Volume 17, Issue 1

In this paper, we show how three often used and seemingly different discourse annotation frameworks – Penn Discourse Treebank (PDTB), Rhetorical Structure Theory (RST), and Segmented Discourse Representation Theory – can be related by using a set of unifying dimensions. These dimensions are taken from the Cognitive approach to Coherence Relations and combined with more fine-grained additional features from the frameworks themselves to yield a posited set of dimensions that can successfully map three frameworks. The resulting interface will allow researchers to find identical or at least closely related relations within sets of annotated corpora, even if they are annotated within different frameworks. Furthermore, we tested our unified dimension (UniDim) approach by comparing PDTB and RST annotations of identical newspaper texts and converting their original end label annotations of relations into the accompanying values per dimension. Subsequently, rates of overlap in the attributed values per dimension were analyzed. Results indicate that the proposed dimensions indeed create an interface that makes existing annotation systems “talk to each other.”

This paper reports on a corpus-based investigation of the verb geben in two alternating ditransitive constructions in present-day German with the Recipient either coded in the dative case (the indirect object construction, abbreviated: IOC) or by means of the PP an + accusative case (the prepositional object construction, POC). The study is based on a quantitative analysis of N =1,301 sentences (712 IOC and 589 POC) drawn from the Deutsche Referenzkorpus (IDS, Mannheim) which were annotated for 20 factors. Using a logistic regression analysis, we found evidence for the effect of 10 predictors (bootstrapped C-index=95%). We discuss our results in comparison to corpus-based research of the English dative alternation and point to some notable differences between both languages. In German, POC appears to be strongly correlated with collective recipient s, collective agent s, passive voice, and concrete and propositional transfer senses of the main verb geben . With regard to the semantics of geben , we argue that the quantitative findings are best accounted for on the basis of a general underspecified verb meaning ‘ geben transfer’ that does not yet differentiate between the three possible verb senses (concrete, abstract, and propositional).

Only a fraction of lexemes are encountered in all their paradigm forms in any corpus or even in the lifetime of any speaker. This raises a question as to how it is that native speakers confidently produce and comprehend word forms that they have never witnessed. We present the results of an experiment using a recurrent neural network computational learning model. In particular, we compare the model’s production of unencountered forms using two types of training data: full paradigms vs. single word forms for Russian nouns, verbs, and adjectives. In the long run, the model displays better performance when exposed to the more naturalistic training on single word forms, even though the other training data is much larger as it includes full paradigms for each and every word. We discuss why “defective” paradigms may be better for human learners as well.

This article presents dependency profiles (DPs) as an empirical method to investigate linguistic elements and their application to the study of 24 discourse connectives in the 3.7-billion token Finnish Internet Parsebank (http://bionlp-www.utu.fi/dep_search/). DPs are based on co-occurrence patterns of the discourse connectives with dependency syntax relations. They follow the assumption of usage-based models, according to which the semantic and functional properties of linguistic expressions arise based on their distributional characteristics. We focus on the typical usage patterns reflected by the DPs and the (dis)similarities among discourse connectives that these patterns reveal. We demonstrate that 1) DPs can be analyzed with clustering to obtain linguistically meaningful groupings among the connectives and that 2) the clustering can be combined with support vector machines to obtain generic and stable linguistic characteristics of the discourse connectives. We show that this data-driven method offers support for previous results and reveals novel tendencies outside the scope of studies on smaller corpora. As the method is based on automatic syntactic analysis following the cross-linguistic universal dependencies, it does not require manual annotation and can be applied to a number of languages and in contrastive studies.

Säily (2016, Sociolinguistic variation in morphological productivity in eighteenth-century English. Corpus Linguistics and Linguistic Theory 12(1). 129–151) offers a very interesting way of gauging the productivity of a word-formation pattern when the corresponding token counts vary. However, I disagree with the operationalization of productivity that she uses. If we employ vocabulary size as an indicator of productivity, later stages in a language’s life will automatically be more productive. As an alternative, I suggest a neologism-based measure, which is preferable for various reasons.

Academic writing in the second half of the twentieth century witnesses a notable decrease in be -passives in British and American English (AmE). This trend is more advanced in the soft than in the hard sciences; with the exception of AmE, moreover, regional variation is not highly significant. This paper aims to discover whether the use of passives is conditioned by the same factors across seven different varieties of English (both as a first and as an institutionalized second language). For this purpose, we automatically retrieve central be -passives and active transitives from syntactically annotated International Corpus of English corpora and code for factors that are likely to play a role in the choice between active and passive (such as the semantics of the participant roles or the length of the constituents). Our results show that, while the same factors predict the choice of a passive over an active verb phrase across first- and second-language varieties, subtle differences are found in the effect size that some factors (animacy, givenness and length of passive subject) have, notably in Hong Kong and Philippine English. Some (but not all) of these find an explanation in substrate influence.

Gilquin (2008, What you think ain’t what you get: Highly polysemous verbs in mind and language. In Jean-Remi Lapaire, Guillaume Desagulier & Jean-Baptiste Guignard (eds.), From gram to mind: Grammar as cognition , 235–255. Bordeaux: Presse Universitaires de Bordeaux) reported that light uses of verbs (e.g. make use ) tend to outnumber concrete uses of the same verbs (e.g. make furniture ) in corpora, whereas concrete senses tend to outnumber light senses in responses to elicitation tests. The differences between corpus frequency and cognitive salience remain an important and much-discussed question (cf. Arppe et al. 2010, Cognitive corpus linguistics: Five points of debate on current theory and methodology. Corpora 5(1). 1–27). The question is particularly complicated because both corpus frequency and cognitive salience are difficult to define, and are often left undefined. Operationalising and defining corpus frequencies are the issues at the heart of the present paper, which includes a close, manual semantic analysis of nearly 6,000 instances of three polysemous verbs with light and concrete uses, make, take , and give , in the British component of the International Corpus of English. The paper compares semasiological frequencies like those measured by Gilquin (2008) to onomasiological frequency measurements (cf. Geeraerts 1997, Diachronic prototype semantics: A contribution to historical lexicology . Oxford: Clarendon Press). Methodologically, the paper demonstrates that these approaches address fundamentally different research questions, and offer dramatically different results. Findings indicate that corpus frequencies in speech may correlate with elicitation test results, if the corpus frequencies are measured onomasiologically rather than semasiologically; I refer to Geeraerts’s (2010, Theories of lexical semantics . Oxford: Oxford University Press) hypothesis of onomasiological salience in explaining this observation.

This corpus study investigated pronunciation variants of word-final obstruent-liquid-schwa (OLS) clusters in nouns in casual Parisian French. Results showed that at least one phoneme was absent in 80.7% of the 291 noun tokens in the dataset, and that the whole cluster was absent (e.g., [mis] for ministre ) in no less than 15.5% of the tokens. We demonstrate that phonemes are not always completely absent, but that they may leave traces on neighbouring phonemes. Further, the clusters display undocumented voice assimilation patterns. Statistical modelling showed that a phoneme is most likely to be absent if the following phoneme is also absent. The durations of the phonemes are conditioned particularly by the position of the word in the prosodic phrase. We argue, on the basis of three different types of evidence, that in French word-final OLS clusters, the absence of obstruents is mainly due to gradient reduction processes, whereas the absence of schwa and liquids may also be due to categorical deletion processes.