Temporal Patterning and “Degrees of Orality” in Occitan and French Oral Narrative

Oral Tradition, 36/1 (2023):91-122


This article explores tense usage and tense-switching in the temporal structuring of Occitan and French oral narratives, drawing on theoretical frameworks in linguistics and sociolinguistics, as well as perspectives from anthropology and folklore studies.1 It forms part of a larger project, ExpressioNarration, financed by a Marie Skłodowska Curie Fellowship, incorporating the creation of a digitized corpus of oral narratives that has served as the dataset for an analysis of three major temporal phenomena in discourse, notably tenses, frame introducers, and connectives. Our objective in this article is to analyze the use of different tenses on the narrative line and the phenomenon of tense-switching in narratives that reflect different degrees of orality, with a particular focus on the relationship between temporal patterns and questions relating to the channel of transmission, the sources of the narratives, and their performance context.2

Tense usage and tense-switching have been investigated from a range of theoretical perspectives in various types of oral narrative in French (conversational, performed, medieval, modern, and so forth) and indeed in many other languages (see Section 1, below). However, despite considerable discussion in the literature regarding the scalar nature of orality (see Section 2.1, below), research on tense patterns in oral narrative has generally not integrated this notion of scalarity into the analysis. Yet for minoritized languages, where a rich oral narrative tradition is often found, sharp divides between neat categories such as “oral” and “written” are not meaningful, given the complexities around questions relating to transmission, sources, and performance of stories. Moreover, a number of other factors can influence linguistic usage, including the status of different languages, the strength of revitalization movements, the impact of language contact, diachronic and diatopic variation, the role of story-type, questions of individual style, as well as language competency in bilingual contexts. In this paper, we take oral narrative in Occitan as a case study through which to explore the relationship between tense

1 The authors gratefully acknowledge funding from Horizon 2020 in the form of a Marie Skłodowska Curie Individual Fellowship (EU grant number 655034). We are extremely grateful for the constructive feedback from the two readers at Oral Tradition.

2 Carruthers and Vergez-Couret (2021) examine frame introducers and connectives.


usage (including tense-switching) and orality, using a model and a corpus (both designed by and for the project) which incorporate a scalar perspective (Section 2, below). We also include a comparator subcorpus of French oral narratives.

Occitan is a Romance language with a rich literary history.3 It is spoken in southern France, in twelve valleys in Italy, and in the Val d’Aran in Spain. There is no agreed standardized variety, and the language is widely thought to have six dialects, namely, Auvernhat, Gascon, Lengadocian, Lemosin, Provençau, and Vivaroaupenc (Bec 1995), two of which, Lengadocian and Gascon, are included in our corpus. As with many of the regional languages of France, usage of Occitan has declined dramatically since the Revolution of 1789 and particularly in the course of the last hundred years. There has been little official support for regional languages in France, where Article 2 of the Constitution states that French is the sole official language. Indeed, laws designed to support regional languages (for example, the Loi Deixonne, enacted in 1951) have had little positive effect, and France has not ratified the European Charter for Regional and Minority Languages. However, it is possible that the provisions of the Loi Molac (Loi relative à la protection patrimoniale des langues régionales et à leur promotion), passed in 2021, will offer increased support to regional languages in the future.4 In practice, the fate of many regional languages is highly dependent on the strength of grassroots movements, with considerable successes in the cases of Breton, Occitan, Alsatian, and Corsican. So-called “native speakers” of Occitan tend to be older, rural citizens, but a revivalist movement has been active since the 1970s, with a network of not-for-profit associations, including the Calandreta schools (bilingual French/Occitan schools which are mainly but not exclusively primary), the Institut d’Estudis Occitans, and the Centre de Formacion Profesionala Occitan.5 The revival is visible on street signs and in newspapers, and is audible on radio and television. Several universities offer major or optional classes, and culturally, there is a substantial network of artists, including writers, singers, and storytellers.

Section 1 will give a brief overview of the tenses on which we will focus in Occitan and French, drawing on previous theoretical research relating to tense usage and tense-switching, and pointing up the differences between the two languages in the context of language contact. Section 2 will open with a discussion of the notion of scalarity in relation to orality and will outline the corpus we have constructed for the project, paying particular attention to sociolinguistic questions relating to speaker background and use of Occitan, channel of transmission, story sources, variation, and performance context. Our research questions are set out at the end of Section 2. The core of the analysis is contained in Sections 3 and 4. These

3 In literary terms, Occitan is perhaps best known as the language of the Troubadour poets (eleventh-thirteenth centuries) and for its more recent revival in the nineteenth century through the work of Frédéric Mistral and the Félibrige.

4 https://www.legifrance.gouv.fr/jorf/id/JORFTEXT000043524722

5 Ronjat (1930) put the number of speakers in 1920 at around ten million. The Enquête sur l’histoire familiale related to the 1999 census puts the current numbers of active users at around 700,000 (Sibille 2010). A recent sociolinguistic study of the regions of Nouvelle Aquitaine and Occitanie (https://www.ofici-occitan.eu/oc/restitucion-de-las-resultas-de-lenquesta-sociolinguistica/), published by l’Ofici Public de la Lenga Occitana (2020), argues that seven percent of the population of these regions are speakers of Occitan. Extrapolating up to the whole Occitan-speaking area would give a figure of around 600,000 speakers. 


sections focus, respectively, on tense usage on the narrative line in relation to the different degrees of orality in our corpus (Section 3) and on the frequency and function of tense-switching across the subcorpora (Section 4). The short section (5) that follows discusses apparently contradictory patterns. The Conclusion will draw together our findings and suggest avenues for future research.

1. Narrative Tenses in Occitan and French: Theoretical Reflections

1.1 Narrative Clauses and the Narrative Line

There are multiple theoretical approaches to defining the term “narrative.”6 For linguists, especially those working on oral narrative, most definitions are predicated on a macro-level assumption that the order of events in a narrative corresponds broadly to their sequence in time. As Labov and Waletzky put it, the translation of experience—real or imagined—into narrative is “one method of recapitulating past experience by matching a verbal sequence of clauses to the sequence of events that actually occurred” (1967:20; see also Labov 1997). At a micro-level, events are contained in “narrative clauses,” although the latter can also contain “states” that are bounded, that is, considered to occur within a bounded time period (Carruthers 2005:13-18). The link between narrative clauses and temporal sequence is captured theoretically in different ways, including frameworks based on discourse structure (Smith 2003), discourse relations (as in Segmented Discourse Representation Theory; Asher and Lascarides 2003), praxematics (Bres 2001), relevance theory (Moeschler 2000), and discourse-pragmatics (Fleischman 1990). All theories account for exceptions to the default assumption of temporal sequence, where events in sequential clauses may not succeed each other in time; in such cases, there is usually linguistic evidence that points to a relation other than sequence, such as simultaneity (Smith 2003:94-95). It is also important to note that stories obviously contain types of clause other than narrative clauses, in particular, descriptive clauses and direct discourse attributable to the characters in the story. The focus in this article is exclusively on tense usage in the narrative clauses that contain events on what we will term the “narrative line.”

In terms of the tenses that can appear as markers of events in narrative clauses, there are three main possibilities in French and Occitan: the Passé Simple/Passat Simple (PS), the Passé Composé/Passat Compausat (PC), and the Présent Narratif/Present Narratiu (Pnarr). However, there are significant differences between French and Occitan in terms of how the PS and PC operate in discourse, a factor that could be significant given the level of contact between the two languages. As background to our analysis in Sections 3 and 4, Sections 1.2-1.4 will give a brief overview of the established temporal and aspectual functions of these tenses in narrative discourse, including the phenomenon of tense-switching.

6 Fludernik (2009:1-12), for example, discusses a range of approaches.


1.2 PS and PC in French

In French, the PS is most frequently found as the narrative tense of many literary narratives. In this context, it references non-habitual, completed past-time events, which, when there are no counter-indications, are assumed to occur in sequence. We will refer to this function as that of marking the preterite.7 The PS is extremely rare in spoken French (especially where there is no link to a written script) but is attested in journalistic French, where it can have key discourse-pragmatic functions in the context of tense-switching in texts where it is not the dominant narrative tense. Waugh and Monville-Burston (1986) highlight journalistic examples where the PS renders certain events or states more salient, referring to this as the PS’s capacity for “detachment” and “dimensionalisation” (that is, creating clear boundaries around the event or state), which is in turn linked to its telicity and aspectual boundedness.

The PC has two main functions in modern French, hence its label as a “temps à deux visages” (Martin 1971:103). The first denotes the present perfect, which is the equivalent of the English “I have done” (j’ai fait), a function fulfilled by the PC in both written and spoken varieties of the language. The strong association of the PC with a present-centered perspective on an event or state is reflected grammatically in the present tense of the auxiliary verb. The second function is as a marker of non-habitual, completed past-time events (in other words, as a preterite) in spoken French and many varieties of the written language, which is the same function as that of the PS in many varieties of literary French. In other words, whereas some written modes, such as literary narratives, distinguish between a preterite (PS: il fit) and a present perfect (PC: il a fait), others, including most oral varieties, use the PC (il a fait) for both these functions.8

1.3 PS and PC in Occitan

The PS is the typical tense for narrative clauses in Occitan stories, whether written or spoken (Sibille 2015). In most varieties of Occitan, the PC does not normally mark events on the narrative line (as in the preterite function) but rather serves mainly as a present perfect, that is, it denotes completion in relation to the present, sometimes creating a “result state,” whereby ara ai manjat (maintenant j’ai mangé) means, “I am now in a state of someone who has eaten and am no longer hungry” (Sibille 2015:255; see also Teulat 2008).9 In other words, the PC/PS opposition in Occitan mirrors the opposition in literary varieties of French but is distinct from the system in spoken French, with which it is in contact. It is also important to note diatopic variation in Occitan, whereby in the Gascon maritime region, the PC is now widely used in the

7 Note that we distinguish form (for example, PS) from function (such as preterite) throughout this paper.

8 See also Waugh (1987), who further subdivides these two main functions of the PC into four.

9 It is also possible to find examples where the event marked by the PC is much less clearly a perfect in relation to the present and functions more like a preterite, which is “recent” and therefore closer to the present, such as (Gairal 2001:81): Arribèt fa cinc jorns, es tornat partir aqueste matin (“he arrived five days ago, he left this morning”). However, examples such as these are not the norm (for a full discussion, see Sibille 2010).


preterite function. Lassalle (2017:82-83)10 dates this development to the late-nineteenth century, citing evidence also of non-standard PS forms that may suggest less widespread use of the PS in the area.11 Since we have some stories in Gascon in the corpus, we will return to the question of possible diatopic variation.

