Planning Assessment
Assessment of Patterns of Spatial Transformations
A Comparative Study of Three Historic City Centres in Algeria
2024 Volume 12 Issue 1 Pages 173-195
Details
2024 Volume 12 Issue 1 Pages 173-195
Sustainable urban design for historical city centres should carefully focus on the transformations in urban morphology and land use aspects as a starting point. It is important to understand the process of an urban space’s evolution to assess its future design. This paper aims to assess the patterns of urban transformation in historical city centres that belong to different periods of their history, to highlight the relationship between the occupation of urban space and urban transformation. Aiming to understand the impact of different factors on urban transformation and identity, the research method adopts space syntax analysis using axial and segment maps to compare the variation of syntactic parameters over time for three city centres in Algeria (Algiers, Constantine and Tlemcen) in three periods that mark the different key stages of urban evolution. The selected case studies represent typical examples of the foundation and evolution of cities in Algeria. Results show that patterns of urban evolution in the three city centres present almost similar variations, despite the global differences between them (size, population, initial morphology, and site topography) in addition to the operations of transforming the urban morphology. The main findings of this research can be stated as a model of urban evolution for the historic city centres in Algeria during the mentioned periods.
Urban form is the result of the successive transformations of the city. Understanding the urban morphology process of evolution is a key concept for sustainable urban design (Braulio-Gonzalo, Bovea et al., 2015; Maretto, 2014; Maretto, Gherri et al., 2017; Sahni, 2018). Urban form evolves in a geographical context (evolution in space) as well as a historical context (evolution in the place). Urban transformation is linked to evolution, it is a proper phenomenon to the city, which results from the juxtaposition or the superposition of heterogeneous urban fractions (Robert-Max, 2016). The analysis of the process of urban evolution through time becomes a necessity for conceiving future scenarios to achieve a sustainable city (Mohareb, 2012; Tiwari, Nigro et al., 2023).
The analysis of the historical process of the generation of urban form aims to determine the impact of the different actors in the process and predict the future potential scenarios according to the objectives of urban development (Wang, Zhang et al., 2023). Until now, studies on historic city centres were focused either on their morphological or functional aspects, and research to understand the link between the morphology resulting from a process of transformations and the origins of these transformations was limited (Mohareb, 2016). The evolution of historic city centres in Algeria has never been analysed deeply to measure similarities and differences. However, typical, standard and generalised urban projects are planned in most cases but generate different reactions for users. It is crucial to understand the process of urban evolution and the position of each actor to develop a global and sustainable vision of the urban composition and its future evolution(Saleh, Ghazali et al., 2023).
We propose in this research a comparative syntactic analysis between three historical centres: Constantine, the Kasbah of Algiers and Tlemcen. This analysis aims to define the common points and differences in the evolution of the historical city centres in Algeria, as well as the possibility of defining a model of the urban evolution for these cities and the resulting morphology of this process. The research is based on space syntax techniques to measure the patterns of urban space in each of the periods of evolution to understand the effect of the operations carried out on the urban space.
The research starts by collecting data on the three case studies in the defined periods (urban plans, geographic, historic, and social occupation), followed by a syntactic analysis of the reported plans using DepthmanX software to measure their syntactic parameters. Then, the results are compared and discussed to understand the similarities and differences. Finally, the effect of each of the transformation factors considered in this study is determined through a discussion of the results.
“Morphological analysis of urban fabrics allows us to understand the logic of their transformation. It allows us to plan, with ‘authentic’ awareness, interventions within the process of change.” (Maretto, 2014)
According to Piron (2001), the evolution of the urban form can be in two ways: an internal change that consists of urban renewal or restructuring of space without exceeding the urban area (example of Haussmann’s works) or an external evolution (growth or extension). As the historic city centre forms the first core of the city, the internal transformations are of a densification or restructuring nature.
The ‘historical centre’ is often considered to be the original core around which agglomerations are formed. “Centrality is the quality, the character of what is central or the centre of something. The notion of the centre is complex, more than confused: it is based on geometric, topographical, historical, functional criteria.” (Toulev, 2016)
Lately, more researchers are interested in the study of the historical evolution of urban form (Aleksandrowicz, Yamu et al., 2018; Cheng and Wang, 2017; Griffiths, 2012; Mohareb, 2012, 2016; Palaiologou and Griffiths, 2019; Shateh, Bin-Dou et al., 2018; Shpuza, 2009). The debate on urban morphology and the continued transformation in historic cities has become the subject of many studies, and researchers are interested in clarifying the different models of urban change, factors influencing the resulting urban form and the causes of the phenomena (Dash, 2023).
Shpuza (2009) studied the urban growth patterns of coastal cities in the Adriatic and Ionian regions regarding three evolution stages by analysing linear maps using syntactic variables of ‘Line Length’, ‘Connectivity’ and ‘Integration’ to define generic rules of urban growth. The study compared 23 cases in three evolutionary stages and highlighted through syntactic measures the trends of the evolution of street networks in the studied region.
