Lists are the logistical core of most journeys: no list, no travel. Even before my departure, I have already used many lists to organize the journey. I might have chosen one of my favorite Airbnb accommodations from my Airbnb wishlist, which collects desirable places. Using a travel guide, I may have created a sublist with sights I would like to visit; and, of course, I have consulted TripAdvisor lists for my destinations. The Michelin guide helps me choose restaurants with one or two stars. Before I leave home, I might make a packing list, helping me not to forget the sunblock. A checklist facilitates a smooth departure. Of course, I take the lists with hotels, sights, and so on with me and update them during the journey, and I might use additional lists, enabling me to adjust plans midtrip and thus travel more spontaneously. When I get back home, I might produce a list with my best photos, adding them to my Tumblr, and I sometimes submit reviews of hotels and restaurants to TripAdvisor. This is only a small selection of a diverse multitude of lists that we use before, during, and after a journey. Lists are crucial but often overlooked devices for organizing a journey; they point at the important fact that a journey starts well before and ends after the bodily departure—and not simply, as is often emphasized, by creating imaginaries of destinations and cultures, but in a much more mundane sense. Lists orchestrate a traveler’s mobility; they are navigational and logistical devices that might even begin a life of their own.

This essay won’t be able to address the whole ensemble of interlocking lists that make traveling possible and perhaps even enjoyable. To understand the role lists play in travel planning, I want to focus on how different sorts of travel lists evolved, and to do this I turn to three historical types of travel list. First I briefly discuss the role of lists in early modern travel, a time before guidebooks or the modern concept of tourism. This historical account shows us how traveling and the generation of knowledge are intrinsically linked. Then, in the main part of the essay, I discuss different types of lists in early travel guides (such as Baedeker) as crucial cultural techniques for organizing and navigation. This leads me to a brief discussion of the contemporary usage of user-generated travel lists in social media (such as TripAdvisor). What is common to these heterogeneous lists is that they are not simply neutral instruments; rather, they are logistical techniques for generating knowledge.

1 Lists in Early Modern Travel

Long before the first guidebook was published in the middle of the nineteenth century, lists played a crucial role for travel. In the sixteenth century, Spanish merchants were given lists with questions about the region they were traveling to (cf. Carey 2012). According to Carey, these questions covered not only natural resources, but also populations and culture. One Spanish questionnaire asked the traveler to look for »all notable things,« including remarkable cultural sights. Similarly, Robert Boyle produced lists for travelers to Virginia, with categories such as beasts, medicine, and insects. Later lists added questions about the climate and the population. These early lists (in the form of questionnaires) had two crucial functions. By suggesting categories and classifications, these lists were able to spark and direct observations: The traveler knew in advance which class of phenomena might be noteworthy, thus prestructuring his or her attention. The ›empty‹ list of the questionnaire was used as a tool for collecting data during the journey. The production of visibility and the collection of data were the two central functions of early travel lists. The use of these lists was well embedded within the production of an »imperial archive« (Richards 1993), gathering data from all over the world. The aim of British colonial politics was to create a »complete taxonomising of the globe […] through empire, the world could be engrossed and enumerated, identified and indexed« (MacKenzie 2005, p. 20).

This creation of a global archive is intrinsically linked to how lists work. Lists have, as Umberto Eco noted, »the capacity to collect the world« (Eco 2009, p. 49; cf. Stäheli 2012). Why are lists able to fulfill this task of becoming the ›imperial archive‹? Jack Goody (1977) has emphasized that lists do not have a syntactic structure. The entries of a list are not connected by pregiven relations; rather, lists are governed by the logic of the and. There is no immanent grammar of the lists; there is no necessary sequence of the items within a list. It is precisely this lack of a discursive or narrative order which makes lists ideal devices for collecting the global. In principle, it is always possible to add an additional item without destroying the order of the list. By combining these three functions—visibility, collection, and globality—lists became ideal techniques for colonial powers. Lists unified the idea of control and openness: The observer is directed by the lists, and she becomes accountable when returning from the journey (thus making traveling a crucial tool for the generation of knowledge); at the same time, lists’ epistemic openness allows them to accommodate unheard-of data. In this sense, the list was a nondogmatic epistemic technique, since it was not dominated by a canon of preexisting knowledge. This does not mean that these questionnaire-lists were neutral devices. But the power of these lists rested not on preexisting rules, but—for example, with the questionnaires of the Royal Society—on the nature of the questions asked (cf. Carey 2012), and the issue of whether classificatory systems should be applied, and, if so, which were most appropriate.

