Language Research Articles

Does ChatGPT have sociolinguistic competence?

Large language models are now able to generate content- and genre-appropriate prose with grammatical sentences. However, these targets do not fully encapsulate human-like language use. For example, set aside is the fact that human language use involves sociolinguistic variation that is regularly constrained by internal and external factors. This article tests whether one widely used LLM application, ChatGPT, is capable of generating such variation. I construct an English corpus of “sociolinguistic interviews” using the application and analyze the generation of seven morphosyntactic features. I show that the application largely fails to generate any variation at all when one variant is prescriptively incorrect, but that it is able to generate variable deletion of the complementizer that that is internally constrained, with variants occurring at human-like rates. ChatGPT fails, however, to properly generate externally constrained complementizer that deletion. I argue that these outcomes reflect bias both in the training data and Reinforcement Learning from Human Feedback. I suggest that testing whether an LLM can properly generate sociolinguistic variation is a useful metric for evaluating if it generates human-like language.

The Geometry and Dynamics of Meaning.

An enigma for human languages is that children learn to understand words in their mother tongue extremely fast. The cognitive sciences have not been able to fully understand the mechanisms behind this highly efficient learning process. In order to provide at least a partial answer to this problem, I have developed a cognitive model of the semantics of natural language in terms of conceptual spaces. I present a background to conceptual spaces and provide a brief summary of their main features, in particular how it handles learning of concepts. I then apply the model to give a geometric account of the semantics of different word classes. In particular, I propose a "single-domain hypotheses" for the semantics of all word classes except nouns. These hypotheses provide a partial answer to the enigma of how words are learned. Next, a dynamic cognitive model of events is introduced that replaces and extends the function of thematic roles. I apply it to analyze the meanings of different kinds of verbs. I argue that the model also explains some aspects of syntactic structure. In particular, I propose that a sentence typically refers to an event. Some further applications of conceptual spaces are briefly presented.

Unique cortical and subcortical activation patterns for different conspecific calls in marmosets.

The common marmoset (Callithrix jacchus) is known for its highly vocal nature, displaying a diverse range of calls. Functional imaging in marmosets has shown that the processing of conspecific calls activates a brain network that includes fronto-temporal areas. It is currently unknown whether different call types activate the same or different networks. In this study, nine adult marmosets (four females) were exposed to four common vocalizations (phee, chatter, trill, and twitter), and their brain responses were recorded using event-related fMRI at 9.4T. We found robust activations in the auditory cortices, encompassing core, belt, and parabelt regions, and in subcortical areas like the inferior colliculus, medial geniculate nucleus, and amygdala in response to these calls. Although a common network was engaged, distinct activity patterns were evident for different vocalizations that could be distinguished by a 3D convolution neural network, indicating unique neural processing for each vocalization. Our findings also indicate the involvement of the cerebellum and medial prefrontal cortex (mPFC) in distinguishing particular vocalizations from others.Significance Statement This study investigates the neural processing of vocal communications in the common marmoset (Callithrix jacchus). Utilizing event-related fMRI at 9.4T, we demonstrate that different calls (phee, chatter, trill, and twitter) elicit distinct brain activation patterns, challenging the notion of a uniform neural network for all vocalizations. Each call type distinctly engages various regions within the auditory cortices and subcortical areas. These findings offer insights into the evolutionary mechanisms of primate vocal perception and provide a foundation for understanding the origins of human speech and language processing.

First-language acquisition in a multimodal language framework: Insights from speech, gesture, and sign

