Geographia Polonica Vol. 93 No. 2 (2020), A new guidebook analysis method for the study of tourist-historic cities: The case of the Maghreb

This paper proposes a new guidebook analysis method for the study of tourist–historic cities, which are well recognized in the Maghreb; hence, six cities from this region were selected for the research sample (Algiers and Constantine in Algeria, Marrakesh and Tangier in Morocco, and Sousse and Tunis in Tunisia). The main purpose of this research paper is to indicate the advantages and disadvantages of this original method.


Introduction
Guidebooks constitute a valuable source of information for the scientist analysing social and cultural phenomena and processes (Bhattacharyya, 1997;McGregor, 2000). In the second decade of the 21st century, they continue to play an important role in studies on tourism (Cordeiro, 2011;Garrod & Kosowska, 2012;Bender, Gidlow, & Fisher, 2013;Roberson, 2015). They are used in long-term historical research, among other things, to create a map of the development of tourist spots around the world from the late 17th to the early 21st century (Antonescu & Stock, 2014).
There are several ways to interpret the content of guidebooks (Peel & Sorenson, 2016). Vajda (2011) highlighted three: geographical, psychoanalytical and lexical. The first examines the changing approach of tourists to space, which is reflected in the graphics of the guidebooks (Verdier, 2011). In the psychoanalytic approach, attention is paid to the content in terms of the methods and purposes of shaping the image of the place (Assoun, 2011). In the lexical approach, the focus is on the vocabulary used to relate to the space (Martinez, 2011). Mazor--Tregerman, Mansfeld, and Elyada (2017) noted that science continues to need new methodological approaches to tourist guidebooks.
In his studies on urban tourism, Liszewski (2014) developed a method for guidebook analysis. It was based on the assumption that once a new spot is considered touristically attractive, it is entered in subsequent editions of guidebooks. This allowed the changing number of places considered tourist attractions to be tracked. A key advantage of such an analysis was the fact that it looked at a long period of time, stretching from the very beginning of tourism in a given city until modern times. Liszewski did not, however, set down the principles for the extraction of relevant tourist attractions from the content of guidebooks. Incorrect identification of such attractions may result in errors in analysis. Moreover, Liszewski did not take objects other than tourist attractions (such as hotels) into consideration; therefore, his analysis could not be comprehensive (Kałaska, 2018).
Tourist-historic city research requires holistic and long-term analysis (Cohen-Hattab, 2004). The potential of tourist guidebooks has not yet been used to analyse such cities. The methods applied thus far do not allow for quantification of the information obtained at the level of individual sections of those cities.
Thus, the aim of this article is to propose a new gu id ebook analysis method for the study of tourist-historic cities. This method contains procedures for obtaining data from guidebooks and processing it in numerical form.

Theoretical background: model of the tourist-historic city
Based on a review of scientific publications up until the end of 2010, Ashworth and Page (2011) distinguished 12 thematic areas of research on urban tourism. Among other things, the studies concentrated on theoretical models. "Whilst critics of models may well argue that the postmodern city is too complex to disaggregate, simplify and reduce to a number of consumption-specific elements, it is a starting point for a more spatially-informed analysis of the city" (Ashworth & Page, 2011: 5). Above all, these models provide answers to questions about city management in the context of tourism development (Ashworth & Tunbridge, 2000). Cohen--Hattab (2004) emphasized the importance of the following models: the accommodation services model (Ashworth, 1989); and Tunbridge and Ashworth's model of the touristhistoric city (2000) -important for analysis in cities with precious historical values. The model of the tourist-historic city based on the analysis of multifunctional medium-sized cities in North-western Europe. The model presents four development phases of the morphological and functional fields of a tourist-historic city. According to Ashworth and Tunbridge (2000: 94): "this, or similar, structural models, can be applied to a number of broad cultural regions other than Northwestern Europe, from where it was derived". Within the framework of the cultural and historical variations of tourist-historic cities, the authors distinguished this variant taking into account the differences in the processes characteristic of Islam and the transformation of pre-industrial urban forms under the influence of western colonization. This led to the creation of dualistic colonial/indigenous cities.
Tourist-historic cities have been mentioned in the studies of the sphere co-developed by Muslims. (Worden, 2001;Sutton & Fahmi, 2002). This term has generally been used only to determine the type of city, without reference to the model. The exception was the Jerusalem variant (Cohen-Hattab, 2004). It has been proved that the use of a theoretical model, omitting historical analysis, does not allow a full understanding of the development mechanisms of tourist-historic cities (Cohen-Hattab, 2004); research should be based on a historical-geographic perspective, combined with theoretical models.