This brief historical sketch shows us that traveling and list-making are intrinsically interwoven. Although lists are mundane written cultural techniques, they are not a simple addendum or supplement to the journey; rather, they are constitutive for the journey itself. It is this constitutive function of lists that interests me when I turn to modern guidebooks, such as Baedeker. We are now confronted with the opposite situation. The guidebook offers neither empty classificatory systems nor open questions; rather, it presents the traveler with a ›full‹ list of tourist attractions and infrastructures. The traveler no longer has to complete the list, but can instead use a ready-made list to select the highlights of her journey. However, this situation changes again with the introduction of flexible and open user-generated lists, such as TripAdvisor.

2 Lists in Early Travel Guides: Baedeker

The beginning of the history of the modern guidebook is often identified with Murray’s Handbook for the Traveller on the Continent (1836) in the United Kingdom and with the first Baedeker Rheinreise (1832) in Germany. In contrast to earlier travel writing, the modern guidebook is not a literary text; it might contain little narrative passages, but it mainly consists of descriptive passages and several types of lists. This points at the logistical and navigational function of guidebooks, marking »what ought to be seen« (Koshar 1998, p. 323).

Guidebooks have to solve several crucial problems that arise from their list-based format. First they have to produce the ›units‹ (such as sights) they are going to order. This point might at first sound banal, but it is actually one of the most contested aspects of guidebooks: How should a guidebook select and ›abstract‹ single items from the fullness of a region?Footnote 1 List-making rests on an implicit everyday theory of what constitutes a sight. As we will see, an easy way of handling this question is to refer to already existing items, for example, in art history (i.e., selecting a famous cathedral or museum). The process of generating entries, however, becomes more difficult if one cannot rely on preconstituted items. Think, for example, of a landscape or a particular point of view.

After the items for a travel guide’s list have been produced and selected, list-making is confronted with a second crucial problem. Although the ›ideal‹ model of a list does logically not presuppose an ordering structure or a classificatory system, empirical lists do always bear marks of minimal ordering structures.Footnote 2 This might be a basic temporal order (e.g., the succession of events) or an arbitrary ordering system (e.g., alphabetical order). The guidebook is confronted with the task of establishing an order within the listing of the sights and attractions it recommends. It would be possible to order the items of a guidebook by different criteria: The writer might simply reproduce the order of his own journey, thus replicating his subjective experiences. Alternatively, one might introduce an ordering principle that facilitates the search for data; or entries might be ranked according to their quality or cultural significance. The first option—using the subjective journey of the writer—was too close to classical travel writing, thus impeding the usability of travel books as an impersonal tool. The entries should not be ordered by subjective experiences and journeys of the travel writer. The ideal guidebook was not a subjective account, but an »impersonal object« (Buzard 2002, p. 48)—Baedeker became famous for presenting its recommendations as objective knowledge. As one writer, A. P. Herbert, noted (1929; quoted in Parsons 2007): »Kings and Governments may err. But never Mr. Baedeker.« To no small degree, this objectivity effect was based on the many lists that were used. They presented the data as necessary in a manner that was supposedly objective, if one did not consider the selection, classification, and ordering processes.

How did early popular guidebooks attain this ›objective‹ and ›neutral‹ mode of representation? Let me turn exemplarily to the Baedeker. The early Baedekers used infrastructures, especially rail- and water-based transportation, as their primary ordering principle (cf. Bruce 2010, p. 93; Müller 2012, p. 38).Footnote 3 Thus before considering the economic or cultural value of sights, guidebooks were organized by their infrastructural ›hardware‹—especially the traffic system, and, as a result, the accommodation and restaurants that were available to the traveler. The rhythm of the routes was dictated by the speed of the transport system—be it a train or a ferry. Often the guidebooks included timetables (the Lonely Planet books include such information even today, which seems a bit anachronistic considering that up-to-date information is easily available on the web). The early Baedeker guides highlighted notable cities, then supplemented them with a listing of restaurants and hotels (cf. Lauterbach 1989, p. 217).

