DOI 10.24053/ VOX-2025-008 Vox Romanica 84 (2025): 183-198 Deconstructing a misconception about the definite article derived from ipse in Pre-Literary Catalan* Camilla Talfani (ILCC-Universitat de Girona) https: / / orcid.org/ 0000-0001-8370-7762 Résumé: Ce travail se propose de revenir sur la question épineuse de la présence d’attestations du déterminant issu du démonstratif latin ipse dans le catalan pré-littéraire, c’est-à-dire dans les premiers textes conservés en langue catalane. Une occurrence de cet article a été identifiée par Joan Coromines dans les Homilies d’Organyà , généralement considéré comme le plus ancien texte littéraire en catalan. Les éditeurs ultérieurs de ce texte, ainsi que de nombreux spécialistes de la langue catalane ancienne, auteurs de grammaires ou d’histoires de la langue, reproduisent cette information de Coromines dans leurs travaux, ce qui a contribué à renforcer l’idée qu’à l’époque ancienne le déterminant issu de ipse était beaucoup plus diffusé dans la langue catalane écrite et orale, contrairement à sa distribution géographique moderne. Ensuite, sa disparition progressive serait due notamment à une connotation négative du point de vue diastratique. Néanmoins, nos recherches récentes permettent de mettre en lumière et de proposer une interprétation alternative pour l’occurrence supposée de l’article dérivé de ipse dans les Homilies d’Organyà , et de démystifier ainsi une croyance qui était ancrée dans l’histoire de la langue catalane depuis environ trois quarts de siècle. Keywords: ipse -derived Definite Article, Pre-Literary Catalan, History of the Old Catalan language, Marked Medieval Syntax, Linguistic Philology 1. The definite article derived from ipse It is well known that some Romance varieties have developed a less common form of the definite article, i.e. the es / so , sa type. In the current Romance linguistic landscape, the es / so , sa article is primarily acknowledged as a peculiar characteristic of Sardinian and Balearic Catalan; however, it is also sporadically found in other minor Romance dialects, albeit in very circumscribed areas. Regarding its origins, we can rely on the rich and detailed contribution by Rosanna Sornicola (2011) 1 , which traces the process of grammaticalization from Latin demonstratives to articles, and provides an exhaustive overview comprising all Ro- * This paper is the result of the research activity I carried out as a post-doctoral researcher at ILCC- Universitat de Girona with funding from the «Juan de la Cierva» of the Ministerio de Ciencia, Innovación y Universidades . 184 Camilla Talfani DOI 10.24053/ VOX-2025-008 Vox Romanica 84 (2025): 183-198 mance languages. The etymology underlying this variant has long been debated, and numerous scholars have taken part in the querelle . Without dwelling at length on the matter, which is much discussed elsewhere, it is particularly worth recalling Meyer-Lübke (M-L, RG / 2: §98), who explained the genesis of the forms so , sa with definite article value through a fusing of the Latin demonstrative hoc and the particle ecce , a development that parallels the one seen in demonstrative pronouns of the type It. ciò , Cat. això , Occ. açò . Meyer-Lübke’s proposal immediately achieved widespread acceptance, so much so that it was adopted by a large portion of experts in Romance philology and even by the FEW . The latter, however, also favoured for the possibility of tracing the variant back to the base ipse (cf. FEW IV, 811a), which has finally been unanimously accepted by scholars. 2. The article derived from ipse in Old Catalan Regarding the linguistic domain of Catalan, which is the focus of the present work, the typically Balearic Islands article es / so, sa , commonly called «article salat», is currently also in use in some dialects of the Costa Brava, more precisely in the area encompassed between Blanes and Begur, and in the continental part between the rivers Tordera and Ter, extending, to a very limited extent, even to the comarca of Empordà (up to Cadaqués, Alt Empordà) and Selva. Despite a very limited distribution in the contemporary era, it is commonly assumed that, at the earliest stages of Catalan, the ipse -derived article was frequent in a far more extensive area. Subsequently, its attestations in written documentation are believed to have diminished considerably around the 12 th -13 th centuries, almost to the point of extinction. This phenomenon has been attributed primarily to the negative connotation, from a diastratic perspective, that would have been associated with the es / so, sa variant, especially in written language. Furthermore, the prestige of the Occitan language and the influence it exerted on Catalan, particularly Literary Catalan, are thought to have contributed to the progressive replacement of this variant - supposedly deemed of lesser prestige - with the ordinary predominant form lo / le , la . Numerous studies, grammars, and histories of the Catalan language have unequivocally adhered to the hypothesis of a primitively widespread use, characteristic of Pre-Literary Catalan and not subordinated to a diatopic dimension, and its subsequent disappearance due to diastratic causes. They have reported and reiterated this without further demonstration or argumentation, to the extent that it has become a commonly accepted assumption. Given the impossibility of compiling a complete list of such publications, it is worthwhile to highlight at least the principal ones: Alomar i Melià 1999: 47, Badia i Margarit 1995: 444, 1999: 119, Bernat Baltrons 1 Without forgetting, of course, other fundamental contributions, like: Aebischer 1948, Renzi 1992, Ledgeway 2012. 