Children across the world acquire their first language(s) naturally, regardless of typology or modality (e.g. sign or spoken). Various attempts have been made to explain the puzzle of language acquisition using several approaches, trying to understand to what extent it can be explained by what children bring to language-learning situations as well as what they learn from the input and the interactive context. However, most of these approaches consider only speech development, thus ignoring the inherently multimodal nature of human language. As a multimodal view of language is becoming more widely adopted for the study of adult language, a multimodal approach to language acquisition is inevitable. Not only do children have the capacity to learn spoken and sign language equally easily, but spoken language acquisition consists of learning to coordinate linguistic expressions in both modalities, that is, in both speech and gesture. To provide a step forward in this direction, this article aims to synthesize findings from research studies that take a multimodal perspective on language acquisition in different sign and spoken languages, including the development of speech and accompanying gestures. Our review shows that while some aspects of language acquisition seem to be modality-independent, others might differ according to the affordances of each modality when used separately as well as together (either in sign, speech, and/or gesture). We argue that these findings need to be integrated into our understanding of language acquisition. We also identify which areas need future research for both spoken and sign language acquisition, taking into account not only multimodal but also cross-linguistic variation.

Understanding public opinion and discussion dynamics of digital humans on social media: an analysis of sentiment, themes and user characteristics

PurposeThe emergence of artificial intelligence-generated content (AIGC) technology has markedly enhanced the capabilities of digital human content generation and natural language processing, thus further advancing the development of digital humans. To enable enterprises and governments to effectively address the challenges and opportunities arising from the rapid development of digital humans, it is imperative to understand the public opinion and discussion dynamics of digital humans.Design/methodology/approachThis study initially analyzed the trends and distribution patterns of public attention to digital humans. By utilizing word cloud technology, we explored the primary focal points of public interest and conducted a topic analysis using latent Dirichlet allocation (LDA) techniques. Subsequently, content analysis was conducted on the popular application domains of digital humans. Finally, this study examined the influence of user characteristics on emotional scores toward digital humans and the presence of differences in focus across user groups.FindingsThe results indicate a sustained increase in public attention toward digital humans, accompanied by notable geographic disparities in the distribution of discussions. Discussions on Weibo are primarily focused on four domains, whereas areas within the digital human application domain that provoke widespread discussion include live streaming, service, cultural entertainment and digital avatars. Significant impacts of user characteristics on sentiment scores were observed, revealing divergent focal points of interest among different user groups toward digital humans.Originality/valueThrough the deep analysis of Weibo data, this study offers new insights into the digital human industry, enabling governments and businesses to understand industry trends and develop targeted digital human customization strategies based on customer characteristics.

Nature and nurture in fruit fly hearing

As for human language learning and birdsong acquisition, fruit flies adjust their auditory perception based on past sound experiences. This phenomenon is known as song preference learning in flies. Recent advancements in omics databases, such as the single-cell transcriptome and brain connectomes, have been integrated into traditional molecular genetics, making the fruit fly an outstanding model for studying the neural basis of “Nature and Nurture” in auditory perception and behaviors. This minireview aims to provide an overview of song preference in flies, including the nature of the phenomenon and its underlying neural mechanisms. Specifically, we focus on the neural circuitry involved in song preference learning, with which auditory experiences shape the song preference of flies. This shaping process depends on an integration hub that processes external sensory stimuli and internal states to enable flexible control of behavior. We also briefly review recent findings on the signals that feed into this integration hub, modulating song preference of flies in an experience-dependent manner.

Leveraging metaphors in qualitative machine translators in the interpretation of emotion

The idea of emotions that is concealed in human language gives rise to metaphor. It is challenging to compute and develop a framework for emotions in people because of its detachment and diversity. Nonetheless, machine translation heavily relies on the modeling and computation of emotions. When emotion metaphors are calculated into machine translation, the language is significantly more colorful and satisfies translating criteria such as truthfulness, creativity and beauty. Emotional metaphor computation often uses artificial intelligence (AI) and the detection of patterns and it needs massive, superior samples in the emotion metaphor collection. To facilitate data-driven emotion metaphor processing through machine translation, the study constructs a bi-lingual database in both Chinese and English that contains extensive emotion metaphors. The fundamental steps involved in generating the emotion metaphor collection are demonstrated, comprising the basis of theory, design concepts, acquiring data, annotating information and index management. This study examines how well the emotion metaphor corpus functions in machine translation by proposing and testing a novel earthworm swarm-tunsed recurrent network (ES-RN) architecture in a Python tool. Additionally, the comparison study is carried out using machine translation datasets that already exist. The findings of this study demonstrated that emotion metaphors might be expressed in machine translation using the emotion metaphor database developed in this research.