For example, Baedeker’s A Handbook for Travellers on the Rhine, from Switzerland to Holland (1864) explains, after a short preface, the ordering principle of the recommended tours. A basic ›skeleton tour‹ is fleshed out in the main part of the book, starting with Bale and ending with three possible day tours from Dusseldorf. This part comprises sixty-two different, interlocking partial tours. The ›short‹ option is a six-week tour from Frankfurt to Dortmund. The description of these tours is based on the railway network and the trains’ speed: »The rapidity of railway communication is taken into account in these calculations. The pedestrian excursions are all within the reach of any tolerable walker« (Baedeker 1864, p. x). This does not mean, however, that the proposed tour simply follows the schedule of public transport. The guide also suggests stops on the routes: The »impressions […] formed from the window of a railway carriage, or even from the deck of a steamer« only provide »cursory glances« (ibid.). The tour introduces a second rhythm, which rests on the official transport schedule, but allows individual variations by combining different smaller tours according to the traveler’s taste and interest.

It is remarkable that this combination of standardization and variation introduced the principle of modularity, long before that principle became a crucial element in industrial production and architecture. Modularity presupposes loosely coupled, independent elements that can be individually recomposed (cf. Stäheli 2017). For modularity to work, it is essential that the different elements are not integrated into an overreaching narrative whole, but instead have an independent existence. Using the combinatory possibilities of modularity, the individual traveler is posited as a partially independent traveler who is given advice about how to use the transport system in her own way: »The railways by which in different directions the Rhine-land is intersected, and the numerous steamboats which navigate its rivers, afford so many opportunities for travelling« (Baedeker 1864, p. xii). The Baedeker solves the ordering problem by referring to a preexisting order of transport networks that served as the basis for the itineraries. It is by this external reference to the hardware of travel infrastructures that the guidebook becomes an ›objective‹ or ›impersonal‹ tool. Traveling is now not so much about the fulfillment of vague subjective desires as, rather more pragmatically, about precise demands that can be fulfilled by an existing infrastructure. It is thus no surprise that the producers of modern travel books deeply disliked elitist travel writing by professors, who often recommended attractions that were hard to reach (cf. Parsons 2007).

This infrastructural foundation of the guidebook worked well until the transport infrastructure changed, notably with the introduction of highways and mass-produced cars. In his reflections about the future of the guidebook, Karl Baedeker remarked:

»My motto ›do away with guide thinking‹ [›Los vom Führungsgedanken!‹] means: away with the Routes, away with the railway, away with the space of traffic [Verkehrsraum]. The transport world is developing so rapidly that it can no longer be fully grasped. In my view, the basic error of any automobile guide, of our treatment of the motorist too, is that the psychology of the driver is totally forgotten. His requirements cannot be served in the same way as those of the railway passenger of old.« (Baedeker 1937, p. 68)

It is noteworthy that during the heyday of fascism, Baedeker’s motto was »Los vom Führungsgedanken! […] überhaupt los vom Verkehrsraum.« This indicated a decisive shift, a decision to abandon the classical structure of the guidebook and its foundation in fixed transport infrastructures. With the car and the highway system, a new conception of space emerged that was no longer organized by fixed timetables and public transportation. Although the driver of a car still depends on a network of roads and highways, the automobile infrastructure affords a decentralized and more individual type of movement.Footnote 4 With the car, the »linear itinerary« came to an end (Bruce 2010, p. 103). Now the bulk of the guide consists of, not itineraries, but an alphabetized catalog of points of interest. The alphabetical order introduced a new idea of space, a non-Euclidean space with an ordering principle of its own, hinting at the organization of topological spaces by lists. While the classical itineraries were still marked by traces of spatial references, the alphabetically ordered places are loosened from any topographical relations of proximity. The partial abandonment of spatial references opened up a whole new horizon for ordering lists, and the list itself is the ideal medium for the creation of new ordering principles. The minimal order of a pure list is a permanent seduction to introduce ordering principles: »Most importantly it encourages the ordering of the items, by number, by initial sound, by category« (Goody 1977, p. 81). These ordering principles do not have an intrinsic relationship with the items they process—and this externality of the ordering mechanism is what eliminates the need to refer to the journey’s topographical space. The alphabetical order increased the entries’ findability; that is, it exploited the proto-database function of lists, enabling to navigate a data space. To do this, it was necessary to sever entries from their contexts, to loosen them from their integration into itineraries or spatial proximity.

