We present a new criterion for proving that a language is not multiple context-free, which we call a Substitution Lemma. We apply it to show a sample selection of languages are not multiple context-free, including the word problem of $F_2\times F_2$. Our result is in contrast to Kanazawa et al. [2014, Theory Comput. Syst.] who proved that it was not possible to generalise the standard pumping lemma for context-free languages to multiple context-free languages, and Kanazawa [2019, Inform. and Comput.] who showed a weak variant of generalised Ogden's lemma does not apply to multiple context-free languages. We also record that groups with multiple context-free word problem have decidable rational subset membership problem.
Relative monads provide a controlled view of computation. We generalise the monadic metalanguage to a relative setting and give a complete semantics with strong relative monads. Adopting this perspective, we generalise two existing program calculi from the literature. We provide a linear-non-linear language for graded monads, LNL-RMM, along with a semantic proof that it is a conservative extension of the graded monadic metalanguage. Additionally, we provide a complete semantics for the arrow calculus, showing it is a restricted relative monadic metalanguage. This motivates the introduction of ARMM, a computational lambda calculus-style language for arrows that conservatively extends the arrow calculus.
Hossain Shaikh Saadi, Minh Duc Bui, Mario Sanz-Guerrero
et al.
This paper presents the JGU Mainz submission to the WMT25 Shared Task on LLMs with Limited Resources for Slavic Languages: Machine Translation and Question Answering, focusing on Ukrainian, Upper Sorbian, and Lower Sorbian. For each language, we jointly fine-tune a Qwen2.5-3B-Instruct model for both tasks with parameter-efficient finetuning. Our pipeline integrates additional translation and multiple-choice question answering (QA) data. For Ukrainian QA, we further use retrieval-augmented generation. We also apply ensembling for QA in Upper and Lower Sorbian. Experiments show that our models outperform the baseline on both tasks.
Lukas Gienapp, Christopher Schröder, Stefan Schweter
et al.
Large language model development relies on large-scale training corpora, yet most contain data of unclear licensing status, limiting the development of truly open models. This problem is exacerbated for non-English languages, where openly licensed text remains critically scarce. We introduce the German Commons, the largest collection of openly licensed German text to date. It compiles data from 41 sources across seven domains, encompassing legal, scientific, cultural, political, news, economic, and web text. Through systematic sourcing from established data providers with verifiable licensing, it yields 154.56 billion tokens of high-quality text for language model training. Our processing pipeline implements comprehensive quality filtering, deduplication, and text formatting fixes, ensuring consistent quality across heterogeneous text sources. All domain subsets feature licenses of at least CC-BY-SA 4.0 or equivalent, ensuring legal compliance for model training and redistribution. The German Commons therefore addresses the critical gap in openly licensed German pretraining data, and enables the development of truly open German language models. We also release code for corpus construction and data filtering tailored to German language text, rendering the German Commons fully reproducible and extensible.
While a language assigns a value of either `yes' or `no' to each word, a lattice language assigns an element of a given lattice to each word. An advantage of lattice languages is that joins and meets of languages can be defined as generalizations of unions and intersections. This fact also allows for the definition of positive varieties -- classes closed under joins, meets, quotients, and inverse homomorphisms -- of lattice languages. In this paper, we extend Pin's positive variety theorem, proving a one-to-one correspondence between positive varieties of regular lattice languages and pseudo-varieties of finite ordered monoids. Additionally, we briefly explore algebraic approaches to finite-state Markov chains as an application of our framework.
In phonological theory there are multiple ways to represent mid vowels. SPE conventions maintain that they are non-[high] and non-[low]. Conversely, frameworks like Element Theory argue that mid vowels are simultaneously [high] and [low]. This article examines eight processes (and groups of processes) within the Germanic language family, which strongly indicate their specification as simultaneously [high] and [low]. That specification is manifest from developments that tease out the [high] and [low] features of a single mid vowel into separate [high] and [low] elements of sound (e.g., [e] > [ja]). It also falls out from changes in which separate [high] and [low] segments coalesce into a single mid vowel (e.g., [au] > [o]).