Methodology
New guidebook analysis method This paper proposes an original guidebook analysis method for the study of touristhistoric cities. Five stages of analysis were outlined (Fig. 1).
1. Selection of the research sample and acquisition of source material. The choice of material stems from the assumptions of the research. The actual selection of travel guides occurs in multiple stages. The standardized way of acquiring the guidebooks involves performing consecutive actions, and thus: • conducting a diagnosis on the history of the publication of guidebooks describing the cities of a given area; • selecting one publishing house which publishes guidebooks in collections periodically, every few years. This means that a relatively similar style of describing cities will be maintained for the longest possible research period. It also ensures the greatest possible continuity in the analysis; • acquiring guidebooks from the selected publisher and series; • using only those guidebooks which generalize content about the research area in the same or a similar way. This step is mandatory in the comparative analysis of cities; • supplementing the sample with guidebooks available in other library collections and archives. This is an indispensable measure if there are long time gaps between editions following consultation at the selected library; • expanding the collected source material with items from other publishing houses. This is done in situations where the continuity of analysis is impossible. It is best to include publications issued in the same country and language, which minimizes the impact of differences between the publishing styles of different cultural circles when interpreting the results of the analysis; • finalizing the list of guidebooks selected for analysis. Only then is it possible to precisely determine the time range of the study, based on the dates of the oldest and most recent publications. 2. Identification of the special features of the guidebooks in the sample. The selected guidebooks are characterized by their structure. The sections devoted to the cities are identified and selections are made based on their usefulness to the research. 3. Identification of the units of analysis. It is assumed that the units of analysis are elements of tourism (after Jansen-Verbeke, 1986). They are identified from the selected sections of the guidebook using a previously prepared list of these elements grouped according to type. In connection with the research objective, field observations in the analysed cities are helpful in addition to the literature of guidebooks and publications on tourism. The method of separation of the units of analysis proceeds according to the rules specified by the researcher. 4. Development of analytical categories.
The analytical categories are determined on the basis of the previously drawn up list of the elements of tourism. The list is modified, striving for maximum simplification. Therefore, the smallest possible number of categories is distinguished in relation to the objective and area of the research. However, they must be exhaustive (each element is assigned to a category) and mutually exclusive (no element can be included in more than one category). Each category receives an index in the form of a letter. 5. Coding and counting of units of analysis.
List of results. The raw data is transformed  Figure 1. Guidebook analysis method for the study of tourist-historic cities into a standardized form. Its coding follows the adopted indexed analytical categories. Next, the codes are counted for each year of the guidebook's publication. The results are presented in the form of tables or graphs.