In his research works, Griffiths (2012) based his idea for the use of space syntax techniques in historical research on the way that space syntax interprets the connection between urban morphology and the use of urban space. He identified four approaches to using space-syntax techniques in historical research: (a) history as ‘background,’ where the historical information is provided to introduce the case, and it serves as an introduction to the following syntactic analysis; (b) history as syntactic ‘growth processes’ where the historical information is the original cartographical materials used as a basis for models of urban growth to understand its social logic; (c) syntactical morphological histories aim to understand the evolution of urban morphology in the social context, and socio-historical logic is considered the generator of the evolution of the spatial configuration and (d) spatial‐locational histories, where Griffith considers space syntax in this approach as an addition to describing the socio-historical phenomena; it is not the main analytical method, but it provides information on morphological or configurational properties to describe the historical process.
Mohareb (2012, 2016) examined the historical growth of numerous Arab walled cities of Arab Mediterranean countries, focusing on the spatial patterns around the wall edges and the spatial relationships between historic edges and their surrounding context. He proposed a model of analysis with two aspects: first, the analysis of spatial configuration patterns using space syntax and land use through traditional methods (gate count) and, second, counting and analysing pedestrian flows in historic centres using GIS representation. This model of analysis is based on axial and segment maps using syntactic measures of connectivity, global and local integration, metric choice and intelligibility. This research aimed to analyse the urban limits of fortified historic Arab cities. This comparative analysis has highlighted the presence of a repetitive pattern of spatial representations, accompanied by land-use activities. This research presents two contributions, firstly, to develop an analytical model that compares and analyses different case studies to understand current situations and predict future scenarios. Second, it extends the scope of the analysis beyond the conventional concern that focuses either on the fortified city as a unit or on the fact that the wall is considered an archaeological artefact.
Aleksandrowicz, Yamu et al. (2018) argued that space syntax is a method that can specify the way an urban system functions; it is also applicable in retrospect to describe social aspects of historical urban systems. In their research on the syntactical growth processes of Jaffa and Tel-Aviv, it confirmed that spatial-syntactical analysis may contribute to the writing of urban history by providing a diachronic outlook on the transformations; “the spatial organisation of a town or a city directly affects the ways by which people perceive, move through, or use space,” according to Aleksandrowicz, Yamu et al. (2018).
History of urban evolution in AlgeriaUrban evolution in Algeria was the subject of numerous dissertations and research (Gsell, 1920; Mercier, 1903; Mimó, 2016; Mostefaï, 2002; Stora, 2010) and many others. The history of human occupation in the Algerian territory dates to antiquity. Urban history began with the first groupings of agricultural villages created by the Imazighen in the interior areas and the Phoenician counters installed along the coast. Amazigh villages were juxtaposed with the counters.
The Romans first set up military settlements near Phoenician counters, juxtaposed and sometimes superimposed on the ruins of Amazigh villages, these colonies quickly turned into cities, sometimes large cities that could accommodate hundreds of thousands of inhabitants (Gsell, 1920; Stora, 2010). Some cities were created in the inner regions, and some colonies were created to control the territory.
After the invasion of the Vandals, most of the Roman cities were destroyed; the Byzantines recovered those cities, rebuilt the defensive walls and the towers and recreated the doors. The imprint of the Byzantines did not exceed these works, and the rest infiltrated the Roman works.
The period of Islamic governance was characterised by a new form of urbanism different from the Roman city (Cote, 2005; Mostefaï, 2002), there are two types of cities in this period.
- New cities were created on the ruins of ancient Roman or Amazigh cities that were destroyed by the effects of nature or wars (the Kasbah of Algiers and Constantine, among others); otherwise, they were juxtaposed.
- Cities were created on new sites, as in the case of the ‘ksour’ of the Sahara and many cities in the highlands.
Two typologies were developed in this period, ‘ksour’ and ‘medinas’ or ‘old towns,’ based on the transmission of accumulated knowledge over the centuries (Cote, 2005). The medina is an old Arab city characterised by its narrow curvilinear streets, alleys and dead ends that give access to the houses, with a religious city centre.
Cities were characterised by defensive walls and organised around the functional centres, which are the mosque, the ‘hammam’ (public bath), the ‘madrasa’ (school) and the ‘souk’ (market)(Mostefaï, 2002). A medina refers to the old part of an Arab-Muslim city, as opposed to modern European-style neighbourhoods. This term is used mainly in the Maghreb countries, Spain, West Africa and East Africa (Bastié and Dézert, 1980). In Algeria, historic districts in several cities are called medinas, citing for example the old city of Tlemcen, the medina of El Oued, the medina of Oran and the medina of Berjaya, among others.
In the Saharan region, several urban cores called ksour were created; the ksar definition is as follows: “[a] ksar (ksour in plural) is a village surrounded by a wall and protected by watchtowers, which includes inside dozens or even hundreds of houses, in addition to the mosque, the streets, collective facilities and a place where the parties were held. All ksour have one or more monumental entrances, sometimes well decorated.”(Mimó, 2016)
With the arrival of the Ottoman regency, they imposed their architectural and urban style, and several medinas were slightly transformed with the construction of houses, schools, baths and palaces in an Ottoman-style shaped ‘Kasbah’; part of the medina is rebuilt in the form of ‘Kasbah’, and sometimes an urban core is built independently as ‘Kasbah’. During the French occupation, the military engineers proceeded by a series of transformations to prepare the ground to lodge the troops, install colonies and facilitate military tasks (control of the urban space and intervention when necessary). According to Mercier (1903), these urban planning operations consisted of the following: The creation of a European neighbourhood juxtaposed with the native city or part of the old city after the expulsion of the inhabitants and restructuring in the form of a checkerboard; the creation of straight breakthroughs to connect the various parts of the city, the gates and the public spaces, facilitating the movements and the control of urban space; the creation of a military zone (camp, barracks, military hospital) in the European neighbourhood or juxtaposed with it; the creation of suburbs for future extensions.
