Grodzinsky interprets linguistic differences between agrammatic comprehension and production symptoms as supporting the hypothesis that the mechanisms underlying grammatical encoding (sentence formulation) and grammatical decoding (syntactic parsing) are at least partially distinct. This inference is shown to be premature. A range of experimentally established similarities between the encoding and decoding processes is highlighted, testifying to the viability of the hypothesis that receptive and productive syntactic tasks are performed by the same syntactic processor.

Theories of language production have long been expressed as connectionist models. We outline the issues and challenges that must be addressed by connectionist models of lexical access and grammatical encoding, and review three recent models. The models illustrate the value of an interactive activation approach to lexical access in production, the need for sequential output in both phonological and grammatical encoding, and the potential for accounting for structural effects on errors and structural priming from learning.

Japanese English learners have difficulty speaking Double Object (DO; give B A) than Prepositional Object (PO; give A to B) structures which neural underpinning is unknown. In speaking, syntactic and phonological processing follow semantic encoding, conversion of non-verbal mental representation into a structure suitable for expression. To test whether DO difficulty lies in linguistic or prelinguistic process, we conducted functional magnetic resonance imaging. Thirty participants described cartoons using DO or PO, or simply named them. Greater reaction times and error (...) rates indicated DO difficulty. DO compared with PO showed parieto-frontal activation including left inferior frontal gyrus, reflecting linguistic process. Psychological priming in PO produced immediately after DO and vice versa compared to after control, indicated shared process between PO and DO. Cross-structural neural repetition suppression was observed in occipito-parietal regions, overlapping the linguistic system in pre-SMA. Thus DO and PO share prelinguistic process, whereas linguistic process imposes overload in DO. (shrink)

In this article, we explore whether cross-linguistic differences in grammatical aspect encoding may give rise to differences in memory and cognition. We compared native speakers of two languages that encode aspect differently (English and Swedish) in four tasks that examined verbal descriptions of stimuli, online triads matching, and memory-based triads matching with and without verbal interference. Results showed between-group differences in verbal descriptions and in memory-based triads matching. However, no differences were found in online triads matching and in (...) memory-based triads matching with verbal interference. These findings need to be interpreted in the context of the overall pattern of performance, which indicated that both groups based their similarity judgments on common perceptual characteristics of motion events. These results show for the first time a cross-linguistic difference in memory as a function of differences in grammatical aspect encoding, but they also contribute to the emerging view that language fine tunes rather than shapes perceptual processes that are likely to be universal and unchanging. (shrink)

This paper aims to establish and identify the linguistic devices through which the niche printed media specifically targeting a young female local audience identify, shape and construct their ad- dressee by acknowledging their subcultural ideology. Our intention is to trace the measure of congruency between the two types of discourse: of the encoder and of the decoder. Such instantiations as were found at the level of text functions, discourse patterns and strategies, rhetorical and linguistic items testify to our conclusion that (...) the niche printed media – specifically teen-oriented music and lifestyle magazines – gratify a local community of young female consumers by pro- grammatically encoding their cultural products in a language of solidarity with their audience. (shrink)

Natural languages exhibit a great variety of verbal paradigms. For instance, in English main verbs are grammatically marked for tense, whereas in the tenseless Eskimo-Aleut language Kalaallisut they are marked for illocutionary mood. Although time is a universal dimension of the human experience and speaking is part of that experience, some languages encode discourse reference to time without any grammatical tense morphology (see e.g. Bittner 2005), or reference to speech acts without any illocutionary mood morphology.

This paper presents a computational model that integrates a dynamically structured holographic memory system into the ACT-R cognitive architecture to explain how linguistic representations are encoded and accessed in memory. ACT-R currently serves as the most precise expression of the moment-by-moment working memory retrievals that support sentence comprehension. The ACT-R model of sentence comprehension is able to capture a range of linguistic phenomena, but there are cases where the model makes the wrong predictions, such as the over-prediction of retrieval interference (...) effects during sentence comprehension. Here, we investigate one such case involving the processing of sentences with negative polarity items and consider how a dynamically structured holographic memory system might provide a cognitively plausible and principled explanation of some previously unexplained effects. Specifically, we show that by replacing ACT-R's declarative memory with a dynamically structured memory, we can explain a wider range of behavioral data involving reading times and judgments of grammaticality. We show that our integrated model provides a better fit to human error rates and response latencies than the original ACT-R model. These results provide proof-of-concept for the unification of two independent computational cognitive frameworks. (shrink)

This investigation draws from research on negative polarity item (NPI) illusions in order to explore a new and interesting instance of misalignment observed for grammatical sentences containing two negative markers. Previous research has shown that unlicensed NPIs can be perceived as acceptable when occurring soon after a structurally inaccessible negation (e.g. ever in *The bills that no senators voted for have ever become law). Here we examine the opposite configuration: grammatical sentences created by substituting the NPI ever with (...) the negative adverb never (e.g. The bills that no senators voted for have never become law). The processing and acceptability of these sentences was studied using three tasks: a speeded acceptability judgment (experiment 1), a self-paced reading task (experiment 2), and an offline acceptability rating (experiment 3). The results are consistent across measures in showing that the integration of the adverb never is disrupted by the linearly preceding but structurally inaccessible negative quantifier no in the relative clause. In our view, this pattern of results is in line with Parker and Phillips’ (2016) proposal that NPI illusions arise when the context containing the inaccessible negation has not been fully encoded by the time the NPI ever is encountered, making the embedded negative quantifier transparently available as a licensor. In a similar vein, the disruption effects observed for grammatical sentences containing two negative elements could arise if the negative quantifier is still being integrated when never is encountered, forcing the parser to deal with two negative elements simultaneously. This interpretation suggests that the same incomplete encodings that could be ameliorating the online perception of unlicensed NPIs could also be responsible for deteriorating the perception of the sentences under investigation here. This would represent an illusion of ungrammaticality. Furthermore, these results provide evidence against the speculation that NPI illusions are the consequence of misrepresenting ever as its near neighbor never, given that continuations with never are judged as unacceptable in spite of their grammaticality. Together, these findings inform the landscape of hypotheses on NPI illusions and offer valuable insights on the complexity of multiple negation and the relation between processing difficulty and acceptability. (shrink)