Study areas
The Maghreb has long been a research area for urban tourism (Widmer-Mü nch, 1990; Bencheikh, Gariépy, Poullaouec-Gonidec, Baillard, & Hardy, 1995;Kałaska, 2015;Sarmento, 2017;Kania & Kałaska, 2019). Specialists in the colonial history of tourism have used guidebooks as a source of information, often the only one at their disposal (Boutbouqalt, 2009). Kałaska (2020) proved that the touristhistoric city model had not been yet applied to study the Maghreb. In this region there are places of ancient and medieval origin, transformed during the colonial period into touristhistoric cities, often influenced by initiatives flowing from metropolises (Cohen-Hattab & Katz, 2001). From the 19th century onwards, the cities of the Maghreb were subject to French influences. The French implemented a Westerndominated model of urban development (Steward, 1999). The authorities identified the European lifestyle and urban development policy as a way to legitimize hegemony over a separate civilization. Thus, a dualistic city was formed, consisting of: (1) the medina, inhabited by indigenous people and 2) the European quarter, referred to as ville européenne in French, where Europeans arrived from the metropolises. The medina was seen by the colonial authorities as the antithesis to the ville européenne (Escher & Schepers, 2008). The medina, the "ghetto musulman" (Abdelkafi, 1989: 251), was treated as an accessory of the European quarter (Lawless, 1981).
For the purposes of this study, in each of the three countries considered the core of the Maghreb region (Algeria, Morocco, Tunisia), the number of dualistic cities was limited to two. The study covered a total of six cities (Algiers and Constantine in Algeria, Marrakesh and Tangier in Morocco, and Sousse and Tunis in Tunisia) that correspond to the different categories of tourist-historic cities distinguished by Ashworth and Tunbridge (2000), depending on: • physico-geographical location -Marrakesh is a circular city that is growing steadily in different directions; Algiers, Tunis, Sousse and Tangier are examples of the waterfront variant, while Constantine, due to the location of its historical part on the edge of the deep ravine of the Rhumel river, can be considered an acropolis city; • population -Algiers and Tunis are big cities, with more than 1 million inhabitants in 2014, while the others had a mediumsized population. The population of the largest city was over ten times that of the smallest (Tab. 1). During both the colonial and post-colonial periods, the selected cities occupied a high position in the hierarchy of the countries' settlement networks.
In this study, it was determined whether, in dualistic cities with different demographic potential and physico-geographical features, the analysis of travel guides could be used in research on the history of tourism.

Results: guidebook analysis method in the Maghreb
In the present study, a procedure was applied based on obtaining and processing data from the selected guidebooks, considering their special features, the characteristics of the area and the research objectives.

Selection of the research sample and acquisition of source material
The analysis was based on guidebooks published by Hachette -the longest-running series, from the 1860s to the present day (Berthonnet, 2006), describing the Maghreb in French (Guide Joanne, and from 1919 onwards Guide Bleu).
In 2014, the research material was acquired at the Bibliothèque du tourisme et des voyages in Paris, which has an abundant collection of guidebooks collated by the Touring Club de France and dating from 1890. Literary items were selected that contained information on at least one of the studied cities with similar content generalization. Publication deficiencies were supplemented with versions from the resources of the National Library of France. Due to the inability to acquire Guide Bleu guidebooks on Algeria published after 1986, English-language items were added to the sample. Ultimately, 35 guidebooks were selected from the period 1882-2014.

Identification of the special features of the guidebooks in the sample
The research sample consisted almost exclusively of French-language classical guidebooks, treated as a compendium of knowledge, and a textbook on art history, presenting the natural, economic and sociodemographic aspects. Descriptions of monuments predominated. Over the years, the structure of these guidebooks did not change substantially. They were split into four sections for each city: • an introductory description (including location), • a historical background, • a list of recommended elements of the touristic facilities, • a basic touristic description, often divided into sightseeing routes, containing information on tourist attractions and tips about landmarks located along the route; in guidebooks issued before 1914, the distinction between tourist attractions and other features is virtually impossible. During the colonial period, the guidebooks were written by scientists, politicians and artists (Hillali, 2009). Barthes (1957) accuses them of succumbing to political pressure. As in the case of opinion-making newspapers (Kałaska & Wites, 2015), the narrative of postcolonial French guidebooks is based on neocolonial relations. However, the commercial purpose of their publication is more important.

