Introduction to the Special Issue Grammaticalization across Languages, Levels, and Frameworks

Over the last few decades, we have witnessed a proliferation in studies on phenomena such as grammaticalization (cf., among many others, Heine et al. 1991; Traugott and Heine 1991; Wischer and Diewald 2002; Hopper and Traugott 2003; Narrog and Heine 2011), lexicalization (among others, Brinton 2002; Hopper and Traugott 2003; Brinton and Traugott 2005), and, more recently, constructionalization (Traugott and Trousdale 2013; Hilpert 2013; Traugott 2014; Boogaart et al. 2014; Barðdal et al. 2015; and contributions in Giacalone Ramat et al. 2013). According to their most basic definitions, these notions refer to different phenomena of language change, which typically begin with a lexical form with a concrete meaning and end, respectively, in a grammatical unit or zero, a new lexical unit, or a fixed grammatical pattern or construction. Extensive bibliographies reflect ongoing discussions concerning every possible aspect of these linguistic processes, including the basic terminology and their internal relationships. Still, a common feature of grammaticalization theories is that they explain particular cases of changing grammar primarily as being a consequence of concrete language use. As a consequence, new knowledge on the nature of overall language change can best be gathered by closely studying particular phenomena in empirical data, coming from different languages, and incorporating insights from different frameworks. Therefore, this Special Issue aims at reporting on new and current interests and developments in this established yet ever-evolving field of research. It features a diverse collection of studies covering a broad spectrum of topics, methodologies, datasets, and languages. Many of these studies were initially presented at the Grammar and Corpora conference held at Ghent University in the summer of 2022.

Since the 1980s, the field of research has seen an increase in studies exploring languages with varied typological profiles and genealogical affiliations. Significant advances have been achieved in identifying common grammaticalization paths across languages (Viberg 1999), leading to the development of various clines or hierarchies based on extensive comparative data (Heine et al. 1991; Haspelmath 2004). However, the majority of these studies focus on specific linguistic phenomena within individual languages, often leveraging increasingly larger corpora. Predictably, the most substantial progress has occurred in languages with abundant historical data, primarily European languages such as those in the Romance and Germanic branches of Indo-European. Despite this focus, the field continues to diversify. For instance, the new millennium has witnessed growing research on grammaticalization in sign languages, which have traditionally been underrepresented in linguistic studies, as well as on comparative analyses of grammaticalization in both spoken and signed languages (e.g., Pfau and Steinbach 2011). The papers included in this volume include well-studied languages like Spanish (with contributions from Garachana and Sansiñena, Torres and Enghels, Salameh Jiménez, and Fuentes Rodríguez), Italian (Morei), and English (Smith). Additionally, the volume features research on less commonly studied languages such as Romanian (Ilioaia) and Old Catalan (Torres-Latorre and Sentí). There are also contrastive studies between different languages (Paoli) and dialects (with Jansegers et al. focusing on Spanish and Deng on French). Moreover, the volume includes research on (German) sign language (Otte et al.), further highlighting its wide linguistic scope.

Next, research on grammaticalization spans various levels of linguistic analysis. Initially, studies predominantly focused on the morphosyntactic dimensions of grammaticalization, viewing it primarily as a process of morphosyntactic reduction (cf., among others, Lehmann [1982] 1995; Hopper 1991; Haspelmath 2004). Conversely, some scholars have emphasized not just the aspect of loss but also the semantic enrichment that accompanies this process (Bybee et al. 1994; Diewald 2002; Heine 2002; Traugott 2003). As a consequence, more recent approaches extend beyond morphosyntax and semantics to explore grammaticalization within the context of discourse. These studies analyze grammaticalization in real spoken conversations and examine its connection to broader cognitive processes involved in the interaction between speaker and addressee. Consequently, both the form and meaning are given equal importance, with ‘meaning’ encompassing not only lexical significance but also discourse function, information structure, and various other pragmatic aspects. Against this background, this volume explores grammaticalization across multiple levels of linguistic analysis.