1.4 Présent Narratif and Tense-Switching12

Use of the présent narratif (Pnarr) as a narrative tense for past-time events, particularly in oral varieties, is a well established phenomenon across a wide range of languages, including French. Early explanations tended to focus on the present-time functions of the Pnarr in making a past event appear more “present” and therefore more vivid, or in making the present of the narrator seem closer to the past event (Ollier 1978). More recent theoretical explanations tend to assume that the present is fundamentally an atemporal form that acquires its temporal meaning in context, thereby opening the way for a host of pragmatically motivated meanings (Fleischman 1990). In the context of research on tense patterning in oral discourse, the Pnarr is most strongly associated with tense-switching, that is, switching between the present and a past tense to mark past-time events on the narrative line (Carruthers 2005). In contemporary conversational French, this is usually a question of alternation between the PC and the Pnarr, since, as noted above, the PC functions as the preterite of informal spoken French. In Old French, there is clear evidence that alternation often involved PS and Pnarr, with PC/Pnarr also a possibility (Fleischman 1990). PS/Pnarr alternation is also possible, if rare, in performed storytelling contexts in contemporary French, as we shall see, but it is not normally found in conversational French. Although the Pnarr is mentioned in grammars and in descriptive accounts of Occitan (for example, Sibille 2015), there is no published research that we know of on tense-switching between past and present in oral varieties of Occitan.

Most research on tense-switching relates to the functions of the Pnarr in oral discourse, but, as mentioned above in Section 1.2, the PS has also been analyzed in journalistic French in alternation with PC and Pnarr. In some contexts, it is the switching itself, and the texture it can create in the narration of events that is paramount, rather than the individual functions of particular tenses. The possible effects and functions of tense-switching are well documented across many languages, including French but not Occitan. It is impossible to do these justice in a few lines, but among the most widely documented are a number of effects and functions that are related to each other in terms of marking phenomena that are linked to saliency and subjectivity (Carruthers 2005:74-97): for example, Pnarr or PS for foregrounding salient events where, as a

10 We are grateful to Jean Sibille for this reference.

11 Interestingly, variation in the conjugation of the PS is also linked by Vaugelas (1647:109), in the early modern period, to increased use of the PC in French.

12 We will use the term “narrative present/présent narratif” where the present marks past events on the narrative line. There is some variation in the terminology used in the literature. Sibille (2010) deploys the term “historic present” for this usage, reserving “narrative present” for contexts where both past events and states are rendered in the present, as found in some contemporary literary contexts. Fleischman (1990) uses the term “narrative present” in the same context as we do but uses “historic present” for contexts where it does not alternate with a past tense but is used consistently for past-time events in large stretches of text.


marked usage relative to the majority narrative tense in the surrounding discourse, the event(s) in question thus stand(s) out from the surrounding narrative; turbulence at narrative peaks, where rapid alternation between present and past occurs at a point of heightened narrative tension; Pnarr and PC/PS structuring the narrative into different sections or blocks with different discourse topics (or episodes), where Pnarr is the narrative tense in one block and PC or PS in the next; Pnarr moving the narrative forward where it occurs at the beginning of a new section or sub-section of the story, in some cases accompanied by frames such as le lendemain matin or by adverbials marking suddenness; closure with the PC or PS, where the aspectual perfectivity of the PC and PS (unlike the aspectual multivalency of the present which has both perfective and imperfective properties) reinforces a sense of completion, either of a section of the narrative or indeed of the entire narrative; and tense-switching to speed up or slow down the narrative, with the Pnarr associated more strongly with the former and PC with the latter. Fleischman (1990) divides such functions of tense-switching into textual (structural), expressive, and metalinguistic categories.

It could be argued that these types of function are brought together in many instances of tense-switching with verbs of speech, notably dire, which is a particularly frequent correlation (Carruthers 2005:70-74): rapid tense-switching with dire often occurs on the borders between direct speech and chunks of narrative text, breaking up the story in structural terms and anticipating or following the performance and subjectivity associated with direct speech.

In this article, we will explore the tenses found on the narrative line in different degrees of orality in Occitan, as well as the phenomenon of tense-switching, including a comparison with French. In practice, the focus will therefore be on PS, PC, and Pnarr in both languages. 

2. The Corpus

2.1 A Scalar Approach

Research in linguistics, literature, anthropology, and folklore has moved well beyond the concept of a binary divide between “oral” and “written.” Amongst the best known models in linguistics is that of Koch and Oesterreicher (2001), where alongside a binary divide between oral and written channels, they propose a scalar polarity in relation to formality between conception écrite and conception orale,as well as a further set of continua between polarities relating to concepts such as private versus public communication, spontaneity versus preparedness, and intimate versus distanced relationship with interlocutors. Several of these parameters will emerge as pertinent for our discussion of oral narrative. Biber and Conrad’s multidimensional account of genre/register (Biber 1988; Biber and Conrad 2009) also maps the complex sets of relationships between linguistic features, situational features (for example, participants, setting, channel, communicative purpose, and topic), and what they term “dimensions” on a continuum between two polarities (such as oral versus literate discourse or narrative versus non-narrative discourse), rather than categorizing discourse in binary terms like “oral” and “written.” Indeed, in Biber and Conrad’s model, such an absolute distinction does not exist (Carruthers 2018).


Folklorists, anthropologists, and literary scholars working on oral literature have also increasingly rejected a binary oral-written divide, both in terms of the way in which broad societal factors relating to orality and literacy are framed, and also in relation to the properties of oral narrative as regards composition, transmission, and performance. Indeed, they argue that such binaries can easily lead to other false dichotomies, with writing becoming all too easily associated with sophistication, rationality, and the urban post-industrial world, and orality with small-scale, non-industrial, communal ways of life (Finnegan 1988). As Finnegan puts it: “how useful is this binary typology when it turns out that most known cultures don’t fit?” (141). Zumthor’s (1983) framework, too, brings out the more nuanced nature of societal dynamics, making a distinction between oralité primaire/pure (where there is no evidence of a writing system), oralité mixte (where a writing system exists but its influence is limited for sociological and/or educational reasons), and oralité seconde (where a writing system is strongly embedded in society, and its influence on orality is therefore inevitable).

Speaking specifically of oral literature, Finnegan (1988:110-22; see also 1977:23-24) demonstrates the “overlap” between oral and written in relation to sources (which can be oral or written or both in a given context), composition (which can involve both written and oral processes, including oral composition-in-performance), and dissemination (which can take a number of different oral and written forms, including publication). In his work on oral poetry, Foley (2002:39) also avoids clear-cut divisions, taking a scalar approach which considers oral, aural, and written dimensions in his model of media categories in relation to composition, performance, and reception. The theoretical debate thus suggests that questions concerning sources/composition, transmission, and performance/dissemination are critical to the dynamic between oral and written in a given storytelling context and against a particular societal backdrop.

Our starting point is therefore that the concept of orality is a relative one, with different degrees of “oral-ness” in different contexts. It follows that our understanding of temporal phenomena in oral narrative has the potential to be greatly enriched and significantly more nuanced if the notion of different degrees of orality is embedded in the analysis, all the more so in a minoritized language where linguistic, societal, and practice-related factors are likely to play a role in the patterns attested. Moreover, beyond the question of scalarity, there is scope for variation across a number of other parameters, not least with regard to diatopic variation, diachronic changes, individual styles of storytellers or writers, different story-types, as well as varying levels of bilingualism and language contact. In this study, we have therefore aimed to strike a balance between devising a taxonomy that allows us to create a corpus which has the potential to show meaningful comparisons across different degrees of orality, while at the same time undertaking an analysis which is sensitive to the multiple possible levels of variation that may be found in the data. As Foley puts it, “the trick is to find a level that is just fine-grained enough to be diagnostic, just comprehensive enough to demonstrate some overall unities and offer some practical bases for comparison, and just flexible enough to accommodate the natural diversity of human expression” (2002:39).


2.2 The Four Subcorpora

The corpus created for the project contains four subcorpora of narratives (three in Occitan and one comparative corpus in French) based on a taxonomy which includes their channel of transmission (oral or written), consideration of story sources (written or oral), and their context of communication (as written texts to be read, as part of a community-based oral tradition, or as a more stylized public performance).13 These communicative contexts involve differing levels of spontaneity: stories recounted as part of an oral tradition are relatively spontaneous and informal, whereas contemporary public performances for an audience generally involve higher levels of pre-planning and a relatively staged context, sometimes with music, props, and costume (cf. Koch and Oesterreicher 2001).

All our stories are drawn either from the European repertoire of traditional stories or from other cultures around the world, although these can incorporate regional and local references. Where possible, in the metadata, we have given the story categorization using Aarne and Thompson’s (1961) widely known classification system, including general categories (such as Animal Tales, Tales of Magic, Anecdotes or Jokes, Formula Tales, Tales of the Stupid Ogre, and so forth), subgroups (for example, Wild Animals and Domestic Animals, Supernatural Adversaries, and The Stupid Man), and tale-types (as in, AT 124, Three Little Pigs; AT 303, The Twin Brothers; AT 1696, What Should I Have Said; and so on). The stories that we could categorize using the Aarne-Thompson system were classified as contes (since they are clearly identifiable as part of the European oral tradition), while those not falling under these categories were classified as récits.14 Given the nature of the data (that is, in a minoritized language with limited sources available), it is not always possible to achieve a perfect balance as regards the story-types included in each subcorpus, although we have tried to do this as far as possible (Carruthers and Vergez-Couret 2018). All available metadata regarding storytellers (age, region, gender, and so forth) and/or publication factors are also recorded in the header for each story. Every attempt was made to balance the corpora as far as possible for gender, although this was not possible for the all-male cohort of writers in the case of the published stories. Any issues relating to diatopic variation will be raised in our discussion (we have stories in both Gascon and Lengadocian), but it was not always possible to balance the subcorpora in this regard. The stories in the different subcorpora also vary in length, and we therefore distinguish between short stories (labeled “S”: < 1,000 words), medium-length (“M”: 1,000-2,000 words), long stories (“L”: 2,000-3,000 words), and very long stories (“VL”: >3,000 words).15

13 The corpus, OcOr (Vergez-Couret and Carruthers 2018), is available for download at https://zenodo.org/record/1451753#.Y3UYUi2cZXg. Carruthers and Vergez-Couret (2018) discuss the methodology employed, including fieldwork, digitization, annotation, and so forth.