Identification of the units of analysis and creation of the database
The basic units of analysis are the elements of tourism identified from two sections of the guidebooks: (1) from the list of the touristic facilities and (2) from the basic touristic description. The elements listed under these types were extracted (Tab. 2).
Three principles were followed in separating out the independent units of analysis. The following were distinguished: • as many individual urban elements as possible. Separate records were created for cemeteries, as well as tombs. Fortifications were separated into city gates and forts. Components located inside a building or constituting fragments of a construction or decoration were not considered independent units of analysis. The furnishings of temples and objects in museums was not itemized; • any urban element with its own name; • groups of urban elements that were given a spatial identification in the guidebooks.
A separate unit of analysis was assigned for residences located on a street named in the text. Certain items were mentioned in a guidebook more than once. Their overrepresentation did not cause any problems. Some units of analysis were both a part of the touristic infrastructure and a tourist attraction. In total, 8,207 units of analysis were included in the database. The tables indicate which editions of the guidebooks they were mentioned in. If a unit of analysis was not described in a year of publication, but was included in the previous and following editions, then it was considered to have been mentioned.

Development of analytical categories
The first category was location. Considering the dual nature of the cities studied, it was necessary to determine in which parts of the cities the separate elements were located: within the medina or outside it (according to the administrative boundaries from 2014). Beyond the borders of the medinas, among other things, lay the European quarter that had been shaped in colonial times. The delimitation of the medinas was performed on the basis of a uniform criterion. Its borders were the city walls depicted on the city map from the beginning of the colonial period. This period was not chosen at random. At that time, the authorities had not yet violated the original tissue of the medina. The remaining categories were defined based on the previously drawn up list of types of touristic elements (Tab. 3).
The elements distinguished from the basic touristic description attract tourists to the historic city, thus constituting its physical setting. The elements taken from the touristic facilities list are referred to as facilities that provide tourist services. The architecture and urban design features were differentiated not according to their original utility function, but according to when they were created. Due to the lack of construction dates, colonial and post-colonial structures have been grouped in one class. The age of buildings and the origin of their architects and investors conditioned the decisions of the authorities to turn them into tourist attractions (Tab. 3). Finally, each category received an index in the form of a letter.