Lastly, our understanding of grammaticalization processes can be enriched by incorporating new insights from relevant theoretical frameworks. The concept of ‘emerging grammar’ (Hopper 1987) (partially) explains why grammaticalization has mainly developed in parallel with the development of cognitive and functional linguistics. Indeed, besides describing the formal paths of grammaticalization, researchers have increasingly focused on the motivations driving these changes and exploring why they occur. Hence increasing attention was paid to the functional meaning of grammaticalization, incorporating various semantic and pragmatic accounts. Since the early 21st century, there have been efforts to examine diachronic change through the lens of construction grammar. Although most studies on ‘constructionalization’ acknowledge that this approach does not fundamentally deviate from traditional grammaticalization studies (see references above), they highlight its utility in providing a comprehensive framework for understanding both lexical and grammatical changes. Furthermore, the adoption of the construction concept allows for more precise measurements of the gradual nature of grammatical change, facilitating detailed investigations into its productivity. The cross-fertilization between grammaticalization and construction grammar, particularly in the notion of constructionalization, is notably present in the papers of Garachana and Sansiñena, Torres and Enghels, Ilioaia, Smith, Salameh Jiménez, and Fuentes Rodríguez.

The volume builds progressively through various levels of linguistic analysis, beginning with two papers on the grammaticalization of morphological phenomena (Paoli and Otte et al.). Advancing to the constructional level, four papers focus on grammaticalization of specific complex syntactic constructions, in particular verbal periphrases (Garachana and Sansiñena, Torres Soler and Enghels, Ilioaia, and Smith). Subsequently, the focus shifts to verbal tenses, thereby adding another layer of complexity (Torres-Latorre and Sentí, Morei). Ultimately, the analysis reaches the discourse level, where grammaticalization processes of pragmatic markers come into play (Jansegers et al., Deng, Salameh Jiménez, and Fuentes Rodríguez).

Sandra Paoli’s paper, titled “Gradualness of Grammaticalization and Abrupt Change Reconciled: Evidence from Microvariation in Romance”, emphasizes the interplay of grammaticalization processes that exhibit both gradual and abrupt characteristics. To this end, her paper presents an in-depth empirical analysis of clitic loss across several Romance languages, namely Brazilian Portuguese, Raeto-Romance, northeastern Italo-Romance varieties, and French. From a theoretical perspective, the concepts of gradience (variation within a category at a given time) and gradualness (the accumulation of small changes over time) are discussed, and it is further shown how localized changes (gradience) contribute to the broader evolutionary trajectories of grammatical structures (gradualness). In conclusion, the study identifies common patterns where certain cells within the clitic paradigms are more susceptible to loss, often influenced by animacy, specificity, and other grammatical features. These patterns are consistent across the studied languages (for instance, third-person and less animate referents are typically the first to exhibit loss). However, by examining microvariation, the paper reveals that what may seem like abrupt changes on a micro level contributes to a perception of gradual change on a macro scale. Overall, this approach allows for a nuanced understanding of the mechanisms driving grammaticalization, aligning well with the Special Issue’s focus on empirical data and theoretical insights in evolving research in this domain.

The paper “Numeral Incorporation as Grammaticalization? A Corpus Study on German Sign Language (DGS)” by Felicitas Otte, Anke Müller, Sabrina Wähl, and Gabriele Langer examines numeral incorporation in German Sign Language (DGS), exploring whether this phenomenon reflects a process of grammaticalization. Based on a DGS corpus, the study recurs to an apparent-time method to infer diachronic trends. In sign languages, numeral incorporation refers to the phenomenon where numeral signs merge with lexical signs (mostly units of time, measurements, or quantities) to form a new, composite sign (e.g., “three weeks”). The analysis of the corpus identifies a progression in the form of these combinations from free morphemes to cliticized morphemes and finally to bound morphemes (affixes), or a movement from independent numeral and base signs toward a unified, grammaticalized form where numeral signs become an integral part of the lexical sign. However, while the study observes some characteristics of grammaticalization, such as increased frequency and fixedness of use among younger signers, it concludes that numeral incorporation does not fully conform to traditional models of grammaticalization because certain expected features of grammaticalization, such as obligatoriness or fossilization, are absent.

