Papers by Katharina Gerhalter
Linguistic Corpora and Big Data in Spanish and Portuguese, edited by Miguel Calderón Campos and Gael Vaamonde, Berlin, Boston: De Gruyter, 2024
AI-Linguistica. Linguistic Studies on AI-Generated Texts and Discourses, 2024
This case study focuses on a specific construction that exists in both Spanish and Portuguese, bu... more This case study focuses on a specific construction that exists in both Spanish and Portuguese, but not in English: topicalized infinitives (=TI), e.g., Sp. comer no come 'as for eating s/he does not eat'. We present three pilot experiments: the first one is a translation task which consists of translating sentences with TI from Spanish to Portuguese and vice versa. DeepL failed in most cases due to contamination by English as a pivot language. The second task is a continuation task: ChatGPT-3.5 was asked to complete sentences that start with a TI. In most cases, natural and adequate continuations starting with pero 'but' were generated. Since this task is based on predicting the most likely continuation, this result is not surprising, as this is exactly how the model works. Contrarily, ChatGPT-3.5 demonstrated a clear inability to perform well on the third task, which consisted of drawing pragmatic inferences from exactly the same examples containing a TI that encodes an adversative implicature.
![Research paper thumbnail of "El silencio no es precisamente una de sus virtudes… Diachronie und Pragmatik der Konstruktion [Negation + Exaktheits-Fokusadverb + X] im Französischen, Portugiesischen und Spanischen"](https://attachments.academia-assets.com/93278337/thumbnails/1.jpg)
"Konstruktionsgrammatische Zugänge zu romanischen Sprachen" (eds. Anja Hennemann & Hans-Jörg Döhla), 2021
This chapter deals with the semi-schematic construction [negation + focus adverb of the exactness... more This chapter deals with the semi-schematic construction [negation + focus adverb of the exactness-domain + X], which shows a similar form and equivalent functions in French, Portuguese, and Spanish. For example, Fr. pas précisément propres, Pt. não exatamente limpo and Sp. no precisamente limpia (‘not exactly clean’) are semantically ambiguous regarding the degree of the intended interpretation ‘(more or less) dirty’. This vagueness is intentional and leads to a wide range of possible contextual inferences. Consequently, the constructions Fr. [(ne) pas précisément/exactement X], Pt. [não precisamente/exatamente X] and Sp. [no precisamente/exactamente X] have different pragmatic functions: e.g., no precisamente limpia may be used for mitigation (‘not really clean’), or as an ironic understatement (hence, intensification and inversion: ‘very dirty’). The analyzed construction is a specific sub-type (or meso-construction) of the stylistic-rhetorical device litotes; the above-mentioned pragmatic functions have been observed for litotes in general.
This chapter analyses the diachrony of the construction [negation + focus adverb of the exactness-domain + X] in Spanish, French and Portuguese based on large corpus samples and shows how these pragmatic functions develop as contextual inferences. First precursors of the construction are found in the 17th century: contrastive negations with an explicitly mentioned alternative. The first examples of litotes, i.e., which lack an explicit alternative and therefore evoke contextual inferences, are found in the 18th century. The evolution is parallel in the three languages, being Sp./Pt. precisamente and Fr. précisément the precursors of Sp. exactamente, Pt. exatamente, and Fr. exactement. The first pragmatic function is mitigation, whereas the ironical understatement (inversion, intensification) is documented later, in the 19th century. Therefore, the evolution leads from a more obvious and accessible pragmatic inference to a cognitively more complex, less transparent, and less compositional one.
The chapter investigates if adding an adverb of the exactness-domain contributes to an easier interpretation of litotes, which are per se vague. The analysis shows that this is not the case: litotes with adverbs of the exactness-domain are as ambiguous as without them. Their function is either mitigation or intensification, but the adverb does not give a cue to which one is the intended interpretation (except for Spanish precisamente). The constructions including an adverb nevertheless have a stable and indispensable function: it evokes a scale of alternatives to the negated element X, and one of the implicit alternatives corresponds to the indirectly intended interpretation of the litotes. The main result of this chapter is that the analyzed construction increases the productivity of litotes because it permits non-scalar elements to enter the slot X, and, by doing so, to acquire an ad-hoc scalarity which is purely subjective and context-depended. Therefore, within this construction, any element in X can be mitigated or intensified as if it was a scalar adjective (such as clean). For example, proper names and non-scalar nouns or prepositional phrases can appear in the slot X: Sp. Quizás Ana no fuera exactamente Serena, Pt. Ele não é exatamente meu amigo, and Fr. j’étais pas exactement aux gâteaux. These sentences would not be interpreted as litotes without the exactness-adverbs. In some cases, the analyzed sentences would not even be semantically possible.
On the one hand, the analyzed construction inherits its pragmatic functions from the litotes. Already in Latin, litotes had both functions: mitigation and ironic understatement (intensification, ironic inversion). On the other hand, the construction expands the applicability of litotes regarding the slot X (“host class expansion”), and therefore increases the productivity of this stylistic-rhetorical device. It makes litotes possible with linguistic elements which outside this construction could not be interpreted as such.
Studia Linguistica Romanica 3, 2020
This paper describes the diachrony of the adverbial locution al justo, and its marginal variants ... more This paper describes the diachrony of the adverbial locution al justo, and its marginal variants a lo justo and al cabal (prepositional adverbials of the pattern “Preposition + Adjective”). These adverbials are seen as competing counterparts to adjective-adverbs (justo) and derived adverbs (justamente, cabalmente). In addition to the data found in Gerhalter (2020, dissertation thesis “Paradigmas…”), this paper relies on more data compiled from CORDE, as well as from Frantext for the French prepositional adverbial au juste.
