Next Article in Journal
Well, If You Talk to Me in Norwegian, I Won’t Answer You: Language Policies and Practices in Latvian Diasporic Families
Previous Article in Journal
Inherently Long Consonants in Contemporary Italian Varieties: Regional Variation and Orthographic Effects
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Languages
Volume 10
Issue 6
10.3390/languages10060119
languages-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Review

From Theoretical Framework to Empirical Investigation: A Bibliometric Analysis of Research Evolution and Emerging Trends in Polarity Sensitivity Studies Between 1980 and 2023

by
Lingda Kong
†,
Yi Li
†,
Yanting Sun
,
Yong Jiang
* and
Xiaoming Jiang
*
Institute of Language Sciences, Shanghai International Studies University, Shanghai 201620, China
*
Authors to whom correspondence should be addressed.
†
These authors contributed equally to this work.
Languages 2025, 10(6), 119; https://doi.org/10.3390/languages10060119
Submission received: 23 January 2025 / Revised: 14 May 2025 / Accepted: 19 May 2025 / Published: 26 May 2025
Browse Figures
Versions Notes

Abstract

:
This study provides a bibliometric analysis of polarity sensitivity research from 1980 to 2023, examining intellectual structure, collaboration patterns, and emerging trends. Analysing 835 documents using Bibliometrix (V.4.1.0), CiteSpace (V.6.1.R6), and VOSviewer (V1.6.18), we identify three evolutionary phases: (1) foundational theoretical development (1980–2000), transitioning from syntactic to semantic-based theories; (2) methodological diversification (2000–2010), incorporating cognitive–pragmatic frameworks and corpus-based studies; and (3) contemporary integration (2010–2023), marked by multidisciplinary approaches. Co-citation analysis reveals three intellectual clusters centred on formal semantics, pragmatic approaches, and minimalist frameworks. Geographic analysis shows the United States as the leading contributor, followed by Germany and the United Kingdom. Collaboration network analysis underscores intensive transatlantic exchanges and emerging computational contributions from Asia. Keyword co-occurrence analysis (165 terms) demonstrates theoretical sophistication and empirical integration, with growing interest in neurocognitive approaches, cross-linguistic variations, and interface phenomena. Challenges include reconciling universal principles with language-specific variations and integrating processing models with formal theories. Promising research directions involve the combination of computational modelling, diachronic studies, and applications in language teaching and natural language processing. This study maps the intellectual landscape of polarity sensitivity research while suggesting future directions toward unified theories that address universal and language-specific patterns.

1. Introduction

The study of polarity sensitivity represents one of the most intriguing and theoretically significant phenomena in linguistic research, lying at the crucial intersection of syntax, semantics, and pragmatics. Polarity-sensitive items (PSIs) are lexical expressions whose distribution is constrained by the polarity of the linguistic environment. These include negative polarity items (NPIs), such as any or ever, which tend to appear only in negative or other licensing environments, and positive polarity items (PPIs), such as already or somewhat, which typically resist such contexts. A key goal of polarity sensitivity research is to understand how such items are licensed or blocked by properties of their syntactic, semantic, and pragmatic contexts. Since Ladusaw’s (1979) seminal work establishing the downward entailment hypothesis—which characterizes a class of linguistic contexts in which general statements entail their subsets (e.g., No student passed the exam entails No linguistics student passed the exam)—research in this domain has evolved from purely syntactic approaches to increasingly sophisticated frameworks incorporating semantic, pragmatic, and cognitive dimensions. The theoretical significance and practical implications of polarity sensitivity research have grown substantially over the past two decades.
The importance of polarity sensitivity research can be appreciated across multiple dimensions. From a theoretical perspective, the study of polarity-sensitive items (PSIs) provides unique insights into the architecture of grammar and the interfaces between different linguistic components (Giannakidou, 1997, 1998). The distribution patterns of negative polarity items (NPIs) and positive polarity items (PPIs) serve as crucial diagnostic tools for understanding how syntactic structure interacts with semantic interpretation and pragmatic constraints. This interface phenomenon challenges traditional modular approaches to grammar and necessitates more sophisticated theoretical frameworks that can account for cross-module interactions. From a cognitive perspective, polarity sensitivity phenomena offer valuable windows into human language processing mechanisms. Recent neurolinguistic studies, particularly those employing ERP and fMRI methodologies, have revealed specific neural substrates involved in processing polarity-sensitive expressions, contributing to our understanding of how the brain handles complex linguistic dependencies (Orth et al., 2021; Schwab et al., 2021; Tieu & Lidz, 2016; Yanilmaz & Drury, 2018). The processing of polarity items involves sophisticated cognitive mechanisms that require real-time integration of multiple levels of linguistic information, making these phenomena particularly valuable for investigating language processing architecture. In the meantime, cross-linguistic research on polarity sensitivity has revealed fascinating patterns of variation and universality in human language. While the specific manifestations of polarity sensitivity differ across languages, as demonstrated by studies spanning the Germanic, Romance, Asian, and Slavic language families, the underlying principles governing these phenomena show remarkable consistency. This tension between universal principles and language-specific instantiations provides crucial evidence for theories about the nature of the human language faculty and its relationship to cognition.
Furthermore, the practical implications of polarity sensitivity research extend beyond theoretical linguistics into areas such as language teaching, natural language processing, and computational linguistics. Understanding how polarity items function is crucial for developing more sophisticated algorithms for sentiment analysis and machine translation, while insights from this research inform pedagogical approaches to teaching these challenging aspects of language use to second language learners.
The present study aims to conduct a comprehensive bibliometric analysis of polarity sensitivity research from 1980 to 2023, focusing specifically on three critical scientific questions:
  • How has the intellectual structure of polarity sensitivity research evolved over the past four decades, particularly concerning the integration of different theoretical approaches and methodological frameworks?
  • What are the primary research clusters and collaboration patterns in polarity sensitivity research, and how have these patterns influenced the development of theoretical frameworks and empirical methodologies?
  • What are the emerging trends and potential future directions in polarity sensitivity research, particularly concerning the integration of traditional theoretical approaches with new experimental and computational methodologies?
To address these questions, our study employs a multi-faceted bibliometric approach, utilizing advanced analytical tools including Bibliometrix, CiteSpace, and VOSviewer. This comprehensive analysis encompasses publication patterns, citation networks, co-word analysis, and collaboration networks, providing a detailed picture of the field’s development and current state. By analysing a dataset of 835 peer-reviewed journal articles and reviews published between 1980 and 2023, we aim to identify key theoretical developments, methodological innovations, and emerging research frontiers in polarity sensitivity studies. Our dataset was intentionally limited to journal articles and reviews in order to ensure methodological consistency and bibliometric reliability. This criterion aligns with established practices in bibliometric research (Aria & Cuccurullo, 2017; Van Eck & Waltman, 2010), which emphasize the importance of comparable publication types in terms of peer review, citation behaviour, and metadata structure. While monographs and book chapters—such as Horn (1989), Kadmon and Landman (1993), Lahiri (1998), and Chierchia (2013)—are not systematically indexed in major bibliographic databases like Web of Science and Scopus, their substantial theoretical influence is captured in our co-citation networks and discussed throughout the study. Thus, although these foundational works are not part of the core dataset, their intellectual significance is acknowledged and integrated into our analysis.
Our study’s significance lies in its comprehensive mapping of the field’s intellectual landscape and its potential to inform future research directions. By identifying emerging trends, methodological innovations, and gaps in current understanding, this analysis can help guide researchers toward promising new areas of investigation and facilitate more effective research collaboration. Furthermore, our findings can contribute to developing more integrated theoretical frameworks that bridge traditional approaches with contemporary methodological innovations. This bibliometric analysis is particularly timely given the field’s current theoretical and methodological transformation state. As revealed by citation patterns and keyword analysis, polarity sensitivity research is moving beyond traditional theoretical divisions toward more integrated approaches that combine formal theoretical analysis with experimental and computational methods. Understanding these transitions and their implications is crucial for the field’s continued development.