Finally, the Algerian cities, after having undergone major restructuring, were composed of two zones called the ‘Arab Quarter’ and the ‘European Quarter,’ with two different urban typologies, usually separated by large streets, and each community lived in its zone under the same colonial authority.
After independence, the city centres were full of components of different periods, superimposed and sometimes juxtaposed, but the whole represents a picturesque heritage. The evolution of Algerian cities in this period went through several forms, and different policies were addressed in trying to fix urban issues. The historic centres have undergone planned and unplanned transformations with the conflicts of the legal status and the management of the private properties.
The urban fabrics of the three city centres date back to antiquity, but the existing historical centres were built on the ruins of the Roman or Amazigh cities they transformed. Until they were taken by the French settlers, they were typical cities (Kasbah or medina) (Figure 1). According to Mercier (1903), during the colonial period, the main transformations consisted of:
- the division of the city into two parts (autochthonous and colonial), and the latter was rebuilt according to a checkerboard pattern,
- the creation of public places in both parts of the city and
- the creation of breakthroughs in the native part linking city gates crossing the public places.
Then, urban growth passed by the creation of ‘faubourgs’. The urban renewal of the historic city centres resumed after independence, with reconstruction and renovation works and urban restructuring, but in general, the urban fabrics kept the overall layouts.
Situation and foundationThe selection of the three cities to be compared in this section was based on the principle of their historical values as well as the diversity between them in their compositions and their sites. The three cities have Amazigh and Roman origins, and they belong to three different sites, mainly their situations and their compositions. They share the same formal characteristics (walled city with gates, the hierarchy of roads, and curved and narrow alleys.
The comparison of the urban evolution of the three city centres shows that the Kasbah of Algiers is located on the coast (Figure 1), a site that was occupied by a Phoenician counter since antiquity. It is a port situated on the Algerian territory; the site was then occupied by the Imazighen, then the Romans, and these settlements were superimposed on the same land. The city was rebuilt by Beni-Mezghenna and then by the Ottomans as a typical Kasbah (Ottoman traditional city). It underwent several restructurings during the colonial occupation by creating European neighbourhoods and new buildings, such as hospitals, army and police facilities following a checkerboard layout (Figure 2).
Constantine is an inner city that occupies the site of ‘The Rock,’ protected by the topography of the site and Oued Rhumel. A defensive site on which a Roman city was built near the Amazigh site Tiddis, with rugged topography and difficult accessibility. The medina occupied the same site with an influence from the Roman layout (Figure 1). The western part was transformed into a Kasbah inside the medina (an area occupied by the beys and their clan), it underwent the same transformations as the Kasbah of Algiers during the colonial occupation (Figure 3).
On the other hand, Tlemcen is a frontier city in the far north-west, based on the Lalla-Setti plateau, near the Mediterranean and a little further from the late Roman city Pomaria (201–235 AD), Agadir. It was rebuilt by the Idrissids and then merged with the new Tagrart city founded by the Almoravids to form Tilimsane. The medina has undergone several transformations under the Almohades, the Zianides and the Mérinides. During the Ottoman period, the city of Tlemcen was limited to the core of Tagrart, surrounded by a wall with a spatial distribution by ethnic group: “the Hadars in the ancient fabrics of Tagrart, the Jews occupied the central district, the Kouloughlis are installed around the palace of Mechouare” (Figure 1). During the colonial era, the city underwent restructuring like all Algerian cities, except that in its case, the western part was totally destroyed and rebuilt following a checkerboard layout (Figure 4) (Hamma, 2011). After independence, property transfers and divisions caused by inheritance led to some internal transformations in the buildings, sometimes opening new dead ends to create secondary entrances. Recently, the Kasbah of Algiers was classified as a World Heritage Site and has undergone restoration work and rehabilitation managed by UNESCO.