The creation of al justo, a lo justo and al cabal is related to the elaboration of variants, i.e., to the lexical “enrichment” tendency of 16th century Spanish. The adverbial al justo originates during the 16th century as a synonym for the short adverb justo, ‘exactly’. It is used as manner adverbial, as well as a focus adverb. Its creation is closely connected to the elaboration of written language (“Sprachausbau”) in technical texts, especially on handicraft-skills. These texts contain instructions for calculating and constructing tools, e.g. for navigation, goldsmithing, or architecture. They are written in an “intermediate” register, which is at the same time elaborated and understandable, since these “manuals” need to be transparent and clear. This may explain why the pattern [a + article + adjective expressing exactness] is only productive with inherited adjectives or those which enter the popular tradition: the locution al justo is copied, during the 16th century, by its former synonym cabal (> al cabal), but not with the learned (and more recently borrowed) adjectives preciso or exacto.
The adverbial locution al justo expands to other text-types (religious and administrative prose, narrative texts, and poetry) and develops new meanings (‘appropriate, adequate’), especially within the collocation venir al justo. The quantitative corpus data shows a short upward trend at the end of the 16th century. Al justo remains the most frequent variant of the lexical group al justo – justo – justamente until the middle of the 17th century. Nevertheless, the frequency of al justo falls during the second half of the 17th century, and the adverbial al cabal even disappears completely. The marginalization of these prepositional adverbials is caused by the competition and selection between several variants that form the paradigm of exactness-adverbs. Since derived adverbs were selected and favored by normative tendencies in 17th and 18th century Spanish, al justo was overtaken and outdated by justamente, which becomes the most frequent adverbial of the lexical group al justo – justo – justamente. Consequently, al justo and al cabal were also omitted during the development of discourse functions led by the adverbs ending in -mente (e.g., justamente, precisamente, cabalmente). Nevertheless, the corpus data shows that the imposition of adverbs ending in -mente had less impact in Latin American Spanish: the prepositional adverbial al justo is still more vital in contemporary Latin American Spanish than in European Spanish.
The contrastive study with French au juste shows that, starting from a similar initial panorama of enriching the adverbial paradigm with new formations, the selection between (synonymous) variants may lead to diverging semantic-functional developments in different languages. French au juste was originally used in the same contexts and with the same meanings as Spanish al justo. Nevertheless, French au juste was not ruled out, since the normative selection-process in French favored “periphrastic adverbs” in order to compensate the excess of adverbs ending in -mente. Therefore, French au juste developed stable uses (e.g. in questions and with epistemic verbs), whereas Spanish al justo was reduced almost exclusively to collocations which do not exist in French (e.g. venir/estar/vivir al justo). As a result, present-day Spanish al justo and French au juste are cognates but not equivalents, because they occupy different functional-semantic “parcels” inside the complex of adverbial functions of the exactness-domain.
Studia Universitatis Babeș-Bolyai. Philologia LXV (4), 2020
This paper focuses on adverbs with an adjectival lexical base in Modern Brazilian Portuguese (=BP... more This paper focuses on adverbs with an adjectival lexical base in Modern Brazilian Portuguese (=BP). We compare the frequencies of three different types of adverbials: adverbs in -mente (e.g. absolutamente), adjective adverbs (e.g. alto in falar alto ‘speak loudly’) and prepositional phrases of the type “Preposition + Adjective” (e.g. de novo), as they often form groups based on the same root (e.g. primeiramente – primeiro – de primeiro; seriamente – sério – a sério). We compare their type- and token-frequencies in spoken and written BP. The data is based on a scrutiny of the Discurso & Gramática-corpus which consists of oral interviews of 171 informants and their written texts on the same topic as the one in the interviews. Hence, this corpus allows to contrast spoken and written language of the same informants.
During the project “Open-Access-Database: Adjective-adverb Interfaces in Romance” (http://gams.uni-graz.at/context:aaif), the corpus was annotated and lemmatized. In total, over 4.000 examples (tokens) of adverbs based on adjectival roots were found in the Discurso & Gramática-corpus. There are 267 different adverbials (types = lemmas), which are based on 221 different base adjectives (lexical roots). The present paper presents a first quantitative analysis of this data, within the framework of a research project on prepositional adverbials. The main findings of the paper, contrasting the three different types of adverbials (adjective-adverbs = AA, adverbs ending in -mente, and prepositional adverbs = PA), are to be found below.
In present-day BP, AA are the most frequent in terms of token-frequency, whereas adverbs in -mente are the most frequent in terms of type-frequency. Hence, there is a smaller inventory of highly frequent AA (types) in comparison to a more diversified inventory of less frequently used -mente-adverbs (types). The data shows that -mente is the most productive pattern to form new adverbs based on adjectives (e.g., half of the lexemes occur just once, i.e., only one token). PA-adverbials are less frequent - both in terms of types and tokens - than AA and -mente. Only one PA, de novo, is used in a considerable manner, whereas the other forms are marginal. Regarding the overlapping of adverbials based on the same lexical root (i.e., groups like seriamente – sério – a sério), we observe a tendency towards lexical differentiation in terms of selecting one adverbial-type or another: most lexical roots (base adjectives) appear either as mente-adverbs or as AA. Only regarding the (scarce) PA-adverbials we observe a tendency towards lexical overlapping and possible synonymy: PA occur mostly (79%) in groups with mente-adverbs or AA based on the same adjectival root.
Regarding the (relative/normalized) frequencies in spoken and written code, AAs - the most frequent formation type in both subcorpora - are more frequent in the oral corpus than in the written corpus. Mente-adverbs, on the other hand, show a clear preference for the written code: their frequency in the written corpus is twice as high as in the spoken corpus. Code-based variation is most salient for mente-adverbs and the study confirms the prevalence of mente-adverbs in written language. As shown by two examples, speakers may substitute AA such as só and PA such as de novo by mente-adverbs (somente and novamente, respectively) when writing down a story they told before.