2. Materials and Methods

2.1. Data Source and Collection

In this paper, we searched WoS (Web of Science) Core Collection databases (including SCI-Expanded, SSCI, AHCI, ESCI, CCR-Expanded, and IC) and Scopus database (the largest abstract and citation database of peer-reviewed literature, including scientific journals, books, and conference proceedings) for the relevant literature (Figure 1). To obtain precise studies on polarity sensitivity research in linguistics, we retrieved data according to the following specified steps:
  • Search terms in WoS Core Collection databases: TS = (Polarity) OR TS = (PPI) OR TS = (NPI); Search terms in Scopus database: TITLE-ABS-KEY (Polarity OR PPI OR NPI)
  • Select subject categories in the WoS Core Collection databases: “Linguistics”, “Language linguistics”, “Psychology Experimental”, “Neurosciences”, and “Psychology”. Explore subject Categories in the Scopus database: “Arts and Humanities”, “Psychology”, and “Neuroscience”.
  • Limit research areas in WoS Core Collection databases: “psychology”, “linguistics”. “behavioural sciences”, “literature”, “communication”, “arts humanities”, and other topics; Limit keywords in Scopus database: Exclude keywords about Physiology, Pathophysiology, and Genetics, such as “Metabolism”, “Startle Reflex”, “Alzheimer Disease”, “Gene Expression Regulation”.
  • Set the period: 1980–2023 [Ladusaw (1979) proposes that the two basic concepts of polarity sensitivity (e.g., licensor and polarity items) are generally regarded as the temporal starting point for theoretical research on the systematization of polar items].
  • Only peer-reviewed journal articles and reviews were included to ensure methodological consistency across bibliometric analyses (Aria & Cuccurullo, 2017; Chen, 2006; Van Eck & Waltman, 2010).
  • Specify language: English
Following the previously specified search steps, an initial pool of 3252 search results related to the polarity sensitivity research was retrieved, including 1465 publications from the WoS Core Collection databases and 1787 publications from the Scopus database. A rigorous screening process was conducted to ensure the dataset’s validity, representativeness, and academic quality. Specifically, we removed duplicate records using the “find duplicates” function in both databases, resulting in 2456 unique entries. We then manually screened the titles and abstracts to exclude studies that were not directly relevant to linguistic polarity sensitivity. In particular, we excluded publications that used the term “polarity” in non-linguistic contexts, such as emotional valence in psychology, cortical or neuronal polarity in neuroscience, and cell polarity in biomedical research. These cases highlight that the presence of the keyword “polarity” was not sufficient for inclusion; only studies addressing phenomena such as negative and positive polarity items, licensing conditions, and related syntactic, semantic, pragmatic, or cognitive aspects of language were retained. This screening process yielded a final dataset of 835 relevant documents, for which we exported the full bibliographic records in both plain text and BibTeX formats.

2.2. Analysis Procedure

We used three tools, Bibliometrix (V.4.1.0), CiteSpace (V.6.1.R6), and VOSviewer (V1.6.18), to explore emerging trends and opportunities. We then identified vital contributors and provided a comprehensive and rigorous perspective on the current research state of polarity items. Bibliometrix is built on the open-source R language and ecosystem and offers a range of statistical algorithms, access to high-quality numerical routines, and integrated data visualisation capabilities. The content presented by the software includes not only some basic information, such as the number of sources, literature types, content, volume, references, and authors, as well as annual scientific output but also some advanced information, including the average number of article citations per year, relevant sources and literature, and most frequently cited authors and contributions by affiliated states (Aria & Cuccurullo, 2017). VOSviewer is an advanced software tool that facilitates the construction and visualisation of document co-citation networks. It is designed to extract relevant data, including country, keyword, author, and other pertinent information from the literature retrieval data, streamline relationship construction, and facilitate visual analysis with this data (Ding & Yang, 2020; Van Eck & Waltman, 2010). CiteSpace is a seminal data standardisation approach grounded in set theory that is leveraged to gauge the similarity of knowledge units and construct time zone and timeline views within time slices as a means of comprehending the developmental trajectory and trends of a given field of inquiry. Central concepts of CiteSpace include burst detection, betweenness centrality, and heterogeneous networks (Chen, 2006). These tools can be employed to visually represent the prevailing status, hotspots, and frontiers of research in a timely and effective manner.
The specific analytic procedures are shown as follows: Firstly, the initial step of our analysis involves a fundamental examination of the basic information about polarity sensitivity research, such as the number and growth trend and the primary countries, organisations, journals, and author analysis. We extract keywords, citations, authors, and other relevant data from bibliographic records in the BibTex format with Bibliometrix. Through Bibliometrix’s capabilities, we can identify the key contributors to polarity sensitivity research and depict the field’s current state. Secondly, we conducted an investigation of research hotspots. To achieve this, we take advantage of the literature co-citation and keyword co-occurrence analyses with VOSviewer. VOSviewer enables the generation of comprehensive visualisations of the literature co-citation networks and keyword co-occurrence networks, which offer a way to identify emerging trends and future frontiers in polarity sensitivity research. We input the data files in plain text format, set the threshold, and generate scientific knowledge maps that exhibit co-occurrence analyses of the literature keywords, subject words, and author information, and provide text mining capabilities. The interactive visualisations generated by VOSviewer present different clusters in various colours, while the connections between points represent cooperative relationships.
In the final stage of our analysis, we examined the development of polarity sensitivity research using CiteSpace. Unlike the previous step, which focused on identifying current research trends, this phase aimed to analyse diachronic changes in polarity sensitivity studies. By inputting data files in plain text format and setting specific parameters, we created keyword cluster maps and timeline maps. These visual knowledge maps utilised nodes and connections, with each node representing a reference. The size of the nodes indicated the frequency of citation, while different colours represented different years. Burst nodes positioned in the centre of the map indicated an increase in co-occurrences or references over time. The thickness and colour of the connections depicted the strength and timing of the relationship. Through these maps, we summarised the temporal evolution of research hotspots and explored the development trajectory of polarity sensitivity research. In summary, the combined use of Bibliometrix, VOSviewer, and CiteSpace has allowed us to provide a robust perspective on the current state of the field, identifying emerging trends, future frontiers, the primary themes, and the knowledge evolution associated with polarity sensitivity research (Figure 2).

