Scholars interested in the history of travel and tourism have long relied on historical guidebooks as a primary source to past patterns of mobility and cultural ideals. The guidebook’s descriptive and prescriptive nature — describing places, peoples, and cultures, and ordering them in a normative hierarchy of attractiveness — is a treasure trove of information about the practices and ideas constituting past travel cultures [Parsons 2007]. Historians have been keen to query guidebooks for insights about aesthetic sensibilities, communications and infrastructures, and representations of the other. The preferred method for studying such themes has been qualitative, close readings of travel guides (classical examples of this include [Buzard 1993]; [Koshar 2000]). Only recently have tourism historians begun to employ digitally enabled methods, for instance Historical Geographic Information Systems (HGIS) exploring the origins of nineteenth-century Baltic spa guests [Antons 2021] and the family networks of Grand Tour travellers to Italy [Anderson et al. 2017], or semantic modelling of Baedeker travel guides [Czeitschner and Krautgartner 2017]. Scholars of the ancient world have also demonstrated the value of HGIS for studying travel times in the Roman Mediterranean [Meeks and Grossner 2012] and for understanding Pausanias’s Periegesis Hellados — a complex cultural geography of Greece often referred to as the first guidebook [Foka et al. 2021]. What we propose in this article, though, is for historians of modern tourism to take advantage of the comparatively homogenous nature of modern guidebooks and harness HGIS for new, computer-assisted tourism histories.

The primary aim of this article is to scrutinize the persistence of older tourist attractions under communism in Czechoslovakia. In doing so, we also demonstrate the promising methodological avenues for “scalable reading” of the tourist guidebook as a historical source, combining traditional close reading with computer-assisted distant reading [Fickers and Clavert 2021]. More specifically, the article illustrates the potential of systematically mapping the spatial information contained in guidebooks diachronically and in combination with qualitative, contextualizing sources. Approaching modern guidebooks to the same destination as time series enables the researcher to chart the rise and fall of touristic sites. The possibility to pinpoint the omission and disappearance of tourist attractions over time and across large geographical areas is particularly promising as such absences easily pass unnoticed during traditional close readings. The method also allows for visualizing attraction hierarchies by comparing the touristic sites contained in comprehensive guidebooks with those in compact guidebooks from the same period. Such observations can then inspire further research into changes and continuities in touristic infrastructure and forms of attraction. In these efforts, the current work thus seeks to bridge the gap between quantitative and qualitative uses of GIS ([Knowles 2014, p. 207]; [Martí-Henneberg 2011, p. 4–5]; [Blevins 2021, p. 4–13]).

This article illustrates some of the methodological opportunities outlined above through a scalable reading of the first guidebook to communist Czechoslovakia for western travellers after the Second World War. The guidebook Czechoslovakia [Chyský 1959] was published eleven years after the communist coup in 1948, and produced by Čedok, the Czechoslovak National Tourist Organization. To provide diachronic contextualization we compare this guidebook with two predecessors similar in terms of audience and scope, but markedly different in terms of production context. Closest in time is the Guide to the Czechoslovak Republic [Král 1928], Čedok’s first travel guide for foreign readers published under the democratic interwar government a decade after the republic’s foundation in 1918. Čedok published no similar volumes between 1928 and 1959, which makes the two publications ideal for a study of how the communist tourist authorities sought to present the country. The continuity in intended audience (foreign, primarily western tourists) combined with the radically differing political circumstances allows for novel insights into the persistence of old travel ideals under communism. The two volumes are further contextualized by comparison with the reputable German guidebook publisher Baedeker’s comprehensive Austria–Hungary including Dalmatia and Bosnia from 1905, which covered the geographical area that became Czechoslovakia after 1918. Taken together, the travel guides allow us to trace the (dis)continuities of tourist attraction distribution in communist Czechoslovakia.