Both in the written and spoken subcorpora, PA are less frequent than mente-adverbs and AA. Regarding the code, there is no remarkable difference in the case of PA, since they occur almost equally in the oral and the written subcorpus and are only slightly more frequent in the written database. Furthermore, in the context of lexical diversification (i.e., different types = different adjectival roots), most PA-types are used in both subcorpora, whereas mente-adverbs (types) are the most diversified in the written subcorpus and AA (types) in the spoken one. These results indicate that PA form a small inventory of lexicalized forms, which are equally used in written and spoken BP as somewhat ‘neutral’ forms that are not marked for any code.
Estudios Humanísticos. Filología, 2018
El objetivo de este trabajo es el estudio diacrónico de dos marcadores de reformulación que prese... more El objetivo de este trabajo es el estudio diacrónico de dos marcadores de reformulación que presentan paralelismos formales (más + adverbio) y semánticos. Sin embargo, la comparación de más exactamente y más precisamente muestra que no surgen paralelamente ni tampoco por los mismos caminos de evolu-ción. Adaptamos el modelo teórico de cooptation para diferenciar entre creación innovadora y analogía. Además, estos marcadores prueban la distinción entre reformulación, paráfrasis y corrección. Palabras claves: Marcadores de reformulación, cooptation, analogía, innovación.
The Spanish reformulation markers más exactamente 'more exactly' and más precisamente 'more precisely' show parallel formal structure and semantics. Nevertheless, the comparative study demonstrates that they did develop neither in a parallel manner nor through the same diachronic clines. To differentiate between innovative creation and analogy, we adapt the theoretical model of cooptation. Additionally, these markers prove the distinction between reformulation, paraphrase and correction.
Enunciado y Discurso: Estructura y Relaciones (eds. María E. Brenes Pena, Marina González Sanz, & Francisco Grande Alija), 2018
El presente estudio se centra en una serie de marcadores de afirmación que tienen su origen en el... more El presente estudio se centra en una serie de marcadores de afirmación que tienen su origen en el concepto de exactitud: exactamente, exacto, justamente, justo y precisamente. El análisis diacrónico muestra que esta función aparece entre finales del siglo XVIII y el siglo XIX, coincidiendo con la aparición de otros marcadores de afirmación. Descartamos la hipótesis de un origen elíptico y aludimos a factores semánticos y sintácticos que indican una gramaticalización incompleta o atípica de estos marcadores.
This paper describes the creation, the annotation process and the model of the Open Access Databa... more This paper describes the creation, the annotation process and the model of the Open Access Database 'Adjective-Adverb Interfaces in Romance' (AAIF) project, with its approach to the creation of a domain-specific ontology. In order to make research data accessible, interoperable, extensible, and transferable, data is annotated in TEI/XML, formalized and enriched with RDF and its conceptual data model is stored in and published via the GAMS digital repository. This produces semantically-enriched, annotated multilingual research data that allows retrieval across heterogeneous corpora. The annotation model expressed in the ontology is offered for further reuse.
CHIMERA: Romance Corpora and Linguistic Studies, 2018
The project Open Access Database: Adjective-Adverb Interfaces in Romance aims at the creation of ... more The project Open Access Database: Adjective-Adverb Interfaces in Romance aims at the creation of an annotated and lemmatised corpus of various linguistic phenomena related to Romance adjectives with adverbial functions. In this paper, we will explain the currently ongoing process of data compilation as well as the morphosyntactic and semantic categories for a thorough annotation by means of some Spanish examples.
Online databases by Katharina Gerhalter
Online Corpus, 2021
The “Adjective-Adverb Interfaces in Romance” database is an annotated and partly-lemmatised corpu... more The “Adjective-Adverb Interfaces in Romance” database is an annotated and partly-lemmatised corpus of Romance adjectives used in adverbial functions (e.g. Fr. voler haut, aller droit, also as modifiers of other word classes than the verb, e.g. the adjective in Sp. es cosa harto seria) and as part of prepositional adverbials (Pt. de novo, Sp. a ciegas). It integrates several corpora of different Romance languages (French, Italian varieties, Portuguese, Romanian, Spanish) and Latin. Currently (version 2021), the database comprises a total of ca. 35,000 annotated examples.
A search interface allows for queries in each separate corpus as well as in freely combined corpora. Adverbs are tagged according to a shared annotation model, using the same categories. Thus, the annotation model offers a cross-linguistic categorization for the multifunctional word-class “adverb” based on its forms, functions and meanings. In the future, the corpus may be enlarged with new datasets using the same annotation model.
This corpus compiles adjective-adverbs in four regional varieties of Southern Italy. The examples... more This corpus compiles adjective-adverbs in four regional varieties of Southern Italy. The examples from Sicilian, Calabrian and Salentino dialects come from contemporaneous texts. The much larger corpus of Neapolitan examples ranges from the the 14th century to the present day. The corpus is based on individual readings of whole texts (the bibliography is listed under sources). The reading and compilation of examples was carried out by Adam Ledgeway (2009-2017) and the examples are analysed in several publications regarding features like adverbial agreement (2009; 2011; 2017). The data have been annotated by Katharina Gerhalter (2019). In order to unify variation in different dialects and to enable easier interrogation of individual lemmas, the verbs and adverbs in the examples were lemmatized with the corresponding standard Italian lemma. In cases where there is no standard Italian equivalent, a reconstructed Latin lemma is proposed preceded by an asterisk (for example, *COMMOGLIARE for the Neapolitan verb forms such as commoglio ‘I cover’).
Books by Katharina Gerhalter
Beihefte zur Zeitschrift für romanische Philologie, 448, 2020
Este estudio basado en datos de corpus ofrece un análisis semasiológico y onomasiológico del para... more Este estudio basado en datos de corpus ofrece un análisis semasiológico y onomasiológico del paradigma de los adjetivos y adverbios de exactitud. En la primera parte del trabajo se rastrea el desarrollo semántico particular de cada par léxico ( justo / justamente, cabal / cabalmente, preciso / precisamente y exacto / exactamente). La segunda parte se centra en el desarrollo diacrónico de tres funciones pragmático-discursivas: focalización, afirmación y reformulación. Los datos analizados muestran que cada uno de estos tres microparadigmas pragmáticos cuenta con un lexema central, que es el que lidera los cambios semánticos y sintácticos que llevan al desarrollo de valores discursivos. Los demás adverbios y/o adjetivos adoptan estos nuevos usos por analogía, para lo que se propone el término «efecto paradigmático». Teniendo en cuenta el carácter polisémico y las distintas trayectorias particulares de cada par léxico, es posible explicar por qué no todos los adjetivos y adverbios de exactitud adoptan cada uno de los usos pragmáticos. Los resultados del análisis se insertan en los modelos teóricos de gramaticalización (más específicamente, pragmaticalización y subjetivización) y de cooptation (thetical grammar), que por sí mismos no son suficientes para explicar factores como las confluencias que se dan dentro de un determinado paradigma.