3. Results

3.1. Analysis of Basic Information

3.1.1. Number and Growth Trend of Annual Publications

Eight hundred thirty-five articles and reviews about natural polarity items between 1980–2023 were concluded in the calculation (Figure 3). The overall number of publications related to polarity items remained very small from 1980 to 1995, ranging from zero to five articles. It showed a modest growth trend in the years 1995–2005. However, the results demonstrated a rapid increase from 2005–2021, suggesting that polarity sensitivity has gained growing attention from scholars in polarity sensitivity and related fields. In 2021, the number of published articles related to polarity reached a peak of 89 articles. Figure 4 shows the average number of article citations per year from 1980 to 2023, with the most citations in 2009 (0.98).

3.1.2. Primary Countries and Journals

This paper presents a comprehensive analysis of research on polarity sensitivity in linguistics, while also incorporating insights from cognitive science and neurolinguistics to address core questions about language structure and processing. A total of 80 countries were found to have contributed to the literature in question. Table 1 displays the top 10 countries, ranked by the number of publications on polarity, which are the United States (298 articles), Germany (155 articles), the United Kingdom (99 articles), the Netherlands (96 articles), France (71 articles), Japan (65 articles), Spain (61 articles), China (58 articles), Israel (43 articles), and Australia (42 articles). The findings indicate that the United States has emerged as the leading contributor to the research on polarity, producing the maximum number of publications. A total of 753 institutions were identified in the dataset; Table 1 reports the top 10 institutions ranked by publication count, with the University of Chicago being the most productive, having published 21 articles between 1980 and 2023. Country-level data in Table 1 are based on the full counting method, meaning that if a publication involved authors from multiple countries, each country represented by at least one affiliated author was credited for that publication. This approach provides a more accurate reflection of international collaboration in polarity sensitivity research (Aria & Cuccurullo, 2017). In addition, Figure 5 depicts a world map of research collaborations, revealing that the United States showed the highest level of collaborative research with the Netherlands, the United Kingdom, and Germany, while China was found to collaborate predominantly with Australia.
Table 2 presents the most productive journals in polarity sensitivity research. Lingua ranks first with 86 publications between 1980 and 2023. It is worth noting that Lingua underwent a major editorial transition in 2015, when much of its editorial board resigned and established Glossa, which has since become a key venue for linguistic research. While Lingua remains active, its academic positioning has evolved, and its prominence in our results primarily reflects its historical output. Natural Language and Linguistic Theory and Nature Language Semantics ranked No. 2 (h-index = 8). The m-index is an h-index variation representing the h-index per year since the first publication. The findings revealed that both the Journal of Psycholinguistic Research and Glossa scored the highest m-index value of 0.667, which indicates that the publications in these journals have consistently maintained a high level of impact throughout the years since their inception. These results show that these journals have exerted a significant influence in the field of polarity, making them notable and essential resources for scholars in this area of study.

3.1.3. Leading Authors

In our data set, a total of 1042 authors participated in this study, including 306 authors of single-authored documents. Table 3 presented the top 10 most productive authors regarding the polarity in linguistics research. These findings reaffirm the significance of the extensive body of work undertaken by these authors in advancing the understanding of polarity sensitivity in linguistics. The evolution of polarity sensitivity research has been shaped by two prominent scholars whose different approaches and theoretical frameworks have generated productive tensions in the field: Anastasia Giannakidou and Jack Hoeksema. Their work represents different yet complementary approaches to understanding polarity phenomena, with Giannakidou focusing on formal semantic analysis and cross-linguistic comparisons, while Hoeksema emphasizes diachronic development and corpus-based investigation. Anastasia Giannakidou, from the University of Chicago, was the most representative author in the polarity sensitivity research, with 31 publications and 125 citations, and has significantly expanded the traditional binary classification of polarity items. The three most cited articles involve the semantics of the subjunctive mood and aspect in Modern Greek (Giannakidou, 2009), the semantics and syntax of free choice items (FCIs) (Giannakidou & Quer, 2013), and negative polarity items (NPIs) with the “even” marker in Greek and Korean languages (Giannakidou & Yoon, 2011). Giannakidou’s (2001) groundbreaking contribution lies in challenging the traditional dichotomy between negative polarity items (NPIs) and positive polarity items (PPIs) by introducing free choice items (FCIs) as a distinct category. NPIs, such as any in “John didn’t see any mistakes”, are licensed in downward-entailing or nonveridical contexts, typically involving negation. PPIs, on the other hand, prefer affirmative contexts and are generally incompatible with negation. FCIs differ from both in that they convey universal-like freedom across alternatives and are licensed in modal, conditional, or imperative environments, as in “Any student can solve this problem.” Unlike NPIs, FCIs are not tied to negation but are sensitive to possibility and nonveridicality, highlighting a separate semantic and licensing behaviour (Giannakidou, 2001; Giannakidou & Yoon, 2011). Although FCIs had been examined in earlier influential work (e.g., Carlson, 1980; Haspelmath, 1997; Horn, 1972; Kadmon & Landman, 1993), Giannakidou offered a more systematic and cross-linguistically grounded account, clarifying their relation to NPIs within a unified polarity framework. Building on this foundation, she introduced the notion of nonveridicality (Giannakidou, 1998, 2001, 2009; Giannakidou et al., 2019) as a general semantic condition for polarity licensing. This framework refines and extends earlier proposals based on downward entailment, anti-additivity, and anti-morphicity (e.g., Ladusaw, 1979; Zwarts, 1996), enabling a broader characterization of licensing environments, including those involving modality, interrogatives, and epistemic weakening.
However, Giannakidou’s categorization has faced significant challenges from other scholars. Chierchia (2006), for instance, contests her separation of FCIs from NPIs, arguing instead for a common origin between negative polarity items and FCIs. This theoretical tension highlights a fundamental debate in the field about the nature of polarity sensitivity: whether it represents discrete categories or exists on a continuum.
Jack Hoeksema, from the University of Groningen (18 publications), offers a different analytical lens that both challenges and complements Giannakidou’s formal approach. While Giannakidou emphasizes synchronic cross-linguistic comparisons, Hoeksema (1994, 1998, 2010) introduces a diachronic perspective through his study of the polarity term fossilization. His research on Dutch demonstrates how words like ooit have evolved from strict NPIs to more flexible usage patterns, suggesting that polarity sensitivity is not a fixed property but rather a dynamic feature that develops over time. Hoeksema’s introduction of “semi-NPIs” (Hoeksema, 1994) represents a significant theoretical advancement that bridges the gap between Giannakidou’s discrete categories and Chierchia’s unified approach. Through corpus analysis, he demonstrates that words like bother and mind exhibit varying degrees of polarity sensitivity, with mind appearing in affirmative contexts only 1% of the time compared to care’s 20%. This gradient approach to polarity sensitivity offers a potential resolution to the theoretical debate between Giannakidou and Chierchia by suggesting that polarity sensitivity exists on a spectrum rather than in discrete categories.
The interaction between these scholars’ work has led to several important theoretical developments: (1) Category flexibility: while Giannakidou’s work establishes clear categorical distinctions, Hoeksema’s research demonstrates how these categories can be fluid over time, suggesting a more nuanced approach to polarity classification. (2) Cross-linguistic validation: Giannakidou’s work on Greek and Korean (Giannakidou & Yoon, 2011) complements Hoeksema’s Dutch studies, showing how polarity sensitivity manifests differently across languages while following similar underlying principles. (3) Methodological integration: the field has benefited from the combination of Giannakidou’s formal semantic analysis and Hoeksema’s corpus-based approach, leading to more robust empirical validation of theoretical claims. Current debates centre on several key questions that emerged from this scholarly dialogue: (1) Whether polarity sensitivity should be viewed as categorical (Giannakidou) or a gradient (Hoeksema). (2) The role of diachronic change in understanding synchronic polarity patterns. (3) The relationship between grammaticalization and polarity sensitivity. (4) The universal versus language-specific aspects of polarity items. These ongoing discussions highlight the need for further research that can integrate formal semantic analysis with diachronic and corpus-based approaches, potentially leading to a more comprehensive theory of polarity sensitivity that can account for both categorical distinctions and gradient effects across languages and time periods.