Coding and counting of units of analysis. List of results
The coding of the units of analysis proceeded according to the indexes of the adopted categories ( Figs. 2 and 3).
A total of ten two-element codes were created. They were counted according to the date of publication of the guidebooks. Changes in the number of the codes are presented in the bar charts. The cities were collated according to four codes relating to facilities (Fig. 2), and six codes relating to physical setting (Fig. 3). Each of the graphs compares four editions of the guidebooks published, as far as possible, at regular intervals (every 22-34 years).
The directions of changes in the number of touristic elements in the dualistic cities of the Maghreb were identified (Figs. 2 and  3). The clarification of these changes was based on the subject literature and, where possible, took into account the subjectivism of the authors of the guidebooks. Two theoretical models -the accommodation services model (Ashworth, 1989) and Tunbridge and Ashworth's model of the tourist-historic city (2000) -proved to be helpful. The interpretation of the results of the analysis are presented on the basis of two types of touristic element represented by the codes AM, and CB.  Accommodation in the medina (AM). In the period 1921-1955, the number of accommodation facilities in the medina was less than ten in all of the analysed cities (1921-1954-Morocco, 1925-1955. This statistic did not include pre-colonial caravanserais (funduqs), which operated along the roads connecting the main mosque with the gates of the city walls. They were overlooked in the guidebooks because they did not offer enough comfort to their readers, who often came from aristocratic families. They were also used by visiting merchants and Muslim pilgrims (Widmer-Mü nch, 1990). They formed the concentration area of A-type hotels, according to Ashworth's classification (1989).
From the 1920s until the end of colonialism in the Maghreb, the number of accommodation facilities increased only in the medina of Morocco. The largest increase was observed in Tangier, frequently visited by merchants and diplomats. Especially in this city, but also in the others analysed, there were small hotels and guest houses with a capacity of no more than 50 rooms. They were invested in by Europeans arriving to stay permanently (Clandermond & MacCarthy, 2013). Families offered tourism accommodation in typically modest conditions, but in new buildings squeezed between compact constructions. They were located on the main square of the medina (for example, Petit T a n g ie r T u n is S o u s s e A lg ie r s C o n s t a n t in e M a r r a k e s h T a n g ie r T u n is S o u s s e A lg ie r s C o n s t a n t in e M a r r a k e s h T a n g ie r T u n is S o u s s e A lg ie r s C o n s t a n t in e M a r r a k e s h T a n g ie r T u n is S o u s s e A lg ie r s  Socco in Tangier) or nearby (for example, in the Zitoun Kedim district in Marrakesh). The s mall hotels near Jemaa el-Fnaa Square (Marrakesh) were focused around the station created by the French serving intercity vehicular traffic. However, this was a concentration area of B-type hotels (Ashworth, 1989), rarely encountered within the walls of the Maghreb medina.
Only inside the city walls of Morocco were individual, luxury, medium-sized hotels (50--100 rooms) created. They were built on the outskirts of the medina, on the side of the European quarter. Hotel La Mamounia was established in a palace in the Arab gardens of Marrakesh (Llanes, 2009). On the other hand, the Continental Hotel was built on the city walls of Tangier.
Even during the colonial period, small hotels and guest houses were losing their profitability due to the small number of facilities for demanding Western clients. Some of the owners decided to close them. Underinvested accommodation facilities also ceased operation in the post-colonial period. Only in the medina of Marrakesh were low-budget hotels being created up until 1980, when the transport station was moved outside the city walls. In Algerian cities, this was observed until 2007.
In the cities of Morocco and Tunisia, at the end of the 20th century, an increase was observed in the number of small hotels and guest houses, most often those of a high standard. This was associated with the renewal of the medina. The process of rehabilitation differed depending on the country, which affected the rate and scale of changes in the accommodation available. In Tunis and Sousse, it was difficult for foreigners to run businesses, and only a few residents had sufficient savings to invest. In these conditions, few accommodation facilities were created (Escher & Schepers, 2008). Meanwhile, in Morocco, foreigners were not legally restricted. Especially in Marrakesh, they could buy dilapidated palaces and other historic structures for next to nothing, renovate them and turn them into guest houses (Kurzac-Souali 2009, Montedoro, 2018. This scheme creates a new type not included in Ashworth's model (1989).
Colonial and post-colonial architecture and urban design feature beyond the medina (CB). With the increasing urbanization caused by the influx of Europeans to the cities of the Maghreb, there was an increase in the number of facilities created according to Western standards outside the medina (1921-1954-Morocco, 1925-1955. In the travel guides, the most frequent examples of colonial architecture were those located in the vicinity of the historic core. Some of these buildings fulfilled service functions for residents and tourists. Others constituted part of the touristic scenery on the route connecting the tourist attractions of the medina with accommodation facilities located mainly in the European quarter. These were secondary attractions, passed on the way to the primary attractions (Swarbrooke, 2002).
The authors of guidebooks from the colonial period documented the development of the European quarter. Residences, diplomatic missions, other public institutions and Christian sacral buildings were built, particularly vigorously for the commemoration of important events. Public space was enriched with small architecture. Fountains, monuments, busts, commemorative plaques, obelisks and music kiosks began to appear (Oulebsir, 2004). Visible signs of European presence were established: monuments of figures related to the politics and history of France and the colonized Maghreb. Celebrated figures from the world of science, culture and sport were commemorated. After World War I, numerous monuments of the fallen appeared, especially in the cities of Algeria (Monmarché , 1938).
There was a decrease in the number of colonial elements outside the city medinas mentioned in the guidebooks (1954-1978-Morocco, 1955-1977. Above all, the description of the sights most associated with colonialism came to a halt. The authorities began to cut themselves off from the past by dismantling the material symbols of foreign rule and culture. Only a few months after the proclaiming of independence, many monuments, busts and commemorative plaques related to the French were removed (Grabar, 2014;Kazdaghli, 2008).
Between 1977 and 2008, a further decrease was observed in the number of colonial features outside the medinas of Algerian cities. During the period of Islamization and civil war, some non-Muslim temples were destroyed (in Algiers, St Anna's Church was replaced with a shopping centre). Most of the structures, however, were preserved, but gradually fell into disrepair. No conservation efforts were made because they were not considered to be part of the national heritage. They were subject to architectural and functional transformations. This applied to churches and synagogues (in Algiers, St. Louis' Church was transformed into a cultural centre, and the synagogue at Chibah Said Street was adapted to house a boarding school). In many cases, it was the job of the authorities to desacralize them by giving them a new usability. Most often they became mosques (Nessark, 2014). Stripped of their original decorations and furnishings, they were no longer of interest to foreign tourists, mostly Christians, and consequently ceased to be described in the guidebooks.
Travel guides focussed on the increasing number of colonial features outside the medinas of cities in Morocco and Tunisia (1977-Tunisia, 1978-2012. From the end of the 1980s, decrees began to be published by virtue of which the first structures from the time of the French protectorate became monuments. 1998 saw the launch of the first rehabilitation project within a European quarter in the Maghreb. It incorporated constructions from the 19th and 20th centuries, tangential to the medina of Tunis (Ben Becher, 2003). These activities resulted in the restoration of splendour to many buildings, which began to be recognized as primary attractions. In Morocco, the approach of the authorities to colonial development also changed (Cattedra, 2003). The decision was made to renovate buildings which, in time, began to be described once again in the guidebooks.