Advancements in Natural Language Processing for Human-Computer Interaction

• This systematic review explores the advancements in Natural Language Processing (NLP) for Human-Computer Interaction (HCI) over the period from 2010 to 2024. It highlights the significant breakthroughs achieved through deep learning models, particularly transformer architectures such as BERT and GPT, which have transformed the ability of machines to understand and generate human language. The integration of multimodal capabilities has further enriched user interactions by enabling the processing of diverse data types, including text, audio, and visual inputs. However, the review also identifies persistent challenges, including maintaining coherence in long dialogues, resolving ambiguous language, addressing bias in training data, and the need for resource-efficient models. Additionally, the paper emphasizes the importance of cross-lingual capabilities for low-resource languages and the necessity of personalized, adaptive systems. The findings underscore the need for ongoing research to overcome existing limitations and enhance the effectiveness and inclusivity of NLP technologies in HCI, ultimately contributing to a more intuitive and accessible user experience.

Handedness and the control of human technology and language

The impact of handedness on neural substrates of common tool processing and tool use skills as proxies to the mechanisms underlying modern reciprocal relationships among humans and their brains, the tools they create and use, and technologies they implement and utilise, has only recently started to be elucidated. Yet, as most of the reports on lateralisation of such processing or skills focused on a single task, and sharp divisions of participants into right-, left-, and mixed-handed, or non-righthanded for the latter two groups combined, little is known about interrelationships between the neuronal underpinnings of different skilled manual actions—e.g., manual praxis skills in the form of disparate hand-tool interactions, including their most common neural phenotypes in the population at large. Here, in 62 individuals with different handedness status, we studied the laterality of two praxic abilities involving common tools. Even though their neural substrates were expected to be closely linked, we identified numerous cases (in 7/28 righthanders [RH] – 25%, and in 6/21 lefthanders [LH] – 29% of participants) with hemispheric dissociations in the underlying mechanisms. They involved both right-lateralised functional grasp planning vs. left-lateralised visual tool use (8% of all tested cases, 3/28 RH: ~11%, and 2/21 LH: ~10%), and vice versa (13% of all cases, 4/28 RH: ~14%, and 4/21 LH: ~19% of participants). The laterality/organisation of these praxic skills was also compared to the laterality/organisation of productive language, with only few cases of dissociations identified. The observed phenotypes are discussed in the context of coevolutionary hypotheses linking ancient toolmaking and tool use skills, the associated cultural evolution, technological innovations, and language to the cortical expansion and functional lateralisation in human evolution.

Eye-movement-prompted large image captioning model

Pretrained large vision-language models have shown outstanding performance on the task of image captioning. However, owing to the insufficient decoding of image features, existing large models sometimes lose important information, such as objects, scenes, and their relationships. In addition, the complex ”black-box” nature of these models makes their mechanisms difficult to explain. Research shows that humans learn richer representations than machines do, which inspires us to improve the accuracy and interpretability of large image captioning models by combining human observation patterns. We built a new dataset, called saliency in image captioning (SIC), to explore relationships between human vision and language representation. One thousand images with rich context information were selected as image data of SIC. Each image was annotated with five caption labels and five eye-movement labels. Through analysis of the eye-movement data, we found that humans efficiently captured comprehensive information for image captioning during their observations. Therefore, we propose an eye-movement-prompted large image captioning model, which is embedded with two carefully designed modules: the eye-movement simulation module (EMS) and the eye-movement analyzing module (EMA). EMS combines the human observation pattern to simulate eye-movement features, including the positions and scan paths of eye fixations. EMA is a graph neural network (GNN) based module, which decodes graphical eye-movement data and abstracts image features as a directed graph. More accurate descriptions can be predicted by decoding the generated graph. Extensive experiments were conducted on the MS-COCO and NoCaps datasets to validate our model. The experimental results showed that our network was interpretable, and could achieve superior results compared with state-of-the-art methods, i.e., 84.2% BLEU-4 and 145.1% CIDEr-D on MS-COCO Karpathy test split, indicating its strong potential for use in image captioning.