3.1.4. Language Coverage

To explore the typological scope of polarity item research, we conducted a comprehensive scan of all titles and abstracts in our dataset using Python (V.3.9)-based keyword matching. Based on a global reference list of language names and their variants, we identified 72 languages that are explicitly mentioned in the metadata, which were then grouped by geographical region (see Table 4)1.
The data reveal a strong concentration on European and East Asian languages. English overwhelmingly dominates with 197 mentions, followed by French (56), Japanese (49), Chinese (48), and German (37). This pattern reflects the historical development of polarity theory in formal semantics, where much foundational work has been rooted in Indo-European languages (e.g., Giannakidou, 1998; Ladusaw, 1979; Zwarts, 1996). East Asian languages such as Japanese, Chinese, and Korean also show significant representation, due in part to their typological relevance and the prominence of researchers working in these linguistic traditions (Cheng & Giannakidou, 2013; Lee, 2003).
By contrast, coverage of languages from Africa, Southeast Asia, and indigenous American language families remains minimal. Only a few studies mention languages such as Zulu, Swahili, Telugu, or Vietnamese (Carstens & Mletshe, 2016; Collins et al., 2018; Dilip & Kumar, 2019). This asymmetry highlights a persistent bias in formal linguistic research, where languages with well-documented grammar, corpus availability, and proximity to dominant research institutions tend to be overrepresented (Bickel, 2014).
The observed distribution underscores the importance of addressing current imbalances by incorporating lesser-studied and typologically diverse languages into future research. Expanding the empirical base of polarity studies may yield new insights into the structural and semantic conditions underlying polarity licensing and serve to test the cross-linguistic robustness of existing theoretical frameworks. Such efforts are essential for developing a more comprehensive and typologically informed account of polarity phenomena.

3.2. Analyses of Research Hotspots

3.2.1. Literature Co-Citation

The literature co-citation network analysis in the polarity domain, generated using VOSviewer with a minimum citation threshold of 20, reveals three distinct intellectual clusters comprising 71 qualifying documents. This network structure illuminates the theoretical evolution and methodological developments in polarity sensitivity research (Figure 6).
A notable feature of the analysis is the divergence between historical influence (shown in the co-citation network) and contemporary impact (reflected in the citation metrics). The three clusters represent distinct theoretical foundations: The red cluster centres around Giannakidou’s (1999, 2001, 2006) works, representing the theoretical core of polarity sensitivity research. Giannakidou’s (1998) seminal contribution introduced the monotonicity hierarchy framework, which marked a significant theoretical advancement by proposing a more nuanced understanding of downward-entailing contexts. This framework expanded beyond simple negation to encompass expressions of scarcity (such as hardly and barely), negative connotation (i.e., difficult and impossible), and conditionals. While these foundational works are heavily co-cited in the network, Giannakidou’s later work (Giannakidou, 2009) appears in the top citation list with 80 citations (normalized citation of 7.65), suggesting an evolution in theoretical frameworks.
The blue cluster, anchored by Krika (1995) and Kadmon and Landman (1993), emphasizes semantic and pragmatic approaches. Kadmon and Landman’s domain-expansion theory challenged prevailing views by arguing that informational strength, rather than logical properties, primarily governs the distribution of polarity items. They demonstrated that free choice items function by expanding referential domains, suggesting that contextual distribution is limited by original meaning and emphasis function rather than universal logical principles.
The green cluster features works by Chomsky (1995) and Laka Mugarza (1990), connecting to minimalist approaches and negation studies. Horn (1989)’s research serves as a crucial bridge between clusters, establishing the field’s theoretical foundation by providing a comprehensive framework for understanding negation in natural language. Horn proposed that polarity sensitivity operates through universal logical principles, though this universal approach was later challenged by subsequent research.
A significant finding emerges when comparing the network structure with current citation impacts: the most highly cited works in Table 5 represent a methodological transformation in the field. Xiang et al. (2009), with 102 citations (normalized 9.75), employed cognitive neuroscientific approaches to examine polarity sensitivity mechanisms. Their groundbreaking research on illusory licensing effects using event-related potentials (ERPs) revealed that NPI licensing exhibits unique grammatical illusion effects absent reflexive binding, challenging both Giannakidou’s formal semantic approach and Kadmon and Landman’s domain-expansion theory. Similarly, Vasishth et al. (2008), with 85 citations (normalized 18.94), contributed to this methodological evolution.
The analysis also highlights important theoretical challenges, such as those posed by Lahiri (1998), whose analysis of Hindi NPIs questioned the universality of polarity sensitivity mechanisms by demonstrating that certain licensing conditions are language-specific rather than universal. This work, together with Linebarger (1987), which connects all three clusters in the network, exemplifies the field’s ongoing negotiation between universal and language-specific approaches.
The disparity between co-citation patterns and raw citation counts reveals a significant transition: while theoretical foundations remain crucial for structuring the discipline, contemporary research increasingly values empirical and experimental contributions. This is evident in works like Steinhauer et al. (2010) and Gajewski (2011), which appear in the top citation list but show different positioning in the co-citation network.
This comprehensive analysis suggests a field in dynamic transition, where historical theoretical foundations continue to structure the discipline while newer, empirically oriented works drive current research directions. The field appears to be moving toward integrated frameworks that can account for both universal principles and language-specific implementations while incorporating insights from processing studies. This trajectory indicates that future advances in polarity sensitivity research will likely emerge from work that bridges theoretical approaches with experimental evidence while accounting for cross-linguistic variation and real-time processing factors.