Words are rarely produced in isolation. Yet, our understanding of multi-word production, and especially its time course, is still rather poor. In this research, we use event-related potentials to examine the production of multi-word noun phrases in the context of overt picture naming. We track the processing costs associated with the production of these noun phrases as compared with the production of bare nouns, from picture onset to articulation. Behavioral results revealed longer naming latencies for French noun phrases with determiners (...) and pre-nominal adjectives (D-A-N, the big cat ) than for noun phrases with a determiner (D-N, the cat ), or bare nouns (N, cat ). The spatio-temporal analysis of the ERPs revealed differences in the duration of stable global electrophysiological patterns as a function of utterance format in two time windows, from ~190 to 300 ms after picture onset, and from ~530 ms after picture onset to 100 ms before articulation. These findings can be accommodated in the following model. During grammatical encoding (here from ~190 to 300 ms), the noun and adjective lemmas are accessed in parallel, followed by the selection of the gender-agreeing determiner. Phonological encoding (after ~530 ms) operates sequentially. As a consequence, the phonological encoding process is longer for longer utterances. In addition, when determiners are repeated across trials, their phonological encoding can be anticipated or primed, resulting in a shortened encoding process. (shrink)

There have been essentially two types of theoretical approaches to account for the grammatical relations associated with the causee argument of causative constructions. Ignoring the specifics of particular theories, there are transitivity based approaches in which the causee is a direct object when the embedded clause is intransitive, and an indirect object or oblique when the embedded clause is transitive. This pattern finds considerable cross-linguistic support. On the other hand, there are languages in which the causee exhibits alternative (...) class='Hi'>grammatical relations irrespective of transitivity: the causee direct object correlates with direct causation, while indirect object or oblique causees are associated with indirect causation. Such phenomena have motivated a semantic approach.Focusing primarily on data from Spanish, we account for both sorts of phenomena by proposing a novel extension of Dowty's [(1991) Language 67, 547–619] proto-property proposal, thereby rendering it a comprehensive model of argument selection for both simple and complex predicates. According to Dowty's original Argument Selection Principle, the most proto-patientive argument in a single argument structure tends to be encoded as a direct object. In the case of causatives with intransitive base predicates, the most proto-patientive argument will be the causee. However, if the embedded clause is transitive, the causee will be less proto-patientive that the embedded patient, and will not be encoded as a direct object. Thus, the Argument Selection Principle, operating in a syntagmatic fashion over co-arguments, effectively derives the transitivity-determined causee encodings. In order to address the effects associated with the semantic approach, we develop paradigmatic interpretation of the Argument Selection Principle. That is, when the causee argument varies in degree of proto-patientivity, the most proto-patientive alternant is encoded as a direct object, and any decrease in proto-patientivity is reflected by relational encodings that are sequentially lower than direct object on the relational hierarchy. Thus, the transitivity and semantic effects of causee encoding are accounted for by the interaction of the syntagmatic and paradigmatic argument selection strategies. We propose that these two strategies represent organizing principles for argument selection information associated with lexical entries. (shrink)

El perfecto es un aspecto de perspectiva retrospectiva que codifica la vigencia en el momento de habla de una eventualidad anterior. El presente trabajo propone la existencia de una categoría semántica y gramatical de antiperfecto que, en contraste con el perfecto, codifica la no vigencia en el momento de habla de una eventualidad pasada. Debido a que, al igual que el perfecto, el antiperfecto es sensible a la vigencia de una eventualidad anterior en el momento de habla, proponemos que se (...) trata de un segundo tipo de aspecto de perspectiva retrospectiva. Tras revisar las nociones de perfecto en Comrie, Dik y Nishiyama y Koenig y la noción de pasado discontinuo en Plungian y Van der Auwera, se observa el modo en que la categoría propuesta se realiza, codificada gramaticalmente, en la lengua mapuche, mediante el morfema -fu, con especial énfasis en su caracterización semántico-pragmática, su relación con la aktionsart y el papel del contexto en su interpretación. The perfect aspect is a retrospective aspect that encodes the current relevance of a previous event. In this work, we propose a semantic and grammatical category labeled as antiperfect that, in contrast to the perfect aspect, encodes the non-validity of a previous eventuality in speech time. Given that, as the perfect aspect, the antiperfect is subject to the current relevance of a previous situation, we propose that this is a second type of retrospective aspect. After reviewing the descriptions of perfect by Comrie, Dik and Nishiyama & Koenig, and the description of discontinuous past by Plungian & van der Auwera, we analyze how the antiperfect is grammatically encoded in Mapudungun by means of the morpheme -fu. We focus the analysis of this category on its semantic-pragmatic description, its relation to aktionsart, and the role of context in its interpretation. (shrink)

This paper investigates the linguistic basis for moral non-cognitivism, the view that sentences containing moral predicates do not have truth conditions. It offers a new argument against this view by pointing out that the view is incompatible with our best empirical theories about the grammatical encoding of illocutionary force potentials. Given that my arguments are based on very general assumptions about the relations between the grammar of natural languages and a sentence's illocutionary function, my arguments are broader in (...) scope than the familiar semantic objections to non-cognitivism relating to the so-called Frege-Geach problem: even if a solution to the Frege-Geach problem has been found, my arguments still stand. (shrink)

Interference effects from semantically similar items are well-known in studies of single word production, where the presence of semantically similar distractor words slows picture naming. This article examines the consequences of this interference in sentence production and tests the hypothesis that in situations of high similarity-based interference, producers are more likely to omit one of the interfering elements than when there is low semantic similarity and thus low interference. This work investigated language production in Mandarin, which allows subject noun phrases (...) to be omitted in discourse contexts in which the subject entity has been previously mentioned in the discourse. We hypothesize that Mandarin speakers omit the subject more often when the subject and the object entities are conceptually similar. A corpus analysis of simple transitive sentences found higher rates of subject omission when both the subject and object were animate (potentially yielding similarity-based interference) than when the subject was animate and object was inanimate. A second study manipulated subject-object animacy in a picture description task and replicated this result: participants omitted the animate subject more often when the object was also animate than when it was inanimate. These results suggest that similarity-based interference affects sentence forms, particularly when the agent of the action is mentioned in the sentence. Alternatives and mechanisms for this effect are discussed. (shrink)

This paper provides a formal analysis of the grammatical encoding of temporal information in Hausa (Chadic, Afro-Asiatic), thereby contributing to the recent debate on temporality in languages without overt tense morphology. By testing the hypothesis of covert tense against recently obtained empirical data, the study yields the result that Hausa is tenseless and that temporal reference is pragmatically inferred from aspectual, modal and contextual information. The second part of the paper addresses the coding of future in particular. It (...) is shown that future time reference in Hausa is realized as a combination of a modal operator and a Prospective aspect marker, involving the modal meaning components of intention and prediction as well as event time shifting. The discussion relates directly to recent approaches to other seemingly tenseless languages such as St’át’imcets (Matthewson, Linguist Philos 29:673–713, 2006) or Paraguayan Guaraní (Tonhauser, Linguist Philos 34:257–303, 2011b) and provides further evidence for the suggested analyses of the future markers in these languages. (shrink)

