1The success of a participatory cultural project depends largely on the vivacity and durability of its community of contributors. It is therefore in the interest of such projects’ initiators to invest seriously in the construction of this community, its renewal and the maintenance of its dynamics. This article examines the case of the Wikimedia Commons free media library, focusing on the Wiki Loves Monuments photography contests. In particular, it will attempt to determine whether these contests are effective and sustainable forms of community animation.
2We will first review the main characteristics of Wikimedia Commons and the forms of community animation that are being developed there. We will then present the main object of this article, the Wiki Loves Monuments contests, as well as the study research methodology that will be implemented. We will then describe the dynamics of participation in these contests, analyze their potential relations of cause and effect, and finally, reflect on the sustainability of community animation.
- 1 Andreas M. Kaplan and Michael Haenlein, “Users of the world, unite! The challenges and opportunitie (...)
3Since the early 2000s, the development of social web tools such as blogs, social networks and file-sharing web sites,1 has facilitated the active participation of Internet users by reducing the technical barriers to the publication of content. Today, these devices are used in a wide variety of ways, especially by communities involved in the creation, sharing and promotion of cultural content.
- 2 Top 500 sites on the web, on the website Alexa: [Online] https://www.alexa.com/topsites [accessed 5 (...)
- 3 Frédéric Kaplan and Nicolas Nova, Le Miracle Wikipédia, Lausanne, Presses polytechniques et univers (...)
- 4 Andreas M Kaplan and Michael Haenlein, “Users of the world, unite! The challenges and opportunities (...)
- 5 Dominique Cardon and Julien Levrel, “La vigilance participative. Une interprétation de la gouvernan (...)
- 6 Pierre-Carl Langlais, “{{Référence nécessaire}} L’émergence d’une norme wikipedienne,” in Barbe, Li (...)
4The free encyclopedia Wikipedia, founded in 2001, is one of the most consulted websites in the world today.2 Its articles, often at the top of search engine results, also feed many other web services.3 However, they are the product of a large-scale collaborative project4 that is resolutely open: anyone can create, complete and modify their content in real time thanks to the MediaWiki software. The quality of the encyclopedia, which is generally comparable to that of traditional encyclopedias, is due to the “participatory vigilance” of the community,5 based on a solid body of rules and guidelines.6
5Wikipedia receives technical, human and financial support from the Wikimedia Foundation, founded in 2003 to “make it easier for everyone to share what they know.”7 The Foundation supports several related projects, such as the Wikidata knowledge base, the Wiktionary multilingual dictionary, the Wikisource free library… which together form the Wikimedia movement. Like the Wikipedia encyclopedia, these projects are collaborative, and their productions are made available under the terms of free licensing agreements, allowing their redistribution and reuse, but preventing any exclusive commercial appropriation. As such, these contents are to be considered as Commons , that is “a resource over which any member of a given community has rights, without having to obtain permission from anyone.”8
6Among the many projects of the Wikimedia movement, this article focuses on Wikimedia Commons, “a collection of […] freely usable media files to which anyone can contribute.”9 Like Wikipedia, Wikimedia Commons relies on the participation of Internet users, which is expressed through the wiki technical infrastructure and is based on a set of rules accepted by all contributors.10 To date, the media library gathers more than 60 million files, accompanied by numerous metadata, under various free licenses.
7Contributors can engage in a wide variety of activities: “Contribute your work” (photographs, animations, illustrations, but also audio or video recordings), “Contribute your know-how” (page translation, image enhancement, etc.) and “Contribute your time” (annotate, verify, classify, select images).11 Unlike Wikipedia, which allows contributions from unregistered users, all forms of participation in Wikimedia Commons require a user account. We know for sure that more than 8 million people have contributed in one way or another to the media library.
- 12 Jenny Preece, Diane Maloney-Krichmar and Chadia Abras, “History of online communities,” Encyclopedi (...)
- 13 Imhoff, Camille, “L’animation de communauté sur le réseau social d’entreprise: injonction à la coll (...)
- 14 Lorna Heaton, Florence Millerand and Serge Proulx, “Émergence d’une communauté épistémique: créatio (...)