3.2.2. Top Keyword Co-Occurrence

From keyword co-occurrence analysis, we were able to evaluate how strongly and tightly related the keywords were to each other and find which keyword appeared most frequently in the polarity sensitivity research field. The current study identified 165 commonly occurring keywords based on the criteria that the occurrence was above five from 2113 (Figure 7). We also put the top 10 co-occurrence times of keywords in Table 6. The keyword co-occurrence analysis (Figure 7) revealed sophisticated interconnections among research themes in polarity sensitivity studies, demonstrating the field’s theoretical depth and methodological diversity.
The close association between “information”, “polarity”, and related terms reflects how polarity sensitivity research has evolved to encompass pragmatic inference and communicative strategies. The high centrality of these co-occurring terms (polarity: 0.30) suggests that polarity items serve as windows into information transmission patterns in discourse. This aligns with Giannakidou and Yoon (2011)’s argument that polarity sensitivity phenomena are deeply connected to speakers’ communicative intentions and information–structural constraints. The network reveals how seemingly simple lexical items (NPIs) can illuminate complex information management patterns in communication, supporting Chierchia (2013)’s proposal that polarity items are sensitive to information–structural configurations.
The visualization reveals extensive cross-linguistic research networks, with significant connections between “polarity” and various language nodes, including Korean, Dutch, German, and English. This cross-linguistic focus, coupled with links to “grammaticalization” and “corpus analysis”, provides robust empirical support for the universality of polarity sensitivity phenomena while highlighting language-specific instantiations. The temporal progression of the network suggests increasing sophistication in typological approaches, particularly in understanding how different languages grammaticalize polarity sensitivity constraints.
A significant cluster emerges around experimental methodologies, connecting “reaction time”, “ERP”, “fMRI”, and “controlled study” with “comprehension” (n = 31, centrality = 0.15). This constellation reflects that neurocognitive approaches have emerged as a crucial dimension in polarity sensitivity research, as evidenced by the dense clustering of methodology-related keywords. The co-occurrence of “brain imaging in language” with “negative polarity” has revealed specific neural substrates involved in polarity processing, while connections between “cerebral cortex” and “brain region” demonstrate successful localization of these processes. The integration of multiple experimental paradigms, including ERP studies (indicated by “event-related potential” and “evoked potential” clusters) and fMRI research, has provided converging evidence about the temporal and spatial dynamics of polarity processing. These methodological advances have been particularly influential in understanding processing constraints, as shown by the connections between “computational modelling”, “capacity”, and “decision making”.
The network reveals increasing theoretical sophistication through the co-occurrence of specialized concepts. The cluster containing “presupposition” (connected to “scope” and “monotonicity”) reflects the field’s engagement with formal semantic principles, supporting Ladusaw (1979)’s seminal work on the relationship between downward entailment and NPI licensing. The presence of the “syntax-semantics interface” and its connections to both “adverb” and “NPI any” demonstrates how research has progressed toward more granular analyses of specific polarity items and their structural contexts. The temporal progression (shown by the colour gradient) indicates that this theoretical refinement has accelerated since 2014, with increasing attention to interface phenomena.
This network analysis suggests that polarity sensitivity research has evolved from purely syntactic accounts toward integrated models that encompass information structure, cross-linguistic variation, processing mechanisms, and refined theoretical frameworks. The high centrality measures of both theoretical (polarity: 0.30) and methodological (human: 0.25) terms indicate a field that successfully combines formal theoretical approaches with empirical investigation.

3.3. Timeline Visualization Analysis

We conducted a clustering and timeline analysis of the literature. Figure 8 and Table 7 display the resulting thematic clusters along with the high-frequency co-words of the top 10 clusters, while Figure 9 traces the temporal development of these clusters. Each horizontal line in the timeline corresponds to a distinct research theme, with keywords positioned according to their first significant appearance; red circles indicate burst keywords identified via burstness analysis, marking emergent topics with rapid citation growth. The temporal evolution of research on polarity sensitivity and negative polarity items (NPIs) reveals distinct developmental phases and research trajectories from 1980 to 2023. The analysis yields a network modularity index (Q) of 0.7877 and a mean silhouette score of 0.9324. The modularity index measures the strength of division of the network into distinct clusters, with values above 0.3 typically indicating a significant and well-structured community organization (Newman, 2006). The mean silhouette indicator assesses the internal consistency of the clusters, ranging from –1 to 1, where higher values reflect greater cohesion within clusters and clearer separation between them (Rousseeuw, 1987). These results indicate a well-defined cluster structure and high internal consistency, which are evidence of coherent and sustained theoretical development within the field.

3.3.1. Emergence of Core Research Areas (1980–2000)

Figure 9 indicates that prior to 2000, the most prominent bursts centred on keywords such as negation, free choice, negative polarity, and human, which were primarily associated with Clusters #0 (language), #1 (male), #3 (polarity), #5 (even), and #7 (negative polarity). These early burst terms reflect the foundational concerns of the field during its formative phase, especially the effort to demarcate the environments that license NPIs and to establish cross-linguistic generalizations grounded in either syntactic or semantic paradigms. The prominence of negation and negative polarity aligns with the field’s initial focus on formal conditions for licensing, while free choice anticipates a theoretical expansion beyond traditional NPI typologies. These conceptual focal points foreshadowed the emergence of competing frameworks in polarity sensitivity research. Ladusaw’s (1979, 1983) influential work establishing the downward entailment hypothesis marked a crucial transition from purely syntactic approaches to semantic-based theories. This period witnessed a productive tension between syntax-oriented perspectives, represented by scholars such as Linebarger (1980, 1987, 1991) and Progovac (1992), who emphasized structural configurations in NPI licensing, and semantic approaches, championed by Hoeksema (1983, 1986), who focused on scopal relationships and downward-entailing operators.
The visualization analysis reveals how this theoretical dialogue manifested in research clusters, particularly evident in the high co-occurrence of terms related to syntactic structure and semantic interpretation. The integration of gradable adjectives into polarity research during this period reflects a growing recognition of the interface between scalar semantics and polarity sensitivity. This connection was particularly illuminated through studies examining how comparative constructions create downward-entailing contexts that license NPIs, contributing to a more nuanced understanding of licensing conditions. In addition to these core theoretical developments, the emergence of human as a burst keyword suggests that researchers also began to explore polarity phenomena from a cognitive processing perspective. This line of work, though less central to formal theory, highlights the growing interest in how polarity sensitivity is realized in language comprehension, drawing on psycholinguistic and experimental methods.