14 In classifying the story-types we give the terms in French so as to distinguish clearly between contes and récits.

15 Given the variations in length of individual stories, the number of stories varies across the subcorpora. OOT = twenty-six stories, approximately 16,600 words; OOC = thirteen stories, approximately 19,000 words; FOC = thirteen stories, approximately 20,000 words; OWT = nineteen stories, approximately 18,200 words.


The four subcorpora are constituted as follows: 

  • OOT (Occitan, oral,16 traditional): The stories in OOT are traditional stories in the sense that they are acquired and disseminated as part of an oral tradition, with most speakers born between the second half of the nineteenth and the early-twentieth century. They were collected by means of fieldwork among native speakers of Occitan, most of whom, given the period in question, were likely bilingual in French and Occitan, with Occitan more likely to be their mother tongue. The recordings are held in the COMDT (Centre Occitan des Musiques et Danses Traditionnelles) in Toulouse and have been subsequently transcribed, digitized, and analyzed by this project.The language used in OOT is thus in every sense “oral”: transmission and sources are oral, with no written influence. The performance is community-based, informal, and spontaneous in nature. Metadata on the storytellers is incomplete, with more information on dates of birth and less on place of birth or work.
  • OOC (Occitan, oral, contemporary): These stories are recounted by contemporary artists, taken from existing recordings and two Toulouse storytelling events organized by the project.17 Story transmission is oral (in front of an audience) and relatively spontaneous; in some cases, there is a strong performative dimension in the sense that the stories are practiced, stylized, and staged (with music, accessories, and so forth) in comparison to those told as part of an oral tradition in OOT. Indeed, the OOC descriptors in relation to performance, stylization, and staging, as well as the tendency to use written sources, mean that most storytellers in OOC would be considered to be “new” rather than “traditional” storytellers (see the discussion of FOC below). However, it is important to point out that some OOC storytellers self-identify as strongly associated with the oral tradition, and of course their use of a regional minoritized language places them closer to this tradition than those in our contemporary French corpus (FOC). Moreover, unlike FOC, oral sources can be used alongside written sources by the storytellers in OOC, especially by those who feel a strong link with the oral tradition. The storytellers in OOC are either native speakers of French who have learned Occitan (not usually through the school system) or are, to varying extents, bilingual in French and Occitan if they have a strong family connection with Occitan. All were born between 1944 and 1981, were educated in French, and are fluent French speakers. The storytellers in OOC are thus a heterogeneous group, sharing a number of descriptors with FOC and some with OOT.
  • FOC (French, oral, contemporary): As a point of comparison, our third subcorpus involving oral transmission is a selection of contemporary oral storytelling in French drawn from the French Oral Narrative Corpus (Carruthers 2013). These stories are recounted by contemporary artists and were recorded at the CLIO (Conservatoire Contemporain de Littérature Orale) and subsequently transcribed and digitized. Stories are performed in a relatively spontaneous context, albeit often with music, accessories, and some staging. In the case of FOC, sources are always written, and stories are worked into an oral performance which is practiced but not memorized. All storytellers

16 “Oral” and “written” in the titles of the subcorpora refer primarily to the channel of transmission.

17 One recording is taken from the DVD Les contes du placard by Florant Mercadier (2013) and used with his permission. All other recordings were made during two events held on September 28, 2016, and October 26, 2016, at l’Ostal d’Occitania, in collaboration with the Institut d’Estudis Occitans (IEO).


in FOC (as with most in OOC) would be termed “new” storytellers; they are not part of an oral tradition but use written sources that blend local, national, and international stories. Although the literature on storytelling draws a clear distinction between “traditional” (OOT) and “new” storytellers, for our corpus the difference is much clearer in FOC than OOC; FOC unequivocally concerns new storytellers, whereas this is the dominant (but not the only) group in OOC. 

OWT (Occitan, written, traditional): The stories in OWT are also traditional in the sense that they have been collected through fieldwork with storytellers who form part of an oral tradition. However, unlike OOT, these are written, published versions of the stories, and there are no existing recordings of the original oral versions collected during fieldwork. This project has digitized the published versions.18 Metadata on the collection process and the original storytellers are of variable quality, but we know from information relating to names, gender, occupation, dates/places of birth, and publication dates that most storytellers were born in the first half of the nineteenth century, with the texts published in the late-nineteenth century. We know nothing about the storytellers’ knowledge of French, although it is likely, given the period in question, that most had little or no French. The writers were certainly educated (including teachers, an archivist, a judge, a poet, and so forth), born also during the nineteenth century; they published both in French and Occitan. We have no information on the nature of the fieldwork or on the relationship between the oral sources and the written published versions of the stories, and this is an issue to which we will return.   

To use Zumthor’s terms (see Section 2.1, above), both OOT and OWT are products of an oralité mixte; although we know that the writers in OWT are literate in Occitan and French, this is not the case for the storytellers in OWT, while those in OOT may have acquired some literacy in French through the school system but not in Occitan. By contrast, both OOC and FOC are products of an oralité seconde, meaning that large swathes of society in general (including the audiences) and the storytellers themselves are strongly influenced by the literate culture around them and by their educational experience in that culture through the medium of French. That said, although the storytellers in OOC can read and write in Occitan, they have differing levels of literacy and fluency: Occitan is invariably somewhat less well embedded linguistically in their background and certainly perceived to be less prestigious than French in wider society.

Table 1 summarizes the descriptors of the four subcorpora (adapted from Carruthers and Vergez-Couret 2018):

18 Carruthers and Vergez-Couret (2018) discuss the complexities of this process.


+ Spontaneous
– Planning
+ Traditional
+ Spontaneous
+ Planning
+/– Traditional
+ Spontaneous
+ Planning
– Traditional
– Spontaneous
+ Planning
+ Traditional
Oral sourcesWritten and oral sourcesWritten sourcesOral sources
Oralité mixteOralité secondeOralité secondeOralité mixte
Speakers born between the second half of the nineteenth and the early-twentieth centuriesSpeakers born from the mid-twentieth century onwardsSpeakers born from the mid-twentieth century onwardsSpeakers born in the first half of the nineteeth century
Table 1. Descriptors for the four subcorpora.

2.3 The Research Questions

In the light of our objective of exploring tense usage and tense-switching in relation to different degrees of orality, our core research questions are:

  • What tenses are used on the narrative line in the four subcorpora, and how do the subcorpora compare with each other? What is the relationship between narrative tense usage and different degrees of orality, taking into account questions around written versus oral transmission, written and oral sources, and performance context in terms of tradition and practice?
  • To what extent is tense-switching found in the four subcorpora, and how, if at all, is this related to different degrees of orality in terms of frequency, rate, and function?
  • To what extent do other factors play a role in the patterns attested, such as individual variation in style, diatopic or diachronic variation, story-type, or language contact between French and Occitan?
  • How clear-cut are the patterns? What factors might be at work in examples that do not fit the dominant patterns?
  • Can the introduction of a scalar approach to orality enhance our understanding of tense usage in oral narrative? If so, what might be the wider implications of this in other domains?


3. Tense Patterning in the Four Subcorpora

3.1 Tenses on the Narrative Line: Overall Patterns

Our analysis of the major tenses occurring in narrative clauses for each subcorpus reveals a variety of patterns, as shown in Table 2.19 In each case, the table gives the number of stories with each of the possible narrative tenses (PS, PC, Pnarr) as the major tense on the narrative line or with a combination of two of these tenses in roughly equal distribution. We have defined a major tense as one that occurs in 60% or more of the narrative clauses in a story; thus, in many contexts, there will nonetheless be a significant presence of at least one other narrative tense, and we will return to this below. In seven stories, where the percentage distributions of two tenses were between 46% and 54%, it seemed arbitrary to speak of a major tense, and in these cases, we classified the story in the category 50/50.

Major tense (by story)OOTOOCFOCOWT
PS/Pnarr (50/50)3103
Table 2. Major tenses on the narrative line by story.

At the two ends of our table (OOT and OWT), the patterns are characterized by a strong preference for one tense as the major narrative tense, that is, Pnarr for OOT and PS for OWT. The picture is more complex for OOC and FOC. In OOC, although the main preterite form (in this case PS) is used most frequently as the major narrative tense, there is nonetheless considerable use also of the Pnarr in that role. As for FOC, the breakdown of the various patterns attested in this range of stories is similar to that found in previous research based on a larger corpus (Carruthers 2005): the PC plays a pivotal role and is quantitatively the most frequent tense, with the Pnarr also used as the major narrative tense. The PS is marginal, as we might expect (see Section 1.2, above). We shall return below to the PC’s absence from OWT and to the stories where PC is the main narrative tense in OOC and OOT. 

Table 2 shows clearly that each subcorpus operates differently in terms of major tenses on the narrative line. These initial observations open the way for further exploration of the relationship between the patterns attested and different degrees of orality, considering, where relevant, factors such as diachronic or diatopic variation, individual variation, and language contact with French. Sections 3.2 through 3.5 will focus on the question of major tenses on the narrative line in each of the subcorpora, as well as the alternation patterns attested, before moving in Section 3.6 towards some interim conclusions. Frequency, rates, and the functions of tense-switching will be considered in detail in Section 4. 

19 See the discussion in Section 2.2 concerning balance across the corpus in terms of numbers of stories and numbers of words in each subcorpus.