8Wikimedia projects such as Wikimedia Commons require the building and maintenance of a connected community, that is, “a group of people who interact in a virtual environment. They have a purpose, rely on technology, and are guided by norms and rules.”12 Discourses praising the spontaneous self-organization of connected communities often invisibilize the work of community managers or animators, in charge of “developing communication and collaboration, federating, encouraging, implicating, but also organizing and capitalizing on the knowledge produced.”13 These actors do not necessarily engage in direct contribution activities, but rather focus on supporting those of other community members: “It is then a matter of ‘cultivating’ the community, in particular by ensuring activities of animation and maintenance of the community (by organizing workshops or field trips, for example).”14
9The Wikipedia encyclopedia has defined many functional roles, giving specific responsibilities and powers to (often elected) users.15 However, these roles mainly concern the management of vandalism, the resolution of conflicting edits, and the development of tools and rules. Community animation, in particular through the organization of edit-a-thons,16 remains rather the prerogative of user groups affiliated with local Wikimedia chapters17 (such as the Wikimedia France association), cultural institutions (museums, archives, libraries), or various associations more or less dedicated to sharing of knowledge.
10This same functional division can be found on Wikimedia Commons: specific roles are assigned to certain designated contributors to ensure administration18 and surveillance19 tasks, while the animation of the community emanates from the various user groups organizing activities and events.
11The Wikimedia Commons library hosts a wide range of media, from video to pictograms to sound recordings, but photographic image occupies a key place. Photography contests are therefore a very popular activity in the community animation. This article will focus on a series of annual contests, entitled Wiki Loves Monuments, which stand out for their scope and their history precedence.
12In June 2009, a first photography contest of this kind was organized by Wikimedia Nederland (the Dutch chapter of Wikimedia), under the title Wiki Loves Arts.20 Participants were invited to take pictures of works of art held by some 40 partner museums and to upload their pictures (through a Flickr group) to Wikimedia Commons. Given the success of this contest (more than 4,000 images uploaded21), Wikimedia Nederland renewed the experience in September 2010, on the occasion of the European Heritage Days, with Wiki Loves Monuments (WLM). The challenge, which this time focuses on all national monuments recognized by the state (Rijksmonuments), has resulted in the publication of more than 12,000 photographs.22 Prizes were granted to the ten best pictures and to the most prolific contributors.
13In 2011, the contest was renewed, taking on a European scale with 18 participating countries,23 then an international scale in September 2012, with 34 countries involved. Wiki Loves Monuments thus becomes a recurrent contest, held every year without exception (until now), in September (in most countries, although some preferred October or November), with a variable but always high number of participating countries. In each country, a national jury makes a first selection of photos, out of which an international jury then awards prizes worth up to 1,500 euros.24 The contest, with a dedicated website (wikilovesmonuments.org, as well as several national sites), a logo and a graphic charter, receives the support of international (Unite4Heritage, Europa Nostra) and national (in France, the association Arkeotopia and the software publishing company Code Lutin) organizations.25
14The study of Wikimedia projects and communities is pleasantly facilitated by the values of openness and transparency that animate them. A wide variety of data and statistics are made available on the websites of these projects. Moreover, the adoption of the MediaWiki software, not only for the Wikipedia encyclopedia but also for most of the connected projects (Wikimedia Commons, Wiktionary, Wikispecies, Wikiquote … as well as a Meta-Wiki related to the Wikimedia movement as a whole) guarantees the preservation of previous information (through a revision history). This wiki system also has programming interfaces (APIs) that allow the automatic extraction and aggregation of information using computer programs.
15This relative transparency, combined with the longevity of the Wikimedia movement (which celebrated its 20th anniversary in January 2021) and the WLM competition (whose tenth edition was held in September 2020), make it a remarkably rich field of study. In this article, all of these sources of information will be used to try to depict the temporal dynamics of the Wiki Loves Monuments photo contests. In particular, we will try to determine if this type of recurring contest is an effective and perennial form of community animation. For practical reasons, some analyses will not be done on all contributions, but only on the French section of the contest. Indications will then be added to inform the generalizations that can be drawn from them.
16The study of the dynamics of participation requires the consideration of several variables: the number of images uploaded within the framework of the contest, the number of people who created these images, but also the place that these images and these people have subsequently found in the Wikimedia galaxy. These analyses must also take into account the international dimension of the contest, of which the global dynamics result from a sum of sometimes contradictory national dynamics.