3.3.2. Modern Research Developments (2000–2010)

Between 2000 and 2010, prominent burst keywords such as comprehension, concord, information structure, semantics, even, scalar implicature, and nonveridicality emerged across nearly all thematic clusters. This widespread distribution of bursts indicates a period of intensified diversification in analytical perspectives—ranging from syntax and semantics to information structure and experimental processing—while simultaneously marking a stage of conceptual consolidation. Together, these trends reflect the transition of polarity sensitivity research into a theoretically mature and methodologically expansive field.
Our bibliometric analysis reveals a significant shift toward integrated cognitive-pragmatic frameworks (Figure 8), as evidenced by the prominence of Clusters #0, #1, and #2. This evolution challenges earlier approaches that treated polarity sensitivity primarily as a syntactic phenomenon. The works of Chierchia (2004, 2013) and Horn (2000) demonstrate how pragmatic factors interact with traditional licensing conditions, showing that NPI distribution patterns cannot be fully explained through syntax or semantics alone.
The emergence of information structure studies in polarity research, represented by a distinct cluster with a high silhouette value (0.975), reflects a growing recognition of discourse-pragmatic factors. This development aligns with research demonstrating that polarity items play crucial roles in conveying speaker attitudes and managing discourse dynamics, as established in the foundational work of Fauconnier (1975a, 1975b) and further developed by contemporary scholars.
Corpus analysis approaches, identified as Cluster #9 in our visualization, have provided empirical support for theoretical claims while uncovering previously unrecognized patterns in NPI distribution. The high frequency of co-occurring terms related to computational modelling and corpus analysis (silhouette = 0.925) indicates a methodological shift toward quantitative approaches, enabling more robust testing of theoretical predictions about polarity item behaviour.
Among the thematic clusters identified, two non-traditional groupings—Cluster #15 (Conversation Analysis) and Cluster #20 (Premotor)—offer valuable perspectives beyond the core syntactic-semantic focus of polarity research. Cluster #15, characterized by terms such as question, response, and negation, includes studies on reversed polarity constructions, particularly reversed polarity questions, which shape semantic environments, like nonveridicality and antiveridicality, that are central to NPI licensing (e.g., Krifka, 2015; Van Rooy, 2003). Cluster #20 centres on embodied and neurocognitive approaches to negation, with studies such as García-Marco et al. (2019) and Alemanno et al. (2012) showing how negative markers modulate motor preparation and neural activation during language processing. Although outside traditional syntactic or semantic domains, these clusters underscore the field’s growing interdisciplinarity, reflecting its extension into interactional, cognitive, and neurobiological dimensions.

3.3.3. Contemporary Research Integration (2010–2023)

Between 2010 and 2023, the most prominent burst keywords—such as quantifier, sentence processing, Chinese, negative quantifier, and Jespersen Cycle—reflect a shift toward more focused lines of inquiry. Research during this period increasingly emphasized fine-grained analyses of specific linguistic structures and greater typological diversity, particularly through in-depth studies of individual languages and historically grounded grammatical phenomena.
For instance, Cluster #14, labelled “Pontic Greek”, exemplifies this trend toward language-specific and diachronically informed analysis. It encompasses research on the Romeyka variety—a highly conservative Greek dialect spoken in the Pontus region—with particular attention paid to the behaviour of infinitival structures under modal and antiveridical environments. These contexts are central to understanding polarity licensing. Representative work by Sitaridou (2014) has demonstrated that the Romeyka infinitive can function as a negative polarity item, thereby contributing important empirical evidence to broader theoretical debates. While the cluster label may appear unexpected at first glance, it reflects a coherent line of inquiry grounded in the interaction between modality, clause structure, and polarity sensitivity in an understudied linguistic variety.
Recent developments in the field, as revealed by our temporal analysis, show an increasing integration of multiple theoretical perspectives (Figure 9). The visualization demonstrates strong connections between computational, experimental, and theoretical clusters, suggesting a movement toward more comprehensive frameworks that can account for both structural and usage-based aspects of polarity sensitivity.
The interaction between scalar implicature and polarity sensitivity has emerged as a particularly productive area of investigation. Research has revealed complex interactions between polarity items and scalar inference patterns, contributing to our understanding of how pragmatic and semantic factors interact in natural language. This development is reflected in the close clustering of terms related to scalar phenomena and polarity sensitivity in our visualization.
Our analysis reveals three significant trends that have implications for both theoretical development and practical applications in linguistics. First, the evolution from purely syntactic approaches to integrated cognitive-pragmatic frameworks represents a crucial theoretical advancement, suggesting that future research must consider multiple linguistic levels simultaneously. Second, the emergence of pragmatic function studies alongside traditional syntactic analyses marks a significant paradigm shift, challenging conventional views about the role of polarity items in grammar. Third, the temporal progression from expletive negation studies through NPI licensing to scalar implication research indicates a movement toward more nuanced understanding of how polarity items operate across different linguistic levels.
These bibliometric findings suggest critical directions for future research, particularly in developing unified theories that can account for cognitive, pragmatic, and computational perspectives while maintaining practical applicability in areas such as language teaching and natural language processing. The field appears to be moving beyond traditional approaches toward a more integrated understanding of polarity sensitivity, as evidenced by the increasing interconnectedness of research clusters in our visualization.

4. Discussion

4.1. Main Findings

The bibliometric analysis of polarity sensitivity research from 1980 to 2023 reveals several significant patterns and developments that warrant detailed discussion. First, the field has shown remarkable growth in both quantity and methodological sophistication, particularly since 2005. This growth pattern suggests not only increasing scholarly interest but also the maturation of polarity sensitivity as a distinct research domain within linguistics. The sharp increase in publications between 2005–2021, culminating in 89 articles in 2021, indicates that polarity sensitivity has evolved from a niche topic to a central concern in linguistic theory.
The geographical distribution of research contributions presents an interesting pattern of both concentration and diversification. While the United States maintains a dominant position with 298 articles, the significant contributions from European institutions, particularly in Germany (155 articles) and the United Kingdom (99 articles), suggest a well-established transatlantic research dialogue. Notably, the Netherlands, despite its relatively small geographic size, shows a high level of institutional participation. This prominence likely reflects its strong academic tradition in linguistics and cognitive science, its emphasis on international collaboration, and the widespread use of English in scholarly communication. Dutch research clusters, particularly around the University of Groningen, have played a visible role. Overall, this geographic distribution highlights how theoretical approaches to polarity sensitivity have been shaped by distinct regional traditions: the American focus on formal semantics, the European emphasis on corpus linguistics, and the emerging Asian contribution driven by computational methods.
The co-citation analysis reveals a fascinating intellectual structure organized around three main clusters, each representing distinct theoretical orientations. The red cluster, centred on Giannakidou (1997, 1998, 1999, 2001, 2006, 2011, 2017); Giannakidou and Cheng (2006); Giannakidou et al. (2019); and Giannakidou and Yoon (2024) emphasizes formal semantic approaches and has shown remarkable staying power in the field. The blue cluster, focusing on pragmatic approaches, suggests a growing recognition of the importance of context and usage patterns in understanding polarity phenomena. The green cluster’s connection to minimalist approaches indicates the continuing relevance of syntactic frameworks, even as the field has broadened its theoretical scope.
Perhaps most striking is the evident methodological transformation revealed by citation patterns. While theoretical works continue to serve as the conceptual foundation of the field—as reflected in the structure of the co-citation network—recent years have seen the highest citation counts increasingly awarded to works employing experimental and quantitative methods. This shift is particularly evident in the success of neurolinguistic approaches, as exemplified by Xiang et al. (2009), whose highly-cited work investigates the online processing of polarity items. This methodological evolution suggests a field-wide move toward closer integration between theoretical modelling and empirical validation. Notably, this trend is not unique to polarity research; similar patterns of methodological convergence have been observed in other areas of linguistic inquiry, such as agreement systems and aspectual classifications, indicating a broader epistemological shift across contemporary linguistic research.
The keyword co-occurrence analysis reveals several intriguing patterns in the research focus. The high centrality measures for both theoretical terms (polarity: 0.30) and methodological terms (human: 0.25) suggest a field that has successfully balanced formal theoretical development with empirical investigation. The emergence of clusters around experimental methodologies, particularly in neurocognitive approaches, indicates a growing sophistication in understanding the processing dimensions of polarity sensitivity.