Utterances are planned and realized incrementally. Which information is salient or attended to prior to initiating an utterance has influences on choices in argument structure and word order, and affects the prosodic prominence of the constituents involved. Many phenomena that the linguistic literature usually treats as reflexes of the grammatical encoding of information structure, such as the early ordering of topics, or the prosodic reduction of old information, are treated in the production literature as a consequence of how (...) contextual salience interacts with production planning. This article reviews information structural effects that arise as a consequence of how syntactic and phonological information is incrementally encoded in the production process, and how we can tell these effects apart from grammatically encoded aspects of information structure that form part of the message. (shrink)

We study how people attend to and memorize endings of events that differ in the degree to which objects in them are affected by an action: Resultative events show objects that undergo a visually salient change in state during the course of the event (peeling a potato), and non‐resultative events involve objects that undergo no, or only partial state change (stirring in a pan). We investigate general cognitive principles, and potential language‐specific influences, in verbal and nonverbal event encoding and (...) memory, across two experiments with Dutch and Estonian participants. Estonian marks a viewer's perspective on an event's result obligatorily via grammatical case on direct object nouns: Objects undergoing a partial/full change in state in an event are marked with partitive/accusative case, respectively. Therefore, we hypothesized increased saliency of object states and event results in Estonian speakers, as compared to speakers of Dutch. Findings show (a) a general cognitive principle of attending carefully to endings of resultative events, implying cognitive saliency of object states in event processing; (b) a language‐specific boost on attention and memory of event results under verbal task demands in Estonian speakers. Results are discussed in relation to theories of event cognition, linguistic relativity, and thinking for speaking. (shrink)

Summary Covert encoding is one of the strategies available to languages for the encoding of motion, in which, in accordance with the laws of Gestalt, the meaning of an expression encoding motion is not coincident with the mere sum of the meanings of each of its constitutive units, relying on the mediation of grammatical and co(n)text­established knowledge for its interpretability. Moving on from a data set gathered for a previous study and adopting a holistic, constructional approach, (...) several strategies were found in a diachronic corpus of Italian, French and Spanish for the covert encoding of motion in such languages, based on whether the motion­interpretation attributed to the covert construction is mediated by either linguistic or extralinguistic knowledge. The diachronic investigation also showed that the use of such patterns is diachronically consistent in frequency, thus proving that such patterns of interaction between language and cognition may be regarded as functional for speakers of all languages and time stages, also confirming that a holistic, constructional approach to language study can help shed light on linguistic–cognitive phenomena on the basis of language variation and change. (shrink)

Semitic templates systematically encode two dimensions of verb meaning: (a) agency, the thematic role of the verb’s external argument, and (b) voice. The assumption that this form-meaning correspondence is mediated by syntax allows the parallel compositional construction of the form and the meaning of a verb from the forms and the meanings of its root and template. The root and its arguments are optionally embedded under a light verb v which introduces the agent (Hale and Keyser 1993; Kratzer 1994). But (...) this is only the unmarked case, which, in Semitic, is encoded by the simple templates. Two dimensions of markedness are introduced by two additional types of syntactic heads: (a) agency heads, which modify agency and are morphologically realized as the intensive and causative templates, and (b) voice heads, which modify voice and are morphologically realized as the passive and middle templates. Causative and middle morphemes are thus accounted for within a unified system, which, first, explains their affinity in language in general (both are found crosslinguistically as markers of transitivity alternations), and which, moreover, sheds new light on problems in the interface of semantics and morphology. One problem is the impossibility, mostly ignored in linguistic theory, of deriving the semantics of middle verbs from that of the corresponding transitive verbs. The second is explaining the identity found crosslinguistically between middle and reflexive morphology. The third is determining the grammatical function of the causee in causative constructions. (shrink)

This paper examines the use of scalar adjectives in two contexts that have played a role in discussions of the subjective/objective distinction: ?faultless disagreement? discourses and the nonfinite complement position of the subjective attitude verb find. I argue that the pattern of distribution and interpretation of scalar adjectives in these contexts provides evidence for two sources for subjectivity, which are distinguished from each other in that one affects the grammatical properties of a predicate and one does not. The first (...) kind, which licenses appearance in the complement position of find, is based in the lexical semantics of predicates that encode qualitative assessments. The second kind, which gives rise to faultless disagreement effects, arises from uncertainty about the dimensions of evaluation that are involved in fixing the extension of a predicate in a context of utterance. (shrink)

Abstract“Force dynamics” refers to a previously neglected semantic category—how entities interact with respect to force. This category includes such concepts as: the exertion of force, resistance to such exertion and the overcoming of such resistance, blockage of a force and the removal of such blockage, and so forth. Force dynamics is a generalization over the traditional linguistic notion of “causative”: it analyzes “causing” into finer primitives and sets it naturally within a framework that also includes “letting,”“hindering,”“helping,” and still further notions. (...) Force dynamics, moreover, appears to be the semantic category that uniquely characterizes the grammatical category of modals, in both their basic and epistemic usages. In addition, on the basis of force dynamic parameters, numerous lexical items fall into systematic semantic patterns, and there exhibit parallelisms between physical and psychosocial reference. Further, from research on the relation of semantic structure to general cognitive structure, it appears that the concepts of force interaction that are encoded within language closely parallel concepts that appear both in early science and in naive physics and psychology. Overall, force dynamics thus emerges as a fundamental notional system that structures conceptual material pertaining to force interaction in a common way across a linguistic range: the physical, psychological, social, inferential, discourse, and mental-model domains of reference and conception. (shrink)

The paper presents a system for construction classification representing multiple levels of specification, such as grammatical functions, grammatically reflected actants, and lexical semantics, aligned with a compositional system of sign combination mediating between a construction perspective and a valence perspective. The system uses a feature structure formalism based on Head-Driven Phrase Structure Grammar but with essential elements from Lexical Functional Grammar, and has as implementation background large scale HPSG grammars. While on the one extreme being able to encode word (...) level selection in multi-word patterns, the system on the other provides a compact format for construction specification, allowing for cross-language comparison both in construction and valence frame inventories. Pivotal in these capacities as well as in sign formalization in general are the grammatical functions. The paper motivates the usefulness of the various functionalities and illustrates the way in which they work together in a formally uniform system. (shrink)