As part of a research on a novel in-process multiprogramming-language interoperability system, this study investigates the interoperability and usage of multiple programming languages within a large dataset of GitHub projects and Stack Overflow Q\&A. It addresses existing multi-lingual development practices and interactions between programming languages, focusing on in-process multi-programming language interoperability. The research examines a dataset of 414,486 GitHub repositories, 22,156,001 Stack Overflow questions from 2008-2021 and 173 interoperability tools. The paper's contributions include a comprehensive dataset, large-scale analysis, and insights into the prevalence, dominant languages, interoperability tools, and related issues in multi-language programming. The paper presents the research results, shows that C is a central pillar in programming language interoperability, and outlines \emph{simple interoperability} guidelines. These findings and guidelines contribute to our multi-programming language interoperability system research, also laying the groundwork for other systems and tools by suggesting key features for future interoperability tools.
In the context of two-player games over graphs, a language $L$ is called positional if, in all games using $L$ as winning objective, the protagonist can play optimally using positional strategies, that is, strategies that do not depend on the history of the play. In this work, we describe the class of parity automata recognising positional languages, providing a complete characterisation of positionality for $ω$-regular languages. As corollaries, we establish decidability of positionality in polynomial time, finite-to-infinite and 1-to-2-players lifts, and show the closure under union of prefix-independent positional objectives, answering a conjecture by Kopczyński in the $ω$-regular case.
We study the class of star-free languages. A long-standing goal is to classify them by the complexity of their descriptions. The most influential research effort involves concatenation hierarchies, which measure alternations between ``complement'' and ``union plus concatenation''. We explore alternative hierarchies that also stratify star-free languages. They are built with an operator $C\mapsto TL(C)$. From an input class $C$, it produces a larger one $TL(C)$, consisting of all languages definable in a variant of unary temporal logic, where temporal modalities depend on $C$. Level $n$ in the navigational hierarchy of basis $C$ is constructed by applying this operator $n$ times to $C$. As bases $G$, we focus on group languages and natural extensions thereof, denoted $G^+$. We prove that the navigational hierarchies of bases $G$ and $G^+$ are strictly intertwined and conduct a thorough investigation of their relationships with concatenation hierarchies. We also look at two problems on classes of languages: membership (decide if a language is in the class) and separation (decide, for two languages $L_1,L_2$, if there is a language $K$ in the class with $L_1\subseteq K$ and $L_2\cap K=\emptyset$). We prove that if separation is decidable for $G$, then so is membership for level \emph{two} in the navigational hierarchies of bases $G$ and $G^+$. We take a look at the trivial class $ST=\{\emptyset,A^*\}$. For the bases $ST$ and $ST^+$, the levels \emph{one} are standard variants of unary temporal logic. The levels \emph{two} correspond to variants of two-variable logic, investigated recently by Krebs, Lodaya, Pandya and Straubing. We solve one of their conjectures. We also prove that for these two bases, level \emph{two} has decidable \emph{separation}. Combined with earlier results on the operator $C\mapsto TL(C)$, this implies that level \emph{three} has decidable membership.
In our paper, we deal with the Germanic–Romance language contact, focusing on Cimbrian, a Germanic minority language spoken in Northern Italy. Specifically, we focus on the violation of the well-known that-trace filter, as it appears to be an interesting case of the superficial convergence that we ascribe to the status of T, which is either too rich (model language) or too weak (replica language) to represent a viable landing site for subject extraction.
In this work, we intend to investigate one fundamental aspect of language contact by comparing the distribution of subjects in German, Northern Italian dialects and Cimbrian. Here, we show that purely syntactic order phenomena are more prone to convergence, i.e., less resilient, while phenomena that have a clearly identifiable morphological counterpart are more resilient. The empirical domain of investigation for our analysis is the morphosyntax of both nominal and pronominal subjects, the agreement pattern and their position in Cimbrian grammar. While agreement patterns display a highly conservative paradigm, the syntax of nominal (vP-peripheral and topicalized) subjects is innovative and mimics the Italian linear word order.
O presente artigo tem o objetivo de discutir e descrever estratégias didáticas durante a realização de um projeto de música em uma disciplina de língua alemã como língua adicional (LA) em contexto universitário. Tendo em vista a importância da inclusão de Tecnologias de Informação e Comunicação (TICs) no ensino de LAs, bem como da promoção do pensamento crítico-reflexivo, focado nos interesses e necessidades da(o)s aluna(o)s de Letras, o projeto de música foi realizado na plataforma Kahoot. Concluímos que as atividades contribuíram para o processo de ensino e aprendizagem em língua alemã, assim como para uma participação mais ativa e crítica por parte da(o)s estudantes.