4.2. Challenges and Future Development

Despite these advances, several significant challenges face the field of polarity sensitivity research. First, there is a persistent tension between universal and language-specific approaches to polarity phenomena. While cross-linguistic research has expanded significantly, as shown by the diverse language nodes in the keyword network, developing theoretical frameworks that can accommodate both universal principles and language-specific variations remains a challenge. Future research needs to address this through more sophisticated typological studies that can account for both commonalities and variations across languages.
Second, the integration of processing models with formal theoretical approaches remains incomplete. While neurolinguistic studies have provided valuable insights into real-time processing of polarity items, connecting these findings to formal semantic and syntactic theories presents both a challenge and an opportunity for future research. The field would benefit from developing integrated models that can account for both competence and performance aspects of polarity sensitivity.
Third, the role of pragmatic factors in polarity sensitivity remains incompletely understood. While the keyword analysis shows increasing attention to information structure and discourse factors, developing formal models that can adequately capture the interaction between semantic licensing conditions and pragmatic constraints represents a significant challenge. Future research should focus on developing more sophisticated frameworks for understanding how context and usage patterns influence polarity item distribution.
Looking forward, several promising directions for future research emerge from this analysis: (1) Computational modelling: The growing prominence of computational approaches in the keyword analysis suggests an opportunity for developing more sophisticated computational models of polarity sensitivity. These models could help bridge the gap between theoretical predictions and empirical observations. (2) Cross-linguistic processing studies: While cross-linguistic variation in polarity sensitivity is well-documented, there is a need for more systematic investigation of processing patterns across languages. This could help determine which aspects of polarity sensitivity processing are universal and which are language-specific. (3) Diachronic development: The timeline analysis reveals relatively limited attention to the historical development of polarity items. Future research could benefit from more systematic investigation of how polarity systems evolve over time, potentially providing insights into the relationship between synchronic patterns and diachronic processes. (4) Interface studies: The increasing recognition of interface phenomena in the keyword analysis suggests a need for more detailed investigation of how polarity sensitivity interacts with other linguistic systems, particularly information structure and discourse organization. (5) Applied research: While theoretical understanding has advanced significantly, applications to language teaching and natural language processing remain underdeveloped. Future research should focus on translating theoretical insights into practical applications.
The field of polarity sensitivity research stands at an exciting juncture. The bibliometric analysis reveals a mature field that has successfully integrated multiple theoretical and methodological approaches. However, significant challenges remain in developing unified theories that can account for both universal patterns and language-specific variations, while also incorporating insights from processing studies and pragmatic approaches. The field’s future development will likely depend on its ability to maintain theoretical sophistication while expanding its empirical base and practical applications.

5. Conclusions

This bibliometric analysis has revealed significant patterns in the evolution of polarity sensitivity research from 1980 to 2023. Through a systematic examination of 835 publications using CiteSpace, VOSviewer, and Bibliometrix, we identified three key developments in the field. First, research output has shown substantial growth, particularly since 2005, reflecting the field’s maturation from a niche topic to a central concern in linguistic theory. Second, co-citation analysis revealed three distinct intellectual clusters representing formal semantic, pragmatic, and syntactic approaches, demonstrating the field’s theoretical sophistication while maintaining productive dialogue across different frameworks. Third, our temporal analysis indicates a significant methodological transformation, with increasing emphasis on experimental and quantitative methods, particularly in neurolinguistic approaches. The keyword co-occurrence analysis (165 terms) highlights the successful integration of theoretical development with empirical investigation, as evidenced by high centrality measures for both theoretical and methodological terms. These findings suggest that polarity sensitivity research has evolved into a mature field characterized by theoretical depth and methodological rigour, pointing to promising future directions in computational modelling and cross-linguistic processing studies.

Author Contributions

Conceptualization, L.K.; methodology, L.K. and Y.L.; software, Y.L.; validation, X.J., Y.J. and Y.S.; formal analysis, L.K.; investigation, L.K.; resources, Y.L.; data curation, Y.L.; writing—original draft preparation, L.K.; writing—review and editing, X.J., Y.J. and Y.S.; visualization, Y.L.; supervision, X.J.; project administration, L.K., X.J., Y.J. and Y.S.; funding acquisition, X.J., Y.J. and Y.S. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by National Natural Science Foundation of China grant number 31971037, National Social Science Foundation of China grant number 19BYY027, the Open Project of the Key Laboratory of Language Science and Multilingual Artificial Intelligence grant number KLSMAI-2023-OP-0004, and Supervisor Academic Guidance Program of Shanghai International Studies University grant number 2023DSYL001.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The original data presented in the study are openly available in OSF at https://osf.io/5z2mg/?view_only=cdae56a42580469d9614b46d1b58dd1f (accessed on 29 October 2024).

Acknowledgments

We would like to express our sincere gratitude to the reviewers for their constructive feedback and valuable suggestions on our manuscript. We also extend our appreciation to the developers of CiteSpace for data visualization.

Conflicts of Interest

The authors declare no conflicts of interest.

Notes

1
It is important to note that this method captures only languages referenced in titles and abstracts; studies where the target language is not explicitly stated may not be accounted for.
2
Cluster labels are automatically assigned based on the highest-ranking keyword (by LLR and p-value) (Chen, 2016).