Do New Words Propagate Like Memes? An Internet Usage-Based Two-Stage Model of the Life Cycle of Neologisms

Neologisms reflect new ideas or new concepts in our life and play an important role in cultural transmission and the vitality of human language. The explosion of neologisms, especially in the past two decades, can also be ascribed to the popularity and accessibility of digital content and social media. In this paper, we focus on the issue of how neologisms arise by looking at the trajectory of developments in terms of their usage over time, i.e., their life cycle. By studying neologisms in vivo, instead of as fait accompli, we hope to better understand the nature of neologisms and to enable better prediction and earlier inclusion of neologisms. To achieve this goal, we examine the memetic model for the life cycle of neologisms and compare it with a recently studied epidemic model. We present a longitudinal modeling of the development of neologisms based on internet usage data aggregated from Google Trends, covering the 90 most influential Chinese neologisms from 2008–2016. Our study verifies that the memetic model can describe and predict the life cycle of the neologisms robustly for the early stages (i.e., the ascending stages) of its cycle, but not for its full life cycle, and crucially cannot predict the inflection point. We conclude that two models are needed for word propagation: a memetic model for the initial stages and an epidemic model for the latter stage, particularly the inflection point. This two-stage/two-model approach allows for neologisms to be more easily identified as potentially new words, as it is easier to write a program to automatically filter for emerging terms using a memetic model.

Exploring Word-Level Representations in Modern Natural Language Processing

Abstract: Natural Language Processing (NLP) is a dynamic and rapidly advancing field at the intersection of artificial intelligence and linguistics, focused on enabling computers to understand, process, and generate human language. Recent advancements intransformer-based models have significantly improved NLP capabilities, enablingmachines to understand and generate human language more effectively. This paper provides a comprehensive overview of NLP, tracing its historical development and recent trends. The discussion includes the different phases of NLP, from text pre- processing and tokenization to syntactic and semantic analysis, along with pragmatic considerations. Text normalization techniques, such as stemming, lemmatization, and removing stopwords, are explored to emphasize their importance in preparing raw text for analysis. Additionally, this paper presents a comparative analysis of popular word-level representation techniques used in NLP, including One-Hot Encoding, Bag of Words (BoW), and Term Frequency-Inverse Document Frequency (TF-IDF).

Parameter-coupled state space models based on quasi-Gaussian fuzzy approximation

The accuracy of a fuzzy system’s approximation is closely tied to the performance of fuzzy control systems design, while this system’s interpretability depends on the description of a mechanical model using human language. This research introduces a quasi-Gaussian membership function characterized by a pair of parameters to achieve the sensitivity of a triangular membership function along with the interpretability of Gaussian membership functions. Consequently, a two-dimensional (2-D) quasi-Gaussian membership function is derived, and a method for establishing quasi-Gaussian fuzzy systems (QGFS) using a rectangular grid is proposed. After validating the approximation properties using the sine function for the one-dimensional (1-D) and 2-D QGFS, the systems are applied to approximate the depyrogenation tunnel, a significant piece of equipment in the pharmaceutical industry with various mechanical designs. Validation results indicate that the 1-D and 2-D QGFS can achieve an approximation error varying within a ± 5% range. Meanwhile, the 1-D and 2-D QGFSs are applied to mechanical models of the depyrogenation tunnel with satisfactory final approximation results. Lastly, the 2-D QGFS is capable of demonstrating an excellent description of models with coupled parameters.