In 1957, more than one hundred diplomats and travel industry representatives from twenty-nine countries converged on Prague in Czechoslovakia for a five-day conference on international tourism. The event took place four years after the death of Stalin in 1953 at a time when the Communist government, in power since the coup in February 1948, was gradually improving its diplomatic relations with the West. Organised by Čedok, the primary aim of the conference was to promote Czechoslovakia as an international tourist destination and ease the mobility of Western tourists across the Cold War divide. The organizers hoped that the global meeting would facilitate contacts and sow the seeds of future collaboration between airlines, railways, and tourist associations of all continents. Shortly before the Prague conference, the travel bureaus of the socialist states had held a separate meeting in Carlsbad where they concluded that “the most effective path to mutual understanding and comprehension is for nations to speak to nations in the most direct manner, by tourism.”[1] The declaration echoed a stance taken by the Czechoslovak Politburo two years earlier as it revised its policy towards Western tourism in the autumn of 1955 [Bechmann Pedersen 2018, p. 132]. Back then, the leading political organ recognized that “at present, the development of tourism is of particular political importance. Foreign trade is becoming one of the instruments of mitigating international tensions.”[2]

In terms of formal tourism policy, the years 1955–56 marked the country’s turning point away from Stalinist isolation and towards the search for a “common ground between East and West in the postwar world order” later epitomised by the award-winning Czechoslovak pavilion at the 1958 Brussels World’s Fair [Giustino 2012, p. 188]. In 1955, Čedok made its first arrangements with western travel companies and the following year it invited forty western travel industry representatives to the country. Starting in March 1956, Czechoslovakia also began to issue regular tourist visas and Čedok prepared thirteen hotels of higher standard specifically for western guests [Bechmann Pedersen 2019, p. 236]. Instead of seeing Western tourists solely as a threat to national security, the Communist regime prepared to capitalise on international tourism as a tool for cultural diplomacy and an invisible export.

To some extent the efforts of the Communist regime to attract foreign visitors and win international recognition through tourism mirrored initiatives of the newly founded Czechoslovak state in the interwar period. After the breakaway from the Austro-Hungarian empire at the end the First World War, the new state utilised international tourism to gain recognition abroad as an independent nation worthy of its sovereignty [Jeschke 2021, p. 68]. Čedok was quickly established as the country’s national travel agency and a semi-official tool of nation branding with several foreign offices ([Štemberk 2009, p. 156–160]; [Mücke 2021, p. 77]). The country also participated in transnational collaboration on tourism, hosting the second international congress of national tourist organizations in 1926 [Le Secrétariat-Général 1926]. These congresses were continued immediately after the Second World War when the International Union of Official Tourist Organizations (IUOTO) was established with Czechoslovakia as a member. Following the February coup in 1948, though, the Stalinist regime in Prague left the organization at the onset of the Cold War.

Nine years later, in 1957, the country returned to the international tourist organization it had once helped found. The same year the international tourist conference in Prague crowned the reorientation of Czechoslovakia’s international tourism policy. Čedok representatives were sent on at least 14 different trips to Western Europe, North America, and Africa where they attended travel fairs and surveyed the local travel industry for potential partners.[3] These efforts to attract foreign tourists raised fundamental questions with significant ideological ramifications. What, in fact, were communist Czechoslovakia’s tourist attractions, and how could they be ordered in a hierarchy of attractiveness? Where should Western visitors be invited to travel and why? Security interests would have to be balanced against economic interests as the spatial distribution of tourist sites for foreigners risked interfering with military zones and state security efforts to thwart western cultural influences and spying. One aspect of the political dilemmas of international tourism to Cold War Czechoslovakia is captured in a US embassy cable sent from Prague to Washington DC in 1957.

Visitors and permanent foreign residents in Prague often bemoan the fact that there is no up-to-date guidebook to the Czechoslovak capital city. Recently an embassy officer asked an employee of ARTIA, the responsible state publishing house, why this was so. “Oh,” the Czech replied, “you have no idea what a difficult task it is, how great is the responsibility involved in determining which buildings and other landmarks should be included in such a guidebook and what reasons should be given for visiting them. So many offices and ministries would have to be consulted by an author, so many decisions would have to be made…”[4]

Disregarding its somewhat satirical tone, the dialogue recounted in the cable illustrates Mary Heiman’s point how, in Czechoslovakia’s “political system that routinely disgraced and punished anyone who stepped out of line with the current policy directives, there was not a lot of scope for independent initiative.” [Heimann 2009, p. 211] Four years earlier, however, amidst Stalinist show trials and campaigns to root out “bourgeois elements,” Čedok had in fact compiled an 80-page guide in Czech to Prague and its environs [Král 1953]. Yet its lengthy sections on “The recreation of Prague workers,” “Prague on the road to socialism,” and “In the footsteps of the workers’ movement and the history of the Communist Party in Prague” did not provide an ideal template for a guidebook geared to a western audience. Complicating matters further and exemplifying the dynamic and sometimes contradictory process of de-Stalinisation in Czechoslovakia, the editor of the 1953 guidebook was the prominent geographer Jiří Král — author of Čedok’s 1928 Guide to the Czechoslovak Republic — who had been forced into retirement after the Communists came to power and now worked as a tourist guide in Prague to make ends meet. The estranged and somewhat eccentric Král was only rehabilitated in 1966, so he and his manuscript would have been unlikely candidates for the symbolically important job of introducing the country to western readers [Martínek 2008, p. 128–130].[5]