"This book addresses recent developments in the study of tense from a cross-paradigm and cross-linguistic point of view. Leading international scholars explore challenging ideas about tense at the interfaces between semantics and syntax as well as syntax and morphology. The book is divided into three main subsections: 1) Tense in tenseless languages; 2) Tense, mood, and modality, and 3) Descriptive approaches to some tense phenonema. Although time is a universal dimension of the human experience, some languages encode reference to time (...) without any grammatical tense morphology of the verb. Some of these exceptional "tenseless" languages are investigated in this volume: Kalaallisut, Paraguayan Guaraní and Movima. Modal verbs are polyfunctional in the sense that they express both tense and modality. In this volume, an untypical modal is analyzed, a modal analysis of imperatives is argued for, and sentential mood, which is closely related to modality, is analyzed. It is always interesting to look at the expression of tense in understudied languages, which is done here for Scottish Gaelic, Austronesian Rukai and German dialects. The volume can be used for graduate and undergraduate level teaching."--Publisher's website. (shrink)

The idea that the form of a word reflects information about its meaning has its roots in Platonic philosophy, and has been experimentally investigated for concrete, sensory-based properties since the early 20th century. Here, we provide evidence for an abstract property of ‘boundedness’ that introduces a systematic, iconic bias on the phonological expectations of a novel lexicon. We show that this abstract property is general across events and objects. In Experiment 1, we show that subjects are systematically more likely to (...) associate sign language signs that end with a gestural boundary with telic verbs (denoting events with temporal boundaries, e.g., die, arrive) and with count nouns (denoting objects with spatial boundaries, e.g., ball, coin). In Experiments 2–3, we show that this iconic mapping acts on conceptual representations, not on grammatical features. Specifically, the mapping does not carry over to psychological nouns (e.g. people are not more likely to associate a gestural boundary with idea than with knowledge). Although these psychological nouns are still syntactically encoded as either count or mass, they do not denote objects that are conceived of as having spatial boundaries. The mapping bias thus breaks down. Experiments 4–5 replicate these findings with a new set of stimuli. Finally, in Experiments 6–11, we explore possible extensions to a similar bias for spoken language stimuli, with mixed results. Generally, the results here suggest that ‘boundedness’ of words' referents (in space or time) has a powerful effect on intuitions regarding the form that the words should take. (shrink)

In this paper, we aim to enhance our understanding about the processing of implicit and explicit temporal chronological relations by investigating the roles of temporal connectives and verbal tenses, separately and in interaction. In particular, we investigate how two temporal connectives (ensuiteandpuis, both meaning ‘then’) and two verbal tenses expressing past time (the simple and compound past) act as processing instructions for chronological relations in French. Theoretical studies have suggested that the simple past encodes the instruction to relate events sequentially, (...) unlike the more flexible compound past, which does not. Using an online experiment with a self-paced reading task, we show that these temporal connectives facilitate the processing of chronological relations when they are expressed with both verbal tenses, and that no significant difference is found between the two verbal tenses, nor between the two connectives. By means of an offline experiment with an evaluation task, we find, contrary to previous studies, that comprehenders prefer chronological relations to be overtly marked rather than implicitly expressed, and prefer to use the connectivepuisin particular. Furthermore, comprehenders prefer it when these relations are expressed using the compound past, rather than the simple past. Instead of using thecontinuityhypothesis (Segal et al. 1991,Murray 1997) to explain the processing of temporal relations, we conclude that a more accurate explanation considers a cluster of factors including linguistic knowledge (connectives, tenses, grammatical and lexical aspect) and world knowledge. (shrink)

Children learn their native language by exposure to their linguistic and communicative environment, but apparently without requiring that their mistakes be corrected. Such learning from “positive evidence” has been viewed as raising “logical” problems for language acquisition. In particular, without correction, how is the child to recover from conjecturing an over-general grammar, which will be consistent with any sentence that the child hears? There have been many proposals concerning how this “logical problem” can be dissolved. In this study, we review (...) recent formal results showing that the learner has sufficient data to learn successfully from positive evidence, if it favors the simplest encoding of the linguistic input. Results include the learnability of linguistic prediction, grammaticality judgments, language production, and form-meaning mappings. The simplicity approach can also be “scaled down” to analyze the learnability of specific linguistic constructions, and it is amenable to empirical testing as a framework for describing human language acquisition. (shrink)

A defining assumption in the debate on contextual influences on truth-conditional content is that such content is often incompletely determined by what is specified in linguistic form. The debate then turns on whether this is evidence for positing a more richly articulated logical form or else a pragmatic process of free enrichment that posits truly unarticulated constituents that are unspecified in linguistic form. Questioning this focus on semantics and pragmatics, this article focuses on the independent grammatical dimensions of the (...) problem. Against the background of a principled account of the different ways in which the lexicon and the grammar, respectively, determine aspects of propositional meaning, and an uncontentious notion of content, nothing turns out to be ‘missing’ in grammatical expressions in order for them to encode complete propositional thoughts. As this predicts, when putatively hidden constituents are made overt or are otherwise added, propositions result that are systematically different from the thoughts originally expressed. Context, while potentially affecting lexically specified aspects of meaning, never affects grammar-determined ones, suggesting a specific role for grammar in the normal cognitive mode. (shrink)

Peirce’s beta graphs are roughly equivalent to our first-order predicate logic. However, Bellucci and Pietarinen have recently argued that the beta graphs are not well-equipped to handle asymmetric relative terms. I survey four proposed solutions to the problem and find them all wanting. I offer a fifth solution according to which Peirce’s beta graphs function at three different levels of abstractness from natural language. I diagnose the problem of asymmetric relative terms as arising when we transition from the first to (...) the second level of abstractness. Making grammatical information encoded in natural language explicit at the first level of abstractness and interpreting graphs at the second level of abstractness as shorn of grammatical information resolves the problem. The solution is both well-motivated by Peirce’s own commitments and increases the expressiveness of the beta graphs. (shrink)

According to classical theories of language and cognition, human cognition is characterized by strong universal commonalities built around notions such as object, space, agency, number, time, and event (Clark, 1973; Miller & Johnson‐Laird, 1976). Languages select from this prelinguistic conceptual repertoire the concepts that become encoded in their lexical and grammatical stock. Language acquisition, on this view, is a mapping process in which the learner needs to figure out which sounds in the language spoken in the environment correspond to (...) which concepts from the ones already in the mind (Fodor, 1975; Gleitman, 1990; Pinker, 1994). (shrink)

