Abandoning the EPP-feature in Polish dual copula clauses by redefining the predication relation

This paper questions the logic behind the presence and the working of the EPP-feature in Polish dual copula clauses (henceforth, DCCs) with the pronominal copula to, the verbal copula być ‘to be’, and two nominative 3rd person DPs, as represented in Bondaruk (2019). The criticism follows from: (i) – Chomsky’s (2000, 2001) downward Agree operation; (ii) – the view that the predicator encodes the predication relation between the pre-copular subject and the post-copular predicate; (iii) – selective multiple Agree, whereby the satisfaction of the EPP- and uφ-features is divorced. Adopting (i)–(iii), Bondaruk’s scrutiny allows either the pre- or the post-copular DP to occupy SpecTP, thereby accounting for DCCs’ agreement and configurational patterns, but, simultaneously, suffering from theoretical shortcomings it creates. We argue for a simpler satisfaction of the subject requirement which does not rely on the troublesome EPP-feature, but is motivated formally by the relation between T and the higher DP. We derive this requirement by following Zeiljstra’s (2012) upward Agree which only takes place once interpretable features c-command uninterpretable features, and Rothstein’s (2004) approach which is based on a neo-Davidsonian event semantics and which argues that be and its complement form a complex predicate, separated from the pre-copular DP both semantically and syntactically.

The event semantics of conjuncts in ‘The Sun Also Rises’

The present paper explores the semantics of Hemingway’s ‘plain style’ in The Sun Also Rises by combining corpus linguistic methodology with event semantics theory. The focus of the study is on how the narrator of the novel segments experienced situations in terms of semantic events. Corpus linguistic analysis shows that the ‘plain style’ of the narrative section of the novel is realized by means of coordinated clauses and that the narrator’s event segmentation is associated with a small set of preferred lexical items. These results are interpreted in terms of event semantics to show that preferred lexical items are indicative of event types typical of the narrative section. The semantic analysis of relations between coordinated clauses indicates that these relations are not simply about the juxtaposition of disjoined events. Finally, the study demonstrates that an approach combining corpus linguistic methodology with insights from event semantics can offer new understanding of the propositional meaning of literary texts, and the way narrators encode experienced situations.

Cumulative readings of "every" and leaks

every surprisingly gives rise to cumulative readings (Schein 1993; Kratzer 2000). The distribution of these readings is governed by scope-related asymmetries (Champollion 2010; Haslinger & Schmitt 2018). In this work, I notice a third property of these readings: cumulative readings of every receive weaker "leaky" truth-conditions under negation, previously thought to be unattested (Bayer 2013). Exploiting this third property, I build an event semantics to deliver these "leaky readings" by default. Within this semantics, it becomes possible to account for cumulative readings of every and their properties, keeping to standard assumption about the denotation for every. I also show how the same analysis predict the scope-related asymmetries and their less studied interaction with overt movement.

Neural Event Semantics for Grounded Language Understanding

We present a new conjunctivist framework, neural event semantics (NES), for compositional grounded language understanding. Our approach treats all words as classifiers that compose to form a sentence meaning by multiplying output scores. These classifiers apply to spatial regions (events) and NES derives its semantic structure from language by routing events to different classifier argument inputs via soft attention. NES is trainable end-to-end by gradient descent with minimal supervision. We evaluate our method on compositional grounded language tasks in controlled synthetic and real-world settings. NES offers stronger generalization capability than standard function-based compositional frameworks, while improving accuracy over state-of-the-art neural methods on real-world language tasks.

Stimulus-independent neural coding of event semantics: Evidence from cross-sentence fMRI decoding

Multivariate neuroimaging studies indicate that the brain represents word and object concepts in a format that readily generalises across stimuli. Here we investigated whether this was true for neural representations of events described using sentences. Participants viewed sentences describing four events in different ways. Multivariate classifiers were trained to discriminate the four events using a subset of sentences, allowing us to test generalisation to novel sentences. We found that neural patterns in a left-lateralised network of frontal, temporal and parietal regions discriminated events in a way that generalised successfully over changes in the syntactic and lexical properties of the sentences used to describe them. In contrast, decoding in visual areas was sentence-specific and failed to generalise to novel sentences. In the reverse analysis, we tested for decoding of syntactic and lexical form, independent of the event being described. Regions displaying this coding were limited and largely fell outside the canonical semantic network. Our results indicate that a distributed neural network represents the meaning of event sentences in a way that is robust to changes in their structure and form. They suggest that the semantic system disregards the surface properties of stimuli in order to represent their underlying conceptual significance.

Negation in Event Semantics

With events as dense as time, negation threatens to be trivial, unless ‘not’ is noughtly, an adverb of quantification. So revised, classical puzzles of negation in natural language are revisited, in which deviation from the logical connective, violating Excluded Middle, appears to prompt a special condition or special meaning. The language of events also contains negative event descriptions—After the flood, it not drying out ruined the basement and one could smell it not drying out—and these appear to founder on the logic of the constructions in which they occur and on reference to suspect negative events, events of not drying out. A language for event semantics with ‘not’ as noughtly resolves the puzzles surveyed—within classical logic, without ambiguity or special conditions on the meaning of ‘not’, and without a metaphysics of negative events.