In their study, “Combinatorial Productivity of Spanish Verbal Periphrases as an Indicator of Their Degree of Grammaticalization”, Mar Garachana and María Sol Sansiñena explore the grammaticalization process of (near-synonymous) Spanish verbal periphrases, focusing on dejar de + INF and parar de + INF. Using a constructionist, usage-based approach, the authors investigate these periphrases’ semantic areas, functional distributions over time, and collostructional patterns to discuss their productivity and how this relates to their degree of grammaticalization. The analysis of the historical data points towards a significant divergence in the evolution of dejar de + INF and parar de + INF, with the latter not paralleling the former in terms of combinatorial patterns, semantic fields, or lexical productivity. While dejar de + INF has reached a stable state in its grammaticalization process (as shown, for instance, by its limited combinatorial productivity), parar de + INF continues to evolve and extend to new contexts, showing variability in its integration into the grammatical system. Overall, the paper provides a nuanced understanding of how these two similar yet distinct periphrases have developed in Spanish, contributing to broader discussions on language change, grammaticalization, and the impact of semantic and functional factors on the evolution of language structures.

The paper “From Motion to Causation: The Diachrony of the Spanish Causative Constructions with traer (‘bring’) and llevar (‘take’)” by Julio Torres Soler and Renata Enghels investigates the historical evolution (between the 13th and 20th centuries) of specific Spanish causative constructions involving the verbs traer and llevar. The study contributes to the understanding of grammaticalization by illustrating how changes at the lexical and semantic levels impact syntactic structures and the productivity of grammatical patterns in a language, resonating with themes of diachronic change and construction grammar. In concrete, the paper identifies a specific linguistic subschema characterized by verbs of ‘caused accompanied motion’ that have semantically specialized to denote indirect causation. This specialization is marked by a low level of syntactic integration between the causative verb and the infinitive form it governs. Moreover, the historical semantic shifts of llevar and traer significantly influenced their integration into causative constructions. Initially, llevar was more associated with unbounded motion, which delayed its adoption into causative uses. From the 16th century onward, as the meanings of the verbs diverged, with llevar becoming more about goal-oriented motion and traer about motion toward the speaker or addressee, their roles in causative constructions also shifted.

Next, the paper by Mihaela Ilioaia, “Constructing Meaning: Historical Changes in MIHI EST and HABEO Constructions in Romanian”, studies the evolution of two competing Latin patterns originally used to express possession, highlighting their different semantic and syntactic trajectories in Romanian compared to other Romance languages. Remarkably, unlike other Romance languages, where habeo replaces mihi est in possessor and experiencer contexts, Romanian preserves both constructions. However, mihi est is more commonly used in experiencer contexts (involving psychological and physiological states), suggesting a specialized functional evolution, whereas habeo tends to cover a broader range of uses, including abstract possession. Ilioaia utilizes a diachronic corpus to track changes over time, focusing on how different state nouns are used within these constructions. The paper further evaluates whether the changes observed in these constructions can be classified under grammaticalization or constructionalization, eventually supporting the idea of constructional change. Overall, in the same line of ideas as Garachana and Sansiñena’s paper, this study emphasized the role of functional differentiation in the grammaticalization process.

