Version classiqueVersion mobile

Studies on Arabic Dialectology and Sociolinguistics

Catherine Miller
Alexandrine Barontini
Marie-Aimée Germanos
et al.

Issues in Sociolinguistics: Contact and Change, Discourse Analysis, Cultural Practises, Mixed Styles and Written Sources

Negotiating Authenticity in Casual Speech: Moroccan Rappers and the /t/ Variable

Sarah Ruth Schwartz


This study undertakes an acoustic analysis of the casual speech of 4 Moroccan rappers in order to: 1) acoustically describe the variation in realizations of the phoneme /t/ in two different dialects of Moroccan Arabic, 2) evaluate the extent to which Moroccan Hip Hop artists do or do not shift to the variant used in the city of Casablanca, 3) understand the role that linguistic style-shifting plays in the performance of authenticity, or “keepin it real”, in the Moroccan Hip Hop community. The study definitively shows regional dialectal variation in realization of the phoneme, but the sociolinguistic implications of this variation were less definitive. While one Meknassi artist does claim that he tries to use Cassaoui speech, this was not borne out by acoustic measurements, suggesting that there may be other, more salient acoustic features that constitute linguistic capital within the Moroccan Hip Hop community. The results of the study also suggest an interesting divide in strategies for accessing power and artistic authenticity within the community – while the aforementioned speaker did affiliate the Cassaoui dialect with the code of the streets, and attempted to adjust his speech thusly, the other Meknassi speaker explicitly preferred to use his native pronunciation in order to remain authentically Meknassi. This reflects the inherent tensions in “keepin it real”, with the first speaker focusing on the authenticity inherent in repping the larger Hip Hop community in Morocco, and the second speaker focusing on the authenticity inherent in repping his hometown.

Texte intégral

1. Introduction

1The following is a sociophonetic analysis of variation in affricated realizations of the voiceless coronal stop /t/ in two dialects of Moroccan Arabic – the coastal Atlantic dialect that stretches from the city of Salé to the city of El-Jadida and realizes /t/ as the palato-alveolar affricate [tš], and the dialect in the central region encompassing Fès, Meknès, and Taza that realizes /t/ as the alveolar affricate [ts] – that seeks to connect pronunciation of this feature to the exchange of symbolic power within the community of Moroccan rappers, as well as the issue of authenticity and “keeping it real” as a Hip Hop artist. This study will answer the questions: Is affricated /t/ a linguistic marker within Moroccan Hip Hop community? Do Moroccan Hip Hop artists strategically use this feature to help position themselves within the Moroccan Hip Hop community? And finally, what implications does this have for the meaning of authenticity within the Moroccan Hip Hop community?

2The paper presents a case study of four Moroccan rappers: two from the coastal region (from the cities of Salé and Casablanca), and two from the city of Meknès. Casablanca and Salé host two of the most prominent rap scenes in the country, and Meknès hosts a rap scene that is prominent, but less prominent than the Cassaoui rap scene. Thus, we may assume that the street speech of Casablanca and Salé hold prestige within the Moroccan Hip Hop community, and we might expect the Meknassi respondents to accommodate to the Cassaoui pronunciation of /t/ by realizing the phoneme as [tš], rather than their native [ts]. However, the results suggest that the relationship between regional dialect and the Hip Hop community is more complicated than this. While the analysis definitively shows regional variation in the affricated realization of /t/, it does not definitively show accommodation to the Cassaoui dialect: one Meknassi speaker claimed to try to use Cassaoui features, but on average produced /t/ in a manner consistent with the Meknassi dialect, and the other Meknassi speaker intentionally refused to use features of the Cassaoui dialect in order to better represent his local heritage. This result indicates two main findings: first, that variation in /t/ affrication may not be a salient linguistic capital in this community, and second, that accessing authenticity in the Moroccan Hip Hop community reflects an inherent tension between accessing the prestige of street code and repping one’s hometown.