185 DOI 10.24053/ VOX-2025-008 Vox Romanica 84 (2025): 183-198 Deconstructing a misconception about the definite article derived from ipse 2014: 2806-07, Blasco Ferrer 1984: 114, Griera 1917a: 56-57, 1931: 77-78, 1965: 77-78, Moll 1952: 197, 2006: 182, Moran 1984: 40-41, 2007: 129, Nadal/ Prats 1992: 284. This notion, however, appears to lack solidity, primarily because it relies heavily on the sedimentation of traces of the ipse -derived article within toponyms 2 , which, in my opinion, constitute very specific case studies and should therefore be examined and evaluated distinctly, rather than as tangible and certain proof of a generalised linguistic situation or a precise ancient period of the language. Secondly, the attestations of this variant in ancient and medieval documentation are inconsistent and do not suggest a unified or coherent boundary that would explain its presence or absence. This lack of a suitable basis precludes the reconstruction of a general overview and context, or the delineation of a concrete distribution map. On the other hand, in contrast to what has occurred for other Romance languages, a significant portion of Medieval Catalan documents, both of a legal-administrative and religious-literary nature, remain unpublished to this day, or have only been partially examined 3 . For these reasons, a more rigorous analysis and definition of this feature would require a comprehensive re-examination of the entirety of preserved, localizable, and datable documentation. In a very recent volume dedicated to Medieval Catalan, Joan Anton Rabella (2023: 88-89) has highlighted the problematic nature of the traditionally accepted hypothesis concerning the presence of the ipse -derived article in early witnesses. He noted how these occurrences have been evaluated by oversimplifying their complexity and, in a certain sense, reducing them to the commonplace notion that the forms es / so, sa were, from a certain period onwards, exclusively relegated to spoken and vernacular language, while the ordinary variant of the article was the only one admitted in written, literary, and chancery language - domains that would subsequently determine the imposition and supremacy of a single linguistic model. Moreover, Rabella (2009) had previously sought to temper the importance given by earlier studies to a diastratic approach, which had led to assigning a pejorative value to the forms es / so, sa . He re-examined the occurrences found in archival documents by applying a quantitative evaluation. This led him to a new interpretation of the data, demonstrating isolated attestations need to be disregarded and considering only those that emerge in the same document with a certain frequency and systematicity. Consequently, it was possible to highlight a diatopic variation in use 4 , which allowed for the conclusion that even in the language of the 11 th -12 th centuries, a clear 2 In the broad panorama of studies, cf., for example, Moran 2007, Bernat Baltrons 2014. 3 All existing works are based solely on edited ancient texts (11 th -13 th centuries), particularly by: Pujol i Tubau 1913, Miret i Sans 1915, Russell-Gebbett 1965, García Rubio 1979, Martí i Castell/ Moran 1986, Bastardas 1995, Moran/ Rabella 2001, 2004. Instead, as regards documents dating back to the 14 th -15 th centuries, most of them remain unpublished and, consequently, unexamined. 4 Cf. also the diatopic factor in the evaluation of the article in Garcia/ Martines (in press: 1166-70). In this regard, I sincerely thank the anonymous reviewers of my paper, who informed me of the forthcoming publication of this new edition of the Gramàtica del català antic , currently in press. 186 Camilla Talfani DOI 10.24053/ VOX-2025-008 Vox Romanica 84 (2025): 183-198 separation already emerged between the block of linguistic varieties pertaining to Eastern Catalan, where the two types of definite article are in competition, and that of Western Catalan, where the article salat is unknown. Finally, still with the aim of re-evaluating commonly accepted interpretations and theories regarding the article form es / so, sa , it is fundamental to recall the interesting work by Bernstein/ Ordóñez/ Roca (2021), which appropriately emphasizes and reflects on a specific connotation that these forms manifest in certain contexts. Their groundbreaking work helps us to propose new, more nuanced interpretations of its use. Indeed, their contribution demonstrates that the two types of articles in many cases co-occur in the same texts and even within a single nominal expression. This is particularly noteworthy and should be explained by hypothesizing the specialization of the definite ipse -derived article under specific syntactic and semantic conditions, meaning it is used as an alternative to the definite iLLe -derived article to reinforce the anaphoric and deictic value linked to referents previously introduced in the discourse 5 . The process of grammaticalization for the ipse -derived variant seems to have occurred incrementally, throughout the course of time. This gradual development is suggested by the fact that a considerable segment of documentation from the 11 th - 12 th centuries bears witness to the undifferentiated use of the dual article system, unconnected to precise semantic or syntactic conditions 6 . This suggests that, in a early stage of the language, both variants coincided in their uses. 3. The article from ipse in the Homilies d’Organyà An especially significant occurrence that has particularly drawn the interest of scholars was identified by Joan Coromines in the well-known Pre-Literary Catalan text, namely the Homilies d’Organyà . This constitutes a small, fragmentary collection of six - as well as the beginning of the seventh - homilies or sermons on the Lenten cycle, corresponding to the liturgical periods of Sexagesima, Quinquagesima, Ash Wednesday, and Lent, with notes on various Gospels and Epistles. The manuscript, presumably datable between the end of the 12 th and the beginning of the 13 th century, was discovered in the rectory of Organyà (Alt Urgell) by the bibliophile Joaquim Miret i Sans and is currently preserved at the Biblioteca de Catalunya (Bar- 5 Cf. especially p. 12-19 and previous research, also for the field of Catalan, in Bernstein/ Ordóñez/ Roca 2019. In the new edition of the Gramàtica del català antic , cf. the paragraphs concerning the anaphoric use of the ipse -derived determinant in Roca (in press: 1276-79 and 1302-04), whom I thank for allowing me to read his work before printing. 