Even before Baedeker introduced the alphabetical ordering of entries, it used another form of non-Euclidean lists in addition to the itinerary. Beginning with the Switzerland guidebook from 1844, Baedeker used a different ordering mechanism: the simple valuation system of asterisks (cf. Bruce 2010).Footnote 5 First only notable sites received a star; later the guidebook introduced a two-star system, which then was applied to hotels and restaurants as well (cf. Koshar 2000, p. 36): »The best hotels and everything particularly worthy of note is indicated by an asterisk« (ibid., p. xxvii). The asterisk valuation system treated a very heterogeneous set of singular objects equally (cf. Karpik 2010 on the valuation of singular goods): Touristic sights and infrastructure were all rated the same way, with either one or two stars.Footnote 6 The asterisk formalized what before had been expressed linguistically, through adjectives such as remarkable or outstanding: »Karl Baedeker’s asterisks served as his laconic substitute for the adjectival raptures of competing authors, and he awarded them with careful and sometimes idiosyncratic discretion« (Mendelson 1985).Footnote 7 Looking back at the introduction of the asterisk system, Karl Baedeker provided a remarkable »poetics« of the guidebook:

»In travel guides, the lyrical moment, poetry, shines mainly by its absence. In this respect, its task is to be exceled by reality. It informs the traveler with universal valuation; it tries to make him see and recognize the true, although always contested quality of things. Thus, its emphasis is not subjective, but as objective as possible in the sense of universal validity. It is here where there are huge difficulties. The well-known Baedeker stars represent most visibly this lyrical moment, including its dangers[.]« (Baedeker 1974; my translationFootnote 8)

This quote points at the problem of the new system of valuation: The itinerary-based ordering received its objectivity from the material infrastructure of transport. The principle of valuation, indicated with Baedeker’s star classification system, is no longer grounded in any form of material objectivity; it is objective »in the sense of universal validity« based on the aesthetic common sense at the time, but also on the popularity of attractions. However, which measures should be used for determining touristic values? The foundation of the star-based valuation system was much more fragile than that of the itinerary-based lists, and even more open to criticism than the alphabetical ordering of entries: How can one explain the selection of some items as more star-worthy than others? And how can one reconcile this selection, which at first glance is highly subjective, with the ideal of the guidebook’s impersonality? The solution Karl Baedeker’s ›poetics‹ retroactively presents is to formalize the lyrical moment of the guidebook: The asterisk becomes the universal, minimalist poem of modern tourism. The asterisk stands for itself; it does not require an explanation or legitimation.

The nondescript asterisk proved to be an extraordinary invention, loosening aesthetic and cultural judgments from transparent reasoning. Most guidebooks have adapted to this practice; even today, it is nearly impossible to find a guidebook without a list of the ten most outstanding attractions, be it national parks, cities, restaurants, or beaches. It comes as no surprise that the introduction of the asterisk system was seen as defining characteristic of the modern travel guide (cf. Koshar 2000): It turned travel into tourism (cf. Lauterbach 2006). Even more than that, the asterisk system Murray and Baedeker devised was a crucial step, preparing the ground for the diffusion of starring systems into many fields, which today include not only hotels and restaurants but also, for example, movies, books, and investment funds. It is remarkable that one of the most popular and prevalent rating tools—the asterisk—was invented by travel guide authors. Yet it is no coincidence that the travel book prompted the invention of this early formalized rating system. The genre of the travel book had already accustomed its readers to a discursive mix of descriptive texts and lists. The isolation of items from their geographical and cultural contexts is a necessary requirement for rating discrete items. These formal preconditions go hand in hand with the intended aim of guidebooks. They are pragmatic manuals, helping to organize the logistics of traveling. They do this by addressing primarily not an expert audience, but the layperson who is not yet familiar with the destinations she intends to visit. In this sense, the starring system was an early example of the popularization of knowledge, which includes a particular style of communication: If one wants to communicate expert knowledge to a lay audience, recommendations should be clearly and unambiguously marked. To understand the importance of a Baedeker star, one does not have to absorb the cultural reasoning that has led to the choice; the system removes those justifications, the arguments supporting the authorities’ choices, and replaces them with an implicit trust in the authorities themselves.