Drafts by Katharina Gerhalter
![Research paper thumbnail of El silencio no es precisamente una de sus virtudes... Diachrony and pragmatics of the construction [negation + focus adverb of the exactness-domain + X] in French, Portuguese, and Spanish](https://attachments.academia-assets.com/100635165/thumbnails/1.jpg)
Translation into English of the German article: ""El silencio no es precisamente una de sus virtu... more Translation into English of the German article: ""El silencio no es precisamente una de sus virtudes… Diachronie und Pragmatik der Konstruktion [Negation + Exaktheits-Fokusadverb + X] im Französischen, Portugiesischen und Spanischen" (2021)
This paper deals with the semi-schematic construction [negation + focus adverb of the exactness-domain + X], which shows a similar form and equivalent functions in French, Portuguese, and Spanish. For example, Fr. pas précisément propres, Pt. não exatamente limpo and Sp. no precisamente limpia (‘not exactly clean’) are semantically ambiguous regarding the degree of the intended interpretation ‘(more or less) dirty’. This vagueness is intentional and leads to a wide range of possible contextual inferences. Consequently, the constructions Fr. [(ne) pas précisément/exactement X], Pt. [não precisamente/exatamente X] and Sp. [no precisamente/exactamente X] have different pragmatic functions: e.g., no precisamente limpia may be used for mitigation (‘not really clean’), or as an ironic understatement (hence, intensification and inversion: ‘very dirty’). The analyzed construction is a specific sub-type (or meso-construction) of the stylistic-rhetorical device litotes; the above-mentioned pragmatic functions have been observed for litotes in general. This chapter analyses the diachrony of the construction [negation + focus adverb of the exactness-domain + X] in Spanish, French and Portuguese based on large corpus samples and shows how these pragmatic functions develop as contextual inferences. First precursors of the construction are found in the 17th century: contrastive negations with an explicitly mentioned alternative. The first examples of litotes, i.e., which lack an explicit alternative and therefore evoke contextual inferences, are found in the 18th century. The evolution is parallel in the three languages, being Sp./Pt. precisamente and Fr. précisément the precursors of Sp. exactamente, Pt. exatamente, and Fr. exactement. The first pragmatic function is mitigation, whereas the ironical understatement (inversion, intensification) is documented later, in the 19th century. Therefore, the evolution leads from a more obvious and accessible pragmatic inference to a cognitively more complex, less transparent, and less compositional one. The chapter investigates if adding an adverb of the exactness-domain contributes to an easier interpretation of litotes, which are per se vague. The analysis shows that this is not the case: litotes with adverbs of the exactness-domain are as ambiguous as without them. Their function is either mitigation or intensification, but the adverb does not give a cue to which one is the intended interpretation (except for Spanish precisamente). The constructions including an adverb nevertheless have a stable and indispensable function: it evokes a scale of alternatives to the negated element X, and one of the implicit alternatives corresponds to the indirectly intended interpretation of the litotes. The main result of this chapter is that the analyzed construction increases the productivity of litotes because it permits non-scalar elements to enter the slot X, and, by doing so, to acquire an ad-hoc scalarity which is purely subjective and context-depended. Therefore, within this construction, any element in X can be mitigated or intensified as if it was a scalar adjective (such as clean). For example, proper names and non-scalar nouns or prepositional phrases can appear in the slot X: Sp. Quizás Ana no fuera exactamente Serena, Pt. Ele não é exatamente meu amigo, and Fr. j’étais pas exactement aux gâteaux. These sentences would not be interpreted as litotes without the exactness-adverbs. In some cases, the analyzed sentences would not even be semantically possible. On the one hand, the analyzed construction inherits its pragmatic functions from the litotes. Already in Latin, litotes had both functions: mitigation and ironic understatement (intensification, ironic inversion). On the other hand, the construction expands the applicability of litotes regarding the slot X (“host class expansion”), and therefore increases the productivity of this stylistic-rhetorical device. It makes litotes possible with linguistic elements which outside this construction could not be interpreted as such.
Resumen: A falta de estudios históricos sobre cabalmente, un adverbio poco usado en la actualidad... more Resumen: A falta de estudios históricos sobre cabalmente, un adverbio poco usado en la actualidad, nuestra investigación aclara su temprana fecha de documentación en los corpus, su origen popular y su desarrollo semántico-pragmático dentro del paradigma de los adverbios de exactitud, que adquieren usos pragmáticodiscursivos de focalización y afirmación. Abstract: The Spanish adverb cabalmente has not yet been studied from a diachronic point of view. Our research brings up its early attestation in the corpus and its popular origin as well as its semantic-pragmatic evolution belonging to the paradigm of adverbs expressing exactness. Cabalmente also develops new pragmatic or discourse functions, e.g. focus and affirmation.
Uploads
Papers by Katharina Gerhalter
This chapter analyses the diachrony of the construction [negation + focus adverb of the exactness-domain + X] in Spanish, French and Portuguese based on large corpus samples and shows how these pragmatic functions develop as contextual inferences. First precursors of the construction are found in the 17th century: contrastive negations with an explicitly mentioned alternative. The first examples of litotes, i.e., which lack an explicit alternative and therefore evoke contextual inferences, are found in the 18th century. The evolution is parallel in the three languages, being Sp./Pt. precisamente and Fr. précisément the precursors of Sp. exactamente, Pt. exatamente, and Fr. exactement. The first pragmatic function is mitigation, whereas the ironical understatement (inversion, intensification) is documented later, in the 19th century. Therefore, the evolution leads from a more obvious and accessible pragmatic inference to a cognitively more complex, less transparent, and less compositional one.