In English, deictic verbs of motion, such ascomecan encode the perspective of the speaker, or another individual, such as the addressee or a narrative protagonist, at a salient reference time and location, in the form of an indexical presupposition. By contrast, Spanish has been claimed to have stricter requirements on licensing conditions forvenir(“to come”), only allowing speaker perspective. An open question is how a bilingual learner acquiring both English and Spanish reconciles these diverging language-specific restrictions. We face this question head (...) on by investigating narrative productions of young Spanish-English bilingual heritage speakers of Spanish, in comparison to English monolingual and Spanish dominant adults and children. We find that the young heritage speakers producevenirin linguistic contexts where most Spanish adult speakersdo not, but where English monolingual speakersdo, and also resemble those of young monolingual Spanish speakers of at least one other Spanish dialect, leading us to generate two mutually-exclusive hypotheses: (a) the encoding of speaker perspective in the young heritage children is cross-linguistically influenced by the more flexible and dominant language (English), resulting in a wider range of productions by these malleable young speakers than the Spanish grammar actually allows, or (b) the young Spanish speakers are exhibiting productions that are in fact licensed in the grammar, but which are pruned away in the adult productions, being supplanted by other forms as the lexicon is enriched. Given independent evidence of the heritage speakers' robust Spanish linguistic competence, we turn to systematically-collected acceptability judgments of three dialectal varieties of monolingual adult Spanish speakers of the distribution of perspective-taking verbs, to assess their competence and adjudicate between (a) and (b). We find that adultsaccept venirin contexts in which they do notproduceit, leading us to argue that (a)venirisnotobligatorily speaker-oriented in Spanish, as has been claimed, (b) adults may not producevenirin these contexts because they instead select more specific motion verbs, and (c) for heritage bilingual children, the more dominant language (English) may support the grammatically licensed but lexically-constrained productions in Spanish. (shrink)

Natural languages exhibit a great variety of grammatical paradigms. For instance, in English verbs are grammatically marked for tense, whereas in the tenseless Eskimo-Aleut language Kalaallisut they are marked for illocutionary mood. Although time is a universal dimension of the human experience and speaking is part of that experience, some languages encode reference to time without any grammatical tense morphology, or reference to speech acts without any illocutionary mood morphology. Nevertheless, different grammatical systems are semantically parallel in (...) certain respects. Specifically, I propose that English tenses form a temporal centering system, which monitors and updates topic times, whereas Kalaallisut moods form a modal centering system, which monitors and updates modal discourse referents. To formalize these centering parallels I define a dynamic logic that represents not only changing information but also changing focus of attention in discourse (Update with Centering, formalizing Grosz et al 1995). Different languages can be translated into this typed logic by directly compositional universal rules of Combinatory Categorial Grammar (CCG) The resulting centering theory of tense and illocutionary mood draws semantic parallels across different grammatical systems. The centering generalizations span the extremes of the typological spectrum, so they are likely to be universal. In addition, the theory accounts for the translation equivalence of tense and illocutionary mood in a given utterance context. Following Stalnaker (1978) I assume that the very act of speaking up has a ‘commonplace effect’ on the context. It focuses attention on the speech act and thereby introduces default modal and temporal topics. These universal defaults complement language-specific grammars, e.g. English tenses and Kalaallisut moods. In a given utterance context the universal discourse-initial defaults plus language-specific grammatical marking may add up to the same truth conditions.. (shrink)

A thorny question surrounding the meaning of ought concerns a felt distinction between deontic uses of ought that seem to evaluate a state of affairs versus those that seem to describe a requirement or obligation to perform an action, as in and, respectively. There ought not be childhood death and disease. You ought to keep that promise. Various accounts have been offered to explain the contrast between “agentive” and “non-agentive” ought sentences. One such account is the Agency-in-the-Prejacent theory, which traces (...) the difference to a particular kind of ambiguity in the prejacent. This theory has been criticized as linguistically unviable. Indeed, I level a few novel complaints against AIP myself in the present paper. But AIP has a kernel of genuine insight which allows us to explain the contrast—that the distinction between agentive and non-agentive ought sentences owes in part to the way natural language encodes information about agency. I develop this idea into a novel account that, like AIP, traces the contrast to an ambiguity in the complement of the modal. However, according to the view I propose, the Coercion View, a linguistically-motivated coercion operation produces the necessary grammatical conditions for agentive ought, which in turn allow a kind of variadic function operator in the style of to produce the semantic effect we see on display in agentive readings of ought. Having explained the mechanism by which we get this structure, I show that it corroborates some of the central intuitions underwriting agentive ought. I submit that the Coercion View offers an explanation of agentive ought to take at least as seriously as any of its competitors. (shrink)

In Gradient Acceptability and Linguistic Theory, Elaine J. Francis examines a challenging problem at the intersection of theoretical linguistics and the psychology of language: the problem of interpreting gradient judgments of sentence acceptability in relation to theories of grammatical knowledge. This problem is important because acceptability judgments constitute the primary source of data on which such theories have been built, despite being susceptible to various extra-grammatical factors. Through a review of experimental and corpus-based research on a variety of (...) syntactic phenomena and an in-depth examination of two case studies, Francis argues for two main positions. The first is that converging evidence from online comprehension tasks, elicited production tasks, and corpora of naturally-occurring discourse can help determine the sources of variation in acceptability judgments and narrow down the range of plausible theoretical interpretations. The second is that the interpretation of judgment data depends crucially on one's theoretical commitments and assumptions, especially with respect to the nature of the syntax-semantics interface and the choice of either a categorical or a gradient notion of grammaticality. The theoretical frameworks considered in this book include derivational theories (e.g. Minimalism, Principles and Parameters), constraint-based theories (e.g. Sign-based Construction Grammar, Simpler Syntax), competition-based theories (e.g. Stochastic Optimality Theory, Decathlon Model), and usage-based approaches. While showing that acceptability judgment data are typically compatible with the assumptions of various theoretical frameworks, Francis argues that some gradient phenomena are best captured within frameworks that permit soft constraints-non-categorical grammatical constraints that encode the conventional preferences of language users. (shrink)

This volume is a comprehensive, up-to-date, and readable introduction to linguistic meaning. While partial to conceptual and typological approaches, the book also presents results from formal approaches. Throughout, the focus is on grammatical meaning -- the way languages delineate universal semantic space and encode it in grammatical form. Subjects covered by the author include: the domain of linguistic semantics and the basic tools, assumptions, and issues of semantic analysis; semantic properties of entities, events, and thematic roles; language and (...) space; tense, aspect, and the internal structure and temporal ordering of events; modality, negation, and the epistemology of the speaker; and modification and attribution. In contrast to most current treatments of semantics, this book is accessible to the beginning student of semantics and linguistics and also useful to the advanced practitioner. A textbook and reference work in a single volume, it can be used in a number of disciplines: psychology, computer science, and anthropology as well as linguistics. (shrink)