This does not mean that the asterisks were not contested. The celebrated stars soon became a primary point of ridicule and critique. The ›lyrical moment‹ of the arbitrary system apparently did not work as smoothly as the ›infrastructural foundation‹ of the itineraries did. Basically, the new ordering principle produced two closely interlinked but different criticisms. Often the independence and the quality of the judgment were questioned. Some asked, what are the criteria for Baedeker’s judgments? But even more important was an argument that was emphasized in cultural critiques. This line of attack can be found as early as 1863, in a cartoon published in Fliegende Blätter featuring two travelers. One traveler is gazing at two beautiful girls; the other hesitates to watch as well: »Hat sie Baedeker besternt? Sonst sehe ich sie nicht an« (Lauterbach 2006, p. 75f.). There is a long continuity of such criticism of the authoritarian voice of travel guides (cf. Koshar 2000, p. 37). These nostalgic critiques do not simply neglect the question of how users of travel guide deal with the starred recommendations. More importantly, they devalue the specific knowledge the travel guides exemplifies: a logistical, navigational knowledge that is a provocation for aesthetics. For example, Roland Barthes, in his essay on the Guide Bleu, criticized its use of asterisks and labeled it an »agent of blindness«:

»To select only monuments suppresses at one stroke the reality of the land and that of its people, it accounts for nothing of the present, that is, nothing historical, and as a consequence, the monuments themselves become undecipherable, therefore senseless. What is to be seen is thus constantly in the process of vanishing, and the Guide becomes, through an operation common to all mystifications, the very opposite of what it advertises, an agent of blindness[.]« (Barthes 1993, p. 75f.)

For many cultural critics, ›star gazing‹ signified a sad loss, the loss of a ›true‹, continuous experience. The American history professor Irwin Abrams, for example, complained: »We don’t want them [our traveling students] star-gazing a la Baedeker. We want them to use their own eyes and behold beauty on their own, to feel a thrill in coming upon the unexpected« (quoted in Gebhard 1965).

What these critics lament is how the star as »judgement device« (Karpik 2010) affects the traveler: She becomes blind, now able to see only what she is supposed to see, only those items that have been marked as outstanding. Thus the stars are criticized not simply for what they point at and what they neglect. Rather, the critics address a more fundamental point: The very translation of ›reality‹ into a ranked list isolates, preselects, and values travel experiences. One of the most famous critics of such ranked lists was Daniel Boorstin, who in the 1960s developed an influential theory of »pseudo-events,« or events that were only staged for mass consumption. (cf. Boorstin 1961) To him, tourist attractions were prime examples of such inauthentic pseudo-events, representing a discontinuous list-like expectation: »Modern tourist guidebooks have helped raise tourist expectations. And they have provided the natives […] with a detailed and itemized list of what is expected of them and when« (Boorstin 1961, p. 104). Thus lists as valuation devices seemed to hold a special power over the traveler. However, these critiques neglect the navigational freedom that the list-based organized knowledge affords. Long before the advent of digital media, the cultural effect of such lists was to digitize traveling—to produce singular items with no necessary connection between them. The tourist not only uses a list, but also starts seeing like a list, losing his autonomous faculty of judgment. Even more importantly, he is no longer, so the criticism continues, able to have a continuous, analogous experience, immersing himself in a different culture. Rather, he hurries from one two-star sight to the next, under time pressure, able to collect only the most important sights. To Boorstin the Baedeker asterisks have created »the uneasy, half-cultivated modern tourist […]. Anyone who has toured with Baedeker knows the complacent feeling of having checked off all the starred attractions in any given place, or the frustration of having gone to great trouble and expense to see a sight only to discover afterward that it had not even rated a single asterisk« (Boorstin 1961, p. 106). Despite Boorstin’s devaluation of list-based knowledge, he points at an important dimension of the use of lists: List-making is affectively structured; generating and working with lists may provide pleasure, but lists may also generate the anxiety of never being complete, of missing out.