The task of producing the first postwar guidebook for western tourists to Czechoslovakia instead fell on Jiří Chyský. The son of Čeněk Chyský, an important figure in the Czech tourist movement and author of guidebooks to Prague ([Chyský 1934]; [Chyský 1948]), Chyský junior had himself edited a Czech tourist guide to Prague in 1955 and written a booklet on the organization of domestic tourism the same year. Jiří Chyský was also an ardent communist and an employee of the Ministry of Justice who had helped the Party seize control of the scout movement in 1948, so he was undoubtedly attuned to the politics of tourism. Among his collaborators on the international guidebook project were Milan Skalník, an accomplice in the purge of the scout movement, and Vladimír Adamec, author of several guidebooks for domestic readers in the 1950s [Lešanovský 2009, p. 16]. Unfortunately the Čedok archive has been lost, which makes further investigations into the production of the two guidebooks difficult [Hais 2008, p. 5]. From other government sources we nevertheless know that in 1958, the Czechoslovak authorities reached an agreement with the Geneva-based guidebook publisher Louis Nagel, who was seeking to establish his publishing house as the leading provider of guides to Eastern Europe. Nagel’s had already published guides to Moscow and Leningrad [Moscou et Léningrad 1956] and Hungary [Schreiber 1957], and according to Čedok’s investigations, Nagel’s guides enjoyed a fine reputation in Switzerland and France as reliable and objective. The manuscript was eventually provided by Čedok in collaboration with Chyský, and seven other authors. Nagel’s published the result in 1959 simultaneously in English, French, and German editions. The print-run amounted to 3,000 copies to be sold through Western channels aiming to earn hard currency profits.[6]

Nagel’s 1959 guidebook to Czechoslovakia is a typical artefact of the cultural Cold War in which the superpowers vied for popular support at home and abroad. It lends itself well to the kind of close readings of descriptions and prescriptions that historians have honed since the cultural turn of the 1980s. However, with its 290 pages and an index of more than 600 entries, this list of toponyms meets the “big data” criteria set forth by Graham et al. to be “big enough” for “computational intervention to make new sense of them” [Graham et al. 2022, p. 3]. HGIS provides another way of exploring the guidebook’s officially approved image of Czechoslovakia as a destination of Western tourism. It also allows for comparisons with the attractions represented by guidebooks from other periods or produced by other publishers. In this way, historical comparisons of guidebooks allow for diachronic studies of a destination’s tourist attractions across political caesuras, thereby querying the relation between politics and tourism from a new perspective.

Building a suitable corpus of guidebooks for exploration with HGIS involves several methodological challenges. First, attempts to build long time series face the problem of genre fluidity as the modern tourist guidebook is barely 200 years old. Before the publication of the first travel guides by Baedeker and Murray in the 1830s, travel writers liberally mixed advice with art history, geology, folklore, and political commentary. It was only with the introduction of authoritative, standardized, and regularly updated texts for travellers in the middle of the nineteenth century that travel writing and travel guide parted ways [Schaff 2022]. To ensure relative conformity of the compared texts, we recommend limiting a travel guide corpus to modern, post-1830s guidebooks.

