UNIVERSITY OF COLOGNE

This talk argues that the diverse causal domains underpinning human language converge and interface at a single privileged locus, a ∼2½-s opportunity for action, called the enchronic envelope. Evidence is given for the existence and nature of the envelope as a universally primary frame for the selection, deployment, and evolution of linguistic structure. The first key argument is that language is a form of action and will therefore be structured similarly to physical actions. The second is that because linguistic actions exploit principles of communication, they are subject to a legibility constraint, which requires speaker–recipient calibration, thus importing strong constraints on the design of linguistic structures from the interpersonal alignment of the dyad in the enchronic time frame of social interaction. The case is made that this envelope is the site at which processes at diverse timescales must be realized. The account has implications for our developing understanding of languages in complex-adaptive-system terms. It seeks to advance the complex-systems idea for language by showing that highly diverse linguistic networks and processes are moored to a single, central interface, where language is processed, learned, transmitted, and conventionalized.

In data from a corpus of American English personal narratives, extremely complex viewpoint embedding is discernible, paralleling in many ways what can be seen in narratologists' understanding of perspective embedding. In oral story-telling, speakers regularly enact their own feelings and reactions, as well as their interactions with the addressee(s). They also enact their characters' actions, utterances and expressions - they speak and gesture "for" the characters. It is not difficult for addressees to track the layers of viewpoint involved: speakers regularly use gaze and partitioning of the surrounding space, as well as the linguistic track (including quoted speech), to indicate who is being represented at any given time. Of particular interest are the ways in which a story-teller's body may simultaneously enact aspects of the story-teller and of a character, or aspects of more than one character. (This kind of complexity is equally to be seen in Signed Language story-telling.) This simultaneous enactment of two viewpoints allows gestural embedded viewpoint to support and complement the embedding markers in the linguistic track of the narration.

Holding a conversation is a complex task: we need to grab someone’s attention, signal understanding, and show a genuine interest in their perspective. In this talk, I present a series of corpus studies of early child-caregiver interactions from the English section of the CHILDES database to trace the developmental stages toward complex interaction involving all these feats. Distributional and prosodic variation of vocatives (Mumey!), expanding functionality in back-channelling and responding (okay?), and confirmation-seeking in invariant (huh?) and invariant tags (isn’t it?) all show that linguistic interaction begins early and grows in complexity. What emerges is a cascading path of development in children aged 2 to 5 that aligns with central milestones of cognitive development: an early availability of turn-taking and attention management, followed by the emergence of preliminary skills of signalling shared beliefs, and finally the arrival of a separation of those beliefs in parallel with a fully developed Theory of Mind. Interactional language can therefore serve as a window into cognitive development and trace how children use language from an early age to join and flourish in a community of speakers through increasingly nuanced contributions.

Heritage languages have long been described through group-level comparisons with baselines, framing their grammars in terms of global convergence, divergence, or stability. This talk argues that such aggregate views may obscure the key locus of variation: the individual speaker. Drawing on corpus data and experimental studies of heritage speakers in Germany and the United States, I show that patterns traditionally interpreted as contact-induced change (such as shifts in clause-combing strategies) mask substantial intra-group heterogeneity, with community-level effects and individual speaker profiles shaping outcomes in distinct ways. Using Bayesian varying effects and individual differences approaches, I demonstrate how factors such as register, language exposure, proficiency and cognitive skills interact to drive variability. I show that in some linguistic phenomena, such as discourse markers, the variability is much larger within groups than between them. These findings invite a reframing of heritage grammars within a native language continuum, where synchronic stability, innovation, and dynamicity co-exist across individuals rather than defining entire communities. I then discuss ongoing shifts in the field of heritage language research and further implications for situating heritage grammars as productive testing grounds for theories of acquisition, processing, and multilingual representation.

It’s difficult to imagine the internet today – and social media platforms especially – without internet memes. Building on humorous forms of indirectness or incongruity, they parcel up often emotional messages in more impactful ways than just words, and allow communicators to bring across viewpoints and elicit responses quickly.

Using tools and inspirations from cognitive linguistics and construction grammar, this talk shows ways in which English itself is changed, its grammar extended, by internet meme usage. Images used in memes are not illustrations of text, but become structural components, making the forms of English adjust to the emerging multimodal rules of meme grammar; at the same time, there is a visual grammar to the way in which memes use the image space available. In terms of the meanings expressed, memes function as shortcuts to stance, making sometimes complex configurations of stance accessible to online interlocutors. Creative variation in the forms used will be shown to confirm meme users’ awareness of the constructional nature of established meme types as form-meaning pairings.

The talk also extends the discussion from memes proper to a wider class of memetically inspired discourse forms. Examples from persuasive discourse types such as advertising suggest that multimodal genres ‘leak’ into each other, and invite broader study of multimodal means of emotional expression. 

Individual differences can provide us with a unique window into human language processing mechanisms, as they allow us to observe the effect of natural variation in cognitive abilities, knowledge and processing strategies. Thanks to recent powerful LLMs, we can even test complex hypotheses of how e.g. working memory limitations and predictive mechanisms might interact. 

In the first part of my talk, I will focus on sentence processing: Based on an experimental study and a reading corpus (InDiCo), I will show that the lossy context surprisal model, which incorporates memory constraints, can successfully predict the effects of differences in human working memory capacity on eye-movements during reading. I will also report an analysis of the PoTeC reading corpus, which demonstrates that a surprisal model that reflects the specific experience and background knowledge of a human can better predict their reading behaviour, compared to a generic model, or even a model that specifically fits the domain of the text. 

In the second part of my talk, I will report recent experimental findings on individual differences in pragmatic processing, and present an ACT-R model that can account for individual differences in pragmatic processing.

More than half a decade after the Chomskyan revolution, Generative Linguistics still presents the only attempt at a comprehensive understanding of grammar as a phenomenon. The major current usage-based alternative, Construction Grammar, downplays the special nature of grammar to the extent that it is the framework is at odds with well-established psycho- and neurolinguistic findings (e.g. Pulvermüller, Cappelle & Shtyrov 2013).

In this presentation I outline an alternative usage-based theory of the nature of grammar which is in all respects compatible with Construction Grammar, but which is precise about what defines grammar as a distinct kind of phenomenon (Boye & Harder 2012; Boye 2023, 2024). Central to this understanding are the notions of attentional prominence and convention.

After presenting the theoretical core, I first exemplify how hypotheses derived from the theory have been tested psycho- and neurolinguistically. I then outline how the theory emphasizes a central role for attentional prominence in agrammatic aphasia. Subsequently, I argue that the theory entails an understanding of grammaticalization as conventionalization of asymmetries pertaining to attentional prominence.

References

• Boye, K. 2023. Grammaticalization as Conventionalization of Discursively Secondary Status: Deconstructing the Lexical–Grammatical Continuum. Transactions of the Philological Society 121(2), 270–292.
• Boye, K. 2024. Evidentiality, discourse prominence and grammaticalization. Studies in language 48.3. 575–607.
• Boye, K. & P. Harder. 2012. A usage-based theory of grammatical status and grammaticalization. Language 88.1. 1–44.
• Pulvermüller, F., B. Cappelle & Y. Shtyrov. 2013. Brain basis of meaning, words, constructions, and grammar. In T. Hoffmann & G. Trousdale (eds.), Oxford Handbook of Construction Grammar, 397–416. Oxford: Oxford University Press, Oxford.