The chapter investigates if adding an adverb of the exactness-domain contributes to an easier interpretation of litotes, which are per se vague. The analysis shows that this is not the case: litotes with adverbs of the exactness-domain are as ambiguous as without them. Their function is either mitigation or intensification, but the adverb does not give a cue to which one is the intended interpretation (except for Spanish precisamente). The constructions including an adverb nevertheless have a stable and indispensable function: it evokes a scale of alternatives to the negated element X, and one of the implicit alternatives corresponds to the indirectly intended interpretation of the litotes. The main result of this chapter is that the analyzed construction increases the productivity of litotes because it permits non-scalar elements to enter the slot X, and, by doing so, to acquire an ad-hoc scalarity which is purely subjective and context-depended. Therefore, within this construction, any element in X can be mitigated or intensified as if it was a scalar adjective (such as clean). For example, proper names and non-scalar nouns or prepositional phrases can appear in the slot X: Sp. Quizás Ana no fuera exactamente Serena, Pt. Ele não é exatamente meu amigo, and Fr. j’étais pas exactement aux gâteaux. These sentences would not be interpreted as litotes without the exactness-adverbs. In some cases, the analyzed sentences would not even be semantically possible.
On the one hand, the analyzed construction inherits its pragmatic functions from the litotes. Already in Latin, litotes had both functions: mitigation and ironic understatement (intensification, ironic inversion). On the other hand, the construction expands the applicability of litotes regarding the slot X (“host class expansion”), and therefore increases the productivity of this stylistic-rhetorical device. It makes litotes possible with linguistic elements which outside this construction could not be interpreted as such.
The creation of al justo, a lo justo and al cabal is related to the elaboration of variants, i.e., to the lexical “enrichment” tendency of 16th century Spanish. The adverbial al justo originates during the 16th century as a synonym for the short adverb justo, ‘exactly’. It is used as manner adverbial, as well as a focus adverb. Its creation is closely connected to the elaboration of written language (“Sprachausbau”) in technical texts, especially on handicraft-skills. These texts contain instructions for calculating and constructing tools, e.g. for navigation, goldsmithing, or architecture. They are written in an “intermediate” register, which is at the same time elaborated and understandable, since these “manuals” need to be transparent and clear. This may explain why the pattern [a + article + adjective expressing exactness] is only productive with inherited adjectives or those which enter the popular tradition: the locution al justo is copied, during the 16th century, by its former synonym cabal (> al cabal), but not with the learned (and more recently borrowed) adjectives preciso or exacto.
The adverbial locution al justo expands to other text-types (religious and administrative prose, narrative texts, and poetry) and develops new meanings (‘appropriate, adequate’), especially within the collocation venir al justo. The quantitative corpus data shows a short upward trend at the end of the 16th century. Al justo remains the most frequent variant of the lexical group al justo – justo – justamente until the middle of the 17th century. Nevertheless, the frequency of al justo falls during the second half of the 17th century, and the adverbial al cabal even disappears completely. The marginalization of these prepositional adverbials is caused by the competition and selection between several variants that form the paradigm of exactness-adverbs. Since derived adverbs were selected and favored by normative tendencies in 17th and 18th century Spanish, al justo was overtaken and outdated by justamente, which becomes the most frequent adverbial of the lexical group al justo – justo – justamente. Consequently, al justo and al cabal were also omitted during the development of discourse functions led by the adverbs ending in -mente (e.g., justamente, precisamente, cabalmente). Nevertheless, the corpus data shows that the imposition of adverbs ending in -mente had less impact in Latin American Spanish: the prepositional adverbial al justo is still more vital in contemporary Latin American Spanish than in European Spanish.
The contrastive study with French au juste shows that, starting from a similar initial panorama of enriching the adverbial paradigm with new formations, the selection between (synonymous) variants may lead to diverging semantic-functional developments in different languages. French au juste was originally used in the same contexts and with the same meanings as Spanish al justo. Nevertheless, French au juste was not ruled out, since the normative selection-process in French favored “periphrastic adverbs” in order to compensate the excess of adverbs ending in -mente. Therefore, French au juste developed stable uses (e.g. in questions and with epistemic verbs), whereas Spanish al justo was reduced almost exclusively to collocations which do not exist in French (e.g. venir/estar/vivir al justo). As a result, present-day Spanish al justo and French au juste are cognates but not equivalents, because they occupy different functional-semantic “parcels” inside the complex of adverbial functions of the exactness-domain.
During the project “Open-Access-Database: Adjective-adverb Interfaces in Romance” (http://gams.uni-graz.at/context:aaif), the corpus was annotated and lemmatized. In total, over 4.000 examples (tokens) of adverbs based on adjectival roots were found in the Discurso & Gramática-corpus. There are 267 different adverbials (types = lemmas), which are based on 221 different base adjectives (lexical roots). The present paper presents a first quantitative analysis of this data, within the framework of a research project on prepositional adverbials. The main findings of the paper, contrasting the three different types of adverbials (adjective-adverbs = AA, adverbs ending in -mente, and prepositional adverbs = PA), are to be found below.
In present-day BP, AA are the most frequent in terms of token-frequency, whereas adverbs in -mente are the most frequent in terms of type-frequency. Hence, there is a smaller inventory of highly frequent AA (types) in comparison to a more diversified inventory of less frequently used -mente-adverbs (types). The data shows that -mente is the most productive pattern to form new adverbs based on adjectives (e.g., half of the lexemes occur just once, i.e., only one token). PA-adverbials are less frequent - both in terms of types and tokens - than AA and -mente. Only one PA, de novo, is used in a considerable manner, whereas the other forms are marginal. Regarding the overlapping of adverbials based on the same lexical root (i.e., groups like seriamente – sério – a sério), we observe a tendency towards lexical differentiation in terms of selecting one adverbial-type or another: most lexical roots (base adjectives) appear either as mente-adverbs or as AA. Only regarding the (scarce) PA-adverbials we observe a tendency towards lexical overlapping and possible synonymy: PA occur mostly (79%) in groups with mente-adverbs or AA based on the same adjectival root.