German literature, Germanic languages. Scandinavian languages
Digital applications to assist language learning are becoming increasingly popular. They typically incorporate one or two dictionaries to improve the service so that users avoid leaving the app to consult external resources. This paper deals with the two dictionaries used in a learning app for Chinese learners of English. Initially, it describes the functioning of the app as well as the two dictionaries that have different roles in the app. It then focuses on the one that is integrated into the course texts and can be activated by clicking on a word or a multiword unit. A number of deficiencies are discussed such as inconsistent treatment of words and senses, data overload, difficult access, and inconvenient location of the pop-up window that displays the lexicographical items. These deficiencies may impact negatively on the learners' motivation and the learning process in general. The paper traces the problems to the database that sustains the dictionary as well as the design of the user interfaces that filter the data offered to the users. Finally, and inspired by the classical Chinese Xun Gu tradition, it suggests an alternative, context-adapted approach that breaks with traditional features of the dictionary article and reduces the content of the pop-up window to an absolute minimum. The idea is to avoid a consultation process that interrupts the learners' reading flow and focus on learning.
Philology. Linguistics, Languages and literature of Eastern Asia, Africa, Oceania
Im Folgenden wird eine Unterrichtsanregung zu literarischem Lernen mit Nora Krugs Graphic Memoir Heimat (2018) entwickelt. Im Zentrum der didaktischen Überlegungen steht die Kompetenz „Mit Fiktionalität bewusst umgehen“ (Spinner 2006) bzw. „Realitätsbezüge von Texten bestimmen“ (Maiwald 2015). Dazu eignet sich Heimat in besonderer Weise, da sie als Graphic Memoir reale und fiktive Inhalte sowie realistische und literarische Darstellungsweisen kunstvoll kombiniert. Aber nicht nur die ästhetische Gestaltung, auch das Thema ist anspruchsvoll. Nora Krug recherchiert ihre Familiengeschichte und verbindet damit die Frage nach der kollektiven Schuld und nationalen Identität Deutschlands nach dem Zweiten Weltkrieg.
Abhängig von Schulart, Klasse und angestrebter inhaltlicher Tiefe eignet sich Heimat für den Literaturunterricht von der späten Mittelstufe (9. oder 10. Jahrgangsstufe) bis zur Oberstufe. Das zentrale Ziel ist es, das Zusammenspiel von Faktualität und Fiktionalität für Schüler:innen durchschaubar und dessen besonderes epistemische Potential begreifbar zu machen. Dafür wird vorgeschlagen das kontrastive Verfahren intratextuelles Vergleichen (vgl. Kepser / Abraham 2016) über eine Progression der drei produktiven Verfahren Rekonstruktion, Restauration und Transformation zu realisieren (vgl. von Brand 2019).
Abstract (english): Nora Krug’s graphic memoir Belonging. Literary learning with a text between factuality and fictionality
This article describes the development of a model for learning with literature. The focus is on the thin line between fact and fiction, factuality and fictionality, respectively. It aims at improving students’ ability to distinguish between those two main categories, but also to discover the epistemic potential of the area in between (cf. Spinner 2006 and Maiwald 2015). Nora Krug’s graphic memoir Belonging (2018) is a challenging book, for that matter. She combines original documents and historic photographs with colored drawings and text. In doing so, she reconstructs the history of her family and at the same time she raises the question of Germany’s ambivalent past and collective guilt.
The developed model is suitable for students from the higher grades of secondary school . The method applied is an intratextual comparison (cf. Kepser/Abraham 2016 and von Brand 2019) to contrast the different forms of presentation in Belonging and to illustrate the special aesthetics of their combination.