| Constantine | Algiers | Tlemcen | |
|---|---|---|---|
| Foundation | Roman | Phoenician, then Roman, city | Roman city built on an Amazigh city |
| Typology Before the French period | Medina (Arab walled city, Figure 1) | Kasbah (Ottoman city Figure 1) | Medina (Arab walled city, Figure 1) |
| Main transformations in the French occupation period |
Restructuring 30% of the city in a checkerboard plan (European city) Urban breakthroughs New urban places (Figure 3) |
The juxtaposition of a checkerboard plan (European city) Urban breakthroughs (Figure 2) |
Demolition and reconstruction of 60% of the city as a checkerboard plan Urban breakthroughs New urban places (Figure 4) |
| Main transformations after independence | Individual reconstructions of private properties with transformations (Figure 5) | Restoration of a major part of the city (Figure 6) | Restoration and partial reconstruction of private proprieties (Figure 7) |
Currently, the three historical city centres form the central core of large cities with territorial importance and urban heritage in the testimony of past periods with their distinct typologies. Algiers is the capital, and the Kasbah of Algiers is preserved and classified as a national heritage site; it includes buildings and houses occupied by their owners and buildings designated as monuments, open to tourists. The last major transformations date back to before 2010 (Figure 5). The city centre of Constantine is still composed of the European district and the traditional one of the Arab medina called Souika; trade is still present on the first floors of buildings, and many heritage monuments are present on site, such as the Bey’s Palace, the old school ‘madrasa’ and Turkish baths, ‘hammam’. The last major transformation was the installation of the teleferic, which required the demolition of several buildings and the creation of a large place in 2010 (Figure 6). In the same way, the city centre of Tlemcen includes the palace of Machouar, buildings of the old medina and colonial constructions. Numerous buildings have been rebuilt and public places created, but the overall structure remains unaffected (Figure 7), and the last urban operation dates to 1998.
This paper adopts space syntax techniques for reading the spatial patterns of urban fabrics. Space syntax is an approach to describing and analysing patterns of urban and architectural space, both at the architectural and urban scales (Kubat, 2010). The analysis of urban morphology and historical analysis of urban fabrics using space syntax technics was the subject of many studies (Aleksandrowicz, Yamu et al., 2018; Cheng and Wang, 2017; Hillier, Bill and Hanson, 1987; Kubat, 2010; Mohareb, 2016; Paul, 2014; Sahni, 2018) focusing on the definition of urban patterns through syntactic parameters. For historical research, space syntax provides a way into conceptualising and thinking about the role of space and its relation to life in the built environment (Griffiths, 2012; Palaiologou and Griffiths, 2019).
The research methodology was structured in three steps. First, scaled maps of the three cities are selected for the two older analysis periods to reproduce them in DXF format using AutoCAD software. For the current period, urban plans (urban development primary plans) were used. Axial and segment maps were generated and analysed using DepthMapX.
Once the syntactic parameters were generated, the second step consisted of the analysis of axial maps for the three case-study subjects of the comparative analysis, focusing on the main characters of the urban fabrics in the first stage and then comparing them to find out the common points and differences in their urban patterns. Subsequently, an analysis of the axial maps of the mentioned urban fabrics in two later periods was conducted, focusing each time on the main changes in syntactic parameters relative to the urban change.
The third step was to draw, for each syntactic parameter, a graph showing its variation for the mentioned cities during the different historical periods. Then, a comparative analysis of these graphs was done to set out the behaviour of each syntactic parameter. The similarities and differences in the syntactic parameter variations between the three cities should be justified by the nature of urban transformations and the conditions of the operations carried out.
According to the explored parameters research, the proposed syntactic analysis focuses on the axial map and the segment map, mentioned by Liu, Batty et al. (2019) as the most essential syntactic tools to analyse the urban area, while axial analysis focuses on the analysis of street patterns using axial lines, while segment angular analysis divides axial lines into segments at the points where the axial lines intersect. The analysis considers the following measures: the number of axial lines is considered an index of the topological size of the urban fabric in terms of the complexity of the street network, in addition to the axial line length. ‘Integration’ is a measure for the interrelations between any space and other spaces in the structure (Hillier, B., 2007) as well as global and local measures with radius R3 to show the local effect of restricted area on the global system. ‘Choice’ indicates the degree to which a line is located on the network’s shortest paths from one line to another (Batty, 2020), where global and local R3 measures are considered. ‘Step depth’ follows the shortest path from the selected root line or segment to all other lines or segments within the system; it shows how deep the street system is (Boumezoued, Bada et al., 2020). ‘Intelligibility’ measures the correlation between connectivity and integration; it is an index of the clarity of the urban system and implies whether the whole can be read from the parts.
The angular syntactic choice and angular integration were normalised to reduce the effect of size. It divides total choice or integration by total depth for each segment in the system. This adjusts choice and integration values according to the depth of each segment in the system (Hillier, B., Yang et al., 2012).
Axial maps were generated using UCL DepthmapX software based on historic maps of the three cities: Constantine (1837, 1937 and 2010 to date), For the Kasbah of Algiers (1831, 1937 and 2010 to date) and Tlemcen (1836, 1942 and 1998 to date); key historic periods were selected according to the availability of historic data and urban transformations. The first period represents the initial typologies before the French occupation, when maps were created by the French military engineering department (1831–1837). The second period represents the end of several restructuring works made by the French (1937–1942), while the third period is the recent one after the major postcolonial transformations, based on the official plans (1998–2010), then verified with recent data using satellite images and open street map data. The study considered the fewest line map analysis, and a total of 45 axial maps and 9 intelligibility graphs were analysed.
Axial maps comparative analysisThis section focuses on the comparative syntactic analysis between the three historic city centres: Constantine, the Kasbah of Algiers and Tlemcen. This analysis allowed us to highlight the patterns of urban change as well as the common points and differences between the historic city centres in Algeria.