Regarding the (relative/normalized) frequencies in spoken and written code, AAs - the most frequent formation type in both subcorpora - are more frequent in the oral corpus than in the written corpus. Mente-adverbs, on the other hand, show a clear preference for the written code: their frequency in the written corpus is twice as high as in the spoken corpus. Code-based variation is most salient for mente-adverbs and the study confirms the prevalence of mente-adverbs in written language. As shown by two examples, speakers may substitute AA such as só and PA such as de novo by mente-adverbs (somente and novamente, respectively) when writing down a story they told before.
Both in the written and spoken subcorpora, PA are less frequent than mente-adverbs and AA. Regarding the code, there is no remarkable difference in the case of PA, since they occur almost equally in the oral and the written subcorpus and are only slightly more frequent in the written database. Furthermore, in the context of lexical diversification (i.e., different types = different adjectival roots), most PA-types are used in both subcorpora, whereas mente-adverbs (types) are the most diversified in the written subcorpus and AA (types) in the spoken one. These results indicate that PA form a small inventory of lexicalized forms, which are equally used in written and spoken BP as somewhat ‘neutral’ forms that are not marked for any code.
The Spanish reformulation markers más exactamente 'more exactly' and más precisamente 'more precisely' show parallel formal structure and semantics. Nevertheless, the comparative study demonstrates that they did develop neither in a parallel manner nor through the same diachronic clines. To differentiate between innovative creation and analogy, we adapt the theoretical model of cooptation. Additionally, these markers prove the distinction between reformulation, paraphrase and correction.
Online databases by Katharina Gerhalter
A search interface allows for queries in each separate corpus as well as in freely combined corpora. Adverbs are tagged according to a shared annotation model, using the same categories. Thus, the annotation model offers a cross-linguistic categorization for the multifunctional word-class “adverb” based on its forms, functions and meanings. In the future, the corpus may be enlarged with new datasets using the same annotation model.
Books by Katharina Gerhalter
Drafts by Katharina Gerhalter
This paper deals with the semi-schematic construction [negation + focus adverb of the exactness-domain + X], which shows a similar form and equivalent functions in French, Portuguese, and Spanish. For example, Fr. pas précisément propres, Pt. não exatamente limpo and Sp. no precisamente limpia (‘not exactly clean’) are semantically ambiguous regarding the degree of the intended interpretation ‘(more or less) dirty’. This vagueness is intentional and leads to a wide range of possible contextual inferences. Consequently, the constructions Fr. [(ne) pas précisément/exactement X], Pt. [não precisamente/exatamente X] and Sp. [no precisamente/exactamente X] have different pragmatic functions: e.g., no precisamente limpia may be used for mitigation (‘not really clean’), or as an ironic understatement (hence, intensification and inversion: ‘very dirty’). The analyzed construction is a specific sub-type (or meso-construction) of the stylistic-rhetorical device litotes; the above-mentioned pragmatic functions have been observed for litotes in general. This chapter analyses the diachrony of the construction [negation + focus adverb of the exactness-domain + X] in Spanish, French and Portuguese based on large corpus samples and shows how these pragmatic functions develop as contextual inferences. First precursors of the construction are found in the 17th century: contrastive negations with an explicitly mentioned alternative. The first examples of litotes, i.e., which lack an explicit alternative and therefore evoke contextual inferences, are found in the 18th century. The evolution is parallel in the three languages, being Sp./Pt. precisamente and Fr. précisément the precursors of Sp. exactamente, Pt. exatamente, and Fr. exactement. The first pragmatic function is mitigation, whereas the ironical understatement (inversion, intensification) is documented later, in the 19th century. Therefore, the evolution leads from a more obvious and accessible pragmatic inference to a cognitively more complex, less transparent, and less compositional one. The chapter investigates if adding an adverb of the exactness-domain contributes to an easier interpretation of litotes, which are per se vague. The analysis shows that this is not the case: litotes with adverbs of the exactness-domain are as ambiguous as without them. Their function is either mitigation or intensification, but the adverb does not give a cue to which one is the intended interpretation (except for Spanish precisamente). The constructions including an adverb nevertheless have a stable and indispensable function: it evokes a scale of alternatives to the negated element X, and one of the implicit alternatives corresponds to the indirectly intended interpretation of the litotes. The main result of this chapter is that the analyzed construction increases the productivity of litotes because it permits non-scalar elements to enter the slot X, and, by doing so, to acquire an ad-hoc scalarity which is purely subjective and context-depended. Therefore, within this construction, any element in X can be mitigated or intensified as if it was a scalar adjective (such as clean). For example, proper names and non-scalar nouns or prepositional phrases can appear in the slot X: Sp. Quizás Ana no fuera exactamente Serena, Pt. Ele não é exatamente meu amigo, and Fr. j’étais pas exactement aux gâteaux. These sentences would not be interpreted as litotes without the exactness-adverbs. In some cases, the analyzed sentences would not even be semantically possible. On the one hand, the analyzed construction inherits its pragmatic functions from the litotes. Already in Latin, litotes had both functions: mitigation and ironic understatement (intensification, ironic inversion). On the other hand, the construction expands the applicability of litotes regarding the slot X (“host class expansion”), and therefore increases the productivity of this stylistic-rhetorical device. It makes litotes possible with linguistic elements which outside this construction could not be interpreted as such.