3.2 Stories in OOT

The patterns in OOT are described in the following table:20

Story TitleStorytellerTense pattern
“L’istòria d’un paisan del caulet” (S)M. BerthoumieuPnarr only
“De passatge a Bordeu” (S)Armand LescouyèresPnarr mainly/Some PS
“Lo ciror e la filha” (S)Germain MontaubricPnarr mainly/Some PS
“La craba qu’èra bièn mentura” (S)Alice MeyratPnarr mainly/Some PS
“Histoires de tours du Drac” (S)Noémie BatifolPnarr mainly/Some PS
“Un dròlle pas interessent” (S)Hermine CalastrencPnarr mainly/Some PS
“Fin volur” (S)Hermine CalastrencPnarr mainly/Some PS
“Josèp l’amic de las bèstias” (S)Céline CalvetPnarr mainly/Some PS
“Lo marat e la marata” (S)Mme GibliPnarr mainly/Some PS
“L’œuf de jument” (S)M. BerthoumieuPnarr mainly/Some PS
“Jan de l’ors” (S)M. BerthoumieuPnarr mainly/Some PS
“La bèstia de 7 caps” (M)M. BerthoumieuPnarr mainly/Some PS
“Le jeune homme et le lion” (M)M. BerthoumieuPnarr mainly/Some PS
“Lo conte del nau lops” (S)Marguerite LagardePnarr mainly/Some PS
“Le mic e mac” (M)André DastePnarr mainly/Very little PS
“Légende des animaux . . .” (S)Elizabeth BarbediennePnarr mainly/Very little PS
“Istòria de Gargantua” (S)Céline CalvetPS only
“Histoire du lac . . .” (S)Noémie BatifolPS only
“Joan d’Auret” (S)Aurélie PhilipotPS/Pnarr
“L’ueu de cisampa” (S)Storyteller_OOT_1PS/Pnarr
“Le conte des trois poulettes” (M)Maria GirbalPS/Pnarr
“L’orra bèstia de la coeta verda” (S)Céline CalvetPS mainly/Some Pnarr
“Istoèra del curè d’Escaudes . . .” (S)Daniel LacampagnePS mainly/Some Pnarr
“Un talhur seguit per un canh” (S)Storyteller_OOT_2PS mainly/Some Pnarr
“La taupa e lo grapaud” (S)Storyteller_OOT_3PC only
“Lo bon diu e lo pleu pleu” (S)Storyteller_OOT_3PC mainly/Very little PS
Table 3. Patterns in OOT.

The most common patterns for OOT are, in order of frequency: Pnarr mainly/PS, PS mainly/Pnarr, and PS/Pnarr in broadly equal proportions. Both Pnarr and PS are therefore central in OOT, even if Pnarr is present in greater proportions across the subcorpus. Broadly speaking, the patterns in OOT are thus very clear. OOT_3 is the only storyteller employing the PC on the narrative line, though she does so almost entirely consistently. There are two possible explanations for this, and it is impossible to know which is valid. We have no biographical

20 S = short; M = medium; L = long; VL = very long—all refer to the length of the story (see Section 2.1, above). Note that we are using approximate terms such as “mainly,” “some,” and “very little” when describing the tense distribution. This is because (i) percentages would be based on low raw numbers in several instances, and therefore reliability would be inconsistent across the corpus; and (ii) raw numbers would not be meaningful in themselves and could obscure the patterns. Our description of the patterns is intended to draw out broad tendencies rather than to offer detailed statistical information.


metadata for the storyteller (who is anonymous), but we know that the story was recorded in 1973, that the speaker was elderly at that point, that (like the other speakers in OOT) she was a native speaker of Occitan who probably also spoke French, and that the recording took place in Moulis-en-Médoc, which is located outside but within fifty kilometers of the Gascon maritime area where the PC is regularly employed with a preterite function (see 1.3 above). There is thus a possibility, given the proximity, that she spent part of her life in the Gascon maritime area, or that she has networks there, and that this is therefore a case of diatopic variation. However, we cannot be certain of this. The other possible explanation is that we are dealing with a case of individual variation whereby the PC is used due to language contact with French. No other speakers in OOT can be located close to the Gascon maritime area, nor does any other storyteller use the PC, so it is difficult to draw reliable conclusions.

We shall return to the widespread use of tense-switching in OOT in Section 4.2, below, but note at this stage that, despite some evidence of individual storyteller preference in relation to the main narrative tense as PS, Pnarr, or PC, patterns are not clear enough and the number of stories per storyteller not sufficiently large to suggest that each storyteller has a preferred pattern. 

3.3 Stories in OOC

The patterns found in OOC are highly diversified:

Story TitleStorytellerTense pattern
“L’ueu de mula” (S)Storyteller_OOC_MCPS only
“La hada e la goja” (S)Storyteller_OOC_MCPS only
“Les fées de la dune de Bombèt” (M)Storyteller_OOC_MCPS only
“Lo peisson-lèbre” (S)Storyteller_OOC_MBPS mainly/Some Pnarr
“Lo gabarròt” (L)Storyteller_OOC_PBPS mainly/Some Pnarr
“L’ase, lo moton, la pola e lo lop” (M)Storyteller_OOC_PBPS mainly/Some Pnarr
“Lo vailet e lo diable” (M)Storyteller_OOC_PBPS/Pnarr
“Lo grelh” (S)Storyteller_OOC_FMPnarr only
“Lo dròlle logat” (M)Storyteller_OOC_MBPnarr only
“La bòta” (L)Storyteller_OOC_MBPnarr mainly/Very little PC
“La caça volenta” (S)Storyteller_OOC_MBPnarr mainly/Very little PC and PS
“Las abelhas” (L)Storyteller_OOC_PVPC mainly/Some Pnarr and PS
“La mair deu conte” (S)Storyteller_OOC_PVPC mainly/Very little Pnarr and PS
Table 4. Patterns in OOC.

First of all, it is less relevant in this case to look at the frequency for each pattern, as it is striking that particular patterns are often strongly—albeit not exclusively—correlated with individual storytellers. Most storytellers from whom we have more than one story seem to have one preferred tense as the main tense on the narrative line: PS for OOC_MC and OOC_PB, PC for OOC_PV. Two storytellers use the PC as a narrative tense, and one of these two (OOC_PV) uses it as his main narrative tense. However, although OOC_PV is a speaker of contemporary Gascon, he is not from the region where the PC in a preterite function is found, that is, the Gascon maritime, or from a neighboring area. Indeed, he is a young speaker who has acquired Occitan as a second language, largely through contact with native speakers, none of whom comes from the


Gascon maritime. It is most likely, therefore, that his systematic use of PC as the main narrative tense is an individual pattern that has developed through language contact with French. OOC_MB uses only a very small amount of PC towards the beginning of her stories. In conversation with her, it became clear that she usually recounts these stories in a bilingual context, where she alternates between French and Occitan. The passages where the PC is found in the corpus are ones that she normally recounts in French; it is most likely, therefore, that her use of PC, like OOC_PV’s, relates to language contact with French, but unlike OOC_PV, OOC_MB makes occasional rather than systematic use of the PC in this function.

We can say with confidence that PS and Pnarr are both important narrative tenses in OOC: one or the other or both appear in all stories, and they are the most common principal narrative tenses in the subcorpus. Perhaps most striking is the fact that each storyteller tends either to have a preferred pattern for tense-switching or tends not to tense-switch. We shall return to this in Section 4.3, below.

3.4 Stories in FOC

The thirteen stories analyzed in our subcorpus reflect the wider patterns across the much larger corpus discussed in previous research (Carruthers 2005). They can be described as follows:

Story TitleStorytellerTense pattern
“Le fils du roi et le corbeau” (M)Storyteller_FOC_ECvPS only
“Ankou” (L)Storyteller_FOC_PCPC only
“Histoire de femmes” (M)Storyteller_FOC_FDPC only
“Le crapaud et la tortue” (S)Storyteller_FOC_ECaPC only
“Mange ma graisse” (S)Storyteller_FOC_ECaPC only
“Pokou 1” (S)Storyteller_FOC_GBPC mainly/Very little Pnarr
“Le roi des pigeons” (M)Storyteller_FOC_ISPC mainly/Very little PS and Pnarr
“Contes de Bourgogne” (M)Storyteller_FOC_CMPC mainly/Very little PS
“Le paysan et la paysanne” (L)Storyletter_FOC_AKPC mainly/Some Pnarr
“Le cordonnier de la ville du Caire” (M)Storyteller_FOC_ABPnarr mainly/Some PC
“La chemise magique” (VL)Storyteller_FOC_RNPnarr mainly/Some PC
“La compagnie des loups” (S)Storyteller_FOC_JGPnarr mainly/Very little PC
“Le mari trompé” (M)Storyteller_FOC_GBPnarr mainly/Very little PC
Table 5. Patterns in FOC.

In terms of frequency, the most common patterns are PC only and PC alternating with Pnarr; in the latter case, there are multiple patterns in terms of quantities of the two tenses, running from mainly PC through to mainly Pnarr. The PS is marginal, both as the sole narrative tense or in alternation with PC, as we might expect in the oral medium; its use is likely to be strongly influenced by literary or other written sources where it is the unmarked past tense in French (see Section 1.2, above). The Pnarr never operates as sole narrative tense; it is strongly associated with tense-switching and therefore always appears alongside either PC (in most cases) or PS and


PC (rarely). The use of PC is thus central to the patterns in contemporary French storytelling, operating as the sole tense on the narrative line or in alternation with Pnarr or occasionally PS (see Sections 1.2 and 1.4, above). We have no examples of PS/Pnarr tense-switching, probably because “a contemporary oralité seconde appears to be uncomfortable with NPR/PS and PS/NPR shifts” (Carruthers 2005:88), since the temporal properties and medium/register associations of the two tenses are totally divergent in French (present versus distant past; spoken discourse/proximity to speaker versus literary narrative/formal discourse).

3.5 Stories in OWT

The patterns found in this subcorpus are described in the table below:

Story TitleAuthorTense pattern
“Lo vielh e los tres volurs” (S)ArnaudinPnarr mainly/Some PS
“Compair Loison e la mair deu vent” (VL)ArnaudinPnarr mainly/Some PS
“Lo bon diu e lo diable” (M)ArnaudinPnarr/PS
“La pelha arrecastada” (S)ArnaudinPnarr/PS
“Lo gojat” (S)LambertPnarr/PS
“Peton-Petet” (S)LambertPS only
“Las sorcièras” (S)LambertPS only
“Joan Lo pigre” (M)BladéPS only
“Los dus bessons” (M)BladéPS only
“La flaüta” (S)BladéPS only
“Pèl-d’Ase” (M)BladéPS only
“Lo voiatge de Joanòt” (M)BladéPS only
“Perqué las fumèlas . . .” (M)LalannePS only
“La formiga que s’enanava a la fèsta de Sabardu” (S)SéréPS only
“Las tres galinetas” (S)LambertPS mainly/Some Pnarr
“Lo lop malaut” (S)BladéPS mainly/Some Pnarr
“Lo maset” (S)MontelPS mainly/Some Pnarr
“Turlendú” (S)MontelPS mainly/Some Pnarr
“La formiga que s’enanava a . . .” (S)SéréPS mainly/Very little Pnarr
Table 6. Patterns in OWT.