By colour spectrum from High to low values of integration: red–orange–yellow–green–light blue–blue– dark blue. The dashed lines represent the shape of the scattered axes to form cores of integration in each city, performed according to the values of integration and represented by colour spectrum from High to low values of integration.
The first period represents the traditional cities before French occupation and transformation, and it consisted of the analysis of the axial maps of Constantine in 1837, the Kasbah of Algiers in 1831 and Tlemcen in 1836 (Figure 8 A–C). The choice of maps to analyse was made according to their availability and according to the history of the cities (the period before the taking of cities by the settlers).
In the second period, after the completion of major restructuring work carried out by the French, the three cities Constantine in 1937, the Kasbah of Algiers in 1937 and Tlemcen in 1942 (Figure 8 D–F) underwent similar transformations, with the division of the cities into two districts. The first was reserved for European settlers with a checkerboard layout, and the second was reserved for the autochthon inhabitants, keeping its traditional layout with some breakthroughs that connected the gates and crossed the public places. The case of Tlemcen differs in that the European quarter was rebuilt (half of the city). In the case of Constantine, it was a partial restructuring, while in the Kasbah of Algiers, it was a bit of both.
The current period is represented in the maps of Constantine from 2010 to date, the Kasbah of Algiers from 2010 to date and Tlemcen from 1998 to date (Figure 8 G–I), which represents the current configuration of the three city centres after the major transformations after independence. The transformations in the city centre of Constantine did not bring major restructuring throughout the space, except for some demolitions, refurbishments and restoration work. In the case of Tlemcen, and according to Hamma (2011), the work done in the historic centre after independence can be summarised in the restoration and maintenance of the old mosques and the restoration and development of the old town in 2009 with the development of several tourist routes. According to Atek (2012), the Kasbah of Algiers experienced only a few projects in the post-independence period, such as the car park, a school, a conservatory and a health centre, which did not influence its overall structure.
| Constantine | Algiers | Tlemcen | |||||||
|---|---|---|---|---|---|---|---|---|---|
| 1837 | 1937 | 2010 to date | 1831 | 1937 | 2010 to date | 1836 | 1942 | 1998 to date | |
| Number of axial lines | 229 | 451 | 294 | 251 | 170 | 374 | 243 | 163 | 193 |
| Line length | 67.32 | 77.84 | 130.4 | 90.6 | 232.7 | 187.1 | 130.4 | 187.1 | 133.7 |
| Connectivity | 3.45 | 4.83 | 4.35 | 3.35 | 6.95 | 3.96 | 3.77 | 5.7 | 4.52 |
| Integration HH | 0.798 | 1.378 | 1.316 | 0.67 | 1.63 | 1.07 | 0.93 | 1.83 | 1.39 |
| Integration HH R3 | 1.5756 | 2.0496 | 1.9165 | 1.52 | 2.38 | 1.82 | 1.672 | 2.269 | 1.948 |
| Norm Integration Rn | 1.482 | 0.95 | 0.829 | 1.412 | 1.05 | 0.833 | 1.392 | 0.845 | 0.83 |
| Choice Rn | 2197.43 | 1046.12 | 1018.47 | 4240.01 | 528.53 | 2216.46 | 1125.11 | 465.38 | 766.74 |
| Choice R3 | 33.13 | 80.984 | 77.535 | 29.79 | 88.7 | 58.28 | 43.28 | 110.99 | 82.25 |
| Norm Choice Rn | 0.05 | 0.033 | 0.046 | 0.041 | 0.037 | 0.03 | 0.058 | 0.034 | 0.04 |
| Norm Choice R3 | 0.102 | 0.044 | 0.063 | 0.115 | 0.037 | 0.067 | 0.074 | 0.03 | 0.07 |
| Step depth | 8.34 | 3.36 | 3.2 | 9.08 | 2.93 | 4.91 | 6.33 | 2.5 | 3.41 |
| Intelligibility | 0.21 | 0.35 | 0.47 | 0.21 | 0.55 | 0.35 | 0.27 | 0.68 | 0.56 |
By analysing the urban evolution of Constantine in 1837, the Kasbah of Algiers in 1831 and Tlemcen in 1836, we noted that the three city centres have gone through similar stages of development, with some differences due to the site and the position of each in its territory. As a result, in this period, the three fabrics had a typical traditional typology, a walled city with a limited number of gates, well-positioned relative to the environment, with mostly narrow curvilinear and hierarchical roads. The dimensions of cities and roads varied from one to another.
Proceeding with the syntactical analysis, a slight difference in the number of axial lines of Kasbah of Algiers, Constantine and Tlemcen was noted; the axial maps consisted of 251, 229 and 243 axial lines, respectively (Table 2). The number of lines can be taken as a size index (Mohareb, 2016), which means that the three cities have very close sizes.
The results of the connectivity analysis showed a great similarity between the axial maps of the three cities, with high and close maximum values. The mean values of integration were 0.798 for the case of Constantine, 0.67 for the case of the Kasbah of Algiers and 0.93 for the case of Tlemcen (Table 2). The three fabrics are similar in that the central streets (connecting access to the centre) were the most connected. Connectivity decreases when going to the periphery and penetrating the alleys then impasses (dead-end roads), according to the geometry of the urban fabric.