Discussion and conclusion
The article presents the author's own, innovative, five-stage guidebook analysis method for the study of tourist-historic cities. Figure 1 is its graphic representation. The proposed comparative studies were conducted on a sample of six tourist-historic cities of the Maghreb -Algiers and Constantine in Algeria, Marrakesh and Tangier in Morocco, and Sousse and Tunis in Tunisia.
The applied method allows for tracking changes in tourist-historic cities from the very beginning of the existence of tourism (which in the case of the Maghreb means since the 1920s or earlier). The advantages of this method include: (1) identifying tendencies and directions of changes in the number of touristic elements; (2) comparing cities with each other, even over the long-term; (3) consistent inference about touristic changes in cities based on one research method. This makes it possible to track changes in the number of (1) all touristic objects (not only tourist attractions); (2) touristic objects within specialized morphological-functional areas of tourist-historic cities (in the Maghreb, e.g. medina).
The disadvantages of the developed guidebook method include: (1) the lengthiness of the stage involving the determination of the units of analysis (touristic elements) identified from sections of the guidebooks using the prepared list; (2) the lack of possibility to reconstruct the condition of tourist facilities for a given year. These elements are selected by their considerable number, especially in guidebooks issued since the end of the 20th century; (3) the lack of information on all the authors of the guidebooks, which is necessary for interpreting the results of the analysis. In many cases, theref ore, their level of ideological commitment cannot be assessed.
The usefulness of the proposed method is limited in the case of towns with a small area and low population. In towns of low demographic potential, a small number of touristic objects can be extracted from the guidebook content, which hinders or prevents tracking tendencies and directions of changes over time. However, it is difficult to establish the lowest value of area and population that serves as a basis for drawing justified conclusions. In order to specify such a limit, it should be helpful to apply the method to a larger number of research studies in different areas of the world.
The research did not involve a thorough delimitation of the tourist-historic cities in the analysed cities. It would be worthwhile using geostatistical analysis for this purpose (for example, hotspots analysis). The implementation of such methods, however, requires the establishment of geographical coordinates, which is not always possible, especially in the case of touristic elements from the colonial period.
The aggregation of data from travel guides could determine a new perspective for the study of tourist-historic cities, especially in the Global South, including the Maghreb. Historical sources are incomplete, and the credibility and accuracy of the official statistics published nowadays arouses reservations in countries with authoritarian rule (Kiregyera, 2015;Plekhanov, 2017). Research on tourist-historic cities is aided by interpretations of the content of travel guides. Bearing in mind the traditionally non-quantitative approach to the topic in this part of the world, thanks to the applied procedure, it was possible to quantify data -that is, to transform it from a qualitative to a quantitative form.
Editors' note: Unless otherwise stated, the sources of tables and figures are the authors', on the basis of their own research.