Андрій Боцман, Ольга Дмитрук, Валерий Григорьевич Бондаренко
Being the remarkable feature of all Germanic languages, the preterite-present group of verbs played a significant role in forming morphological units of analytical type. The process of analytisation caused the introduction of some verb forms, which appeared as a result of gradual grammaticalisation when a preterite-present verb fused with a verbal form (participle or infinitive) creating a new morphological category of finite verb (future tense, oblique mood). During the process of their development, transformation and coexistence, the preterite-present verbs were separated into a stable morpho-semantic group which generated the field of modality and made some verbs follow the same way of development and transformation. As a result, a new morpho-semantic sub-group of analogous verbs was formed when the group of preterite-present verbs received its stability and potentiality in the Germanic verb corpus. The inductive force of the preterite-present verbs was so powerful that even an individual verb was involved into the process of grammaticalisation following the patterns of analogous and preterite present verbs, too. The verbs of preterite-present group, analogous subgroup and the individual verb functioned and coexisted within the East, West and North (Scandinavian) Germanic languages. As a result of that historical coexistence all these verbs got not only the common features in morphological and semantic aspects, they simultaneously gained a set of individual features differing semantically. Only after a detailed comparison of morphological and semantic individual features of related (according to the Proto-Germanic nature of origin) verbs, is it possible to explain their specific involvement into the process of grammaticalisation or their further disappearance from usage on the edge of Old and Middle or Middle and New periods of language transformation and development. Remarks on the paradigm of preterite-present verbs help to reveal the specifics of the verb functions and trace the way of the analytical form creation in some separate Germanic languages. Paradigmatic rows demonstrate either close or distant relations between the Germanic languages in their geographical groups or even intergroup relations. These relations are really essential for further investigation. Only the East Germanic group with its main representative of the Gothic language disappeared without leaving any close relatives in the Middle and New Germanic periods. The Scandinavian languages where the reflection of Old Norse. The West Germanic languages had a really entangled way of development. Old High German and Old Saxon were reflected in the German language. Anglo-Saxon, Old Saxon and Old Frisian were reflected in the English language. Old Frisian and Old Saxon were reflected in the Dutch and Frisian languages. Inner mutual semantic and morphological correlations of the Germanic preterite-present verbs help to indicate leading verbs which were involved into the process of the Germanic language analytisation.
Many Germanic languages have a finite-clause complementizer that resembles a demonstrative, e.g. English that, Dutch dat, German dass. No Romance language does. The traditional view of complementizers as simplex projecting heads that take IP or some comparable category as a complement has no way of accounting for this difference between Germanic and Romance. In this chapter, I will attempt to make progress toward an account, in part by reinterpreting finite-clause complementizers as relative pronouns.
In the absence of a stranded auxiliary or modal, VP-topicalization in most Germanic languages gives rise to the presence of a dummy verb meaning ‘do’. Cross-linguistically, this is a rather uncommon strategy as comparable VP-fronting constructions in other languages, e.g. Hebrew, Polish, and Portuguese, among many others, exhibit verb doubling. A comparison of several recent approaches to verb doubling in VP-fronting reveals that it is the consequence of VP-evacuating head movement of the verb to some higher functional head, which saves the (low copy of the) verb from undergoing copy deletion as part of the low VP copy in the VP-topicalization dependency. Given that almost all Germanic languages have such V-salvaging head movement, namely V-to-C movement, but do not show verb doubling, this paper suggests that V-raising is exceptionally impossible in VP-topicalization clauses and addresses the question of why it is blocked. After discussing and rejecting some conceivable explanations for the lack of verb doubling, I propose that the blocking effect arises from a bleeding interaction between V-to-C movement and VP-to-SpecCP movement. As both operations are triggered by the same head, i.e. C, the VP is always encountered first by a downward search algorithm. Movement of VP then freezes it and its lower copies for subextraction precluding subsequent V-raising. Crucially, this implies that there is no V-to-T raising in most Germanic languages. V2 languages with V-to-T raising, e.g. Yiddish, are correctly predicted to not exhibit the blocking effect.
In English and Dutch, listeners entrain to prosodic contours to predict where focus will fall in an utterance. Here, we ask whether this strategy is universally available, even in languages with very different phonological systems (e.g., tone versus non-tone languages). In a phoneme detection experiment, we examined whether prosodic entrainment also occurs in Mandarin Chinese, a tone language, where the use of various suprasegmental cues to lexical identity may take precedence over their use in salience. Consistent with the results from Germanic languages, response times were facilitated when preceding intonation predicted high stress on the target-bearing word, and the lexical tone of the target word (i.e., rising versus falling) did not affect the Mandarin listeners' response. Further, the extent to which prosodic entrainment was used to detect the target phoneme was the same in both English and Mandarin listeners. Nevertheless, native Mandarin speakers did not adopt an entrainment strategy when the sentences were presented in English, consistent with the suggestion that L2 listening may be strained by additional functional load from prosodic processing. These findings have implications for how universal and language-specific mechanisms interact in the perception of focus structure in everyday discourse.