2. Background

2.1. Language, Style, and the Global Hip Hop Nation

3The connection between street language and Hip Hop culture is critical – Hip Hop is an essentially urban art form, and the language of the streets is the language of the art. Alim states in his (2003) study of the casual speech and performance styles of the rappers Eve and Juvenile that rappers must cultivate a “street-conscious identity” in their lyrics in order to create a sense of familiarity between themselves and the urban African American community. He compares rates of copula deletion, the “showcase variable” of AAE (2003: 46), between the respective artists’ casual speech and their lyrics, and finds significant differences in the rate of copula deletion between both artists’ respective casual speech and performance speech. Alim posits that this represents “street-conscious copula deletion”: a conscious shift away from Standard American English (SAE) in order to express kinship with the African American community. Maintaining a street-conscious identity, he concluded, is crucial to rappers’ success within the Hip Hop community, because it was a sign that they could authentically identify with Hip Hop’s main audience.

4The concept of authenticity, or “keepin it real”, is a significant feature of Hip Hop culture. However, in the context of the Global Hip Hop Nation (GHHN) “authenticity” becomes a complex matter – while rappers across the globe are still expected to pay homage to Hip Hop’s origins in the US Black community, a rapper from Korea, for example, could not claim authenticity if they only repped US Hip Hop culture and not their own culture. As such, on one level maintaining authenticity in Hip Hop manifests as a dialectic between the global and the local, or the “glocal”, to use a term common in Hip Hop scholarship (Androutsopoulos 2009: 44). Furthermore, as Pennycook (2007: 103) explains, neither can authenticity be limited to “an individualist obsession with the self”. Instead, authenticity to oneself requires a “dialogical engagement with community” – balancing repping oneself and repping one’s community.

5As one may expect, this process is highly variable – the balance between the global and the local, and the individual and the communal, that rappers must attain in order to access authenticity differs widely between Hip Hop communities. In the Hip Hop literature, the process of negotiating these tensions and defining authenticity is called “localization” (Pennycook 2007). For example, Hassa (2010) describes the use of code-switching within French rap as a form of localization – the Arab-French rappers she studies use Arabic terms in order to represent their home cultures and the French streets, evoking the connective marginality between Arab immigrants in France and African Americans in the US (Osumare 2007 cited in Terkourafi 2010a: 3), and use Verlan in order to specifically represent their local southern French identity. In Morocco, as I will explain below, localization involves both language and subject matter, resulting in a highly diverse Hip Hop community.

2.2. Setting the Scene: The Moroccan City and Hip Hop Culture

  • 1 Three Hip Hop scenes, to be precise, but because of the demographic and cultural similarities betwe (...)

6This study centers on speakers from two dialect regions in Morocco representing two prominent Hip Hop scenes1. The first area I will consider is the dialect region between the western coastal cities of Casablanca and Salé. Both Casablanca and Salé represent a type of new urbanism that is reflective of the urban culture to which global Hip Hop culture pays homage. Since the rapid urbanization of 1970’s, Casablanca has become the economic powerhouse of the country, with the economic expansion serving wealthy capital owners while creating the conditions for the type of poverty and economic depression amongst the lower classes that mirror the conditions under which Hip Hop was born in the Bronx in the 1970’s and 1980’s (Miller 2007). Salé underwent a similar process of rapid urbanization in the 1970’s (Triki 1991). Unsurprisingly, Casablanca and Salé are home to two of the oldest Hip Hop scenes in Morocco.

7Meknès, on the other hand, is both a relatively smaller city and an imperial city, with deep ties to traditional Moroccan culture; one of the city’s main claims to fame is the ʿIssawa Sufi brotherhood, which is also the subject of Meknassi rap crew H-Kayne’s most famous track, “Issawa Style”. H-Kayne produces taqlīdī (‘traditional’) rap: rap about traditional Moroccan values (Almeida 2015). Here we see localization at work – whereas the music of Hip Hop artists in Casablanca is typically more reflective of broader themes found in Hip Hop throughout the GHHN, taqlīdī rap consists of a call to respect the dominant Muslim nationalist culture. While Meknès has a history of being a center for Hip Hop culture in Morocco, it hosts a less prominent rap scene than larger urban centers, with rappers and producers often leaving the city in order to find more opportunities in Casablanca.