The experience of the traveler, therefore, replicates the isolation of entities that the list has already performed. The list disembeds sights from their spatial, cultural, and historical contexts. This isolation of items is the very prerequisite for ordering them anew, be it on the basis of individual travel plans or on the basis of a—very often opaque—measure of valuation. The list, then, is a tool that helps the tourist to navigate and to plan his journey economically, but it also transforms the traveling subject by reorganizing that subject’s experience. Moreover, the ranked list creates a new desire, one that the itinerary and the alphabetical list did not evoke, or only marginally evoked: the desire to see the best, to collect the highlights. The list’s highly ranked items have acquired a heightened visibility and desirability, creating FOMO (fear of missing out) well before our digital culture. Even though I concur with the analysis that traveling by lists punctuates logistically the travel experience, I do not embrace the cultural criticism of early and contemporary critiques of lists. Such critiques presuppose that a proper touristic experience is a narrative experience, and thus regard more impressionistic modes of experience as inferior.Footnote 9

The starred valuation promises to improve the selection of sights, even though this might create a quite ruptured itinerary, connecting highlights that might be far apart from each other. Temples, hotels, mountains, or cities all become comparable by being translated into (non)starred items. By breaking with what sometimes has been called the ›democratic‹ character of the guidebook (cf. Sears 1989 quoted in Koshar 2000, p. 36)—which has always been a somewhat misleading idea, considering the choices any guidebook has to make—ranked classes of distinct travel experience are created. Now the traveler might decide to go to places where a particularly high number of two-star experiences may be found: an avant-garde restaurant, a luxury hotel, and a UNESCO-protected city.Footnote 10

To summarize the argument so far: With the advent of guidebooks, the function of travel lists changes. The traveler no longer uses ›empty‹ lists such as Boyle’s questionnaires, but ready-made, full lists. The history of the Baedeker illustrates how guidebooks assemble different types of lists—and how this composition has changed historically. Most important is the shift from topographical lists, grounded in the infrastructures of traveling, to nontopographical lists based on formal sorting criteria or on valuation. These lists open up a new virtual space of calculation—that is, a space where the isolated, discontinuous entities of the list can be manipulated independently of their geographical location. I have distinguished two different types of such nontopographical lists. Alphabetically organized lists are optimized for navigating and searching information. Valuation lists—such as asterisks, stars, or ›Best Of‹ lists—turn the list into a judgment device. These valuations profit from lists’ facticity; the very structure of lists makes it unnecessary to include the reason an item is on the list, the reason an item has been chosen as especially noteworthy. Baedeker was right to call the valuation markers such as asterisks the remaining »lyrical moment« of the list. The asterisk combines the ›objectivity‹ of a formal valuation system with the ›subjectivity‹ of cultural and aesthetic judgments. Its strength lies precisely in the fact that the asterisks speak for themselves.

These three types of lists follow different epistemological principles: reference in topographical space, formal ordering, and valuation. However, they are not neatly separated; rather, they coexist within the space of the same travel guide. They are interlinked, so that an asterisk may be included into an itinerary. This interlinking of lists is especially interesting, since it allows the articulation of the different ordering principles. In this way, the valuation, which potentially violates the impersonality of the travel guide, can become re-naturalized, for example, by being part of a well-grounded itinerary.

3 Social Media Travel Lists

Let me now briefly turn to user-generated recommendation systems such as TripAdvisor lists. TripAdvisor generates lists of hotels, restaurants, and sights for different locations, but it also offers ›Best Of‹ lists that are not restricted to a particular place (such as Best Beaches of the World and other World Travel Awards). These lists find a new solution to what Baedeker called the »lyrical moment« in his travel guides—that is, the recommendation of travel attractions. Sometimes Baedeker’s writers and editorial teams have struggled with the criteria for their recommendations. It was necessary to strike a balance with the contemporary aesthetic consensus and the attractions’ popularity. These struggles were usually hidden from the user of the book, who ideally accepted the stars without question, choosing not to dispute the guidebook’s ›impersonality‹.

Online recommendation systems such as TripAdvisor tackle this problem by separating the subjective and objective dimension of valuation. Users submit their review to TripAdvisor; this review consists of a written part and a formal valuation. The written part often includes personal experiences, even small narratives, and often also a statement legitimizing the review (e.g., »I am well traveled, but I have never been at such a bad 5‑star hotel before«). This discursive and narrative part is independent from the formal valuation and does not directly affect the overall ranking. The formal valuation is based on a pregiven classification system, using the criteria of service, cleanliness, and accommodation, as well as an overall valuation (each with 5 possible stars). The format of the review separates the subjective narrative account from the objective valuation. To the user, the original connection of these two parts becomes partially invisible. Looking at a review, we see the traveler’s overall rating, but not the more detailed rating in categories such as cleanliness. Thus, the problem of the »lyrical moment« is tackled by a separation between the review’s subjective and objective elements.