6 The observation reported here is based on the documents edited and analysed by Moran (1984: 36-40). 187 DOI 10.24053/ VOX-2025-008 Vox Romanica 84 (2025): 183-198 Deconstructing a misconception about the definite article derived from ipse celona, BC, 289) 7 . The Homilies do not actually constitute the oldest extant text written in the Catalan language; however, owing to their renown and the process of mythologization to which they have been subjected, they have ultimately become so in common perception 8 . The initial editors, Miret i Sans (1904, 1915), Griera (1917b), and Molho (1961), did not focus on the linguistic study of the text. It was only Coromines’s edition (1976/ 1) that offered the first linguistic analysis, drawing attention to the presence of the article salat (1976/ 1: 148). More precisely, in the passage: «aquesta femna gran fe ag, per que quan sa fila nomenà, e crezeg que de sola za paraula de .N. S. garia sa fila». In this textual segment, the form za would thus represent the feminine definite article, referring to the noun paraula , derived from the Latin demonstrative ipsA . In this regard, it also seems pertinent to add that this would constitute the sole occurrence of the ipse -derived article throughout the entire text, as no further evidence of it is discernible in the remaining eight folios of the manuscript. Merely a few years prior to Coromines’s edition, Russell-Gebbet (1965) had published several excerpts of Pre-Literary Catalan texts accompanied by a linguistic description. In the section dedicated to the definite article, he also noted the manifestation of the ipse -derived variant of the feminine article in the sixth and final homily of the Homilies d’Organyà . Thus, both scholars concurred and shared the interpretation of the form za in the Homilies , presenting their respective proposals in very close chronological proximity. Notably following Coromines, all specialists who have subsequently addressed the Homilies have accepted and supported this hypothesis, often without unduly concerning themselves with challenging or at least verifying it, given the illustrious origin of the assertion. First and foremost, Gökçen’s linguistic analysis of the Homilies (1977: 66), published only one year after Coromines’s edition (1976/ 1), reiterates the emergence of a single instance of the ipse -derived article and justifies this element with the usual presumption of a wider diffusion of the variant in the medieval period. Even the editors Soberanas/ Rossinyol (2001, 2004), authors of what is currently regarded as the reference edition for the text of the Homilies , unhesitatingly embrace the theory regarding the ipse -derived article, clarifying in a footnote that the identification of a single occurrence does not delegitimize the possibility of its manifestation (2001: 102-03). Confirmation of this interpretation for the form za is found in the translations of the text produced by Soberanas/ Rossinyol into modern Catalan («Nostre Senyor la guariria tan sols amb la paraula » 2004: 59), Spanish («Nuestro Señor la curaría solo con la palabra » 2004: 76), and English («Our Lord would help 7 The editors of the text are rather numerous; accordingly, here I only mention the main ones: Miret i Sans (1904, 1915), Griera (1917b), Molho (1961), Coromines (1976/ 1), Moran (1990), Soberanas/ Rossinyol (2001). Cf. also Moran (2003), which provides an overview of all editions published after 1990. 8 As evidenced, for example, by Moran/ Rabella (2004, 2007: 21-22), Gimeno Betí (2005: 90) and Rabella (2023: 39). 188 Camilla Talfani DOI 10.24053/ VOX-2025-008 Vox Romanica 84 (2025): 183-198 her by words alone» 2004: 92), as well as in popularizing translations, readily accessible online, which are responsible for conveying information to the general public 9 . Leaping forward to more recent years, it is necessary to revisit Rabella’s contribution on the diatopic variation of the article in Old Catalan (2009). Drawing upon Russell-Gebbet (1965), Rabella identifies the form za from the Homilies as the sole example of an ipse -derived article in texts later than the 12 th century for the entire Western Catalan domain (2009: 156-57). This information is still reported, with reference to Rabella, by the forthcoming Gramàtica del català antic (Garcia/ Martines in press: 1168) 10 . Finally, stepping back a little, it is imperative to recall perhaps the most significant instance of adherence to the hypothesis of Russell-Gebbet (1965) and Coromines (1976/ 1): the eminent history of the Catalan language by Nadal/ Prats, an indispensable international reference work. Indeed, Nadal/ Prats adopt the language of the Homilies as a benchmark for the evaluation and comprehension of the Pre- Literary Catalan, and they illustrate its principal features. Among these, the form za precisely stands out as an ipsA -derived variant for the definite article (1992: 284) 11 . In this manner, the news of the presence of the ipse -derived article in the Homilies has resonated in numerous works and has thus contributed to supporting and strengthening the traditionally accepted idea that the variant was widely documented in Pre-Literary Catalan. Despite this, it is worth reflecting on the common absence of the article salat in the origin area of the Homilies , namely the comarca of Urgell, located in the Western Catalan. For this reason, I checked the medieval documents from the same geographical area edited by Rasico (2006). In all analysed cases, dating from a chronological span between the 11 th and the 14 th centuries, the only traces of the Latin demonstrative ipse emerge exclusively in placenames, which, as already stated, constitute a specific field of investigation, while all occurrences of so, sa recorded in the texts unequivocally correspond to the atonic possessive adjective 12 . Furthermore, the situation of the phrase « za paraula» in the Homilies does not even conform to the anaphoric marking circumstances highlighted by Bernstein/ Ordóñez/ Roca (2021), as the element to which the form za refers does not appear previously, and, concurrently, the passage in question does not seem to be characterised by specific factors that require the accentuation of the deictic. 