The integration analysis also shows similar values between the three city centres, with some geometric differences. In all three cases, we noticed the presence of a strongly integrated central core (the red dashed lines), and the values decreased gradually towards the peripheries where the axes are more segregated (each one presents a different geometry for the central core). This was as in Hillier’s deformed wheel geometry, where the values were closer to the local measurements, where the central core became wider, and there were several integration cores in the centres of the districts, which necessarily made a considerable difference between the navigation of inhabitants and foreigners (Hillier, B, Penn et al.)
In terms of choice, the three fabrics show similar results. There are low values of choice and always the same central spaces that offer more sustained values. Local values are closer than global values, the centres of neighbourhoods show higher values of choice, and we noticed the same pattern of variation of local and global choice in the three fabrics.
The values of intelligibility and synergy were very close for the cases of Constantine and the Kasbah of Algiers, 0.21 for intelligibility and 0.40 for synergy, which means that these fabrics were not intelligible as a security system for their inhabitants to control the accessibility of foreigners, with a weak effect on local areas where each part presents an autonomous structure, but they were homogeneous. The case of Tlemcen shows higher values, with 0.27 as the intelligibility coefficient and 0.53 as the synergy value. This may be due to the values of the axis lengths resulting from site topography (Table 2).
Thus, the three fabrics present the same phenomena in their configurational measurements, and the structure of the urban space was based on procedures of security and control of the accessibility of foreigners in traditional fabrics, and we noted important values of connectivity, integration and choice in the same central spaces (central core) that links the gates to the centres (usually containing the public buildings: markets, mosques and hammams). The centres of neighbourhoods were more segregated and offered low values of choice and connectivity. At the local scale, values were more sustained. The fabrics presented fractal geometry, and the navigation was limited according to the degree of knowledge of the system.
Second period of analysis (late occupation 1937-1942)The comparison of the axial maps of the first period (Constantine 1837, the Kasbah of Algiers 1831 and Tlemcen 1836) and those of the second period (Constantine 1937, the Kasbah of Algiers 1937 and Tlemcen 1942) set out the impact of spatial transformations in historical centres on their configurational parameters. The similarities and differences found between the three cases can be summarised in the following points.
-The three fabrics are similar in their initial characters, and the transformations made had the same objectives and principles. The results are similar in their general characteristics and different in their procedures.
-The studied fabrics are traditional typologies, walled cities with controlled gates, one or two main roads, curved and narrow alleys and impasses.
Following these two parameters, the transformations in the three cities produced new fabrics with more integrated axes, globally and locally, and that were more intelligible, with a very moderate local effect (Table 2). The global integration values increased in this period from 0.798 to 1.378 in the case of Constantine, from 0.67 to 1.63 in the case of Algiers and from 0.93 to 1.83 in the case of Tlemcen. Therefore, the growth rates were 173, 243 and 196%, respectively, for global integration and 129, 156 and 135% for the local integration R3.
Simultaneously, the global and local choices decreased from the first period to the second, while the local choice increased, but the influence of these transformations is noticeable only in the transformed spaces. The global choice values decreased by 48, 12 and 41%, while the local values increased by 244, 297 and 256%. The values of Constantine and Tlemcen were very close, and the case of Algiers was strongly influenced by the transformations in this period.
The European neighbourhoods resulting from the restructuring of a part of the city were open spaces that were more connected and integrated, whereas the autochthonous neighbourhoods remained the most segregated and offered fewer choices, except in the breakthroughs that penetrated them. The logic of social segregation was apparent in the local parameters of the two neighbourhoods.
The most important noticeable differences among the three cases were the rates of variations of configurational values. Tlemcen’s fabric, which was more flexible to transformations at the beginning, showed a significant variation than in the case of the Kasbah of Algiers in terms of integration. The case of the Kasbah of Algiers was less flexible, with a greater variation than that recorded in Constantine, the latter of which was more rigid than the two previous cases.
This difference comes back to the parameters of the site of Constantine, which was very hard and limited in surface area, as well as the nature of urban transformations. It is necessary to recall that in the case of Tlemcen, it is a total reconstruction, whereas, in the case of Constantine, it was much more of a restructuring. The case of the Kasbah of Algiers combined both phenomena.
Third period of analysis (post-independence to date)The analysis of axial maps in the third period showed that the slight transformations applied to the traditional fabrics in the post-colonial period modified their configurations. This may be due to several factors.
- The change in the actors of the urban space after independence, whereby the inhabitants of the colonial districts left the places that were invaded by new inhabitants as well as the installation of new inhabitants resulting from the rural exodus. The authority did not control their interventions in the early days.
- The renovation and restoration work may have resumed the initial state of the traditional space before the transformations of the settlers, a hypothesis that remains to be verified, but these transformations indeed converged on the initial configuration values.
Consequently, the spatial configuration of the three fabrics did not present the same characteristics as they did during the two previous periods. We noticed different variations in all configurational parameters except choice, which was not highly influenced from the beginning by the transformations, excluding the creation of breakthroughs (main straight streets) (Table 2). These influences will have different impacts on the perception of space and its use. However, to define the causes of these variations, it will be necessary to make a thorough analysis of the history of urban evolution for the studied fabrics. The advantage of these differences is that each of these three fabrics had a specific identity due to its differential evolution.