The second challenge concerns the creation of a meaningful sample from the heaps of modern guidebooks churned out by a vast number of publishing houses around the world as leisure travel developed from an elite privilege to a mass phenomenon. In this study, the focus is on the officially approved image of Czechoslovakia as represented by travel guides produced by the state tourist agency targeted at foreigners. We therefore chose to focus on two travel guides produced by the same organization (Čedok), geared towards the same audience of relatively affluent foreigners with time to spare for a longer journey abroad; Guide to the Czechoslovak Republic [Král 1928], and Czechoslovakia [Chyský 1959]. Čedok did not publish any similar travel guides for foreign tourists between 1928 and 1959, only some small brochures. To provide further historical contextualization we also compare with the 1905 Baedeker guide to Austria–Hungary. While the German version of this Baedeker guidebook was regularly updated until the outbreak of the First World War, the edition from 1905 was the final one published in English before the eventual dissolution of the Habsburg Empire. Čedok’s 1928 guidebook was published simultaneously in German and English, while the Čedok/Nagel guidebook from 1959 was published simultaneously in German, English, and French. We chose to compare the English editions of the three guidebooks since minimizing the number of German toponyms in the corpus reduced the need for disambiguating place names.

Working with twentieth century guidebooks entails a third methodological challenge. As the personal idiosyncrasies of travel writing gave way to the formulaic descriptions of frequently updated guidebooks the literary status of the genre diminished. Twentieth century guidebooks attained an ephemeral quality similar to grey literature with lasting consequences for preservation and access. Given the short half-life and low literary standing of these books, library and archive collections are often patchy. And even if one successfully locates a complete series of a guidebook title, a fourth challenge emerges: twentieth century guidebooks rank just about rock bottom of current digitization priorities. This is unsurprising, as the preserved books are often in good conditions, oftentimes under copyright, and perceived to be of marginal relevance to the national community’s cultural heritage. Since we did not possess the means required for producing digitized editions with high quality machine-readable text we opted for a simpler and more efficient method to prepare the data for computer-assisted analysis. Instead of working with complete guidebook texts we digitized, cleaned, and mapped only the toponyms included in the guidebook indexes. This cost-saving method builds on the reasonable assumption that the guidebook editors included all the places they deemed most relevant to foreign tourists in the three indexes. Our assumption is supported by the fact that the guidebooks contained comprehensive indexes (see Table 1) referring to an average of 2–3 unique toponym per content page. However, some of the toponyms are duplicates as the English-language indexes also included alternate names in Czech, Slovak, German, and Hungarian. The precise number of toponyms and unique positions for each guidebook is listed in table 1. Note that these numbers do not include indexed toponyms outside the 1959 borders of Czechoslovakia, nor do they include geographic entities that are not meaningfully represented by a single coordinate (e.g. rivers, mountain ranges, large forests). Such toponyms are also excluded from the maps.

The method excludes all toponyms considered too marginal by the editors to feature in the index. However, this apparent weakness works to our advantage, as the point of the analysis is to identify continuity and change in tourist attractions — not to map every town or village mentioned in the travel guide merely to help readers navigate from one attraction to another. In the process we excluded toponyms listed as subentries under the bigger cities in the 1905 and 1959 guidebooks (such as palaces, churches, and museums) since our focus is on the countrywide range of attractions. Mapping the changing tourist geography of a city like Prague would undoubtedly yield interesting results. However, precise geocoding at street level during a period of rapid urban development would require manual labour beyond the scope of this project. For the heatmap visualizations (Figure 2) we still assigned approximate weight to the cities based on the subentries (or the page span in the case of the 1928 guidebook, which did not have subentries).[7]

To geolocate the lists of toponyms included in each index we matched them against a gazetteer with toponyms and coordinates compiled from the open-source geographical database GeoNames, supplemented with toponyms from WikiData. GeoNames boasts of having “27 million geographical names” compiled from 402 different sources, including official sources such as the Czech Office for Surveying, Mapping and Cadastre, and the Geodetic and Cartographic Institute Bratislava.[8] A persistent challenge for HGIS is disambiguation ([Salvoldi 2017]; [Horne 2020]). To capture all the possible places and toponyms inside the borders of communist Czechoslovakia we combined the gazetteers for Czechia, Slovakia, and Hungary and linked them with GeoNames’ own list of alternate names. To ease the disambiguation of historical toponyms featured in the travel guide indexes we supplemented the gazetteer with English and German spelling variations from WikiData, bringing the total number of languages to five. From the onset, our gazetteer thus contained 156,932 toponyms across 68,947 coordinates stored in a relational database.