9 For example, URL: https: / / sites.google.com/ site/ magisterhumanitatis/ lenguas-romanicas/ las-homilias-de-organya-texto [13.10.2025]. 10 In this regard, I would like to thank the editors of the book for providing me with the chapters relevant to my research. 11 Even beyond the strictly Catalan sphere, the news of the ipse derived article in the Homilies has also been reported by scholars who have examined the variant in Occitan texts. Cf., for instance, Thomas (1974 [ 1 1925]: XXVI), Hoepffner (1926/ 1: 118) concerning the Canso de Santa Fe . It must be admitted that, precisely due to the prestige of the listed specialists, the author of the present article also alluded to the presence of the article salat in the Homilies , in an earlier publication (Talfani 2025: 53). However, the subsequent doubt fortunately gave rise to the present article. 12 Cf. the linguistic commentary of the texts (Rasico 2006: 290, 327-28, 389, 414-15, 441-42). 189 DOI 10.24053/ VOX-2025-008 Vox Romanica 84 (2025): 183-198 Deconstructing a misconception about the definite article derived from ipse 4. A new interpretation for the «za» in the Homilies d’Organyà While the possibility of an attestation of the definite article derived from ipse in the Homilies d’Organyà is of considerable interest, the value of za as an article, specifically within the context « za paraula de Nostre Seinor», seems dubious. Moreover, the uncertainty is further heightened by its solitary attestation within a text of a certain length, spanning eight folios, in which the remaining 94 occurrences of the feminine definite article show the standard form la . Therefore, I suggest that we are more likely dealing with a compound locution consisting of a redundant atonic possessive adjective ( za ), referring to the noun paraula , which anticipates the subsequent genitive complement «de Nostre Seinor». The formal aspect of the occurrence, specifically its graphic representation featuring the grapheme <z> instead of the more ordinary grapheme <s> 13 , does not invalidate the plausibility of our proposal. Firstly, it is crucial to recall that the landscape of graphic usages in the earliest stages of Catalan remains highly uncertain due to a lack of unequivocal information and, more importantly, an insufficient amount of data. Secondly, judging by the graphic habits of the scribe of the only witness of the Homilies , an abundant use of the grapheme <z> is observable in all positions and without a precise, unequivocal relationship to its Latin etymological base. Therefore, the notation in question appears to represent various phonetic realizations in the document: / s/ , / z/ , / ts/ 14 . Thirdly, the flexibility in graphic usages exhibited by the scribe responsible for the Homilies is also attested in other documents of the same period and seems to correspond to a generalised situation in the early phase of the language. The lack of specialization and distinction in graphic solutions for the outcomes of Latin bases s / c e / c i in the oldest documentation, particularly in an intervocalic position, has been observed, for instance, by Moll (2006: 55, 82, 87) and Rasico (2006: 170, 192-93). Undoubtedly, the best-known forms are the determinants derived from deictics reinforced by the Latin ecce 15 . For these, several equivalent graphic variants coexisted, such as a(i)cell / a(i)zell / a(i)çell , a(i)cella / a(i)zella / a(i)çella . As further proof of the oscillation between the graphemes <s> and <z> in the same etymological and phonetic contexts, I conducted multiple form searches in the CICA corpus, where competing formal variants alternating <c>/ <ç>/ <s>/ <z> for the results of Latin s / c e / c i can be readily identified. For example: (1) a. mer c è b. mer ç è c. mer s è d. mer z è 13 Cf. the strong (tonic) and weak (atonic) forms of possessives in the current reference historical grammar for Catalan, i.e. Batlle/ Martí i Castell/ Moran/ Rabella (2016). 14 Cf. the data provided by the linguistic commentaries of Gökçen (1977: 62-63) and Gimeno Betí (2005: 95). 15 Cf. the use of demonstrative cell , survived until the late Middle Ages, in Roca (in press: 1288-93). 190 Camilla Talfani DOI 10.24053/ VOX-2025-008 Vox Romanica 84 (2025): 183-198 (2) a. c elar b. s elar c. z elar In a number of cases, I also found graphic solutions such as ço / çon , ça for the atonic possessive, in locutions where it is unequivocal that an article derived from ipse is not involved. For example: (3) a. en ço cor b. tot çon poder c. demana çon depòsit (4) a. tota ça gent b. tota ça part c. pagar ça part d. en ça volentat Returning now to the occurrence of za in the Homilies and its origin, it should be noted that, while the place of compilation of the document presumably coincides with its discovery site - namely Organyà, in the comarca of Alt Urgell -, it has been ascertained that the text derives from an Occitan original, which was translated into Catalan by an anonymous individual 16 . For these reasons, the syntactic pleonasm I have highlighted could be explained on the basis of two alternative hypotheses. A first possibility would consist of a calque from the Occitan original; that is, the anonymous translator would have literally and faithfully transferred the Occitan locution into Catalan without altering its elements. The second solution, which I deem more plausible, is that the pleonastic construction can be directly linked to a common use