In “Productivity from a Metapragmatic Perspective: Measuring the Diachronic Coverage of the Low Level Lexico-Grammatical Construction Have the N (Body Part/Attitude) to ↔ <Metapragmatic Comment> Using the COHA”, Chris A. Smith addresses the issue of constructional change and how to determine its emergence, as well as the productivity of such changes. The overarching question of the paper is how semantics affects productivity, in the generative sense of extensibility of a construction (a form-meaning pairing). The study focuses more precisely on the metapragmatic reaction to perceived insolence via the expression have the N to (e.g., have the guts to). Theoretically, it adopts the constructional network approach, implying that constructions form a network of interrelated constructions at a higher and lower level that can affect each other. Methodologically, this paper presents a lexicalist bottom-up approach to the construction and contributes to the question of how to overcome methodological issues relating to a qualitative rather than quantitative approach to constructional architecture and the relative productivity of constructions. In doing so, it relies on a fine-grained distributional semantic analysis using the Corpus of Historical American English (COHA). It shows that studying low-level lexico-grammatical patterns can provide a more reliable understanding of the emergence and development of a family of constructions and that a global view of constructional architecture at multiple levels should be pertinent to identify the extensibility potential of the construction.

The paper “Clitic Placement and the Grammaticalization of the Future and the Conditional in Old Catalan” by Aina Torres-Latorre and Andreu Sentí focuses on a well-known process of morphological change in Romance languages—the grammaticalization of Latin periphrasis cantāre habeō towards the Romance future and conditional tenses—but concentrates on a less-studied language, namely Catalan. The study adopts a descriptive, empirical, and quantitative approach and constitutes the first exhaustive depiction of the frequencies of the different possibilities for clitic placement (proclisis, mesoclisis, and enclisis) associated with Catalan future/conditional constructions throughout the Middle Ages and the 16th century. As such, it completes the panorama of Ibero-Romance varieties on this phenomenon and allows for a comparison as to the placement of Catalan within the Romance continuum. As it turns out, the grammaticalization of future and conditional in the 13th and 14th centuries was more advanced in the eastern languages of the Iberian Peninsula, such as Catalan, than in the western ones. The results obtained for Old Catalan appear to confirm a language contact phenomenon in which Catalan (and probably Occitan) spread the more grammaticalized future/conditional to Navarro-Aragonese and Castilian. The authors prove these significant quantitative differences through a detailed analysis of all data in a well-balanced corpus of selected texts from the Catalan historical digital corpus (CICA).

A similar focus on the grammaticalization of tense is present in the paper “Not Only Anteriority in the Past: The Functions of the Pluperfect in Spoken Italian”, where Eleonora Morei conducts a corpus-based approach on the Italian pluperfect to empirically evaluate the hypothesis of the existence of an aoristic (i.e., deictic) use of this tense. To this end, a dataset of spoken spontaneous speech was compiled, based on the (northern) Italian ParlaTO corpus. She shows that the Italian pluperfect has developed secondary functions that can be explained by (inter)subjectification paths of grammaticalization and demonstrates how tense, aspect, and modality intertwine in defining the Pluperfect’s semantics.

The use of spontaneous spoken data also characterizes the study conducted by Marlies Jansegers, Chantal Melis, and Elenor Arrington, “Diverging Grammaticalization Patterns across Spanish Varieties: The Case of perdón in Mexican and Peninsular Spanish”. By means of a corpus-based comparative analysis based on spoken conversations and interviews in both varieties of Spanish, this study empirically investigates the contemporary grammaticalized uses of the apologetic marker perdón (‘sorry’). Contrary to what usually happens to other pragmatic markers, the point of departure for perdón rests in an element that already has an illocutionary force at the onset, the original illocutionary act of regret. It is argued that this original illocutionary material is recruited for further illocutionary and discursive uses, determining the contemporary grammaticalized uses and values of perdón. That is, the grammaticalization process of perdón embodies a process that affects the illocutionary potential of the speech act itself. Therefore, rather than semantic bleaching, this can be best described as a case of progressive weakening of its illocutionary force. Interestingly, this even seems to give rise to diverging patterns of grammaticalization across varieties of the same language, since evidence from both corpora suggests a more advanced stage in the grammaticalization process of perdón in Mexican Spanish.