Creative minds like ours? Large Language Models and the creative aspect of language use

Descartes famously constructed a language test to determine the existence of other minds. The test made critical observations about how humans use language that purportedly distinguishes them from animals and machines. These observations were carried into the generative (and later biolinguistic) enterprise under what Chomsky in his Cartesian Linguistics, terms the “creative aspect of language use” (CALU). CALU refers to the stimulus-free, unbounded, yet appropriate use of language—a tripartite depiction whose function in biolinguistics is to highlight a species-specific form of intellectual freedom. This paper argues that CALU provides a set of facts that have significant downstream effects on explanatory theory-construction. These include the internalist orientation of linguistics, the invocation of a competence-performance distinction, and the postulation of a generative language faculty that makes possible—but does not explain—CALU. It contrasts the biolinguistic approach to CALU with the recent wave of enthusiasm for the use of Transformer-based Large Language Models (LLMs) as tools, models, or theories of human language, arguing that such uses neglect these fundamental insights to their detriment. It argues that, in the absence of replication, identification, or accounting of CALU, LLMs do not match the explanatory depth of the biolinguistic framework, thereby limiting their theoretical usefulness.

Syntax in animal communication: its study in songbirds and other taxa

Many building blocks of human language can be found within the vocal communication systems of other species, most notably songbirds. One of the most prominent of these building blocks is syntax. While studies of syntax are abundant, a lack of consensus on the definition of syntax in non‐human animal communication studies has led to much debate. Consistent and deliberate use of terminology is needed to facilitate understanding across disciplines. In addition, new terminology may better describe syntactic structure found in vocal signals that are devoid of semantic associations, such as birdsong. Here, we propose two terms to describe the types of syntax commonly found in birdsong: sequential syntax and dialectical syntax. Sequential syntax can be defined as the rules that govern the patterns of sound without regard to semantic meaning. Dialectic syntax can be defined as sequential syntax that is distinct among different populations or groups with behavioral significance for those groups. Taken together, these two terms can describe the type of syntax seen in ornamental signals, such as birdsong.

Aircraft human‐machine interaction assistant design: A novel multimodal data processing and application framework

AbstractDuring aircraft operations, pilots rely on human‐machine interaction platforms to access essential information services. However, the development of a highly usable aerial assistant necessitates the incorporation of two interaction modes: active‐command and passive‐response modes, along with three input modes: voice inputs, situation inputs, and plan inputs. This research focuses on the design of an aircraft human‐machine interaction assistant (AHMIA), which serves as a multimodal data processing and application framework for human‐to‐machine interaction in a fully voice‐controlled manner. For the voice mode, a finetuned FunASR model is employed, leveraging private aeronautical datasets to enable specific aeronautical speech recognition. For the situation mode, a hierarchical situation events extraction model is proposed, facilitating the utilization of high‐level situational features. For the plan mode, a multi‐formations double‐code network plan diagram with a timeline is utilized to effectively represent plan information. Notably, to bridge the gap between human language and machine language, a hierarchical knowledge engine named process‐event‐condition‐order‐skill (PECOS) is introduced. PECOS provides three distinct products: the PECOS model, the PECOS state chart, and the PECOS knowledge description. Simulation results within the air confrontation scenario demonstrate that AHMIA enables active‐command and passive‐response interactions with pilots, thereby enhancing the overall interaction modality.

Complexity in Chick‐a‐Dee Calls of Mountain Chickadees (Poecile gambeli): Call Variation Associated With Flock Size and Flight

ABSTRACTThe chick‐a‐dee call of chickadees, tits, and titmice is a vocal system used in a wide range of social contexts by both sexes throughout the year and is one of the more structurally complicated vocal systems outside of human language. Relatively little is known about the chick‐a‐dee calls of mountain chickadees, Poecile gambeli, however. This is an important species for increasing our comparative understanding of variation in chick‐a‐dee calls as they are one of the chickadee species with the largest naturally occurring flock sizes. Flock size relates to the social complexity of flocks, and the social complexity hypothesis for communication predicts that individuals in more complex social groups should communicate with greater complexity than individuals in simpler social groups. Correlational and experimental evidence in support of the hypothesis has been found in the calls of a wide range of species, including Carolina chickadees, P. carolinensis. Here, we provide the first description of the variation in note composition and note‐ordering rules in calls from mountain chickadee flocks in California and Colorado. California flocks were found to be significantly larger than Colorado flocks. Analysis of note‐type usage and transition probabilities between note types found that calls of California birds were more complex than calls of Colorado birds, supporting a key prediction of the social complexity hypothesis for communication. We also found relatively high rates of reversals of note‐ordering rules in mountain chickadee calls, which might help explain the complexity of the chick‐a‐dee calls of this species. Additionally, birds in flight produced calls with different note compositions when compared to perched birds. Generally, the note‐type ordering and transition probabilities of calls of mountain chickadees seem comparable to other better‐studied chickadee species, although their frequent note‐type order rule reversals suggest potential syntax‐like properties in this call system.