in Medieval Catalan (cf. §6). I have also identified an analogous formulation within the Homilies themselves, though it is based on the strong form of the possessive preceded by an article: «Nostre Seiner posà la sua benedita mà sobre∙ls ulls del ceg». It should be noted how, in this passage, the possessive sua appears redundant with respect to the subject of the action. Some cases referring to very different circumstances, but which can nevertheless be related to our typology of constructions with redundant and unexpected possessives, are illustrated by Badia i Margarit (1995: 508) for contemporary oral language. I reproduce below Examples (10) and (12b): Va treure el seu llibre de la seva cartera No sabia fer el nus de la seva corbata Both formulations use of strong possessive forms, preceded by an article, in a pleonastic construction. Concurrently, the possessor is not explicitly stated by a genitive complement but is rather the subject of the sentence itself. Nevertheless, it appears 16 Cf. at least Bruguera (1986), Moran (1982, 1990, 1994), Benigne (1994-1995), Tavani (1994: 45-47). 191 DOI 10.24053/ VOX-2025-008 Vox Romanica 84 (2025): 183-198 Deconstructing a misconception about the definite article derived from ipse evident that the pleonastic use of the possessive precisely addresses the need to reinforce the subject involvement in the action or the possessive relationship (Badia i Margarit 1995: 507), as observed in the two cases from the Homilies . I conclude this analysis by referring to an interesting study by Pérez Saldanya/ Martines (2009) on the particular values assumed by the possessive in constructions where it is syntactically pleonastic. The aspects highlighted in their contribution appear to confirm my interpretative proposal for the phrase « za paraula de Nostre Seinor» in the Homilies . The two scholars reflect on three peculiar connotations of strong possessive forms preceded by an article. The latter two have particularly drawn our attention, as they are attributable to the purpose of highlighting the elevated status of the possessor or reinforcing the relationship between the possessor and the object of possession (Pérez Saldanya/ Martines 2009: 14). If we draw a parallel between the situations described here and the two passages from the Homilies , the status of the possessor subject is indeed undeniably elevated, as it refers to the Lord Jesus Christ: 1) « Nostre Seiner posà la sua benedita mà», 2) « za paraula de Nostre Seinor ». 5. The duplication of the possessive in Old Catalan Most Romance languages exhibit the particular syntactic construction at the heart of this work, which is composed of a pleonastic possessive adjective followed by an explicit genitive complement. While this feature was well-known in languages such as Old French and Old Italian, I have recently brought to light several cases involving Old Occitan as well (cf. §6). For the purposes of my investigations, I also deemed it useful and pertinent to consult historical grammars of the Catalan language, with the aim of finding references to any pleonastic use of the possessive, as well as in the other languages. However, I found no mention of it, for example, in the most recent historical grammar by Batlle/ Martí i Castell/ Moran/ Rabella, which, in the chapters dedicated to syntax (2016: 475-692), report on pleonastic uses characteristic of medieval marked syntax, yet contain no allusions to the situation under discussion. Following the submission of this paper for publication, I had the chance to learn from the anonymous reviewers about the forthcoming publication of a new edition of the Gramàtica del català antic . In this volume, specifically in the section dedicated to possessives, a paragraph is precisely devoted to the duplication of the possessive (Brucart/ Claveria in press: 1410-11). The two authors of this chapter, Brucart and Claveria, apply the concept of duplication to the cases where the possessive appears in tandem with a noun complement that is coreferential with it. This type of structure is found primarily in connection with personal names and is generally ex- 192 Camilla Talfani DOI 10.24053/ VOX-2025-008 Vox Romanica 84 (2025): 183-198 plained by the need to avoid the referential ambiguity inherent to the third-person possessive when it has a human referent. The examples they have collected demonstrate that this practice is attested in Medieval Catalan documentation from the earliest texts. Excluding now the examples already cited by Brucart/ Claveria (in press: 1410-11), for which the reader is referred to their study, I list here some examples completed with their date, which I have retrieved with the assistance of the CICA database 17 . (5) « sa fila d’en Clereçó » ( Llibre de Cort de Justícia d’Alcoi , 1263-1265) (6) «de son marit d’ela , en Bernat de Vals» ( Llibre de Cort de Justícia de Cocentaina ,1269-1290) (7) «a son dan de les sues possessions » ( Furs de València , 1325-1349) (8) «de son avi e de son pare de Corralí » (Bernat Desclot, Crònica , vol. II, end of the 13 th c.) (9) «havia sa fiyla de Don Lop d’Alvero » (Llibre dels fets del rei en Jaume , end of the 13 th c.) (10) « son nom d’aquel maestre » (Llibre dels fets del rei en Jaume , end of the 13 th c.) With reference to the last two cases from the Llibre dels fets del rei en Jaume , it is worth remembering that some traces of the article salat can be found in this text, as has been shown by Russell-Gebbet (1965: 44). His assertion is based on passages where the forms so , sa can certainly not be considered possessives, like «en sa mar», «en es dos». 