2.3. Affrication of /t/

8The two regions described in this study also differ with respect to dialect. This study focuses in particular in variation in the production of the phoneme /t/, with the two regions varying in place of articulation of the frication period. In Meknès (as well as in neighboring Fès) /t/ is typically realized as the alveolar affricate [ts], and throughout the Atlantic coast, from Salé to as far south as El-Jadida, /t/ is typically realized as the palato-alveolar affricate [tš] (Caubet, personal communication). According to Naciri (2014), affrication may be blocked by a preceding sibilant or following sonorant consonant, but the affricated realizations of the phoneme are otherwise universal in these dialects. In general, these two features are not well described in the literature on Moroccan dialectology. Furthermore, no prior study has described these variants acoustically: this study constitutes the first acoustic analysis of affricated /t/ in Moroccan dialects.

9Affrication of /t/ is well described in the sociolinguistic literature on urban dialects of French. Several works cite the affricated realization of /t/ as being a prominent feature of cité or beur speech, or the French equivalent of what Alim (2004) describes as “the code of the streets” – in French society the affricated realization of /t/ is generally stigmatized, but has covert prestige largely because of the connection between street speech and Hip Hop culture, similar to AAE in the United States (Gasquet-Cyrus 2013, Jamin, Trimaille, and Gasquet-Cyrus 2006, Jamin 2004, Gadet and Hambye 2014).

3. Data and Methodology

3.1. Data Collection

10The data included in the study was collected over a 2-month period of fieldwork in Morocco in 2016. I collected 14 sociolinguistic interviews in total (7 with slam poets and 7 with rappers), and the current study includes data from 4 interviews. The 4 informants were chosen and labeled in this study based on their city of origin – informants M1 and M2 are from the city of Meknès, and informants C and S are from the cities of Casablanca and Salé, respectively. All of them are male rappers between the ages of 22 and 27 at the time of interview, and educated at the university level. Informants M1, S, and C have spent their entire lives primarily living in the cities of their birth, with some travel, and M2 moved to Casablanca from Meknès in 2009 for university and has lived there ever since.

11All interviews were recorded with a Tascam DR-100mkIII Linear PCM Recorder and a Shure A-10 head mounted microphone, sampled at a rate of 44.1 kHz. The elicitation involved a variation on the traditional variationist sociolinguistic model following Labov (1984) with a semi-structured interview intended to elicit casual speech, reading speech (from a word list), and a performance register. The semi-structured interview was adapted from the traditional Labovian interview to include questions about the informants’ experience as artists, their political opinions, and metalinguistic questions about their perception of dialect variation in Morocco. This study focuses on data from the semi-structured interviews.

3.2. Data Analysis

12All data was annotated and measured in Praat 6.0.14 (Boersma and Weenik 2016). The primary acoustic metric used in this study was center of gravity (henceforth CoG). Given that the primary goal is to differentiate between tokens of [ts] and tokens of [tš] in the data, I focused on the CoG of the frication duration – that is, the alveolar sibilant [s] versus the palato-alveolar sibilant [š]. Measurements were taken using a Praat script written by Wendy Elvira-Garcia (2015) that measures the average CoG over the entire frication period. Following Kolgjini’s (2004) thesis on affrication in Albanian, each token of /t/ was separated into the closure period and the frication period, with the frication period measured from the onset of aperiodic high-amplitude frication noise until the offset thereof. Figure 1 below shows this coding system, where “f” on the first and second tiers of the text grid indicates the frication period following the stop closure. Tokens of /t/ following a sibilant or preceding a sonorant consonant were coded but excluded from analysis, because these phonological environments block the affrication of /t/ (Naciri 2014: 70-72).

Fig. 1 Example of coding system for /nta/ (“you”)

Fig. 1 Example of coding system for /nta/ (“you”)

13For statistical analysis of the data set, a linear mixed effects model using the lmer function of the lme4 package was fit with the continuous dependent variable of CoG, and presence of closure period (Closure) and city of origin (OriginMeknes and OriginCasa-Sale) as binary fixed effects, an interaction between closure and origin (Origin*Closure), and Speaker as a random effect. P values were obtained using the mixed function of the Matrix package. After determining the significant predictors of CoG in the mixed effects model, inter-speaker variation was tested with unpaired two-tailed t tests between each of the speakers.

3.3. Hypotheses

141. There will be a main effect for Origin.
2. While Origin will be a significant predictor of CoG, the Meknassi informant M1’s average CoG values will be closer to those for [š] than those for [s].