Overall, there is strong preference for PS as the major narrative tense in OWT, the other possible tense being Pnarr. The most common patterns in order of frequency are PS only, PS mainly/Pnarr, and Pnarr/PS. In theory, as outlined in Section 2, above, OWT contains stories in print that were taken from fieldwork with oral storytellers, and indeed the two corpora closest to the oral tradition and oral sources (OOT and OWT) share the same two dominant tenses. However, the Pnarr is much more frequent in OOT, while the PS is the major tense in OWT. It is unlikely that this difference could be explained in terms of diachrony, despite the fact that OOT is a later corpus than OWT: there is no known evidence of a diachronic change of this sort. On the other hand, there is a vast amount of evidence, both diachronic and synchronic (Section 1.4, above), that use of the Pnarr, especially where tense-switching is attested, is strongly associated with the oral medium, so it is not surprising that the oral


versions of the stories contain high levels of Pnarr. We have, of course, no way of knowing which tenses precisely were used in the oral versions of the stories published in written form by the writers in OWT. It is very possible, though, that there may have been much heavier use of the Pnarr in the oral versions, and that in some instances, it may have been replaced by PS in the published written version, in line with the Occitan tense system and paralleled in literary texts in French (with which our writers would have been familiar), both of which may therefore have reinforced the choice of PS. Individual variations, even though they are not as strong as in OOC, suggest at this stage of the analysis that two writers, Bladé and Lambert, use the PS as the narrative tense more frequently in their collected stories than Arnaudin, where both PS and Pnarr are frequent. Stories in Arnaudin’s collection share two strong features with those in OOT: the use of Pnarr as major tense and the numbers of stories where we find both the PS and the Pnarr. We shall return to the implications of this for questions relating to orality alongside our detailed analysis of tense-switching in Section 4.4.

Finally, we note that there is no use of the PC as a narrative tense in OWT, despite evidence in the metadata that four storytellers/stories and one of the authors (Arnaudin) are from the Gascon maritime area. It is possible, given the period in question (that is, collection in the late-nineteenth century), that some of these speakers may have used the PC as a narrative tense—we have no way of knowing, since there are no recordings. Lassalle (2017:83) notes that Arnaudin must have been aware of the loss of the PS in this area; however, he clearly opted for the normative PS in the published versions, alongside the Pnarr, the latter reflecting their oral sources.       

3.6 Interim Conclusions: “Major Tense” Patterns in Relation to Orality

The major tense patterns outlined in Sections 3.2 through 3.5 would suggest that factors relating to transmission, sources, and performance context play a crucial role, alongside wider societal factors relating to orality and literacy. The two subcorpora with the clearest patterning are OOT and OWT. The stories in OOT are the most “oral” in the full corpus, with oral transmission, sources, and belonging firmly to an oral tradition. High use of Pnarr and high numbers of stories with more than one narrative tense are in evidence, both of which are strongly associated with oral narrative. Indeed, the storytellers in OOT are unlikely to have been influenced by written forms of the stories; they are native speakers of Occitan (who may or may not also have French) in a context of oralité mixte, that is, they have acquired their stories through an oral tradition, even if the society in which they live has developed a writing system in both languages. Not surprisingly, therefore, there is very little evidence of influence from French, with the possible exception of OOT_3, although it is not impossible that use of the PC by OOT_3 is a case of diatopic variation. OWT, on the other hand, as the only published collection in the corpus, shows much more standard use of Occitan tenses, with frequent use of PS as the narrative tense of choice and higher numbers of stories with only one narrative tense. Use of the Pnarr in OWT is likely to be an indicator of the oral sources of these stories: we will explore the possible link with tense-switching in Section 4.4.

Both FOC and OOC are more complex and more hybrid. Both involve contemporary oral performance, FOC drawing on written sources and OOC predominantly written but sometimes oral sources. Both are more formal and staged than OOT, particularly FOC, since it has no direct


connection with an oral tradition. The major tense patterning in FOC is dominated by the narrative tenses found in oral French, PC especially and also Pnarr, but there is nonetheless a presence of the more formal, literary PS, probably due to the prominence of PS as the narrative tense in source texts and to the strong inculcation of a written literary norm in some storytellers. Tense-switching is attested, but it is striking that most stories have either one sole narrative tense or have one narrative tense that dominates. There is evidence of the influence of language contact in the use of the PC in certain instances in OOC, both systematically (by OOC_PV) and occasionally (by OOC_MB). OOC is the corpus with the most individual variation, perhaps a reflection of the mix of influences found: some storytellers appear to be closer to the oral tradition in both sources and style (and thus perhaps more likely to tense-switch), while others are closer to the new storytelling patterns found in FOC in the sense that stories are dominated by the past tense form (PS in Occitan, PC in French), with some switching to Pnarr. We shall return to detailed questions relating to the frequency of tense-switching in OOC in Section 4.3.

4. Tense-Switching in Occitan and French

In this section, we will discuss the frequency and rates as well as the functions of tense-switching across the four subcorpora, working from the hypothesis that these are likely to be related to the different degrees of orality but taking into account other possible influential factors.

4.1 Overview

Tense-switching is attested in the four subcorpora in variable proportions:

Percentage of narratives containing tense-switching (texts containing tense-switching /number of texts)84.6% (22/26)61.5% (8/13)61.5% (8/13)52.6% (10/19)
Average rate of tense-switching relative to number of narrative clauses containing verbs (narratives containing tense-shifting only)1 switch per 4.1 clauses1 switch per 6.7 clauses1 switch per 9.7 clauses1 switch per 3.6 clauses
Range of tense-switching rate across corpus (i.e., 1 switch per X clauses—narratives containing tense-shifting only)X= between 2 and 13X= between 4 and 50X= between 4 and 55X = between 2 and 9
Table 7. Tense-switching in the four subcorpora.
TypePercentage of narratives containing tense-switching (text containing tense-switching / number of texts)Average rate of tense-switching relative to number of narrative clauses containing verbs (narratives containing tense-switching only)
Conte d’animaux80% (8/10)1 switch per 4.8 clauses
Conte de l’ogre dupé75% (3/4)1 switch per 3.8 clauses
Conte facétieux70% (7/10)1 switch per 4.3 clauses
Conte formulaire75% (3/4)1 switch per 3.9 clauses
Conte merveilleux69% (11/16)1 switch per 5.4 clauses
Total Conte73% (32/44)1 switch per 4.8 clauses
Récit contemporain100% (4/4)1 switch per 9.3 clauses


Récit facétieux100% (5/5)1 switch per 9.6 clauses
Récit légendaire58% (7/12)1 switch per 4.8 clauses
Total Récit76% (16/21)1 switch per 7.6 clauses
Table 8. Tense-switching and story-type.21

Table 7 shows the percentage of narratives in each of the subcorpora containing tense-switching (on the top row); in the instances of tense-switching, the average rate of tense-switching across the stories (in the second row); and finally, the range of tense-switching rates (in the bottom row). Table 8 shows the percentage of narratives containing tense-switching in the major story-types and the average rate of tense-switching across these stories.

It is important to note at the outset of this discussion that initial figures suggest that story-type is not significant: the four major categories of conte in the corpus (contes d’animaux, de l’ogre dupé, facétieux, and merveilleux) all have a similar percentage of stories with tense-switching (between 69% and 80%), and there is no noteworthy difference between contes and récits (see Section 2, above) at 73% and 76%, respectively. There are some differences in the rates of tense-switching, with contes showing one shift every 4.8 clauses and récits one every 7.6 clauses. Although the difference is not strikingly large, we will nonetheless keep this in mind in the discussion below. We note also that within the contes category, while there are some differences between the story-types (ranging from changes every 3.8 clauses for the ogre dupé category to every 5.4 for the contes merveilleux), there is not enough evidence to suggest major differences, and in any case, the numbers are too small to draw conclusions. Within the récit category, the lower percentage of récits légendaires with tense-switching may be linked to the fact that they tend to be shorter, while the higher rate is attributable to the influence of one particular subcorpus, OOT, where there are relatively high numbers of récits légendaires, as we shall see in Section 4.2.

In Sections 4.2 through 4.5, we will discuss the implications of these figures for the link between tense-switching and degrees of orality. The functions and textual effects of tense-switching across the four subcorpora are broadly familiar from previous research and echo those discussed in Section 1, above. In our analysis of each of the subcorpora, we will therefore focus on any striking tendencies in terms of the function of tense-switching in particular subcorpora and any links to the degree of orality in question.

4.2 OOT

Perhaps most striking is that the percentage of stories in OOT where we find tense-switching (84.6%) is considerably higher than in any other subcorpus. In fact, only four stories have no tense-switching, and these are all of short-length. Where tense-switching is attested, the rate is very high, averaging one switch every 4.1 clauses, with the lowest possible rates of tense-switching still relatively frequent, at one switch every thirteen clauses. Indeed, as shown in Table 2, above, the proportions of each tense reach 50/50 in three stories. In other words, most stories

21 Conte d’animaux = Animal Tale; conte de l’ogre dupé = Tale of the Stupid Ogre; conte facétieux = Anecdote or Joke; conte formulaire = Formula Tale; conte merveilleux = Tale of Magic; récit contemporain = contemporary story; récit facétieux = anecdote (in the broadest sense); récit légendaire = legend.


contain tense-switching, and within these stories, the rate is high. Again, while we might point to a small difference between contes and récits (15/16 contes contain tense-switching, and the rate is one switch every 4.1 clauses, while 7/10 récits contain tense-switching, and the rate is one every 4.4 clauses), this is not sufficiently large to be significant. The key finding with regard to OOT is that switching frequency and rates are strongly correlated with a high degree of orality, since OOT is the most strongly oral of our subcorpora in terms of transmission, sources, and performance tradition.