6. The analogous case of Old Occitan Before concluding, it might be useful and pertinent to rapidly revisit here certain points from my previous research (Talfani 2025), which have allowed me to identify the same type of structure involving an expressed complement of specification and a pleonastic possessive adjective within the domain of Old Occitan. Also in this case, I was the first to put forth a novel interpretive hypothesis for the occurrences illustrated 18 . After re-examining the presumed attestations of the definite article so, sa , derived from the Latin demonstrative ipse , in the troubadour vidas and razos , the cases brought to light clearly demonstrated the necessity of somehow emphasizing possession or the relationship between characters. For instance, in locutions such as: «plazian li fort sas chansos d’en Bernat », «amava sa molher de son senhor », «mas sos amics d’el li deron». These elements, however, had previously been explained by earlier studies as remnants of an ancient Pyrenean article and, based on this theory, attributed to a supposed Gascon component within the manuscript tradition of these 17 I also thanked the reviewer for their generosity in pointing this out. 18 The syntax of Old Occitan by Frede Jensen dedicates a paragraph to the «emploi pléonastique du possessif» (1994: 120-21); however, scholars do not seem to have focused on the importance of this aspect, and this oversight prevented them from appropriately deciphering the attestations that were the subject of my paper. 193 DOI 10.24053/ VOX-2025-008 Vox Romanica 84 (2025): 183-198 Deconstructing a misconception about the definite article derived from ipse texts. A new analysis of the variant’s sporadic attestations in Old Occitan literature led me to reject this proposal. The collected data allowed for a redefinition of both the feature’s geographical distribution and its dating. This redefinition debunks the misconception that this article fell into premature disuse and was exclusive to the Gascon region (hence its «Pyrenean» designation). A number of its occurrences are, in fact, attributable to an area between the Alps of Upper Provence and the Maritime Alps, where the ipse -derived variant still survives in the modern language. To further confirm the legitimacy of interpreting the cases drawn from the vidas and razos as representative examples of pleonasm in medieval marked syntax, I extended my investigations to the entire troubadour corpus and demonstrated that the redundant use of the possessive is not an isolated phenomenon exclusive to the narrative stylistic register of biographies. On the contrary, it is also documented extensively in lyric poetry, among troubadours of diverse geographical origins and from different generations. The recorded examples primarily stem from the work of Bertran de Born, but also from Aimeric de Pegulhan, Elias de Barjols, Gaucelm Faidit, Guiraut de Salaignac, and the Monge de Montaudo. Outside the troubadour domain, yet remaining within the field of Old Occitan texts and manuscripts, I have identified numerous additional attestations. On this basis I have observed that the redundant employment of the possessive adjective, followed by an expressed complement of specification, characterizes particularly the formulary language typical of the legal register and, once again, the expository register of the biographical textual genre, in this second case especially hagiographies. Furthermore, I have ascertained at least three other particularly relevant factors to which constructions of this type are linked: 1) the pleonasm of the possessive is often clearly attributable to the intention of emphasizing a relationship of possession or between characters, or the agent of an action; 2) the redundant possessive can also precede or follow a relative clause; 3) the locutions in which pleonasm occurs often also feature terms belonging to the lexical families of words connoted as voluntat and conoisensa . Let us examine some examples for each typology. (11) a. «esser sa molher de son filh » b. « sa mainada d’ell » c. « sos amics d’el » d. « son engan de Raimun de Medollo » (12) a. « son anel que portava» b. « s’ amor qu’el a» c. « sa badia don era abatz» d. «amic de cui non ama son castic» (13) a. «seguet sa voluntat de son cors » b. «de far e dir sa volontat de sel » c. «tota sa conoizceuza d’en Gauter » 194 Camilla Talfani DOI 10.24053/ VOX-2025-008 Vox Romanica 84 (2025): 183-198 Moreover, it must be specified that the textual sources of the previous examples are localizable in various regions of the Occitan area and systematically exhibit the ordinary form of the definite article lo / le , la . Concurrently, forms featuring a mobile n in final position, such as those in examples 11d and 12a, corroborate the possessive adjective son as the sole possible interpretation, as they exclude a priori the hypothesis of an article. Finally, by undertaking a swift comparison among various medieval Romance syntaxes, I was able to ascertain elsewhere the legitimacy of the pleonastic use of the possessive adjective, which is reliably documented in both Old Italian and Old French and attributable to the extensive series of marked elements characteristic of medieval syntax. For these comparisons, the current reference grammars for the two aforementioned Romance languages were used: Ménard (1976: 37) and Buridant (2000: §126, §158, 2019: §141, §230-31) for French, and Salvi/ Renzi (2010/ I: 374-75) for Italian, respectively. 