This chapter analyses the diachrony of the construction [negation + focus adverb of the exactness-domain + X] in Spanish, French and Portuguese based on large corpus samples and shows how these pragmatic functions develop as contextual inferences. First precursors of the construction are found in the 17th century: contrastive negations with an explicitly mentioned alternative. The first examples of litotes, i.e., which lack an explicit alternative and therefore evoke contextual inferences, are found in the 18th century. The evolution is parallel in the three languages, being Sp./Pt. precisamente and Fr. précisément the precursors of Sp. exactamente, Pt. exatamente, and Fr. exactement. The first pragmatic function is mitigation, whereas the ironical understatement (inversion, intensification) is documented later, in the 19th century. Therefore, the evolution leads from a more obvious and accessible pragmatic inference to a cognitively more complex, less transparent, and less compositional one.
The chapter investigates if adding an adverb of the exactness-domain contributes to an easier interpretation of litotes, which are per se vague. The analysis shows that this is not the case: litotes with adverbs of the exactness-domain are as ambiguous as without them. Their function is either mitigation or intensification, but the adverb does not give a cue to which one is the intended interpretation (except for Spanish precisamente). The constructions including an adverb nevertheless have a stable and indispensable function: it evokes a scale of alternatives to the negated element X, and one of the implicit alternatives corresponds to the indirectly intended interpretation of the litotes. The main result of this chapter is that the analyzed construction increases the productivity of litotes because it permits non-scalar elements to enter the slot X, and, by doing so, to acquire an ad-hoc scalarity which is purely subjective and context-depended. Therefore, within this construction, any element in X can be mitigated or intensified as if it was a scalar adjective (such as clean). For example, proper names and non-scalar nouns or prepositional phrases can appear in the slot X: Sp. Quizás Ana no fuera exactamente Serena, Pt. Ele não é exatamente meu amigo, and Fr. j’étais pas exactement aux gâteaux. These sentences would not be interpreted as litotes without the exactness-adverbs. In some cases, the analyzed sentences would not even be semantically possible.
On the one hand, the analyzed construction inherits its pragmatic functions from the litotes. Already in Latin, litotes had both functions: mitigation and ironic understatement (intensification, ironic inversion). On the other hand, the construction expands the applicability of litotes regarding the slot X (“host class expansion”), and therefore increases the productivity of this stylistic-rhetorical device. It makes litotes possible with linguistic elements which outside this construction could not be interpreted as such.
The creation of al justo, a lo justo and al cabal is related to the elaboration of variants, i.e., to the lexical “enrichment” tendency of 16th century Spanish. The adverbial al justo originates during the 16th century as a synonym for the short adverb justo, ‘exactly’. It is used as manner adverbial, as well as a focus adverb. Its creation is closely connected to the elaboration of written language (“Sprachausbau”) in technical texts, especially on handicraft-skills. These texts contain instructions for calculating and constructing tools, e.g. for navigation, goldsmithing, or architecture. They are written in an “intermediate” register, which is at the same time elaborated and understandable, since these “manuals” need to be transparent and clear. This may explain why the pattern [a + article + adjective expressing exactness] is only productive with inherited adjectives or those which enter the popular tradition: the locution al justo is copied, during the 16th century, by its former synonym cabal (> al cabal), but not with the learned (and more recently borrowed) adjectives preciso or exacto.
The adverbial locution al justo expands to other text-types (religious and administrative prose, narrative texts, and poetry) and develops new meanings (‘appropriate, adequate’), especially within the collocation venir al justo. The quantitative corpus data shows a short upward trend at the end of the 16th century. Al justo remains the most frequent variant of the lexical group al justo – justo – justamente until the middle of the 17th century. Nevertheless, the frequency of al justo falls during the second half of the 17th century, and the adverbial al cabal even disappears completely. The marginalization of these prepositional adverbials is caused by the competition and selection between several variants that form the paradigm of exactness-adverbs. Since derived adverbs were selected and favored by normative tendencies in 17th and 18th century Spanish, al justo was overtaken and outdated by justamente, which becomes the most frequent adverbial of the lexical group al justo – justo – justamente. Consequently, al justo and al cabal were also omitted during the development of discourse functions led by the adverbs ending in -mente (e.g., justamente, precisamente, cabalmente). Nevertheless, the corpus data shows that the imposition of adverbs ending in -mente had less impact in Latin American Spanish: the prepositional adverbial al justo is still more vital in contemporary Latin American Spanish than in European Spanish.
The contrastive study with French au juste shows that, starting from a similar initial panorama of enriching the adverbial paradigm with new formations, the selection between (synonymous) variants may lead to diverging semantic-functional developments in different languages. French au juste was originally used in the same contexts and with the same meanings as Spanish al justo. Nevertheless, French au juste was not ruled out, since the normative selection-process in French favored “periphrastic adverbs” in order to compensate the excess of adverbs ending in -mente. Therefore, French au juste developed stable uses (e.g. in questions and with epistemic verbs), whereas Spanish al justo was reduced almost exclusively to collocations which do not exist in French (e.g. venir/estar/vivir al justo). As a result, present-day Spanish al justo and French au juste are cognates but not equivalents, because they occupy different functional-semantic “parcels” inside the complex of adverbial functions of the exactness-domain.
During the project “Open-Access-Database: Adjective-adverb Interfaces in Romance” (http://gams.uni-graz.at/context:aaif), the corpus was annotated and lemmatized. In total, over 4.000 examples (tokens) of adverbs based on adjectival roots were found in the Discurso & Gramática-corpus. There are 267 different adverbials (types = lemmas), which are based on 221 different base adjectives (lexical roots). The present paper presents a first quantitative analysis of this data, within the framework of a research project on prepositional adverbials. The main findings of the paper, contrasting the three different types of adverbials (adjective-adverbs = AA, adverbs ending in -mente, and prepositional adverbs = PA), are to be found below.