All familiar functions of tense-switching mentioned in Section 2, above, are attested, although it is important to point out that, given the highly spontaneous and relatively informal nature of OOT, the patterns are not always neat and clear-cut. One striking feature, perhaps not surprising given the high rates, is the sustained use of switching throughout the story, or at least throughout the Complicating Action (the central part of the narrative in terms of events; Labov and Waletzky 1967). A second striking feature of this subcorpus is the use of tense-switching in parallelisms across episodes and sub-episodes. Parallel episodes are a well known feature of stories in the oral tradition (Guézennec 2010); these are sequences, usually three or more, where a broadly similar series of events is repeated but with variations, either each time and/or where the final series results in a different outcome than the previous series. In our corpus, they are particularly frequent in the three Occitan corpora which are more strongly attached to an oral tradition than FOC, even if OWT involves published versions.

In Example 1, where the story is recounted largely in the Pnarr, a king sets a series of challenges to a group of suitors who are seeking his daughter’s hand in marriage. Only one suitor, Josèp, passes the first test (1a) and is then required to undertake two more tests (1b and 1c), constituting three episodes (a, b, and c). The beginning of each episode is marked by the use of the PS, first with the verb dire and then for a significant event where Josèp thinks (pensar) of how he might solve the challenge set by the king. In the first two cases (a and b), this act is accompanied by the adverbial tot d’un còp (“suddenly”) and in the third (c) by the connector puis (“then”). Within each episode, there is then a return to the Pnarr for a series of broadly similar events, notably Josèp calling (apelar) the carp in episode (a), ants in episode (b), crows in episode (c), and then Josèp arriving (arribar):22

Example 1
a) e lo rei lor diguètPS [. . .] (and the king said to them)
e tot d’un còp Josèp pensètPS a la carpa/ diguètPS [. . .] / e la sonètPS /la carpa arribaPnarr / e li diguètPS [. . .] (all of a sudden Josèp thought of the carp / he said [. . .] / he called it / the carp arrives / and he said [. . .])

22 Our examples include only the narrative line, for expedience. Descriptions and direct speech are not included; their absence is marked by [. . .]. The clauses in PS are in black and bold, while the clauses in Pnarr are black (not bold). Translations in English are given in grey and in brackets, while summaries of events are reported in grey and in square brackets. Finally, the tenses of all verbs are indicated in subscript. Transcription of oral material in Occitan and French does not contain punctuation as such (as is standard practice), although in order to facilitate comprehension, we have used a forward slash in both the transcription and the translation to indicate a change of narrative clause (see Section 1.1). We have also used a forward slash in OWT in Example (4), which of course contains punctuation as it is a written text. Summary text in square brackets contains punctuation for ease of comprehension.


b) lo rei lidiguètPS [. . .] (the king said to him)
[Josèp starts the test but he fails]
tot d’un còp pensètPS als formilhs / e apèlaPnarr les formilhs / les formilhs arribanPnarr (all of a sudden he thought of the ants / and he calls the ants / the ants arrive)
[the ants solve the test and Josèp takes credit for it with the king]
c) lo rei diguètPS [. . .] (the king said)
[Josèp starts the test but fails]
puis pensètPS als gòrbs / e apèlaPnarr les gòrbs / les gòrbs arribanPnarr (then he thought of the   crows / and he calls the crows / the crows arrive)
[the crows solve the test, and Josèp takes credit for it with the king]

At the end of the story, the king’s decision that Josèp can marry his daughter is again introduced by dire in the PS, and the final significant act of marrying is also in the PS:

Example 2
lo rei li diguètPS [. . .] / (the king said to him) /
e voilà cossí Josèp se maridètPS ambe la filha del rei (and this is how Josèp married the king’s daughter)

There are at least three functions of tense-switching at work in this story. Most strikingly, structural parallelism in the story is mirrored by parallel use of tenses, with (almost) consistent use of the Pnarr and PS for particular events in the sub-episodes, the “almost” in this analysis reflecting the spontaneous nature of OOT. Second, although both PS and Pnarr have the capacity to highlight particular events (the PS through dimensionalization/detachment and the Pnarr through foregrounding; see Section 1, above), it is the PS that does so in this story, with the unmarked dominant tense being the Pnarr. Finally, the capacity of the PS to emphasize closure at the end of the story is also attested: it reinforces a sense of completion (see Section 1, above). Indeed, of the fifteen OOT texts with Pnarr/PS alternation where Pnarr is the dominant narrative tense, eight close the story with the PS.

4.3 FOC and OOC

FOC and OOC have the same percentage of stories containing tense-switching (61.5%), a figure situated in between the lower levels in OWT (see Section 4.4) and the higher levels in OOT. This position is not surprising, since we are dealing with two subcorpora where transmission is oral but sources are usually (though for OOC not always) written and with highly educated storytellers who are working in a context of oralité seconde. Both corpora also demonstrate a similar range of tense-switching rates where switching is found (4-50 in OOC and 4-55 in FOC), suggesting that there is considerable variation within each subcorpus. In FOC we cannot explore individual storyteller patterns, as there is only one story for each storyteller. However in OOC, where we have two or more stories from several storytellers, it is clear that the two storytellers who claim to use oral sources and to be close to the oral tradition (Storyteller_OOC_PB and Storyteller_OOC_PV) are those with the highest tense-switching rates


in all their stories, whether contes or récits. Their influence on overall figures is probably the reason why there is a considerable difference in the average rate of tense-switching across FOC and OOC where it is attested: it is higher for OOC (1 switch every 6.7 clauses) than for FOC (1 every 9.7), despite the fact that there are proportionally more contes in FOC than récits.23 In other words, for both FOC and OOC, the degree of orality is again the dominant factor in shaping the frequency of tense-switching, with considerable complexities related to the continuum between new and traditional storytelling in the case of OOC; factors such as récit versus conte are more variable and certainly secondary.24 

In terms of the function and effects of tense-switching, both FOC and OOC demonstrate a wide range of familiar functions, for example, foregrounding salient events, turbulence at narrative peaks, structural roles, and closure with the PC (FOC) or PS (OOC). Perhaps not surprisingly, OOC contains substantial numbers of parallelisms where tense-switching exhibits patterns that are similar to Example 1, above. This is particularly the case where storytellers are close to the oral tradition (for example, it occurs in three stories by Storyteller_OOC_PB), but it is also found with one storyteller who is in the “new storyteller” category (OOC_MB).

For example, in “Gabarròt” (Table 9), a fantastic tale recounted by Storyteller_OOC_PB, PS is the main narrative tense, and Pnarr is exploited for structural and foregrounding purposes. In this story, a king announces that he will give his daughter in marriage to whomever can build him a boat that can sail on land as well as at sea, and three brothers attempt the mission. Each of the three episodes is introduced by verbs of speech in the PS, and in the first two episodes, the early events (preparation of the bread, finding the old man at the crossroads, discussion with the old man to whom the young man declines to give a piece of his bread) are foregrounded through use of the Pnarr, with the remainder of the episode (his failure to build the boat) in the PS. The third episode is structured similarly up to the point where, instead of refusing to give the old man a piece of his bread, the young man gives him the bread. From that point onwards, the same events that lead to failure in the first two episodes (continuing his journey, arriving in a wood, knocking down a tree) lead to success in the final episode, where the sequence of events previously marked by the PS is marked by the Pnarr:

Episode 1Episode 2Episode 3Translations
Block 1diguètPS
demandètPS… faguètPS

diguèt (said)
siguèt (was)
demandèt (asked)
faguèt (did)
partiguèt (left)
trobèt (found)

23 See Table 8, which shows that rates are generally higher for contes than for récits.

24 For example, of the three storytellers in OOC who recount récits, two use tense-switching. OOC_MB (who is firmly in the “new storyteller” category) has a very low rate (one every 26.2 clauses), whereas for OOC_PV, who identifies strongly with the oral tradition, the rate is very high (one every 5.3 clauses).


Block 2aprèstaPnarr… tròbaPnarr… demandaPnarr… ditzPnarr… rapèlaPnarr… demandaPnarr… ditzPnarr… rapèlaPnarr… ditzPnarraprèstaPnarr… ditzPnarr… ditzPnarrditzPnarraprèsta (prepares) tròba (finds) demanda (asks) rapèla (calls back) ditz (says)
Block 3diguètPS… contunhètPSdiguètPS… donètPS… contunhètPSdonguètPS… diguètPSdiguèt (said) contunhèt (continued) donèt/donguèt (gave)
Block 4arribètPS… tombètPS assagètPS… demorètPS… diguètPS s’en tornètPSarribètPS… tombètPS se metètPS… demorètPS… comprenguètPS s’en tornètPSarribaPnarr… tòmbaPnarr… tòmbaPnarr… se viraPnarr… tòrnaPnarr… virarPnarr… montaPnarr… s’en vaiPnarrarribèt (arrived)/arriba (arrives) tombèt (fell)/tomba (falls) assagèt (tried) demorèt (remained) se metèt (started) comprenguèt (understood) s’en tornèt (went back) se vira (turns) tòrna vira (turns back) monta (goes up) s’en vai (leaves)
Table 9. Tense patterning in “Gabarròt” (OOC).

In “Gabarròt,” the transformational outcome of the story in the final episode is conveyed in the Pnarr rather than the PS of the previous episodes, foregrounding success; the tense switch to Pnarr is sustained across the final seven narrative clauses, marking the fact that the young man’s luck has changed such that he can head towards the king’s castle and claim the king’s daughter.

A second tendency in OOC and FOC which distinguishes them from OOT (where rapid tense-switching is widespread) is that we find more instances of “small blocks” of text exhibiting a tense shift, that is, shifts of tense for between one and four clauses. The surrounding text can involve smaller or larger sections of text in another tense, and this variability can be seen in the use of the PS in blocks 1 and 3 in Table 9, above, from OOC. Additionally, in FOC, there are more instances of one or two isolated shifts in a longer chunk of text where the narrative tense is stable. For instance, in Example 3, below, the story is reported in the PC,25 and only one clause is foregrounded at the peak of the story (when the sorcerer picks out Pokou’s child as the one who is going to be sacrificed) with a switch to Pnarr:

Example 3
puis le sorcier a écartéPC tout ça / il a regardéPC Pokou / il l’a montréePC du  doigt / il lui a ditPC c’est cela que nous avons de plus cher / et il désignePnarr l’enfant dans son dos / alors elle a dénoué son boubou / elle a pris son fils dans ses bras / et elle s’est mise juste au-dessus (de la) du26 fleuve agité

25 Note that in stories where the narrative past tense is the PC, black and bold are used for these clauses. The Pnarr clauses remain in black only.