17The number of images submitted to the competition increased rapidly during the first years of the contest (table 1). This growth can be directly linked the increasing number of countries associated with the competition, multiplicating both the number of monuments that can be photographed and the number of potential photographers. Participation peaked in 2013, with more than 360,000 images published, then decreased slightly before stabilising in 2015 in a sort of plateau. This plateau is nevertheless characterised by significant annual variations, with little correlation, in the number of associated countries, published images and participating individuals.
Table 1
Year
|
Countries
|
Images
|
Participants
|
2010
|
1
|
11,985
|
199
|
2011
|
18
|
156,484
|
5,497
|
2012
|
34
|
351,665
|
15,188
|
2013
|
53
|
362,839
|
12,379
|
2014
|
41
|
304,382
|
9,192
|
2015
|
33
|
225,214
|
6,740
|
2016
|
42
|
273,557
|
10,844
|
2017
|
52
|
244,728
|
9,953
|
2018
|
56
|
257,370
|
13,940
|
2019
|
48
|
212,598
|
7,240
|
2020
|
51
|
230,425
|
7,709
|
Number of countries, contributions and participants for each edition of the Wiki Loves Monuments contest.
Source: https://wikiloves.toolforge.org/
18This plateau and its variations mask diverse participation dynamics specific to the various participating countries (fig. 1). Most are characterised by one or more peaks in the number of contributions, followed by a more or less sharp decline. Some, such as the Netherlands, did not participate in all editions of the competition, resulting in blank years. Countries that joined the initiative later on, such as Azerbaijan or Brazil, often show a later peak in participation than the others. These peaks tend to offset the marked erosion seen in the majority of countries that began participating early.
Figure 1
Evolution of the number of contributions to the WLM competition in different countries.
Source: https://commons.wikimedia.org/wiki/Commons:Wiki_Loves_Monuments
19Thus, while overall participation is relatively stable, participation at the national level seems to systematically weaken after a more or less long period of high activity. As long as new countries continue to join the competition, their initial drive may compensate for the slow decline of others, maintaining the previously observed participation plateau. But it seems that unless this trend is reversed, WLM is inexorably bound to lose momentum.
- 26 Sylvain Machefert and Benoît Evellin, “Wiki Loves Monuments. Un concours pour constituer une banque (...)
20If the Wiki Loves Monuments contests are primarily intended to enrich the image repository, they are also an important tool to promote the community and to recruit new contributors: “through the placement of banners on Wikipedia during the month of September, the visibility of the contest led to nearly 1,400 new contributors to Wikimedia Commons in three years (for the French edition).”26
21The API data tends to confirm the important (even preponderant) role of new subscribers in the contest. By crosschecking users' registration dates with their participation in Wiki Loves Monuments, we can count the people who registered in the same month as their first participation in the contest. For the French edition, several hundred people have probably joined the community each year through the contest (fig. 2). These figures confirm the significant erosion of participation after several very active years. However, the share of new subscribers remains roughly the same: despite this loss of momentum, the contest remains an effective tool to attract new contributors.
Figure 2
Evolution of the number of participants in the French edition of the contest, registered at the time of their participation or before.
22However, the registration of new contributors is not the only metric to take into account in order to assess the development of the community. Analysis of these new subscribers’ number of contributions allows us to determine the extent to which they got involved in Wikimedia Commons. It appears that a large majority of users have only made a limited number of contributions, probably without getting involved beyond a single edition of the contest. However, a small number of users who registered through the contest have made more than 50 contributions (fig. 3), which might indicate a significant involvement in the community. Despite the decline in the number of participants, this proportion does not change much, suggesting that Wiki Loves Monuments is still able to attract active contributors. Interviews with these long-term contributors (or with those who dropped out after a first contest) would allow us to better understand the factors that determine such careers.
Figure 3
Evolution of the number of contributions of subscribers who joined Wikimedia Commons through the WLM contest, sorted by year of registration.
- 27 Jorge Oceja and Ángel Obregón Sierra, “Gamifiying Wikipedia?,” in ECGBL 2018 - Proceedings of the 1 (...)
- 28 Sylvain Machefert and Benoît Evellin, “Wiki Loves Monuments. Un concours pour constituer une banque (...)