Comparative analysis of variation of parameters Axial line lengthsBetween the first and the second period, the mean values of axial line lengths increased following the operations of the restructuring and the creation of urban breakthroughs between the main gates and the public places. During the third period, the values of axial line lengths increased in the case of Tlemcen due to the new breakthroughts , but they decreased in the two other cases (Figure 9-A). The different operations of reconstruction did not respect the alignment of the late urban breakthroughs. This decrease in axes’ lengths influenced the other configurational parameters, such as integration, so it influenced all the other parameters of the second degree.
According to the graphs in Figure (9-B), there was an increase in the mean values of connectivity in the second period and a decrease in the third period of the three analysed cases, and the decrease in the case of the Kasbah of Algiers is the strongest (almost returned to the initial state before the transformations). This phenomenon pushes us to assume that the transformations after the independence took again the initial state of the urban fabrics, like reference, in particular with the decrease of the mean value of the axis lengths, which indicates that the overall structure is modified by minimising the connections.
IntegrationThe mean values of integration showed a similar behaviour as the connectivity, with an increase in the second period and a decrease in the third period (Figure 9-C). The changes in the three urban fabrics were similar, with different rates of variation. In the third period, the case of Constantine was almost constant, while Tlemcen and the Kasbah of Algiers underwent a remarkable decrease. This had a strong influence on the measures of the second degree: intelligibility and synergy.
The analysis of integration on a restricted topological scale, Radius 3 (Figure 9-D), shows that the case of the Kasbah of Algiers responded more to the changes.
ChoiceChoice was not very influenced in its overall distribution by transformations, except in the urban breakthroughs that created high-value streets and dominated the choice in the three fabrics. The variation of the mean values indicates a strong decrease in the second period for the case of the Kasbah of Algiers against an average decrease for the two other cases (Figure 9-E). During the third period, we noted a sharp increase in mean choice value for the case of the Kasbah of Algiers, a slight increase for Tlemcen and an almost stable value for the case of Constantine. The variations of local measures on radius R3 (Figure 9-F) between the three periods showed a contrary behaviour to the global measures, with similar phenomena between the three case studies.
These differential variations showed the instability of the choice parameter concerning spatial transformations; it should be recalled that Hillier et al. (1993) described it as a predictor of “through-movement” and a guide for “inhabitants” in their navigation since it requires knowledge of the site.
Normalised choiceNormalised measures avoid the effect of size differences between the urban fabrics being compared. Therefore, mean values of normalised choice presented a high similarity in their variations during the three studied periods (Figure 9-G) or the local values R3 (Table 2), except for the average value of the global normalised choice in the case of Algiers between 1937 and 2010 (Table 2). Considering global measures, Constantine and Tlemcen presented typically similar graphs of variation with a decrease of 66% and 58.62%, respectively, of the values from the first period of analysis to the second, then an increase of 139.39% and 117,64% respectively, while the case of Algiers presented a decrease of 90.24% between the first period and the second and continues to decrease to 81.08% for the third period. Consequently, mean values of local normalised choice R3 presented typically similar variations, with slight differences in variation ratios, decreasing by 0.43, 0.32 and 0.40% in the second period and then increasing by 1.43, 1.81 and 2.33% for the third period for Constantine, Tlemcen and Algiers, respectively.
Normalised integrationThe mean values of normalised angular integration presented a high similarity in their variations during the three studied periods (Figure 9-H and Table 2). The tree cities presented close values in the first and third periods, the values decreased to 0.65%, 0.76% and 0.61% between the first period and the second for Constantine, Algiers and Tlemcen, respectively. And continue decreasing by 0.88%, 0.78% and 0.98% between the second and the third period.
Step depthThe step depth measure decreased considerably in the second period for the three analysed cases. The urban restructurings aimed to facilitate accessibility to the different urban fabrics. It was a general character in the cities after the colonial interventions; however, in the third period, it increased in the cases of the Kasbah of Algiers and Tlemcen and slightly decreased in the case of Constantine (Figure 9-I).
IntelligibilityThe diagrams of variation of intelligibility for the three cases showed different attitudes (Figure 9-J). The case of Constantine maintained its growth in the three periods, and the mean values in the other two cases increased then decreased, with a very strong decrease in the case of the Kasbah of Algiers. This is due to the difference between the connectivity values in the latter case, while the case of Constantine has maintained the high connectivity values of the same most connected spaces.
The purpose of this study was to identify the patterns of urban change in the city centres in Algeria and to verify the possibility of classifying them as a model of urban evolution, which may help urban designers and decision-makers to better understand the impact of urban operations on old urban fabrics. The comparison of the evolution of the mean values of each syntactic measure in three key historical periods showed that there are similarities and differences in the configurational parameters of urban fabrics and consequently the impacts of urban operation carried on, these differences being due to several factors that were highlighted in this research.