4. Results

15The study analyzed 2 hours of interview data, resulting in 673 tokens of affricated /t/. Table 1 describes the data, divided by speaker. Tables 2 and 3 give the results of the linear mixed effects model with a dependent variable of CoG. The model found main effects (p < .001) for both Closure and Origin, and no significant effect for the interaction between Closure and Origin. CoG values for OriginMeknès are estimated to be around 739Hz higher than CoG values for OriginCasa-Sale, and CoG values for ClosureYes are estimated to be around 393Hz higher than CoG values for ClosureNo. Thus, within the linear mixed effects model both Closure and Origin (but not the interaction between Closure and Origin) have been found to be reliable predictors of CoG, with Speaker as a significant source of variance. This suggests that the speakers’ pronunciation of /t/ most closely resembled that of other speakers with the same value for Origin.

Table 1: Total tokens of affricated /t





















Table 2: Fixed effects in linear mixed effects model

Fixed Effects

Estimate Std

t value

Std Error

Degrees of Freedom

p value






< 2e-16 ***






1.51e-08 ***






2.86e-05 ***







Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Table 3: Random effects in linear fixed effects model

Random Effects


Standard Deviation







16Having established that city of origin is a reliable predictor of CoG in this data set, which suggests that each informant for the most part produced CoG values consistent with the other informant native to his region, it would be pertinent to understand how close the CoG values were within the two groups of informants and between the groups, still keeping in mind that Speaker was not a fixed variable in the regression model. Figure 2 and Table 4 depict the average CoG values produced by each of the speakers in the data set. C, who has lived in Casablanca his entire life, grew up in the old city, and considers himself authentically Cassaoui, has the lowest median CoG value and the shortest interquartile range. M2, the Meknassi who moved to Casablanca in 2009, has the highest median CoG – higher than the Meknassi speaker (M1) who had never lived in Casablanca. Pair-wise comparisons confirm the results of the mixed effects model (see Table 5). As previously established, Origin was a significant predictor for CoG, which suggests that speakers with the same Origin value should not have a significant difference in their CoG values. This is the case between C and S and between M1 and M2. The t values for each pair further confirms this result – the t values for the M1-S and M1-C pairs are significantly lower than the t values for the M2-S and M2-C, respectively.

Fig. 2: Plot of CoG values per Speaker

Fig. 2: Plot of CoG values per Speaker

Table 4: Mean CoG values and Std Deviation per speaker








CoG (Hz)





Std. Deviation





Table 5: Unpaired t tests between speakers


t value

Std Error

Degrees of Freedom

p value (adjusted with bonferroni correction)















< 0.0001***





< 0.0001***





< 0.0001***





< 0.0001***

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

5. Discussion

5.1. Response to Metalinguistic Questions

17It is pertinent to discuss the informants’ self-evaluation, given that inter-speaker variation is a major focus of this case study and the speakers’ perceptions of their own speech is informative of their general position vis-à-vis the phoneme in question and phonetic variation as a marker of dialect in general. I will draw attention to two metalinguistic questions that informed this study:
1. How can you differentiate accents from around Morocco?
2. Have you ever changed the way you speak for any reason?

18When asked to explain different dialects in Morocco, none of the informants explicitly mentioned the variable examined in this study. Furthermore, none of the informants specified differing pronunciation of individual phonemes as significant dialect markers. In fact, when asked about pronunciation, C stated that sounds are neither important to the way one speaks nor to the way one raps. M1 alluded to the palatalized realization of /t/ when asked to imitate the Cassaoui accent, with very exaggerated [tš] pronunciation, but did not mention it outright. In responding to the question about changing dialect, both S and C mentioned shifting to French in formal environments. M2 mentioned that he changed his speech to sound “more Cassaoui” when he moved, but in the previous few years he had tried to shift his accent back to what he evaluated as his native dialect. Finally, M1, who had never lived outside of Meknès long-term, did not mention shifting his regional pronunciation, but when questioned after the interview about his production of [tš] during the course of the interview, he responded, “Oh, I guess sometimes I try to sound more like my Cassaoui friends.”