The automated toponym disambiguation is a particular variant of Approximate String Matching (ASM): the problem of declaring whether two non-identical strings of characters match. It is a classical computer science problem that has inspired numerous solutions in the form of metrics declaring a pair a match or not based on a predetermined threshold ([Levenshtein 1966]; [Navarro 2001]; [Cohen, Ravikumar, and Fienberg 2003]; [Davis Jr. and De Salles 2007]). Machine Learning models offer another solution, but while they tend to perform better, they also require training on large amounts of annotated data and typically lead to static models, needing retrained whenever new pairings are discovered ([Recchia and Louwerse 2013]; [Hosseini, Nanni, and Ardanuy 2020]). In many toponym disambiguation cases, a simple metric approach thus suffices to yield reliable results efficiently ([Recchia and Louwerse 2013]; [Kılınç 2016]; [Santos, Murrieta-Flores, and Martins 2018]). With more than 3,000 toponyms to geolocate and a need to minimize false positives, we opted for a very restrictive threshold to ensure that only highly similar toponyms were auto-matched (Jaro-similarity of 0.9).

For easy data entry, we constructed a place name disambiguation and geocoding application (CITADEL) composed of two parts: a web application with a graphical user interface (built with Anvil), linked with a server-side application (written in Python) to run the automated disambiguation and matching against the toponym database.[9] When users enter index toponyms in the GUI they are matched against the gazetteer in a series of successively looser matching conditions, from requiring a perfect match to Jaro 0.9, to generate a list of probable positions. CITADEL geolocated the majority of the index toponyms and provided a short list of suggestions for most of the remaining toponyms to aid in their manual disambiguation. To assign a coordinate to the remaining toponyms we cross-referenced the indexes, the guidebook descriptions, modern and historical maps to approximate the coordinates. The toponyms we geolocated manually were added to the gazetteer to improve disambiguation and geolocation of other historical sources in the future. Finally, we mapped the toponyms[10] using the open-source GIS tool QGIS and importing historical borders from the CShapes 2.0 Dataset [Schvitz 2022].[11]

To provide a baseline for our interpretation of the 1959 travel guide we began by mapping its indexed toponyms along with the indexes of the 1905 and 1928 guidebooks. Figure 1 shows the results, with each dot representing the coordinate of a toponym within the borders of Cold War Czechoslovakia. Given our analytical focus on the tourist attractions in 1959 we chose to map all guidebooks on the same projection of communist Czechoslovakia’s political borders. Our emphasis on the country’s cold war borders admittedly yields a somewhat misleading impression of the earlier guidebooks since the area’s political geography appeared entirely different during the Austro-Hungarian Empire and the interwar period. Most importantly, in the Dual Monarchy before the Great War the Czech lands belonged to Austria whereas Slovakia constituted a part of Upper Hungary. Moreover, interwar Czechoslovakia also included the area east of Slovakia known as Carpathian Ruthenia (part of present-day Ukraine). We nevertheless decided to exclude Ruthenian toponyms from our analysis since this poor and remote region cursorily mentioned in the 1928 guidebook was transferred to the Soviet Union after the Second World War and thus did not feature in the 1959 guidebook.

At first sight, the three maps in Figure 1 show a similar, fairly even distribution of attractions throughout the elongated country. Upon closer inspection, however, the 1905 and 1928 maps include more toponyms in the western part of the country compared to the 1959 map. To make the differences clearer, we generated heatmaps based on the same toponym data and standardized across the three datasets to allow for comparison (Figure 2).[12]

The heatmaps show a continuity of toponym density in four areas: The cities of Prague and Brno, which had long been the urban centres of Bohemia and Moravia; the Bohemian Paradise (Český ráj) northeast of Prague, famous for its medieval castles and distinctive sandstone rock formations; and the High Tatra Mountains on Slovakia’s northern periphery, a popular area for hiking, hunting, mountaineering, and medical retreats since the late nineteenth century. The heatmaps also show a clear shift in emphasis from the older guidebooks’ denser distribution of toponyms in Bohemia (the western half of present-day Czechia) and thinner distribution in the mountainous Slovakia compared to the 1959 guidebook’s more balanced representation of Czech and Slovak attractions. (Note here that the overall “lower temperature” of the 1959 map owes to its relatively smaller number of indexed toponyms compared to the older guidebooks.)