Previous studies have described a range of placeholder (PH) items. A PH fills in the grammatical slot of a target that a speaker is unable or unwilling to produce. This paper argues that Japanese unnun, an expression wholly underdescribed in the literature, serves as a PH and that it may also be used as a general extender (GE). Unlike previously known PHs, unnun is regarded as a ‘meta-discourse’ PH; it replaces a discourse segment, rather than a linguistic form. I (...) develop a cognitive-pragmatic account in Relevance theory, arguing that unnun encodes procedural meaning and that the PH and GE functions emerge through interaction among encoded meaning, pragmatic principles, and contextual assumptions. Further, the account situates unnun in the broader picture of vague language in a theoretically coherent manner. I also suggest that some of the previously described PHs in other languages may be re-classified as meta-discourse PHs. (shrink)

This paper takes a stand on Hamblin semantics and its relation to the semantics-to-pragmatics interface. Biased questions, where the speaker finds one of the options more likely and expects the confirmation that p is true, raise a concern about the limits of Hamblin semantics. I argue that biased questions have modified Hamblin semantics, while unbiased questions have unconstrained Hamblin semantics. The optional bias feature explains compositionally. It is triggered by likelihood presuppositions ranging Hamblin sets and highlighting the preferred alternative(s). Biased (...) effects are exceptionally possible in alternative and wh-questions, but only in polar questions, they can be encoded lexically and grammatically. I argue that Hamblin semantics covers all core types of questions, while some of its applications for non-questions are problematic. (shrink)

Inflectional morphology has been taken as a paradigmatic example of rule-governed grammatical knowledge (Pinker, 1999). The plausibility of this claim may be related to the fact that it is mainly based on studies of English, which has a very simple inflectional system. We examined the representation of inflectional morphology in Serbian, which encodes number, gender, and case for nouns. Linguists standardly characterize this system as a complex set of rules, with disagreements about their exact form. We present analyses of (...) a large corpus of nouns which showed that, as in English, Serbian inflectional morphology is quasiregular: It exhibits numerous partial regularities creating neighborhoods that vary in size and consistency. We then asked whether a simple connectionist network could encode this statistical information in a manner that also supported generalization. A network trained on 3,244 Serbian nouns learned to produce correctly inflected phonological forms from a specification of a word's lemma, gender, number, and case, and generalized to untrained cases. The model's performance was sensitive to variables that also influence human performance, including surface and lemma frequency. It was also influenced by inflectional neighborhood size, a novel measure of the consistency of meaning to form mapping. A word-naming experiment with native Serbian speakers showed that this measure also affects human performance. The results suggest that, as in English, generating correctly inflected forms involves satisfying a small number of simultaneous probabilistic constraints relating form and meaning. Thus, common computational mechanisms may govern the representation and use of inflectional information across typologically diverse languages. (shrink)