5.2. Discussion of Statistical Results

19The results confirm one of the two hypotheses. Origin was shown to be a reliable predictor of CoG value, indicating that the respondents largely produced affricated realizations of /t/ consistent with respondents from the same region. This would also indicate that CoG is a reliable acoustic correlate of regional variation in the realization of this phoneme – the respondents from the coastal area stretching between Salé and Casablanca reliably produced CoG values on the lower end of the spectrum, and the speakers from Meknès reliably produced CoG values on the higher end of the spectrum. This result constitutes an acoustic confirmation of regional variation in production of affrication, which I intend follow up with a more extensive acoustic study of affrication of /t/ in collaboration with colleagues in Morocco.

20Pair-wise comparisons between the informants confirm this relationship between Origin and CoG, refuting my hypothesis that M1’s CoG values would skew closer to those from S and C. M1’s CoG values were not significantly lower than those of M2, the other Meknassi rapper who intentionally modified his speech to sound more Meknassi. None of the informants specifically noted alveolar or palato-alveolar affrication of /t/ as a salient feature of either dialect, though M1 recognized it when prompted.

5.3. Affrication and the Street-Conscious Identity

21But what do these results mean from a broader perspective? The data for this study was taken entirely from the semi-structured interview, the most casual style of speech tested in the sociolinguistic interview, unlike Alim’s (2003, 2006) study of the casual speech and rap lyrics of Eve and Juvenile, which proposed the cultivation of a “street-conscious identity” (2006: 113) within these artists’ performance registers but not in their casual speech. What are the implications of these respondents consciously choosing, or not choosing, to shift their casual speech style? To put this in context, we must look at it from the dynamics of this particular community of practice – that is, the Moroccan Hip Hop community – and examine the role of linguistic capital both from the perspective of the role it plays in the community’s power dynamics and from the perspective of artistic authenticity.

22There is myriad evidence to point to Casablanca as a central location in Moroccan Hip Hop culture, from its status as Morocco’s largest economic center and the site of rapid post-independence urbanization (Miller 2007), to its role in hosting major music festivals such as L’Boulevard, to its symbolic role as the birthplace of Nass el Ghiwane (Almeida 2013). However, M2 instead makes an active effort to rep his hometown of Meknès. This raises the question of “keepin it real” in the Moroccan context. As various chapters in Terkourafi (2010b) explore, the particular balance of these two elements that will allow Hip Hop artists to become successful is dependent upon the local Hip Hop culture – from the broader regional level, to the national level, to the community level.

23Within Morocco in particular, different local rap scenes and even different styles of rap will use combinations of linguistic codes, depending on their message and audience. For example, “commercial” rappers in Morocco will rap primarily in English in order to directly invoke the global authenticity of American Hip Hop, whereas “conscious” rappers will rap almost exclusively in Moroccan Arabic (Harrouchi 2015). Furthermore, the degree to which localization of themes and language is concentrated at the community level or at the broader national level changes from city to city. All of this indicates that the process of localization of Moroccan rap is not uniform within the Moroccan Hip Hop scene, nor is the Moroccan Hip Hop scene itself a monolith – different artists with different styles localize their language to varying degrees in order to best represent their notion of authenticity. M1 adopts the street language of the country’s largest urban center to access the authenticity of that city’s Hip Hop artists, but M2 reverts to the speech of his hometown, to access the authenticity of keeping it local. This reflects the inherent tensions in “keepin it real,” within a system as complex as the Moroccan HHN. In future research, I hope to determine and measure what the so-called “showcase” variables are in Moroccan street speech, to more fully understand the specific capital of this linguistic market.


Alim, H. Samy. 2003. “‘We Are the Streets’: African American Language and the Strategic Construction of a Street Identity", Makoni, Sinfree, Smitherman, Geneva, Ball, Arnetha F., and Spears, Arthur K. (eds.), Black Linguistics: Language, Society, and Politics in Africa and the Americas. New York: Routledge. 40-59.

Alim, H. Samy. 2004. You Know My Steez : An Ethnographic and Sociolinguistic Study of Styleshifting in a Black American Speech Community. Durham: Duke University Press for the American Dialect Society.

Alim, H. Samy. 2006. Roc the Mic Right: The Language of Hip Hop Culture. New York: Routledge.

Almeida, Cristina Moreno. 2013. “Unravelling Distinct Voices in Moroccan Rap: Evading Control, Weaving Solidarities, and Building New Spaces for Self-Expression”, Journal of African Cultural Studies 25 (3). 319–32.