How the Rejection of Incompatible Speech Acts Transforms Human Cognition

AbstractEngaging with the literature on transformative conceptions of rationality, I argue for the following position on the way reason transforms human cognition: when the capacity for knowing that one ought to do something is directed at one's own speech acts, an initially domain‐specific and practical grasp of genus/species relations – manifest in the ability to select among the various permitted ways to do as one judges one ought – becomes a mechanism through which the reflective study of genus/species relations hones the domain‐general classificatory abilities that accompany adult human language use. In this fashion, our instinctive behaviors may be transformed: we might cease to respond to events simply as (e.g.) fearful or enraging, recognize these motivations as cases of cowardice and recklessness, and begin to treat them as opportunities for courage and restraint.

Thinning of Balti Script : Way Forward to Balti OCR

Natural language is one of the applications of Artificial Intelligence, which trains machines to do the jobs in human language. OCR is one of the fields where the writing efforts are omitted and text images are converted into editable text. An OCR may have post and preprocessing to enhance the text image more suitable for the rest of the OCR process. Thinning is the preprocessing approach in which the characters, words and text is thinned to its one-pixel skeleton. Much of the work has been done in the various languages of the world as well as Pakistani languages. The work on Balti OCR is nonexistent. In this study, a thinning algorithm is proposed for the Balti language, a language spoken in the northern areas of Pakistan and India. Many of the Balti images were tested with the proposed algorithm and the proposed system produced accurate results by giving a one pixel skeleton of input image. The proposed algorithm tested with hundreds of Balti language images and selected results are presented in this paper. The current research has many directions including the way forward to building Balti OCR, Balti ICR (both segmentation based and segmentation free).

Syntax and the brain: language evolution as the missing link(ing theory)?

This paper provides proof of concept that neurolinguistic research on human language syntax would benefit greatly by expanding its scope to include evolutionary considerations, as well as non-propositional functions of language, including naming/nicknaming and verbal aggression. In particular, an evolutionary approach can help circumvent the so-called granularity problem in studying the processing of syntax in the brain, that is, the apparent mismatch between the abstract postulates of syntax (e.g. Tense Phrase (TP), Determiner Phrase (DP), etc.) and the concrete units of neurobiology (neurons, axons, etc.). First, we decompose syntax into its evolutionary primitives, identifying one of the earliest stages as a simple, flat combination of just one verb and one noun. Next, we identify proxies ("living fossils") of such a stage in present-day languages, including compounds and small clauses, lacking at least some layers of structure, e.g. TPs and DPs. These proxies of ancestral language have been subjected to fMRI neuroimaging experiments. We discuss the finding that less hierarchical small clauses, in contrast to full sentences with TPs and DPs, show reduced activation in the left Broca's area (BA) 44 and the right basal ganglia, consistent with the hypothesis that more recent, more elaborate syntax requires more connectivity in the Broca's-basal ganglia network, whose neuronal density has been significantly enhanced in recent evolution, implicating mutations in FOXP2 and other genes. We also discuss the finding that the processing of ancestral verb-noun compounds, which are typically used for (derogatory) naming and nicknaming, shows enhanced activation in the right fusiform gyrus area (BA 37), the area that is implicated in the processing of metaphoricity and imageability, but also in naming and face recognition, opening up an intriguing possibility that the enhanced face recognition in humans was facilitated by the early emergence of a simple syntactic strategy for naming. The considerations in this paper are consistent with the hypothesis of a gradual gene-culture co-evolution of syntax and the brain, targeting cortico-striatal brain networks. It is also of note that a sound grounding in neurobiology of language should in turn inform syntactic theories themselves.