23The simplicity of the gamification strategies of Wiki Loves Monuments has already been identified as a potential limit to the participants’ engagement.27 However, the loss of momentum of the French edition of the contest and, probably, of all editions after a certain number of years of existence, could also be explained by the progressive saturation of the contest’s theme. The number of listed monuments in a given country is limited and, once a monument is widely documented in the Wikimedia Commons media library, the motivation to contribute may decrease and the production of original images becomes more complex. This hypothesis was already mentioned after only a few editions of the contest: “we tend towards exhaustiveness concerning the accessible monuments, which makes it difficult to conceive the perpetuation of the contest.”28 The analysis of recent data seems to confirm it.
- 29 Cultural Heritage Monuments in France with known Ids, on Wikimedia Commons: [Online] https://common (...)
- 30 Since 2019 or 2020, the rules of the French edition of the contest include, in addition to the Méri (...)
24Wikimedia Commons has a category, based on the national database Mérimée, that lists all the classified historical monuments in France.29 Of the 45,025 monuments listed, only 598 (or 1.3%) have no illustration at all. The notion of “accessible monument”, which implies that some monuments that are difficult to access cannot be photographed by amateurs, shows that a complete coverage of historical monuments would be difficult to achieve. Thus, with more than 98% of the monuments illustrated, the hypothesis of the saturation of the theme seems plausible.30 The distribution of the number of images ( whether or not resulting from the WLM competition) shows a “long tail”: while the majority of monuments are documented with only a few images (the median value is five images per monument), some of them have more than 700 photographs (fig. 4). Thus, the wide coverage of French monuments is not equal, with some monuments attracting much more attention from photographers than others.
Figure 4
Distribution of the number of images associated with each French historical monument.
25The data on the reuse of images produced in the WLM contest also provides interesting insights. Through the Wikimedia Commons APIs, it is possible to count, for each image, the number of times it has been reused on all Wikimedia projects (the Wikipedia encyclopedia in all languages, but also the other wikis of the movement). The reuse numbers of the French contest images also show a “long tail” distribution: while some images have been reused several hundred times, an overwhelming majority has never been reused, and this trend seems to intensify in recent editions of the contest (fig. 5). The time factor plays an obvious role: the more recent an image is, the less it has been seen and therefore the less it had the opportunity to be reused.
Figure 5
Distribution of the number of reuses of photographs produced during the different French editions of the contest.
26It should be noted that the quality of the images is not the cause of this decrease in the reuse of images: the rate of images identified by the community as “quality images”31 tends to increase over the editions of the contest (table 2). The saturation effect still remains: when the article about a monument is already illustrated by several suitable photographs, the new images produced are less likely to be used, even if they are of better quality. A detailed exploration of the dynamics of illustration substitutions in Wikipedia articles could confirm this assumption.
Table 2
|
2011
|
2012
|
2013
|
2014
|
2015
|
2016
|
2017
|
2018
|
2019
|
Uploaded images
|
24,838
|
27,048
|
21,135
|
17,839
|
11,903
|
7,760
|
9,512
|
9,566
|
8,297
|
Quality images
|
118
|
366
|
182
|
65
|
66
|
15
|
279
|
126
|
120
|
Percentage of quality images
|
0,5
|
1,4
|
0,9
|
0,4
|
0,6
|
0,2
|
2,9
|
1,3
|
1,4
|
Total of reuses
|
33,909
|
33,854
|
18,159
|
12,920
|
5,327
|
2,597
|
2,613
|
2,124
|
1,904
|
Average reuses
|
1,4
|
1,3
|
0,9
|
0,7
|
0,4
|
0,3
|
0,3
|
0,2
|
0,2
|
Evolution of the share of “quality images” and the number of reuses for the photographs produced in the different French editions of the WLM competition.
27The analyses presented in the previous section seem to indicate that the organisation of a recurrent photographic contest constitutes an effective method of animating the Wikimedia Commons community, although the effects of this activity tend to diminish with time. The study of this community also shows its great creativity in order to overcome these limitations, with a multiplication of the animation mechanisms to renew its dynamic.