While the Baedeker still had to balance popularity and quality, TripAdvisor is based on the popularity of the ranked items. However, these lists are not generated by a simple mathematical averaging of travelers’ votes. Rather, the information is ordered by a secret algorithm that creates the popularity index (cf. Jeacle/Carter 2011, p. 199). This algorithm has often been discussed and criticized; many forums try to figure out how it works, since TripAdvisor provides only partial information about it. Three aspects seem to be crucial for how the popularity index works: the reviews’ quality (i.e., the formal ranking), quantity, and freshness.Footnote 11 One of the crucial effects of this logic is to produce as many reviews as possible. Consequently, hotels and restaurants have started to encourage their customers to write reviews—they even urge returning guests to write a new review. The former intransparency of the aesthetic values of an author of a guidebook has been replaced with the interplay between the transparency of popular votes and the opacity of the algorithm. Such recommendation systems have received a lot of attention in marketing and tourism research, mainly focusing on the question of how these systems generate trust. The perspective of this essay differs somewhat—although the question of trust is certainly important—since I am interested in the epistemic and spatial effects.

This interest made me focus on the question of how lists are organized and how different types of list may coexist. TripAdvisor lists have a different type of valuation and accounting than travel books, and the effects of this valuation are different as well (cf. Scott/Orlikowski 2012). The primary form of these lists is no longer a rating of two or three classes, but a numerical ranking of items. These rankings are much more volatile than the star system; it is even possible to personalize the ranking by including the traveler’s preferences. The ranked list of the classical travel guide multiplies itself in different personalized lists. The volatility and multiplication of rankings might be one reason that TripAdvisor in 2010 introduced the »Certificate of Excellence« as a more stable rating of items. These certificates require a consistently outstanding performance, thereby stabilizing the movements of the ever-changing rankings.

Looking back at the history of traveling lists, we can see that user-/algorithm-generated lists have profited from earlier travel lists. In a sense, these user-/algorithm-generated lists return in one crucial way to the mode of functioning of early travel lists: These lists are open to new information that the traveler has collected. The traveler is supposed to use these lists not only for her travel planning, but also to generate knowledge. Again she is confronted with categories that she is supposed to fill with her travel experiences. As with the colonial lists, the traveler does not know how her reviews will affect the global knowledge she generates; however, she is not confronted with an empty list, since she has access to the reviews of other travelers. Nevertheless, the list works again as a device that not only orders items, but actively collects and structures knowledge.

In highlighting the similarity between early modern and contemporary open lists I do not mean to neglect crucial differences. For example, though these lists depend on user-generated knowledge, these data become the material for algorithms, best exemplified by TripAdvisor’s popularity indexes. They are the material for the often opaque generation of a multitude of new travel lists based on ranking (e.g., ›Best Of‹ lists). In contrast to the early modern travel lists, these lists create value judgments, thereby drawing from tools of valuation that had already been developed in a rudimentary form: that is, rating tools such as the asterisk in modern guidebooks. However, it is no longer the often criticized authoritative voice of the guidebook authors, but an anonymous and ›democratic‹ multitude of standardized voices embedded in a socio-technical valuation assemblage. Elements of this assemblage include the ordering algorithms, the pregiven classificatory systems, the personalization of the lists, and the data of the travelers.

TripAdvisor lists thus exploit consistently the ordering potential that lists offer. One effect of this function is that these lists become ever more independent of the geographical space of traveling. This is best exemplified with the Traveler’s Choice lists, which consist of the best of the best of the world (although it is still possible to create such lists based on a particular region). Although the primary classificatory system still relies on geographical places, the logic of comparison that is inherent to rating lists seems to call for an ever-wider scope of comparisons.Footnote 12 Early modern lists, seeing the list as a tool for collecting items, attempted to generate a global archive, and now the world-making quality of lists has become ever more important. The travel list carries an implicit universalism that to some degree conflicts with artificially imposed limits such as certain regions. With the introduction of ranking mechanisms comes the desire for an all-encompassing comparison and ranking—a global ›Best Of‹ list which is always in flux.