In the general framework, three-city centres were analysed using space syntax methods and techniques. Two types of analysis were applied: accessibility and visibility analysis, focusing on the spatial changes in three historic periods. The three city centres showed common points and differences. Two main differences can be summarised as follows.
First, the three city centres belonged to three different typologies: Kasbah (Ottoman traditional city), Medina (Tlemcen, Arab walled city) and the old city of Constantine. Traditional urbanism was influenced by the medina’s typology but also influenced by its Roman origins; inside the city, there was a part of a Kasbah.
The second main difference is that the analysed cities were founded on different sites: a coastal site for the Kasbah of Algiers with a considerable slope, a hilltop site for Constantine, which is very rough and does not offer large flat surfaces, and a site in the highlands with a moderately low slope for Tlemcen.
The nature of urban transformations carried out during the colonial period in these cities varied from restructuring to partial and total reconstruction.
Thus, the three analysed cities have common points, their foundations and the periods of their evolution that do not separate from the characteristics of the urban evolution of all the other cities in Algeria. During their colonisation by the French, they were composed of typical traditional fabrics superimposed in the same sites on the Roman cities, except for Tlemcen, in which the cities were juxtaposed. They have undergone colonial transformations with the same objectives but different proceedings.
In the results of the comparative syntactic analysis, we noted a great similarity between the three city centres in the major part of configurational parameters, which confirmed the hypothesis that the traditional fabrics in their origins were built according to a logic of management of the urban space through its configurational parameters and control of access to spaces, especially when it comes to the navigation of foreigners. The parameters that influence the navigation of foreigners had very limited values, except in the central spaces. A hierarchy was applied inside the neighbourhoods to the dwellings’ entrances, and the presence of central cores of neighbourhoods and their forms confirmed that.
The second period showed several similarities as well as differences in the configurational parameters of the three cities, and we noted the same phenomena in the evolution of the syntactic parameters with different variations in their values. The different actions of transformations in the three cities had the same objectives and similar processes, and they gave very similar results despite the differences in site nature and the original typology of the cities. In the three fabrics, the space was divided into two distinct neighbourhoods where the European quarter was more connected, presented more integration as a whole and had more intelligibility. Autochthons neighbourhoods remained the most segregated, were less connected and offered fewer choices except in the breakthroughs that penetrated them. The logic of social segregation was apparent in the local parameters of the three neighbourhoods.
The different rates of variation of configurational parameters showed that some urban fabrics were more flexible to transformations, such as the case of Tlemcen, and others are more rigid, as in the case of Constantine. These differences are due to the points already mentioned: the site and the facilities it offers (therefore, the fabric’s typology and the dimensions of the streets) and the conducted urban transformations, in addition to the social framework.
The third period presented very diversified results; this can be justified by the fact that in this period individual unplanned transformations were done in the traditional fabrics without programming or common objectives, as well as the logic of restoration which tends toward the initial state of the city centre. This period produced various fabrics, and each had a different identity.
As a result, the comparative syntactic analysis allowed us to notice the similarities and differences in the state and evolution of the historical city centres of three cities in Algeria. We tested the variations of the configurational parameters in the face of several types of spatial transformations that can be considered variable from one urban fabric to another according to site parameters, urban typology and the nature of the transformations.
These results can be useful at several levels, mainly in the operations to be carried out in historic centres. First, they facilitate understanding the urban structure because of the process of urban evolution, the principles of its composition and the impact of each on the formulation of the urban fabric. Consequently, they facilitate the understanding of the consequences of any type of transformation in these historical centres and the possibility of predicting their results through a syntactic analysis, which has proven its effectiveness through the analysis model proposed in this research. The influence of the different components of the urban space (the nature of the site and the typology of the urban fabric) and any transformation in their parameters proved by the syntactic parameters allows for remedying the problems of circulation, accessibility in the whole urban space and revaluing the devalued spaces.
Previous researchers in the field were interested in the different models of urban change and influencing factors. Shpuza (2009) studied the urban growth patterns of coastal cities to define generic rules of urban growth. Others tried to interpret the connection between urban morphology and the use of urban space Griffiths (2012). This paper focused on the evolution of cities in Algeria and the results of different parameters (site, land use and urban transformations) to clarify the effect of each one on the resulting urban form, but also draw a model of urban evolution for the Algerian cities that underwent particular circumstances during the colonial period. Previous research on those cities has traditionally focused on morphology and historic descriptions (Gsell, 1920; Mercier, 1903; Mimó, 2016; Mostefaï, 2002; Stora, 2010)
Although the analysis took into consideration three key historic periods, due to the availability of plans, the research remains limited in the mentioned periods, an extension to other periods may give better results. The analysis of the actual state was based on official primary plans established by urban services and dating to 1998–2010; however, they were verified using satellite images and onsite visits.
Conceptualization, S.F.; methodology, S.F; software, S.F.; investigation, S.F.; resources, S.F.; data curation, S.F.; writing—original draft preparation, S.F.; writing—review and editing, S.F.; supervision, S.F. All authors have read and agreed to the published version of the manuscript.
The author declare that he has no conflicts of interest regarding the publication of the paper.
The authors would like to thank D. Nabil Mohareb for his advice during the preparation of the document.
This research received no external funding.