7. Conclusion I can conclude that the single occurrence of the form za in the Homilies d’Organyà cannot be considered the ipse -derived definite article called salat . This proposal is supported by two key factors: 1) the anomalous nature of its solitary attestation within a lengthy text spanning eight folios, which runs contrary to the expected frequency; 2) the Western Catalan origin of the Homilies , a dialectal area where the ipse -derived article is otherwise undocumented. My analysis, therefore, posits that this instance of za functions as a pleonastic possessive, which anticipates the subsequent genitive complement «de Nostre Seinor». This hypothesis is further substantiated by the fact that such syntactic constructions were not only possible but also widely attested in Catalan and other Romance languages. Moreover, as numerous scholars have demonstrated, a pleonastic use of possessive often served a deliberate pragmatic function, to emphasize the possessor’s involvement in an action or to highlight their elevated status. Building on the methodology developed in my prior work on Medieval Occitan, I intend to extend this line of investigation to a broader corpus of Medieval Catalan texts. My future research will specifically concentrate on two document typologies that have already proven particularly fruitful for Occitan: legal-administrative texts and hagiographies. The former will allow me to ascertain whether this construction is a recurrent feature of a specific formulaic language. The latter will provide a means to test the hypothesis that the duplication of possessive expressions is more frequent in cases involving high-status possessors, such as saints or divine figures. 195 DOI 10.24053/ VOX-2025-008 Vox Romanica 84 (2025): 183-198 Deconstructing a misconception about the definite article derived from ipse Bibliography A eBischer , P. 1948: «Contribution à la protohistoire des articles ‘ille’ et ‘ipse’ dans les langues romanes», CN 8: 181-203. A LoMAr i M eLià , A. 2002: «L’article ‘mallorquí’ a Catalunya», Línia directa 10: 7-11. B AdiA i M ArgArit , A. M. 1995: Gramàtica de la llengua catalana: descriptiva, normativa, diatòpica, diastràtica , Barcelona, Proa. B AdiA i M ArgArit , A. M. 1999: Les Regles de esquivar vocables i «la qüestió de la llengua» , Barcelona, IEC. B AstArdAs , J. 1995: La llengua catalana mil anys enrere , Barcelona, Curial. B AtLLe , M./ M Artí i c AsteLL , J./ M orAn , J./ r ABeLLA , J. A. 2016: Gramàtica històrica de la llengua catalana, Barcelona, Publicacions de l’Abadia de Montserrat. B enigne , M. 1994-1995: «La transmissió manuscrita del text de les Homilies d’Organyà», Urgellia 12: 415-24. B ernAt B ALtrons , F. 2014: «L’article salat a la toponímia de la Costa Brava: alguna cosa més que un dialectalisme? », in: J. t ort i d onAdA / M. M ontAgut i M ontAgut (ed.), Els noms en la vita quotidiana. Actes del XXIV Congrés Internacional d’ICOS sobre Ciències Onomàstiques , Publicació electrònica del Departament de Cultura de la Generalitat de Catalunya: 2806-15. URL: http: / / www.gencat.cat/ llengua/ BTPL/ ICOS2011/ 270.pdf [27/ 6/ 2025] B ernstein , J. B./ o rdóñez , F./ r ocA , F. 2019: «On the emergence of personal articles in the history of catalan», in: M. B ouzouitA / A. B reitBArth / L. d AnckAert / E. w itzenhAusen (ed.), Cycles in language change , New York/ Oxford, Oxford University Press: 88-108. B ernstein , J. B./ o rdóñez , F./ r ocA , F. 2021: «Emphatic elements and the development of definite articles. Evidence for a layered DP in early Romance», Journal of Historical Syntax 5: 1-32. B LAsco F errer , e. 1984: Grammatica Storica del Catalano e dei suoi dialetti con speciale riguardo all’Algherese , Tübingen, G. Narr. B rucArt , J. M./ c LAveriA , g. in press: «Els possessius», in: M. p érez s ALdAnyA / J. M Artines / G. r igAu / J. A ntoLí (ed.), Gramàtica del català antic , València, Publicacions de la Universitat de València: 1351-425. B ruguerA , J. 1986: «Les Homilies d’Organyà i els seus possibles occitanismes», in: J.-C. B ouvier (ed.), Actes du XVII ème Congrès International de Linguistique et Philologie Romanes. Vol. 3: Linguistique descriptive, phonétique, morphologie et lexique , Marseille, Université de Provence: 253-61. B uridAnt , c. 2000: Le moyen français. Le traitement du texte , Strasbourg, Presses Universitaires. B uridAnt , c. 2019: Grammaire du français médiéval (XI e -XIV e siècles) , Strasbourg, ELiPhi. CICA = J. t orrueLLA / M. p érez s ALdAnyA / J. M Artines (ed.), Corpus Informatitzat del Català Antic . URL: http: / / cica.cat/ index.php [30.9.2025] c oroMines , J. 1976: «Les Homilies d’Organyà. Edició crítica millorada i anotada», in id ., Entre dos llenguatges , vol. 1, Barcelona, Curial: 127-51. FEW = w ArtBurg , w. von . et al. 1922-2002: Französisches etymologisches Wörterbuch. Eine Darstellung des galloromanischen Sprachschatzes , Bonn etc., Klop etc. g ArciA s eBAstià , J. v./ M Artines , J. in press: «L’article», in: M. p érez s ALdAnyA / J. M Artines / G. r igAu / J. A ntoLí (ed.), Gramàtica del català antic , València, Publicacions de la Universitat de València: 1153-258. g rierA , A. 1917a: «L’article en català i la llengua literària», BDC 5: 50-60. g rierA , A. 1917b: «Les Homilíes d’Organyà», Vida Cristiana 4/ 21: 132-38, 4/ 22: 172-78. g rierA , A. 1931: Gramàtica històrica del català antic , Barcelona, Casa de la Caritat. g rierA , A. 1965: Gramática histórica catalana , Barcelona, Instituto Internacional de Cultura Romànica/ Abadia de San Cugat del Vallès. h oepFFner , e. 1926: La chanson de Sainte Foy. Fac-similé du manuscrit et texte critique. Introduction et commentaire philologique , 2 vol., Paris, Les belles Lettres. J ensen , F. 1994: Syntaxe de l’ancien occitan , Tübingen, Niemeyer. DOI 10.24053/ VOX-2025-008 Vox Romanica 84 (2025): 183-198 196 Camilla Talfani L edgewAy , A. 2012: From Latin to Romance: Morphosyntactic typology and change , Oxford, University Press. g iMeno B etí , L. 2005: Aproximació lingüística als inicis de la llengua catalana (segles VIII al XIII) , Barcelona, Universitat Jaume I. g ökçen , A. 1977: «The language of the Homilies d’Organyà», in: J. g uLsoy / J. M. s oLà -s oLé (ed.), Catalan studies. Volume in memory of Josephine de Boer , Barcelona, Hispam. M Artí i c AsteLL , J./ M orAn , J. 1986: Documents d’història de la llengua catalana. Dels orígens a Fabra , Barcelona, Empúries. M énArd , p. 1976. Manuel du Français du Moyen Âge. Partie 1. Syntaxe de l’ancien français , Bordeaux, Sobodi. M iret i s Ans , J. 1904: «El més antic text literari català», Revista de Bibliografia Catalana 4: 30-47, 215-20. M iret i s Ans , J. 1915: Antics documents de llengua catalana i reimpressió de les Homilies d’Organyà , Barcelona, Casa de la Caritat. M-L, RG = M eyer -L üBke , W. 1890-1902: Grammatik der