In present-day BP, AA are the most frequent in terms of token-frequency, whereas adverbs in -mente are the most frequent in terms of type-frequency. Hence, there is a smaller inventory of highly frequent AA (types) in comparison to a more diversified inventory of less frequently used -mente-adverbs (types). The data shows that -mente is the most productive pattern to form new adverbs based on adjectives (e.g., half of the lexemes occur just once, i.e., only one token). PA-adverbials are less frequent - both in terms of types and tokens - than AA and -mente. Only one PA, de novo, is used in a considerable manner, whereas the other forms are marginal. Regarding the overlapping of adverbials based on the same lexical root (i.e., groups like seriamente – sério – a sério), we observe a tendency towards lexical differentiation in terms of selecting one adverbial-type or another: most lexical roots (base adjectives) appear either as mente-adverbs or as AA. Only regarding the (scarce) PA-adverbials we observe a tendency towards lexical overlapping and possible synonymy: PA occur mostly (79%) in groups with mente-adverbs or AA based on the same adjectival root.
Regarding the (relative/normalized) frequencies in spoken and written code, AAs - the most frequent formation type in both subcorpora - are more frequent in the oral corpus than in the written corpus. Mente-adverbs, on the other hand, show a clear preference for the written code: their frequency in the written corpus is twice as high as in the spoken corpus. Code-based variation is most salient for mente-adverbs and the study confirms the prevalence of mente-adverbs in written language. As shown by two examples, speakers may substitute AA such as só and PA such as de novo by mente-adverbs (somente and novamente, respectively) when writing down a story they told before.
Both in the written and spoken subcorpora, PA are less frequent than mente-adverbs and AA. Regarding the code, there is no remarkable difference in the case of PA, since they occur almost equally in the oral and the written subcorpus and are only slightly more frequent in the written database. Furthermore, in the context of lexical diversification (i.e., different types = different adjectival roots), most PA-types are used in both subcorpora, whereas mente-adverbs (types) are the most diversified in the written subcorpus and AA (types) in the spoken one. These results indicate that PA form a small inventory of lexicalized forms, which are equally used in written and spoken BP as somewhat ‘neutral’ forms that are not marked for any code.
The Spanish reformulation markers más exactamente 'more exactly' and más precisamente 'more precisely' show parallel formal structure and semantics. Nevertheless, the comparative study demonstrates that they did develop neither in a parallel manner nor through the same diachronic clines. To differentiate between innovative creation and analogy, we adapt the theoretical model of cooptation. Additionally, these markers prove the distinction between reformulation, paraphrase and correction.
A search interface allows for queries in each separate corpus as well as in freely combined corpora. Adverbs are tagged according to a shared annotation model, using the same categories. Thus, the annotation model offers a cross-linguistic categorization for the multifunctional word-class “adverb” based on its forms, functions and meanings. In the future, the corpus may be enlarged with new datasets using the same annotation model.
This paper deals with the semi-schematic construction [negation + focus adverb of the exactness-domain + X], which shows a similar form and equivalent functions in French, Portuguese, and Spanish. For example, Fr. pas précisément propres, Pt. não exatamente limpo and Sp. no precisamente limpia (‘not exactly clean’) are semantically ambiguous regarding the degree of the intended interpretation ‘(more or less) dirty’. This vagueness is intentional and leads to a wide range of possible contextual inferences. Consequently, the constructions Fr. [(ne) pas précisément/exactement X], Pt. [não precisamente/exatamente X] and Sp. [no precisamente/exactamente X] have different pragmatic functions: e.g., no precisamente limpia may be used for mitigation (‘not really clean’), or as an ironic understatement (hence, intensification and inversion: ‘very dirty’). The analyzed construction is a specific sub-type (or meso-construction) of the stylistic-rhetorical device litotes; the above-mentioned pragmatic functions have been observed for litotes in general. This chapter analyses the diachrony of the construction [negation + focus adverb of the exactness-domain + X] in Spanish, French and Portuguese based on large corpus samples and shows how these pragmatic functions develop as contextual inferences. First precursors of the construction are found in the 17th century: contrastive negations with an explicitly mentioned alternative. The first examples of litotes, i.e., which lack an explicit alternative and therefore evoke contextual inferences, are found in the 18th century. The evolution is parallel in the three languages, being Sp./Pt. precisamente and Fr. précisément the precursors of Sp. exactamente, Pt. exatamente, and Fr. exactement. The first pragmatic function is mitigation, whereas the ironical understatement (inversion, intensification) is documented later, in the 19th century. Therefore, the evolution leads from a more obvious and accessible pragmatic inference to a cognitively more complex, less transparent, and less compositional one. The chapter investigates if adding an adverb of the exactness-domain contributes to an easier interpretation of litotes, which are per se vague. The analysis shows that this is not the case: litotes with adverbs of the exactness-domain are as ambiguous as without them. Their function is either mitigation or intensification, but the adverb does not give a cue to which one is the intended interpretation (except for Spanish precisamente). The constructions including an adverb nevertheless have a stable and indispensable function: it evokes a scale of alternatives to the negated element X, and one of the implicit alternatives corresponds to the indirectly intended interpretation of the litotes. The main result of this chapter is that the analyzed construction increases the productivity of litotes because it permits non-scalar elements to enter the slot X, and, by doing so, to acquire an ad-hoc scalarity which is purely subjective and context-depended. Therefore, within this construction, any element in X can be mitigated or intensified as if it was a scalar adjective (such as clean). For example, proper names and non-scalar nouns or prepositional phrases can appear in the slot X: Sp. Quizás Ana no fuera exactamente Serena, Pt. Ele não é exatamente meu amigo, and Fr. j’étais pas exactement aux gâteaux. These sentences would not be interpreted as litotes without the exactness-adverbs. In some cases, the analyzed sentences would not even be semantically possible. On the one hand, the analyzed construction inherits its pragmatic functions from the litotes. Already in Latin, litotes had both functions: mitigation and ironic understatement (intensification, ironic inversion). On the other hand, the construction expands the applicability of litotes regarding the slot X (“host class expansion”), and therefore increases the productivity of this stylistic-rhetorical device. It makes litotes possible with linguistic elements which outside this construction could not be interpreted as such.