The high density of attractions in Bohemia on the 1905 and 1928 maps is an expected finding. Bohemia had traditionally been the richest and most populous part of the country and the West Bohemian spa triangle outlined by Carlsbad (Karlovy Vary), Marienbad (Mariánské Lázně), and Franzensbad (Františkovy Lázně) had a long history of attracting visitors from far-away lands [Large 2015]. The borderland’s spa towns continued to cater to an international clientele after the First World War while numerous historical castles perched on hilltops and charming chateaux surrounded by romantic gardens provided spectacular tourist sights. The pre-WWII density of attractions along the borders with present-day Austria, Germany, and Poland illustrates the transformation of areas characterized by high hills and deep forests into recreational sites for hiking, camping, and skiing in the early twentieth century. These parts of the country, Bohemian Switzerland, the Bohemian Forest (Šumava), and the Giant Mountains, were deemed such important tourist attractions that they merited special fold-out maps in the 1928 guidebook, underlining their touristic value and accessibility. Both the 1905 Baedeker and the 1928 Čedok guide also included detailed maps of the High Tatra Mountains, and the area’s many summits, saddles, valleys, lakes, hiking routes, and huts were meticulously described and indexed in all three guidebooks, which explains the “high temperature” there.

More surprising, though, is the relatively thin representation of western Bohemia in the 1959 guidebook and the more balanced representation of Czech and Slovak attractions. The first difference is a likely effect of the cold war militarization of this region where the Warsaw Pact bordered directly with NATO. For security reasons, the Czechoslovak authorities did not want western tourists roaming too close to the state border, which at this point was not yet fully demarcated and fortified [Bechmann Pedersen and Noack 2019, p. 4]. The second difference indicates an attempt to present the two parts of the country, the Czech lands and Slovakia, as equally attractive, despite the latter’s more remote location from the western tourist’s point of the view (which the guidebooks all adopted). Writing on behalf of the new state’s official tourist organisation, Král seems to have recognized the political obligation to give Slovakia a more detailed treatment already in 1928; a concern that did not burden a commercial guidebook publisher like Baedeker at the turn of the century. In his introduction, Král explained that

The greatest attention is of course devoted to the world famous Spas. But also those regions, which nature has endowed with such uncommon attractions — notably in Slovakia and Subcarpathian Russia — and where so far, the foreigner has only rarely found his way, are dealt with here, allowing at least a rough idea to be formed by him, who would penetrate into districts where both people and nature have a character of their own, even if conditions are sometimes below the standard of comfortable tourism.  [Král 1928, p. 5]

Král’s guidebook devoted just 48 of its 272 pages (18%) to Slovakia but it was nevertheless considered exemplary compared to contemporary handbooks on Czechoslovakia. Writing a few years later, a reviewer of two English-language travel guides to Czechoslovakia (published outside the country) held up the reviewed books against Kral’s and found Slovakia treated “in a very perfunctory way” and “rushed through in a very cursory manner” [J. H. R. 1931, p. 371]. Three decades later, the authors of the 1959 guidebook were careful to steer clear of the exoticizing language enveloping Slovakia and the historical bias towards attractions in Bohemia and Moravia. Numerically, 85 of its 290 pages (29%) covered Slovakia.

To further probe the (dis)continuities between the three guidebooks we analysed the indexed toponyms with the help of a clustering algorithm. This method allowed us to merge coordinates referring to the same tourist site. For instance, the 1959 index contained an entry for “Bouzov, castle” while the 1928 index had an entry for “Bouzov”, the Moravian village next to the castle. The geocoding process assigned different coordinates for the two entries, but since the attraction is the same, the clustering algorithm is useful to group such entries under a single coordinate. The clustering process also eliminates the problem that our multilingual gazetteer would sometimes contain near-identical coordinates for identical toponyms. The clustering algorithm merged any points within a radius of roughly 4 km calculated as the great-circle distance of the coordinates into the same cluster. The low radius was chosen to avoid an avalanche of clustering creating meaningless superclusters. When all positions within the radius had been placed in the same cluster, the cluster’s coordinates were calculated as the arithmetic mean of the member coordinates. The method allows for the subtraction of clusters between maps, which renders the (dis)continuities more visible.