romanischen Sprachen , 4 vol., Leipzig, Fues/ Reisland. M oLho , M. 1961: «Les homélies d’Organyà», BHisp 63/ 3-4: 186-210. M oLL , F. d . B. 1952: Gramática histórica catalana , Madrid, Gredos. M oLL , F. d . B. 2006: Gramàtica històrica catalana. Edicio corregida i anotada per Joaquim Marti Mestre, amb la collaboracio de Jesus Jiménez , València, Universitat de València. M orAn , J. 1982: Les homilies d’Organyà en relació amb les homilies provençals de Tortosa , in: A. M. B AdiA i M ArgArit / p. B ohigAs (ed.), Miscel·lània Pere Bohigas , vol. 1, Barcelona, Publicacions de l’Abadia de Montserrat: 67-84. M orAn , J. 1984: El capbreu de Castellbisbal estudi filològic i lingùístic d’un text català antic , Barcelona, Universitat de Barcelona. M orAn , J. 1990: Les Homilies d’Organyà , Barcelona, Curial/ Publicacions de l’Abadia de Montserrat ( Textos i Estudis de Cultura Catalana 20). M orAn , J. 1994: «L’ús de la llengua vulgar per a fins religiosos i catequètics en els orígens», Estudis de llengua i literatura catalanes 28 (Miscel·lània Germà Colón 1) , Barcelona, Publicacions de l’Abadia de Montserrat: 81-87. M orAn , J. 2003: « Homilies d’Organyà . Edicions i estudis nous», Llengua i literatura: Revista anual de la Societat Catalana de Llengua i literatura 14: 417-40. M orAn , J. 2007: «L’article en els topònims: teoria i pràctica», Butlletí de la Societat d’Onomàstica 106/ 107: 126-33. M orAn , J./ r ABeLLA , J. A. 2001: Primers textos de la llengua catalana , Barcelona, Proa. M orAn , J./ r ABeLLA , J. A. 2004: «Els primers textos en català. Textos anteriors a les Homilies d’Organyà », Les Faldades de Lo Banyut 2: 23-25. M orAn , J./ r ABeLLA , J. A. 2007: «El procés d’escripturació del català», in: El llibre i la lectura: una revolució en la història de la humanitat. Actes del seminari del CUIMPB-CEL 2005 , Barcelona, IEC: 19-29. n AdAL , J. M./ p rAts , M. 1992: Història de la llengua catalana. 1/ Dels inicis fins al segle XV , Barcelona, Edicions 62. p érez s ALdAnyA , M./ M Artines J. 2009: «Per a una Gràmatica del català antic: objectius, metodologia i un tast», in: id . (ed.), Per a una gramàtica del català antic , València, Institut Interuniversitari de Filologia Valenciana: 7-39. p uJoL i t uBAu , p. 1913: Documents en llengua vulgar dels segles XI, XII, XIII, procedents del bisbat de la Seu d’Urgell , Barcelona, IEC/ Diputació. r ABeLLA , J. A. 2009: «L’article literari i l’article salat: variable diatòpica o diastràtica», in: M. p érez s ALdAnyA / J. M Artines (ed.), Per a una gramàtica del català antic , València, Institut Interuniversitari de Filologia Valenciana: 151-61. r ABeLLA , J. A. 2023. «Notes linguistiques», in: M. z iMMerMAnn (ed.), Le Catalan médiéval , Turnhout, Brepols: 81-90. 197 DOI 10.24053/ VOX-2025-008 Vox Romanica 84 (2025): 183-198 Deconstructing a misconception about the definite article derived from ipse r Asico p. d. 2006 : El català antic , Girona, CCG. r enzi , L. 1992: «Le developpement de l’article en roman», RRLi 37: 161-76. r ocA u rgeLL , F. in press: «Els demostratius», in: M. p érez s ALdAnyA / J. M Artines / G. r igAu / J. A ntoLí (ed.), Gramàtica del català antic , València, Publicacions de la Universitat de València: 1261-347. r uBio g ArcíA , L. 1979: Documentos lingüísticos catalanes (s. X-XII) , Murcia, Universidad de Murcia. r usseLL -g eBBett , p. 1965: Medieval Catalan linguistic texts , Oxford, Dolphin. s ALvi , g./ r enzi , L. 2010: Grammatica dell’italiano antico , 2 vol., Bologna, Il Mulino. s oBerAnAs A. J./ r ossinyoL A. 2001: Homilies d’Organyà, facsímil del manuscrit, edicions diplomàtica i crítica, amb un estudi d’Armand Puig i Tàrrech , Barcelona, Barcino («Els Nostres Clàssics»). s oBerAnAs A. J./ r ossinyoL A. 2004: Homilies d’Organyà. Edició trilingüe, amb el facsímil del manuscrit , Barcelona, Barcino. s ornicoLA , r. 2011: «Per la storia dei dimostrativi romanzi: i tipi neutri [tso], [so], [ço] e la diacronia dei dimostrativi latini», ZrP 127/ 2: 220-314. t ALFAni , c. 2025: «Articolo o aggettivo possessivo? Alcune occorrenze ambigue nell’occitano medievale», in: c. M enichetti / F. F usAroLi (ed.), Perspectives en philologie linguistique galloromane et italienne , Strasbourg, ELiPhi: 51-71. t AvAni , g. 1994: Breu història de la llengua catalana , Barcelona, Edicions 62. t hoMAs , A. 1974 [ 1 1925]: La Chanson de Sainte Foi d’Agen , Paris, Champion. DOI 10.24053/ VOX-2025-008 Vox Romanica 84 (2025): 183-198 Deconstructing a misconception about the definite article derived from ipse in Pre-Literary Catalan Abstract: This paper aims to revisit the thorny question of the presence of evidence for the determinant derived from the Latin demonstrative ipse in Pre-Literary Catalan, meaning in the earliest texts preserved in the Catalan language. An occurrence of this article was identified by Joan Coromines in the Homilies d’Organyà , generally considered the oldest literary text in Catalan. All subsequent editors of this text, as well as many specialists in Old Catalan, authors of grammars or histories of the language, followed Coromines by including this information in their works. Consequently, by treating this data as an axiom, it helped reinforce the idea that, in the ancient era, the ipse -derived determinant was much more widespread in the Catalan language, written and spoken, contrary to its modern geographical distribution. Its progressive disappearance would then be due, notably, to a negative connotation from a diastratic point of view. Nevertheless, our recent research allows us to shed light on and propose an alternative interpretation of the supposed occurrence of the ipse -derived article in the Homilies d’Organyà , thereby deconstructing a misconception which had been entrenched for about three-quarters of a century in the history of the Catalan language. Keywords: ipse -derived Definite Article, Pre-Literary Catalan, History of the Old Catalan language, Marked Medieval Syntax, Linguistic Philology 198 Camilla Talfani