26 The storyteller self-corrects here from de la to du, but both are included in our literal transcription.


(then the sorcerer moved everything aside / he looked at Pokou / he pointed his finger at her / he said to her / that is the most precious thing we have / he points to the child on her back / so she undid her sling / she took her son in her arms / and she placed herself just above the fast-flowing river)

4.4 OWT

Relative to the other subcorpora, OWT contains fewer texts with tense-switching (52.6%), although where it occurs, the average rate is in fact higher than in OOT (one switch every 3.6 clauses), and indeed the lowest rate still indicates frequent tense-switching (one switch per nine clauses). This may be a reflection of the fact, on the one hand, that the stories in OWT were collected from oral performances (and therefore some may retain similar patterns to OOT), and, on the other, that OWT is nonetheless a written published collection destined for a reading public, and certain writers may opt to eliminate or reduce particular oral features (hence the lower percentage of stories with tense-switching). Indeed, two of the most represented writers, Arnaudin and Bladé, have clear individual preferences: all of Arnaudin’s stories (he is closest to the oral tradition) contain switches, whereas five of Bladé’s six stories have no switching, with only one switch in the remaining story. Caution is required in relation to the other authors because of the small quantity of stories, but Montel uses tense-switching in both of his texts, and Lalanne not at all in his one text. Two authors, Lambert and Séré, have more mixed patterns, and there is no obvious rationale for the choice in terms of story-type or length. As with the other subcorpora, the question of contes versus récits is not decisive: one out of two récits contain tense-switching, and nine out of seventeen contes, but the récit in question has a rate of switching that is higher than the average for all contes, because the author in question, Arnaudin, opts consistently for patterns that reflect the oral origins of the stories. In short, patterns appear to be individual, with some close to strongly written norms and others closer to OOT, most especially Arnaudin.

All tense-switching functions are attested in OWT. However, two tendencies stand out. The first, already hinted at, is that Arnaudin’s stories have distinctly oral patterns, with rapid switching for a wide variety of functions (turbulence at narrative peaks, structural functions, and so forth). The second is that across OWT, while there are many instances of blocks of one to four clauses, there is a higher number of longer blocks than in the stories in FOC and OOC (where blocks of one to four clauses dominate), and there are far fewer individual tense switches. The blocks in OWT can be found in a variety of contexts, including peaks or sections of the story where we can have, for example, a run of several narrative clauses in the Pnarr with surrounding text in the PS. For instance, the story entitled “Lo Gojat” (where the main tense is the PS), can be divided into nine blocks, with three of them containing seven to fourteen clauses, the longest of which (the peak of the story) is in the Pnarr:


BlockTenseNumber of verbsSummary of events in blocks
1PS1[A man said to his wife that he was going to do some business at the market.]
2Pnarr3[He sets off, arrives at the greengrocer’s stall, and has a conversation with the owner.]
3PS7[During the conversation, he explained that he wanted to buy a foal,28 having seen the lovely foals that his neighbor apparently bought there. The greengrocer told him that he had run out of foals but could sell him a mare’s egg which was ready to hatch, but that the man would need to be careful not to drop it on the way home. He chose the nicest pumpkin and bought it from the greengrocer.]
4Pnarr1[He sets off on the road home.]
5PS8[The pumpkin fell, split in pieces, and in the process, hit a hare and caused it to run away. The man thought it was the foal that had escaped from the egg.]
6Pnarr14(peak)[The man goes back again to buy another egg. The greengrocer can only give him an egg that will require him to brood without speaking (otherwise the egg will be lost). He goes back home and gets into bed to brood, without saying a word.]
7PS4[The man’s wife, who was starting to get worried, called in the neighbors and brought them up to the bedroom.]
8Pnarr2[To make the man react, one neighbor goes up to the man’s wife, strokes her cheek, and gives her kisses.]
9PS2[The man ended up reacting, speaking, and realizing the egg will not hatch. The story ends with the man blaming his neighbor and his wife realizing her husband is a fool.]
Table 10. Tense patterning in “Lo Gojat” (OWT).27

This mixture of rapid tense-switching (where it occurs) and longer blocks of text is most likely a reflection of the oral sources, on the one hand, and the fact that these are written published texts, on the other. Individual variation across authors, as noted above, is also likely to be a major influential factor. 

4.5 Interim Conclusions

High levels of tense-switching—both in terms of the volume of stories and the rate of switching—are strongly correlated with higher degrees of orality. Our most “oral” subcorpus (namely, OOT), where both transmission and sources are oral and form part of an oral tradition, shows the highest levels of tense-switching by every measure. OOC resembles FOC in many respects in terms of frequency, but it is clear that some storytellers who consider themselves connected to an oral tradition demonstrate patterns that are closer to OOT. OWT shows different patterns for different writers, with some containing little or no tense-switching (and thus conforming to written norms) and others showing patterns that have features in common with OOT, reflecting perhaps a wish on certain authors’ parts to convey the oral nature of the stories. The functions and effects of tense-switching are familiar from previous research, and all types of function can occur in the different subcorpora, with some discernible patterns in particular subcorpora. One function stands out as characteristic of the conte in the oral tradition: the use of switching to structure parallel episodes in ways that highlight crucial similarities and differences across parallel episodes. Story-type and conte versus récit status do not appear to strongly influence patterns. By far the most important factor in determining frequency and rates of tense-

27 This story is built on a farce where the main character is duped by his neighbors and the greengrocer.


switching is the degree of orality in question, with OOT, the most strongly oral subcorpus by all measures (transmission, sources, performance tradition), exhibiting the highest levels, and OWT, the least oral subcorpus, exhibiting the lowest.

5. Apparent Contradictions and Complexities

As noted in Section 2, above, in this type of corpus, where there are high levels of variation for a variety of reasons, data often do not show entirely clear-cut results. As our analysis above suggests, we are dealing much of the time with tendencies and patterns. Some elements of the analysis can even seem contradictory, pointing up the complexity of the interaction of different factors in some instances.

For example, OWT generally has low levels of tense-switching and therefore fewer instances of Pnarr, whereas OOC tends to have a certain level of tense-switching and therefore relatively frequent use of Pnarr, albeit in the context of highly individualized patterns. In each of these subcorpora, we can identify one writer (in OWT) and one storyteller (in OOC) where the patterns contradict the general tendency: the stories published by Arnaudin (OWT) demonstrate a relatively high level of tense-switching between PS and Pnarr, whereas the stories recounted by Storyteller_OOC_MC contain no tense-switching and therefore no Pnarr. We know from discussion with OOC_MC that she draws on Arnaudin’s published stories as a major source of her narratives and that, in fact, she has drawn on a story published in the volume we have used to construct OWT, thus allowing us to make a direct comparison between the same narrative in each of two subcorpora. The similarities between the two versions are indeed striking with regard to the choice of verbs, tenses, vocabulary, structure, and frames (for example, un matin): this is clear both in Example 4, which gives a short passage from the beginning of the story from each storyteller, and in Table 11, which gives the tense patterning across the whole story in each version:

Example 4
un matin la goja que devóPS anar tirar aiga mei de d’òra que de costuma / e quan arribèPS a la hont dab la soa pinga / que trobèPS la hada lo peu tot desligat asseduda au bòrd de la hont (OOC)
Un matin, que calotPS anar tirar l’aiga mei a bona ora que de costuma, / e quan arribètPS dab sas pingas, / que trobètPS la hada seitada au bòrd de le hont, lo peu tot desligat / (OWT)
(one morning the servant was supposed to go and fetch water earlier than usual / and when she arrived at the fountain with her pail(s) / she found the fairy with her hair undone sitting at the edge of the fountain)


TenseOOC version (MC)OWT version (Arnaudin)English translation
PS in OOC/
que devóPS
arribèPS… que
trobèPS… que
PS… que
PS… que li
que calotPS
PS… que
PS… que la
PS… que
PS… qui li
devó/calot (had to)
arribè/arribèt (arrived)
trobè/trobèt (found)
saludè/saludèt (greeted)
digó/dichot (said)
demandè/demandèt (asked)
contè/condèt (told)
PS in OOC/
Pnarr in OWT
e que se
e que s’en vanPnarr
anèn (went) / van (go)
orbín (opened) / aubrissen (open)
Table 11. Comparison of the same story, “La Hada e la Goja,” between OOC and OWT.

However, in spite of the fact that there is clear tense-switching in Arnaudin’s printed version (that is, the last two verbs appear in Pnarr), there is absolutely no tense-switching in Storyteller_OOC_MC’s version of the story, with all verbs in PS. We cannot be certain of the rationale behind these patterns. It is possible, though, that on the one hand, Arnaudin’s texts are more faithful to the oral versions of the stories than either Bladé’s or Lambert’s (hence the switching), while on the other hand, OOC_MC’s version shows a high degree of normalization, perhaps due to her own training as an Occitan language teacher and her background as an educated French speaker in a context of oralité seconde. Indeed, all three of her stories in our OOC subcorpus follow the same pattern of PS as the major tense with no switching. This example of complexity demonstrates two things. First, patterns more typical of written published texts can be found in oral versions and vice versa. Second, the relationship between a written source and an oral performance, even where we know that a particular storyteller has drawn on a particular written source, is not always direct: many complex factors impact linguistic and stylistic choices.