Figure 3 shows the 174 clusters comprised of 573 unique, close-neighbouring positions included in all three guidebooks. A quick glance at the map suffices to note that the country’s main border posts and all the bigger cities are represented. A closer reading of the clusters’ locations shows that the shared content includes a number of spas, many of which had been popular with foreign guests since the nineteenth century, e.g. Carlsbad, Marienbad, Franzensbad, Bad Königswart (Lázně Kynžvart), Bad Gräfenberg (Lázně Jeseník), Pistyan (Kúpele Piešťany), and Kúpele Korytnica. The clusters also contain some of the country’s picturesque old towns and famous castles including Český Krumlov, Karlštejn, Kroměříž, Křivoklát, and Loket. Natural landmarks such as peculiar rock formations, scenic cliffs, caves, forests, valleys, hills, and mountains also feature in all three guidebooks. Among the historical landmarks with an established pedigree as tourist attractions included in the books is the Austerlitz battlefield near Brno where Napoleon defeated the Austrian and Russian armies in 1806.

To check for the continuities between the Czechoslovak guidebooks produced by the country’s national tourist organisation, we also mapped the toponyms clusters shared by these two guidebooks (Figure 4) using the same clustering method and radius as outline above. The map includes 290 clusters comprised of 680 unique, close-neighbouring positions. The map is largely identical to Figure 3, though it includes even more famous castles and chateux (e.g. Pernštejn, Kost, and Zvíkov), some of which had possibly been closed to the public or in poor condition around 1905. The shared clusters between 1928 and 1959 also contain some remotely located dwellings such as the small Moravian town of Kralice where the humanistic scholar Comenius worked on a famous bible translation in the seventeenth century. By 1959 a Kralice Bible Museum had opened, which the communist-era guidebook hastened to stress “is also of great interest to laymen because many very beautiful impressions have been produced by the Kralice press” [Chyský 1959, p. 145].

We then mapped the clusters unique to 1928 and 1959 individually, thus identifying the attractions unique to each guidebook (Figures 5 and 6). Using the same clustering algorithm we eliminated clusters containing toponyms from both indexes so that the remaining points were attractions at least 4 km away from any attraction listed in the other guide. Figure 5 thus shows the 480 toponym clusters (636 positions) included in the 1928 guidebook but not present in the 1959 guidebooks. The figure confirms the observation that the 1928 guidebook contained many more references to attractions along the western borders and a heavier emphasis on Bohemia. In comparison, the attractions unique to the 1959 guidebook (Figure 6) are overwhelmingly located in Slovakia, especially around Komárno on the border to Hungary and in the mountains of Central Slovakia. The Moravian Karst north of Brno and the Haná region around Olomouc where the Javoříčské caves were discovered in 1938 also contains toponym clusters unique to the 1959 guidebook. Other unique toponyms represent recently constructed dams and the new recreational areas that had sprung up in their vicinity (e.g. Brno and Slapy).

Returning to the question of the relation between politics and tourism, we can thus conclude based on the toponym distribution maps that the 1959 Čedok editors departed from their predecessors’ emphasis on Bohemia to present a more balanced picture of Czechoslovakia. Otherwise, the result of this distant reading of guidebooks points to the continuity. With a 4 km clustering radius, the communist-era guidebook contains 290 clusters shared with the interwar guidebook and 173 unique clusters (193 positions). A closer look at the unique clusters show that many of them contain attractions unconnected with the Czechoslovak railroad network. The editor of the 1928 guidebook had deliberately organized the text in accordance with the railways, which explains why these harder-to-reach attractions feature exclusively in the 1959 guidebook. As Felix Jeschke writes about editor Jiří Král’s selection criteria, “the railways were the network that made the country accessible, offering a pre-ordered system of routes. Places not served by the railway were not presented as destinations and had to be sought out by intrepid travellers themselves.” [Jeschke 2021, p. 81] Examples of such attractions added to the 1959 edition include the medieval castle of Smolenice, the baroque castle at Milotice, the excavation of fossilised trilobites at Skryje, the remote Slovak hunting grounds at Remetské Hámre, and the small wooden churches built in Byzantine style near Svidník. The addition of sights like these support our argument that the 1959 guidebook presented a country rich in romantic attractions dating back centuries.