continent. 1.2 (2011): 136-140. In early 2011, Cow Heavy Books published The Official Catalog of the Library of Potential Literature , a compendium of catalog 'blurbs' for non-existent desired or ideal texts. Along with Erinrose Mager, I edited the project, in a process that was more like curation as it mainly entailed asking a range of contemporary writers, theorists, and text-makers to send us an entry. What resulted was a creative/critical hybrid anthology, a small book in which each page opens (...) to a new iteration of textual desire. These texts explore the material possibilities of the book. Somewhat parallel to the call of N. Katherine Hayles who, in her book Writing Machines , urges literary theorists to take up the practice of Medium Specific Analysis (to account for the way the medium in which it is presented conditions or at least bears on a literary text). I see in the imagined works of The Official Catalog a call for the innovative writers of today to become Medium-Responsive. This would mean thinking through the specific (materially constrained) possibilities offered by the media in which texts are presented, and in thinking of the literary text as a kind of art in the greater context of other arts and the book as a medium situated within the context of many other media. In doing so, the contemporary writer refutes the chorus of critics who lament the death of the book by consistently reinvigorating literary innovation. The following are selections from The Official Catalog of the Library of Potential Literature that show possible paths for (thinking about) new writing that engages with its medium. —Ben Segal, Editor THE CUBE Even the most radical non-linear texts have tended to exploit or subvert only the sequential possibilities of print—from the continuous loop of Joyce's Finnegans Wake to the shuffled cards of Marc Saporta's Composition No. 1 —but The Cube takes such multiplicities to an entirely new level. Set in a grid, the book's words can be read conventionally, across the page, as well as down each column—with either route making complete grammatical sense. But they can also be read as stacked strata and mined like lexical core samples through the layered pages of the book. Each path tells the same story from a different perspective (the narrative, naturally, hinges on the potential outcomes of a throw of cubed dice). By opening up the z-axis to reading in this way, The Cube recognizes the book as a three-dimensional sculptural space. Taking its lead from Armand Schwerner's (If Personal) and Raymond Queneau's Cent mille milliards de poèmes , The Cube reads like a experiment by Christian Bök precision printed by Emily McVarish. Craig Dworkin is the editor, most recently, of The Consequence of Innovation: 21st Century Poetics (Roof Books, 2008), The Sound of Poetry/The Poetry of Sound , with Marjorie Perloff (U. Chicago Press, 2009), and Against Expression: An Anthology of Conceptual Writing , with Kenneth Goldsmith (Northwestern UP, 2010). He teaches at the University of Utah. HE GOES In He Goes , we read notes, letters and e-mails from a scholar father to his novelist daughter. We read of the father's musings on Beckett, on Pinter, on Anne Frank; his description of a woman hanging laundry from a line. We read about his journey toward dying, followed by a brief, third person account of his death, and his obituary. Then a long series of blank pages that demand to be read in real time, non-sentence by non-sentence, blank page by blank page. Finally—and it is here that this peculiar little book begins to soar—the dead father writes to his as-of-yet-still-living daughter. He does not write from death. He does not write from life. The words unprint, unstamp, unkindle. Still, they require no translation. The father "writes" (for lack of a better word) about the serendipitous, the commonplace; he recommends another book. He jokes. He asks his daughter how her stomach is. He says forget about presence in absence, darling; screw words as memorial and the guys in garbage cans and loss as redemption and I can't go on I must go on. He goes, "Love, Fodder." He goes, "incidentally." He goes, "I thought you might like to know." Elizabeth Graver is the author of a story collection, Have You Seen Me? , and three novels: Unravelling ; The Honey Thief ; and Awake . Her work has been anthologized in Best American Short Stories , Best American Essays , and Prize Stories: the O. Henry Awards . She teaches at Boston College. THE PAPER ARCHIVIST A stunning package and a triumph of imagination, The Paper Archivist at times looks to be less a book than an abstract expressionist painting. Softly bound, its contents unfold to a single sheet of uneven thickness and texture—a canvas splattered with colored lines, stickers, broken sentences, and nonsense pictographs. But by following the directions to fold, dip, smell, rub, scratch, and tear the sheet according to the contingencies of the weather and using only the objects at hand, the reader slowly brings the forces hidden in the noise into a glorious sculptural convergence, processing a different story and shape each time. This is the rare book that continues to stir, whirl, and pop on every new reading. Sean Higgins blogs at BOMBlog where he is responsible for the column Volumes and Territories, as well as Ghost Island , a fledgling collaborative intellectual collective. THE SLOW BOOK The Slow Book , written by an anonymous author at the dawn of literacy, on a minor planet (otherwise notable only as the source of that exceptionally hardy, not very tasty grain called “shef” sowed on hostile planets as an early step to colonization), and encoded into a series of punctures on a strip of copper coiled inside a clever device, something between a player piano and an old-fashioned film projector, is being released into print, as was the author’s intention, at a rate of one word per century (local time). Each word is, across the Forty Galaxies, agreed to be uncannily apt for the century in which it appears—even “of,” in a century during which the highest value was attached to fidelity, whether to ideals, worlds, or romantic love; even “the,” which governed two centuries, one extraordinarily materialistic, during which advances in propulsion and navigation accelerated the exchange of exceptional objects between the remotest planets of the Forty, and one in which the central concern, both of philosophers and the common man, was whether, in an age of rapidly proliferating hypothetical worlds, anyone or anything concretely existed at all. Even those words published long before interstellar contact can be seen in retrospect to have transgalactic pertinence. As a result, attempts to abstract the machine from its publishers, Hobson & Hui, in order to “predict the future” for insight or gain by “fast- forwarding” the copper strip have been many and ingenious. While, in centuries of skepticism (“maybe”), or of unrest (“go”) the book has been nearly forgotten, in others it seems to haunt every thought, every deed, despite the fact that the subject of The Slow Book is still unclear. So far only a few sentences exist in print; everyone knows them, can quote them, offer the standard exegeses and assorted heresies; yet certainties are the stuff of adolescence; mature readers are forced to acknowledge that these sentences are probably only a preamble to the main argument. They contain no proper nouns, nor can we identify any definite theme. There is even disagreement about their tone, whether coolly ironic, as some insist, or ardent. The appearance of an unusual grammatical case, sometimes called the future pluperfect continuous, used to describe events that at some future point will have always been true (but are not yet)—hitherto known to appear only in the synthetic dogmas of the Thanatographical Society, and in certain highly circumscribed religious contexts—has suggested to some scholars that the Slow Book was originally intended for ritual use, but the proximity of the usage to a term designating a small hand plow that, as Pott and Mielcke have convincingly shown, would have borne a distinctly obscene double meaning in its culture of origin in the author’s time, argues otherwise. It is likewise unclear whether the situation that seems to be—with teasing incompleteness—sketched out in these few lines is intended as an illustration of general principles, a case study, a dramatic scene, or an extended metaphor. In short, we have no idea what The Slow Book is about. In our own time, we believe that it is almost certainly a work of fiction, but that may be because we live in the century of “if”. In each age, perhaps, we see the book we most need to read. Some have dared to suggest that the metal strip is blank until, with millennial fanfare, it advances into its new position, that no ur-text exists, that the book itself is brought into being—written—by our need. But that is exactly the sort of thing we would believe, in 7645. Shelley Jackson is the author of the story collection The Melancholy of Anatomy , the novel Half Life , and hypertexts including Patchwork Girl . The recipient of a Howard Foundation grant, a Pushcart Prize, and the 2006 James Tiptree Jr Award, she has also written and illustrated several children's books, including The Old Woman and the Wave ; Sophia, the Alchemist's Dog ; and Mimi's Dada Catifesto . Her stories and essays have appeared in Conjunctions , McSweeney's , The Paris Review , and Cabinet Magazine . In 2004 she launched her project SKIN , a story published in tattoos on 2095 volunteers. THE BOOK OF SOUNDS The Book of Sounds is just that: a book of sounds made when letters are construed in new ways to bring forth out of the alphabet new forms of speech. A book meant to be read out loud, The Book of Sounds is not unlike Laurie Anderson's O Superman or Brian Eno's Music for Airports in its attempt to make music out of the most primary and simplest of methods. It breaks language down to its barest bones and makes out of the page a drum that has never before been beaten upon. Peter Markus is the author of a novel, Bob, or Man on Boat (Dzanc Books) as well as two books of short fiction, Good, Brother and The Singing Fish , both of which were published by Calamari Press. A new collection of stories, We Make Mud , is now available from Dzanc Books. PARADISE OF THE BLIND by Celan Solen Although the reclusive Celan Solen published his first and only book in 1963—paying out-of-pocket for a limited edition of the slim collection No One May Have the Same Knowledge Again —he remained in American obscurity for almost three decades. In 1992 a micro-press in Istanbul brought out No One in Turkish. A German translation followed in 1995. Soon it became clear in literary circles Solen was a world-class (if highly unclassifiable) artist—lyrical, dense, enigmatic—who could undo the conventional short story in 397 words by inventing impossible worlds housed in impossible whirls (in “Small Sadnesses,” a single chartreuse tree frog in Borneo unknowingly holds time together by its very presence in the universe, while each letter of its tale refers, not to itself, but to the one preceding it in the alphabet). By his disappearance last year, Solen was considered master by a generation of writers and critics (except, alas, for those gentlemen in the Swedish Academy). Imagine, then, that generation’s delight at the discovery, locked away in the author’s safe-deposit box, of his second and final composition. Had Lynch’s Lost Highway been book instead of film, and had it been penned by Beckett at his least certain, revised by Barthelme at his most formally deranged, and typeset by Derrida at his most semiotically catastrophic, the result might have been something like Paradise of the Blind : interlacing narratives of a man composed of borrowed organs (whose most cheerless and difficult to locate, god, could only have been invented by an empty heart), a nonexistent medieval painting blamed for the ruin of future hope, and the spread of a philosophy that holds earth a mistake constantly recurring in the dream of a fish lying on the floor of the Atlantic (if the fish wakes, our world winks off)—all contained in a text packed with typed-over passages, torn postcards, poems that can be deciphered only when held up to a mirror, pages ornamented with trompe-l'œil paperclips and coffee stains and buzzing houseflies, some busy with illegible runes that dissolve when exposed to light, three that smell like roses or lemons (depending on whether a man or woman is reading), two that stain with the bloody fingerprints of the those who handle them, one that ignites when brushed with breath, thirteen sewn from baby skin, one that moans when touched, and one that screams—yet all without mass, unimaginable, and invisible. Lance Olsen is author of more than 20 books of and about innovative fiction, including, most recently, the novels Calendar of Regrets (2010) and Head in Flames (2009). He teaches experimental narrative theory and practice at the University of Utah. SUPERSTRUCTURE! by Barbara D'Albi As soon as I opened the third drawer of Barbara D'Albi's wooden novel, everything became hopeless. Now in Ithaca, there was no going back. And it wasn't just the intricate series of shelves, hinged doors and locked drawers which D'Albi layered into the book, no, lo, I was constructed anew by the story. Who else but D'Albi to imagine a God who becomes a carpenter and gets killed?! And makes it good! You want stories? D'Albi is a skyscraper, built with planes and levers. Momentarily I wondered where I could shelve this book, and then I thought: no matter; I couldn't put it down. Adam Robinson lives in Baltimore, where he runs Publishing Genius. He is the author of Adam Robison and Other Poems and Say, Poem. (shrink)

