriented to the south due to traditional production methods.

As a result of this study, it was determined that traditional Safranbolu houses were built by the environmental design data of the parcel, far from randomness, in quite a pragmatist order of rules. The buildings were positioned in the parcel according to the parcel use, the direction of the sunlight period, the street relationship, the dominant view, the orientation of the street providing the entrance to the parcel, and the elevation difference within the parcel. Elements such as the parcel entrance and the number of entrances, the slope usage in the building, and the location and orientation of the projections provided diversity. Finally, it was emphasized that the orientation of the street providing the entrance to the parcel affected the arrangement of the interior of the house. Each unit in the house was placed according to the sunlight period depending on the orientation of the street to which the building was related. In addition, privacy was provided with windowless walls facing each other in neighboring buildings. The placement of wet areas on these facades was the result of the most efficient usage of all the dynamics of the parcel at the urban and building scale. The study draws attention to the traditional design approach by referring to the parcel dynamics in the Turkish housing construction culture. According to the results of this study, the traditional house-building practice has centuries of accumulation with its design capability based on a utilitarian approach of parcel-based, site-specific, contextual dynamics.

Today’s modern building culture has become quite different from traditionally accepted building principles. The main reasons for this can be summarized as social changes, technological advances, and significant increases in the search for comfort. With this research on historical cities and traditional houses representing unchanging design principles in a changing world, it is possible to rethink the building practices of the past.

This study, which analyzed the traditional dwellings of Safranbolu, aimed to provide a framework for future studies. The spatial design approach of conventional architecture from the urban scale to the building scale was explained through visual analyses, statistical information, and related classifications. Houses provide useful information for the infrastructure of design language/grammar that can be used in computational design approaches. In this context, the number of studies that are attempting to adapt the existing or traditional building culture to today’s conditions is increasing [38,39,40]. It is thought that with the development of the design approach in this study, the possibility of producing modern houses compatible with the context within the historical environment of Safranbolu will increase. For this purpose, it is necessary to increase the depth of the analyses in the study with various quantitative approaches. The consistency of the information obtained from the study can be tested, and its scope can be expanded with the space syntax method, which reveals the relationships between structures and societies by investigating the correspondences between form and functions [41,42,43]. In future studies, it is possible to obtain experimental findings by examining Safranbolu houses with space syntax and visibility graph analyses.

The natural and environmental design data of the parcel emphasized throughout the study will undoubtedly make the built environment more livable by being taken into account in building designs. The criteria listed under headings within the framework of this approach offer serious potential in creating design rules. The new architecture developed in the urban space with various derivation methods will not repeat the building culture it inherits, but will make qualified references to its past. With master plans prepared with this sensitivity, the building culture will be interpreted, developed, and passed on to future generations.