Towards Conclusions and Future Research

Our analysis builds on a substantial body of work on temporality in medieval texts, conversational oral narrative, and performed oral storytelling in contemporary French. Our findings underscore not only the well established connection between tense-switching and orality but also the discourse functions that tense-switching can assume in context. However, we have sought to move beyond this, primarily through embedding a scalar approach to orality and building a corpus that has allowed us to analyze tense usage and tense-switching in relation to what we have termed different “degrees of orality.” In the context of storytelling in Occitan, the main factors we have considered concern the channel of transmission, the sources of the stories, and the nature of the performance (as part of an oral tradition involving a relatively intimate context or as a more staged public performance). Our corpus-based analysis, including a comparison with a contemporary French corpus, shows clearly that both tense patterning on the narrative line and the use of tense-switching are directly related to the parameters involved in these different degrees of orality: the greater the degree of orality in terms of transmission, sources, and performance tradition, the higher the use of the narrative present as a main narrative tense on the narrative line and the greater the extent of tense-switching as measured by the


combination of frequency across each subcorpus and rate of switching per number of narrative clauses. Moreover, in addition to familiar functions of tense-switching, our data show different types of distribution across the subcorpora (for example, blocks of discourse versus one-off switches) and reveal a particular structural usage in the context of parallelisms in the oral tradition. In terms of future research, there is more to be done around the notion of “performance” in relation to temporality. Previous research suggests that higher levels of performance markers (for example, higher quantities of direct speech, sound effects, asides, gestures, and so forth) tend to raise the levels of tense-switching (Wolfson 1982); comparative work across different degrees of orality could shed new light on this issue. 

Notwithstanding the discernible patterns, the context of a minoritized language such as Occitan, alongside what Foley has termed “the natural diversity of human expression” (2002:39; see Section 2.1, above), unquestionably introduces complexities and even apparent contradictions into our findings. These include questions relating to language contact (for example, there is some evidence of the influence of French on Occitan tense patterns in OOC and possibly in OOT), differences between “new storytellers” and those self-identifying as closely linked to an oral tradition (within OOC), and, crucially, individual preferences on the part of writers or storytellers (this is particularly clear in OWT but also in OOC). Additionally there is one possible but unconfirmed case of diatopic variation in OOT. Although we have included a discussion of story-type at several points in this article, there is not enough evidence of clear patterning to draw any substantial conclusions in this regard, and the nature of the data makes it difficult to obtain a balanced corpus for an investigation of this nature. A larger corpus involving more storytellers/writers and more stories from each storyteller would facilitate further exploration of some of these complexities, particularly the question of individual variation.

In short, our findings suggest that any investigation of the relation between temporal phenomena and orality can be greatly enriched and nuanced by integrating a scalar approach not only into theoretical discussion but also into corpus building and analysis. In this instance, our object of study is a minoritized language which operates alongside a powerful national language; the precise way in which a scalar approach could be applied elsewhere will vary according to both context and language. It is hoped that our approach could serve to underpin analyses of temporality, or indeed of other linguistic features, in other minoritized languages, other language contact situations, or other types of discourse that require a theoretical framework and a corpus that move beyond an oral/written divide.

Queen’s University Belfast (Janice Carruthers)
University of Poitiers (Marianne Vergez-Couret)


Aarne and Thompson 1961           
Antti Aarne and Stith Thompson (trans. and ed.). The Types of the Folk-Tale: A Classification and Bibliography. 2nd rev. ed. FF Communications, 184. Helsinki: Suomalainen Tiedeakatemia, Academia Scientiarum Fennica.


Asher and Lascarides 2003
Nicholas Asher and Alex Lascarides. Logics of Conversation. Cambridge: Cambridge University Press.

Bec 1995
Pierre Bec. La langue occitane. 6th ed. Paris: Que sais-je.

Biber 1988
Douglas Biber. Variation across Speech and Writing. Cambridge: Cambridge University Press.

Biber and Conrad 2009
Douglas Biber and Susan Conrad. Register, Genre and Style. Cambridge: Cambridge University Press.

Bres 2001
Jacques Bres. “De la textualité narrative en récit oral: L’enchaînement des propositions.” Revue québécoise de linguistique, 29.1:23-49.

Carruthers 2005
Janice Carruthers. Oral Narration in Modern French: A Linguistic Analysis of Temporal Patterns. Research Monographs in French Studies, 19. Oxford: Legenda.

Carruthers 2013
______. French Oral Narrative Corpus. Conservatoire Contemporain de Littérature Orale and Oxford Text Archive. http://frenchoralnarrative.qub.ac.uk

Carruthers 2018
______. “Oral Genres: Concepts and Complexities.” In Manual of Romance Sociolinguistics. Ed. by Wendy Ayres-Bennett and Janice Carruthers. Manuals of Romance Linguistics, 18. Berlin: De Gruyter. pp. 335-61.

Carruthers and Vergez-Couret
Janice Carruthers and Marianne Vergez-Couret. “Méthodologie pour la       _____2018___________________ constitution d’un corpus comparatif de narration orale en Occitan: Objectifs, ____________________________  défis, solutions.” Corpus, 18. http://journals.openedition.org/corpus/3490

Carruthers and Vergez-Couret
______. “Temporal Structures in Occitan and French Oral Narrative: The Role _____2021___________________  of Frames and Connectives.” Lingvisticae Investigationes, 44.1:1-36.

Finnegan 1977
Ruth H. Finnegan. Oral Poetry: Its Nature, Significance, and Social Context. Cambridge: Cambridge University Press.

Finnegan 1988
______. Literacy and Orality: Studies in the Technology of Communication. Oxford: Blackwell.

Fleischman 1990
Suzanne Fleischman. Tense and Narrativity: From Medieval Performance to Modern Fiction. Austin: University of Texas Press.

Fludernik 2009
Monica Fludernik. An Introduction to Narratology. London: Routledge.


Foley 2002
John Miles Foley. How to Read an Oral Poem. Urbana: University of Illinois Press.

Gairal 2001
Sèrgi Gairal. Un estiu sus la talvera. Puylaurens, France: Institut d’Estudis Occitans.

Guézennec 2010
Nathalie Guézennec. “Parallélismes et contes oraux bas-bretons: Niveaux, rôles et fonctions.” In The Conte: Oral and Written Dynamics. Ed. by Janice Carruthers and Maeve McCusker. Bern: Peter Lang. pp. 261-87.

Koch and Oesterreicher 2001
Peter Koch and Wulf Oesterreicher. “Gesprochene Sprache und geschriebene Sprache (Langage parlé et langage écrit).” In Lexikon der romanistischen Linguistik, I.2: Méthodologie (Sprache in der Gesellschaft/Sprache und Klassifikation/Datensammlung und -verarbeitung). Ed. by Günter Holtus, Michael Metzeltin, and Christian Schmitt. Tübingen: Niemeyer. pp. 584-627.

Labov 1997
William Labov. “Some Further Steps in Narrative Analysis.” Journal of Narrative and Life History, 7.1-4:395-415.

Labov and Waletzky 1967
William Labov and Joshua Waletzky. “Narrative Analysis: Oral Versions of Personal Experience.” In Essays on the Verbal and Visual Arts: Proceedings of the 1966 Annual Spring Meeting of the American Ethnological Society. Ed. by June Helm (McNeish). Seattle: University of Washington Press. pp. 12-44.

Lassalle 2017
Renaud Lassalle. Grammaire gasconne du parler de la Grande-Lande et du Born: Au travers des écrits de Félix Arnaudin. Colleccion Occitania e Gasconha de oei, 68. Cressé: Editions des Régionalismes.

Martin 1971
Robert Martin. Temps et aspect: Essai sur l’emploi des temps narratifs en moyen français. Bibliothèque française et romane, Série A: Manuels et études linguistiques, 20.Paris: Klincksieck.

Mercadier 2013
Florant Mercadier. Les contes du placard. Sévérac-le-Château, France: Sirventés.

Moeschler 2000
Jacques Moeschler. “L’ordre temporel dans le discours: Le modèle des inférences directionnelles.” In Passé et parfait. Ed. by Anne Carlier, Véronique Lagae, and Céline Benninger. Cahiers Chronos, 6. Leiden: Brill. pp. 1-11.

Ofici Public de la Lenga Occitana 2020
Ofici Public de la Lenga Occitana. Resultats de l’enquèsta sociolingüistica sus la practica e las representacions de la lenga occitana en Novèla-Aquitània, Occitània e Val d’Aran. https://www.ofici-occitan.eu/oc/restitucion-de-las-resultas-de-lenquesta-sociolinguistica/


Ollier 1978
Marie-Louise Ollier. “Le présent du récit: Temporalité et roman en vers.” Langue française, 40:99-112.

Ronjat 1930
Jules Ronjat. Grammaire istorique [sic] des parlers provençaux modernes. Vol. 1. Montpellier: Société des Langues Romanes.

Sibille 2010
Jean Sibille. “Les langues autochtones de France métropolitaine: Pratiques et savoirs.” In Autour du mot: Pratiques et compétences. Séminaire du Centre du français moderne, Tome II, 2006-2009.Ed. by Claude Gruaz and Christine Jacquet-Pfau. Limoges: Lambert-Lucas. pp. 69-85.

Sibille 2015
______. Description de l’occitan parlé à Sénaillac-Lauzès (Lot) et dans les communes voisines. Limoges: Lambert-Lucas.

Smith 2003
Carlota S. Smith. Modes of Discourse: The Local Structure of Texts. Cambridge Studies in Linguistics, 103. Cambridge: Cambridge University Press.

Teulat 2008
Robert Teulat. Éléments de syntaxe occitane référentielle. Cressé, France: Edition des régionalismes.

Vaugelas 1647
Claude Favre de Vaugelas. Remarques sur la langue françoise: Utiles à ceux qui veulent bien parler et bien escrire. Paris: La Veuve Jean Camusat et Pierre le Petit.

Vergez-Couret and Carruthers 2018
Marianne Vergez-Couret and Janice Carruthers. OcOr: A Corpus of Occitan Oral Narratives(Version 0.0.1). Zenodo. http://doi.org/10.5281/zenodo.1451753

Waugh 1987
Linda R. Waugh. “Marking Time with the Passé Composé: Toward a Theory of the Perfect.” Lingvisticae Investigationes, 11.1:1-47.

Waugh and Monville-Burston 1986
Linda R. Waugh and Monique Monville-Burston. “Aspect and Discourse Function: The French Simple Past in Newspaper Usage.” Language,62.4:846-77.

Wolfson 1982
Nessa Wolfson. The Conversational Historical Present in American English Narrative. Dordrecht: Foris.

Zumthor 1983
Paul Zumthor. Introduction à la poésie orale. Paris: Seuil.


This page is intentionally left blank.