Almeida, Cristina Moreno. 2015. “The Politics of Taqlidi Rap: Reimagining Moroccanness in the Era of Global Flows”, The Journal of North African Studies, 1–16.

Androutsopoulos, Jannis, Awad Ibrahim, and Alistair Pennycook. 2009. “Language and the Three Spheres of Hip Hop”, Alim, H. Samy (ed.), Global Linguistic Flows : Hip Hop Cultures, Youth Identities, and the Politics of Language. New York: Routledge. 43-62.

Boersma, Paul, and David Weenik. 2016. Praat: Doing Phonetics by Computer (version 6.0.14).

Elvira-Garcia, Wendy. 2014. Zero-Crossings-and-Spectral-Moments (version v.1.3).

Gadet, Françoise, & Hambye, Philippe. 2014. “Contact and Ethnicity in ‘Youth Language’ Description: In Search of Specificity”, Nicolaï, Robert (ed.), Questioning Language Contact: Limits of Contact, Contact at Its Limits. Leiden: Brill. 183-217.

Haeri, Nilofar. 1997. The Sociolinguistic Market of Cairo: Gender, Class, and Education. London: Kegan Paul International.

Harrouchi, Zineb. 2015. “Les Fonctions Des Languages Dans Le Rap Marocain”, Langues, Cultures, et Sociétés 1 (1).

Hassa, Samira. 2010. “Kiff My Zikmu: Symbolic Dimensions of Arabic, English and Verlan in French Rap Texts”, Terkourafi, Marina (ed.) Languages of Global Hip-Hop. London: Continuum International Publishing Group. 44-66.

Jamin, Mikaël. 2004. “‘Beurs’ and Accent Des Cités: A Case Study of Linguistic Diffusion in La Courneuve”, Contemporary French and Francophone Studies 8 (2). 169–76.

Jamin, Mikaël, Trimaille, Cyril, & Gasquet-Cyrus, Médéric. 2006. “De La Convergence Dans La Divergence: Le Cas Des Quartiers Pluri-Ethniques En France”, French Language Studies 16. 335–56.

Kolgjini, Julie M. 2004. “Palatalization in Albanian: An Acoustic Investigation of Stops and Affricates”, PhD, Arlington: University of Texas at Arlington.

Labov, William. 1984. “Field Methods of the Project on Linguistic Change and Variation”, Baugh, J., & Sherzer, J. (eds.) Language in Use: Readings in Sociolinguistics. Englewood Cliffs, NJ: Prentice Hall. 28-53.

Miller, Catherine. 2007. “Arabic Urban Vernaculars: Development and Change”, Catherine Miller, Al-Wer, Enam, Caubet, Dominique, & Watson, Janet C.E. Arabic in the City: Issues in Dialect Contact and Language Variation. New York: Routledge. 1-31.

Naciri, Youssef. 2014. “Spirantization of the Voiceless Dental Stop in Rbati-Moroccan Arabic”, BA, Rabat: Mohammed V University Agdal.

Pennycook, Alistair. 2007. “Language, Localization, and the Real: Hip-Hop and the Global Spread of Authenticity”, Journal of Language, Identity & Education 6 (2). 101–15.

Terkourafi, Marina. 2010a. “Introduction”, Terkourafi, Marina (ed.) Languages of Global Hip-Hop. Advances in Sociolinguistics. London: Continuum International Publishing Group. 1-18.

Terkourafi, Marina (ed.). 2010b. Languages of Global Hip-Hop. London: Continuum International Publishing Group.

Triki, Laïla. 1991. “Réaménagement de Salé: La Ville-Dortoir Se Réveille”, L’Économiste


1 Three Hip Hop scenes, to be precise, but because of the demographic and cultural similarities between Casablanca and Salé I am subsuming them under one category.

Table des illustrations

Titre Fig. 1 Example of coding system for /nta/ (“you”)
Fichier image/jpeg, 392k
Titre Fig. 2: Plot of CoG values per Speaker
Fichier image/jpeg, 120k


University of Texas at Austin

Le texte et les autres éléments (illustrations, fichiers annexes importés) sont sous Licence OpenEdition Books, sauf mention contraire.


Open access


Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search