28In the years following the creation of Wiki Loves Monuments we can notice the emergence of several other contests designed on the same model,32 and generally named similarly: Wiki Loves Earth (launched in 2013 focusing on natural heritage), Wiki Loves Africa (created in 2014), but also the recent Wiki Loves Folklore (initially named Wiki Loves Love in 2019), or Wiki Loves Folk (organised since 2016 without succeeding in developing beyond Spanish borders). Participation statistics of these contests seem to indicate trajectories similar to those of Wiki Loves Monuments, characterised by an initial rapid progression, driven by an increase in the number of participating countries, followed by a plateau or slow decline when the number of countries reaches a static threshold. As the themes of these new competitions—which are not limited by finite lists, contrary to classified historic monuments—are less likely to reach a saturation point, it is therefore not to be considered the only obstacle to the sustainability of recurrent photographic competitions. Further work would be necessary to complete the analysis of the causes of this phenomenon.
Table 3
|
Wiki Loves Earth
|
Wiki Loves Africa
|
|
Images
|
Participants
|
Countries
|
Images
|
Participants
|
Countries
|
2013
|
9,695
|
346
|
1
|
|
|
|
2014
|
19,470
|
2,370
|
14
|
5,868
|
873
|
47
|
2015
|
104,201
|
8,837
|
25
|
7,352
|
722
|
48
|
2016
|
109,936
|
13,187
|
24
|
7,768
|
836
|
49
|
2017
|
130,482
|
15,074
|
36
|
17,874
|
2,435
|
55
|
2018
|
89,683
|
7,645
|
32
|
-
|
-
|
-
|
2019
|
94,699
|
9,699
|
37
|
8,212
|
1,350
|
53
|
2020
|
106,020
|
9,095
|
34
|
16,982
|
1,904
|
53
|
Evolution of participation in the Wiki Loves Earth and Wiki Loves Africa contests.
29Although none of these contests has apparently found a solution to the gradual decline in momentum, their multiplication is in itself an interesting response to this obstacle. Indeed, even if each contest has only a limited period of intense participation (a few years, nonetheless), the succession of contests on different themes seems to ensure a regular renewal of the community’s dynamics. Since December, monthly contests have also taken off,33 each time focusing on narrower themes (for example: bridges in February 2021, umbrellas in January 2021). There are also initiatives to promote using in Wikipedia photographs produced during the contests, much as the Monuments Challenge organised by Wikimedia France in November 2020. Such events, also characterised by precise rules of participation and the attribution of rewards to the most active contributors, contribute in a complementary way to the mobilisation of the community and the promotion of its productions.
30This article has focused on describing the dynamics of the Wiki Loves Monuments contests, organised in relation to the Wikimedia Commons media library. Since 2011, these annual contests dedicated to photographing listed historical monuments have been widely successful around the world. This article aimed to determine whether such contests are an effective and sustainable form of community animation. It was based on the analysis of different data from the contest pages but also from the APIs of the wiki system.
31The results show that Wiki Loves Monuments generated a large number of quality contributions, some of which have been widely reused in other projects of the Wikimedia movement. It also involved many participants across dozens of countries, attracting many new contributors, some of whom continued to contribute after the contest ended. However, the study of the temporal dynamics shows a progressive erosion of participation which seems inevitable, although it is still partially compensated by the last countries that joined the contest. Several clues suggest that the progressive saturation of the contest’s topic could contribute to this decrease in participation, without this being the only factor. But the Wikimedia Commons community seems to have found a solution, through the regular organisation of new contests on various topics, following similar methods. If no contest can be considered as a truly permanent form of community animation, this continuous process of renewal and diversification could be the key.
32The Wikimedia movement is a very rich scientific object, whose exploration is facilitated by the values of transparency that animate it and the technical characteristics of the tools on which it is based. Above all, the multiplicity and diversity of the communities it gathers, as well as their long duration of existence, constitute an extraordinary field for studying cultural participation dynamics. Wikimedia is like a huge public laboratory, in which people from all over the world have been experimenting for twenty years with new ways of working collectively to produce and share knowledge. Methods that produce positive results are retained and refined by the community, while others are abandoned or transformed. Maintaining the dynamics of participation over such a long period of time demonstrates the effectiveness of this process, as well as the value of the lessons that can be learned, both by scholars who study these movements and by cultural institutions that would like to follow their lead.