Further scrutiny of the guidebook text substantiates our claim that the 1959 guidebook emphasized the country’s inherent appeal to tourists attracted by nature and historical architecture: “The very many natural beauties of Czechoslovakia make the country a tourist’s paradise” [Chyský 1959, p. 22]. The Czechoslovak landscape has a “fresh and green appearance” [Chyský 1959, p. 10], “Slovakia is a paradise for all mountain lovers” [Chyský 1959, p. 133], and the numerous mineral springs are “of world-wide reputation” [Chyský 1959, p. 10]. As we have already shown, the country’s numerous castles and chateaux featured prominently. Pernštejn is “one of the most valuable architectural monuments” [Chyský 1959, p. 133]. Karlštejn is “out of a fairy story” [Chyský 1959, p. 62] and Konopiště was “sumptuously rebuilt” [Chyský 1959, p. 65]. Readers had to look carefully to find tourist sites more explicitly related to contemporary politics or the history of communism. For instance, the Slovak National Uprising against the German occupants and their fascist collaborationists in August 1944 is relegated to a historical sidenote in small print in the section on Banská Bystrica, the city where the uprising took place. The museum dedicated to this event of considerable importance to the party’s self-understanding [Radonic 2014, p. 492] is only mentioned after two sizeable paragraphs about renaissance “burgher houses” and “the old castle buildings” [Chyský 1959, p. 219]. The scalable reading of the 1959 guidebook produced by communist Czechoslovakia’s National Tourist Organization thus demonstrates the continuity with the interwar branding of the country’s many tourist sights.

In her work on heritage management in communist Czechoslovakia, Cathleen Giustino [Giustino 2013, p. 50] has shown how an ideological ambiguity was at play in the country’s castles and chateaux. After the Second World War, dozens of former aristocratic residences were confiscated and turned into state-owned museums, which became popular sites of domestic tourism. And as party loyalists often noted, the sites lacked clearly encoded ideological messages. The translation of communist cultural policy into heritage management and tourist site administration did not result in a simple eradication of the aristocratic material culture. Communist leaders had more pressing concerns than exhibition design at backwater tourist sites, leaving art historians and tour guides some discretion to promote their site’s high-cultural attractions. Our scalable reading of the 1959 guidebook suggests that its editors also had the liberty to steer clear of awkward ideological messages that could rub western readers the wrong way. Instead, the book appears tailored to the educated and historically interested international audience. Returning to the guidebook author’s dilemma recounted in the US embassy cable, our analysis shows that the editors eventually opted for continuity, including many of the attractions already mentioned in the 1905 and 1928 guidebooks. This tendency can be traced back to stylish photo books published by Artia, the state publishing house, since 1953, showcasing the beauties of Prague, as well as castles and chateaux, interspersed with images of socialist society [Paulouš 1953]. Books like these were sold at the Brussels Expo in 1958 where the Czechoslovak pavilion also introduced visitors to historic monuments with photos of castles and chateaux, hoping that it would attract foreign tourists to Czechoslovakia [Giustino 2012, p. 208].[13] This hope, however, remained frustrated for a few more years. It was only by the mid-1960s that Western tourists started to arrive in Czechoslovakia in significant numbers and not until 1978 did the total number of visitors from non-socialist countries exceed one million [Franke 1984, p. 380]. Still, the attempts to “normalize” the image of communist Czechoslovakia in the West and reassert its status as a worthy destination of international tourism began already in the 1950s. By building on the legacy of romantic attractions established in the nineteenth century, Čedok tailored the guidebook to western audiences and showcased the entire country as ready for consumption by foreign visitors.

There is an important distinction to be made between cold war tourism and tourism in the Cold War as Christian Noack and Sune Bechmann Pedersen have argued. While the Cold War was at the centre of the first form of tourism, it was a contextual chronological marker in the second [Bechmann Pedersen and Noack 2019, p. 4–8]. The visualization of toponyms presented in this article allows for a distant reading of tourist attractions in Czechoslovakia and illustrates the symbolic continuity with the pre-war era, undisturbed by the Cold War, which our close reading also supported. Naturally, this was only symbolic. It goes without mentioning that the experiences of western tourists on the ground differed greatly from those of the interwar period, and that other sources than guidebooks are needed for research into this matter. Still, our scalable reading reveals the symbolic attempt by Czechoslovak authorities to normalize tourist relations across the Iron Curtain and re-integrate communist Czechoslovakia in the international tourist industry by perpetuating a pre-socialist tourist legacy. Our study has thus shown the efficacy of analysing travel guides with HGIS and pointed out a promising new methodological avenue for the field of tourism history. To enable similar research for other areas or travel guides, we have made CITADEL publicly available and prepared to be set-up for any combination of countries and languages.

This work was supported by Erik Philip Sörensens stiftelse (grant no. H2019-039) and conducted at the DigitalHistory@Lund research platform, Media History, Department of Communication and Media, Lund University