We examine two connectionist networks—a fractal learning neural network (FLNN) and a Simple Recurrent Network (SRN)—that are trained to process center-embedded symbol sequences. Previous work provides evidence that connectionist networks trained on infinite-state languages tend to form fractal encodings. Most such work focuses on simple counting recursion cases (e.g., anbn), which are not comparable to the complex recursive patterns seen in natural language syntax. Here, we consider exponential state growth cases (including mirror recursion), describe a new training scheme that seems (...) to facilitate learning, and note that the connectionist learning of these cases has a continuous metamorphosis property that looks very different from what is achievable with symbolic encodings. We identify a property—ragged progressive generalization—which helps make this difference clearer. We suggest two conclusions. First, the fractal analysis of these more complex learning cases reveals the possibility of comparing connectionist networks and symbolic models of grammatical structure in a principled way—this helps remove the black box character of connectionist networks and indicates how the theory they support is different from symbolic approaches. Second, the findings indicate the value of future, linked mathematical and empirical work on these models—something that is more possible now than it was 10 years ago. (shrink)

Words in Semitic texts often consist of a concatenation of word segments, each corresponding to a Part-of-Speech (POS) category. Semitic words may be ambiguous with regard to their segmentation as well as to the POS tags assigned to each segment. When designing POS taggers for Semitic languages, a major architectural decision concerns the choice of the atomic input tokens (terminal symbols). If the tokenization is at the word level the output tags must be complex, and represent both the segmentation of (...) the word and the POS tag assigned to each word segment. If the tokenization is at the segment level, the input itself must encode the different alternative segmentations of the words, while the output consists of standard POS tags. Comparing these two alternatives is not trivial, as the choice between them may have global effects on the grammatical model. Moreover, intermediate levels of tokenization between these two extremes are conceivable, and, as we will aim to show, beneficial. To the best of our knowledge, the problem of tokenization for POS tagging of Semitic languages has not been addressed before in full generality. In this paper, we study this problem for the purpose of POS tagging of Modern Hebrew texts. After extensive error analysis of the two simple tokenization models, we propose a novel, linguistically-motivated, intermediate tokenization model that gives better performance for Hebrew over the two initial architectures. Our study is based on the well-known Hidden Markov Models (HMMs). We start out from a manually devised morphological analyzer and a very small annotated corpus, and describe how to adapt an HMM-based POS tagger for both tokenization architectures. We present an effective technique for smoothing the lexical probabilities using an untagged corpus, and a novel transform for casting the segment-level tagger in terms of a standard, word-level, HMM implementation. The results obtained using our model are on par with the best published results on Modern Standard Arabic, despite the much smaller annotated corpus available for Modern Hebrew.. (shrink)

This paper offers a review of a current understanding of the content and the form of linguistic roots. It first updates and buttresses the case against semantic content of uncategorised roots and for Late Insertion of roots; then it investigates how native speakers identify roots. More specifically, the idea that roots may be polysemous or may encode the shadow of a denotation, namely the common denominator of the denotations of words derived from it, is refuted on the basis of conceptual (...) and empirical arguments from a number of languages. Subsequently, the existence of a spectrum of content to which roots belong, with roots ranging from contentless to semantically specific and concrete, is also shown to be illusory, and to result from the actual productivity, hence diversity, of the words derived from it. Arguments for Late Insertion of roots are then reviewed and updated, divorcing roots from the forms that realise them. These arguments are systematically combined with the semantic contentlessness of roots in support of Acquaviva’s analysis of them as abstract indices, i.e. as the syntax-internal criteria of lexical identity. This account is taken to its logical conclusion in the final section: if roots are indeed abstract indices, then they cannot be identified either by the semantic content they realise within grammatical structures or by their forms. An account is therefore advanced according to which roots are identified just once by native speakers over their lexicon at a given moment and on the basis of three heuristic principles: one form-based, one based on the feature content and the exponence of the structures in which roots are embedded, and one taking care of root suppletion. (shrink)

This paper demonstrates that the progressive interpretation in Ga is an effect of the interaction between the imperfective aspect and a definite description of events. Crucially, the data from Ga point to the consequences of the view that definite descriptions of events encode the familiarity of the discourse referent and its uniqueness in bearing the property in question. Namely, they yield direct evidentiality and the necessary ongoingness of the event at the topic time. Thus, the paper identifies previously unattested variation (...) in the semantics of the progressive in a cross-linguistic perspective and shows that not only lexical but also grammatical aspect exhibits striking parallelisms with the nominal domain. (shrink)

Stojnić (2021) argues that the content of linguistic utterances is determined by the rules of natural language grammar more stringently than what is generally assumed. She proposes specifically that coherence relations are encoded by the linguistic structures and determine what individuals count as most prominent, thereby serving as the referents of free (“demonstrative”) pronouns. In this paper, I take a close look at the empirical evidence from English and Serbian that she offers in support of this position. Considering these data (...) points in connection with additional linguistic data (also from German and Japanese), I argue that there is no compelling evidence for the assumption that coherence relations directly determine the resolution of pronouns. Instead, grammatical restrictions imposed by different types of pronouns and tenses have a larger impact on the meaning conventionally expressed by complex utterances than what is generally assumed in the literature on coherence relations. (shrink)

A lot of effort is currently made to provide methods to analyze and understand deep neural network impressive performances for tasks such as image or text classification. These methods are mainly based on visualizing the important input features taken into account by the network to build a decision. However these techniques, let us cite LIME, SHAP, Grad-CAM, or TDS, require extra effort to interpret the visualization with respect to expert knowledge. In this paper, we propose a novel approach to inspect (...) the hidden layers of a fitted CNN in order to extract interpretable linguistic objects from texts exploiting classification process. In particular, we detail a weighted extension of the Text Deconvolution Saliency (wTDS) measure which can be used to highlight the relevant features used by the CNN to perform the classification task. We empirically demonstrate the efficiency of our approach on corpora from two different languages: English and French. On all datasets, wTDS automatically encodes complex linguistic objects based on co-occurrences and possibly on grammatical and syntax analysis. (shrink)