A similar process of diverging grammaticalization patterns across varieties of the same language was discovered by Delin Deng in her paper “The Grammaticalization of the Discourse Marker genre in Swiss French”. The study focuses more precisely on the discourse marker genre in Swiss French. By conducting an apparent-time variationist analysis based on oral data taken from the OFROM corpus (le corpus Oral de Français de Suisse Romande), the study offers a quantitative approach to the grammaticalization status of genre in Swiss French and its correlation with social factors, such as age, gender, sociolinguistic situation, and socio-educational status of the speakers. It also examines phonological reduction to support the claim that genre in Swiss French is undergoing an independent process of grammaticalization from that in Hexagonal French documented in the literature. Moreover, the study provides new insights into the development of the same particle in different regions being conditioned by different social factors: while the ongoing change of genre is led by female speakers in Swiss French, it is first led by male speakers, then quickly spread to both gender groups in Hexagonal French.

The relevance of studying processes of grammaticalization and constructionalization in conversational formulas is further illustrated in the study by Shima Salameh Jiménez, “Paths of Constructionalization in Peninsular Spanish: The Development of “Pues Eso”. A 20th Century Case”. The study traces the recent diachronic evolution of Peninsular Spanish pues eso in the 20th century as a construction by analyzing linguistic patterns related to pre- and post-constructional processes, including syntactic, semantic, and pragmatic features. Methodologically, the proposal incorporates units and positions from the Val.Es.Co. model of discourse segmentation, and the data were retrieved from the CDH corpus (Corpus del Diccionario histórico de la lengua Española, Real Academia Española), with particular attention to written texts reproducing orality. By adopting a constructional approach, the study reveals that the current functions covered by pues eso were mainly consolidated during the 20th century, but with some important antecedents in the 16th and 17th centuries, when agreement and self-reinforcement functions seem to have originated.

Catalina Fuentes Rodríguez’s paper “From Peripheral Structure to Discourse Operator: No Veas” studies the grammaticalization and cooptation processes of the Spanish expression no veas. This expression, originally part of more syntactically bound structures, has evolved into a more flexible discourse operator with various pragmatic functions. The study uses corpus data from both the CORPES XXI and the MEsA corpus and focuses on instances where no veas is used in different contexts, starting from a literal verb phrase (with the meaning of seeing or watching something), progressively used for intensification, to a peripheral usage as a discourse operator indicating surprise or emphasis. It is argued that the evolution of no veas reflects cooptation, where it gains new pragmatic functions without undergoing complete grammaticalization, which would typically involve more extensive semantic bleaching and syntactic fixation. Moreover, it illustrates a process of subjectification as the expression transitions from describing external events to expressing the speaker’s stance. Overall, the paper contributes to the theoretical discussion on how linguistic elements evolve from syntactic components to playing roles in discourse structuring, adapting to the communicative needs of speakers.

In conclusion, as has been showcased by the papers included in this volume, the exploration of grammaticalization has yielded substantial insights into the evolution of language at multiple levels of analysis. However, we believe that future research should increasingly focus on the social factors that facilitate or inhibit grammaticalization processes. Specifically, the role of social networks in shaping linguistic change warrants closer examination. Recent studies have expanded the scope of grammaticalization to include discourse-level phenomena, highlighting the intricate interplay between structure and meaning. Moving forward, it is imperative to integrate social dimensions into this framework. Alongside lexical, semantic, and pragmatic information, social information such as the dynamics of speaker communities, patterns of language contact, and the influence of social hierarchies could provide a more comprehensive understanding of grammaticalization. By incorporating these social factors, future research can offer a more holistic view of how grammatical structures emerge and evolve within their sociocultural contexts, thereby enriching our understanding of language change and development.