References

  1. Alemanno, F., Houdayer, E., Cursi, M., Velikova, S., Tettamanti, M., Comi, G., Cappa, S. F., & Leocani, L. (2012). Action-related semantic content and negation polarity modulate motor areas during sentence reading: An event-related desynchronization study. Brain Research, 1484, 39–49. [Google Scholar] [CrossRef] [PubMed]
  2. Aria, M., & Cuccurullo, C. (2017). Bibliometrix: An R-tool for comprehensive science mapping analysis. Journal of Informetrics, 11(4), 959–975. [Google Scholar] [CrossRef]
  3. Bickel, B. (2014). Linguistic diversity and universals. In N. J. Enfield, P. Kockelman, & J. Sidnell (Eds.), The Cambridge handbook of linguistic anthropology (pp. 101–124). Cambridge University Press. [Google Scholar]
  4. Carlson, G. N. (1980). Polarity any is existential. Linguistic Inquiry, 11(4), 799–804. [Google Scholar]
  5. Carstens, V., & Mletshe, L. (2016). Negative concord and nominal licensing in Xhosa and Zulu. Natural Language & Linguistic Theory, 34, 761–804. [Google Scholar]
  6. Chen, C. (2006). CiteSpace II: Detecting and visualizing emerging trends and transient patterns in scientific literature. Journal of the American Society for Information Science and Technology, 57(3), 359–377. [Google Scholar] [CrossRef]
  7. Chen, C. (2016). CiteSpace: A practical guide for mapping scientific literature. Nova Science Publishers Hauppauge. [Google Scholar]
  8. Cheng, L. L.-S., & Giannakidou, A. (2013). The non-uniformity of wh-indeterminates with polarity and free choice in Chinese. Strategies of Quantification, 44, 123. [Google Scholar] [CrossRef]
  9. Chierchia, G. (2004). Scalar implicatures, polarity, and the syntax-pragmatics interface. In A. Belletti (Ed.), Structures and beyond: The cartography of syntactic structures (Vol. 3, pp. 39–103). Oxford University Press. [Google Scholar]
  10. Chierchia, G. (2006). Broaden your views: Implicatures of domain widening and the “logicality” of language. Linguistic Inquiry, 37(4), 535–590. [Google Scholar] [CrossRef]
  11. Chierchia, G. (2013). Logic in grammar: Polarity, free choice, and intervention. Oxford University Press. [Google Scholar]
  12. Chomsky, N. (1995). The Minimalist Program. The MIT Press. [Google Scholar]
  13. Collins, C., Postal, P. M., & Yevudey, E. (2018). Negative polarity items in Ewe. Journal of Linguistics, 54(2), 331–365. [Google Scholar] [CrossRef]
  14. Dilip, M. J., & Kumar, R. (2019). Negative polarity items in Telugu. Acta Linguistica Asiatica, 9(1), 9–28. [Google Scholar] [CrossRef]
  15. Ding, X., & Yang, Z. (2020). Knowledge mapping of platform research: A visual analysis using VOSviewer and CiteSpace. Electronic Commerce Research, 22, 787–809. [Google Scholar] [CrossRef]
  16. Ernst, T. (2009). Speaker-oriented adverbs. Natural Language & Linguistic Theory, 27, 497–544. [Google Scholar]
  17. Fauconnier, G. (1975a). Polarity and the scale principle. Chicago Linguistics Society, 11, 188–199. [Google Scholar]
  18. Fauconnier, G. (1975b). Pragmatic scales and logical structure. Linguistic Inquiry, 6(3), 353–375. [Google Scholar]
  19. Gajewski, J. R. (2011). Licensing strong npis. Natural Language Semantics, 19, 109–148. [Google Scholar] [CrossRef]
  20. García-Marco, E., Morera, Y., Beltrán, D., de Vega, M., Herrera, E., Sedeño, L., Ibáñez, A., & García, A. M. (2019). Negation markers inhibit motor routines during typing of manual action verbs. Cognition, 182, 286–293. [Google Scholar] [CrossRef]
  21. Giannakidou, A. (1997). The landscape of polarity items [Ph.D. Dissertation, University of Groningen]. [Google Scholar]
  22. Giannakidou, A. (1998). Polarity sensitivity as (non) veridical dependency. John Benjamins Publishing Company. [Google Scholar]
  23. Giannakidou, A. (1999). Affective dependencies. Linguistics and Philosophy, 22, 367–421. [Google Scholar] [CrossRef]
  24. Giannakidou, A. (2001). The meaning of free choice. Linguistics and Philosophy, 24(6), 659–735. [Google Scholar] [CrossRef]
  25. Giannakidou, A. (2006). Only, emotive factive verbs, and the dual nature of polarity dependency. Language, 82(3), 575–603. [Google Scholar] [CrossRef]
  26. Giannakidou, A. (2009). The dependency of the subjunctive revisited: Temporal semantics and polarity. Lingua, 119(12), 1883–1908. [Google Scholar] [CrossRef]
  27. Giannakidou, A. (2011). Negative and positive polarity items: Variation, licensing, and compositionality. In C. Maienborn, K. von Heusinger, & P. Portner (Eds.), Semantics: An international handbook of natural language meaning. Mouton de Gruyter. [Google Scholar]
  28. Giannakidou, A. (2017). Evaluative subjunctive and nonveridicality. In Tense, mood, and modality: New answers to old questions (pp. 177–217). University of Chicago Press. [Google Scholar]
  29. Giannakidou, A., & Cheng, L. L.-S. (2006). (In) definiteness, polarity, and the role of wh-morphology in free choice. Journal of Semantics, 23(2), 135–183. [Google Scholar] [CrossRef]
  30. Giannakidou, A., & Quer, J. (2013). Exhaustive and non-exhaustive variation with free choice and referential vagueness: Evidence from Greek, Catalan, and Spanish. Lingua, 126, 120–149. [Google Scholar] [CrossRef]
  31. Giannakidou, A., von Heusinger, K., Maienborn, C., & Portner, P. (2019). Negative and positive polarity items. In P. Portner, C. Maienborn, & K. von Heusinger (Eds.), Semantics—Sentence and information structure (pp. 69–134). De Gruyter Mouton. [Google Scholar]
  32. Giannakidou, A., & Yoon, S. (2011). The subjective mode of comparison: Metalinguistic comparatives in Greek and Korean. Natural Language & Linguistic Theory, 29, 621–655. [Google Scholar]
  33. Giannakidou, A., & Yoon, S. (2024). Rethinking negative polarity and free choice in comparatives: A crosslinguistic perspective. TABU Festschrift for Jack Hoeksema, 38, 30–56. [Google Scholar] [CrossRef]
  34. Haspelmath, M. (1997). Indefinite pronouns. Oxford University Press. [Google Scholar]
  35. Hoeksema, J. (1983). Negative polarity and the comparative. Natural Language & Linguistic Theory, 1, 403–434. [Google Scholar]
  36. Hoeksema, J. (1986). Monotonicity phenomena in natural language. In J. Gutiérrez-Rexach (Ed.), Semantics: Critical concepts in linguistics (pp. 121–135). Routledge. [Google Scholar]
  37. Hoeksema, J. (1994). On the grammaticalization of negative polarity items. In S. Gahl, A. Dolbey, & C. Johnson (Eds.), Annual Meeting of the Berkeley Linguistics Society (pp. 273–282). Berkeley Linguistics Society. [Google Scholar]
  38. Hoeksema, J. (1998). On the (non)loss of polarity sensitivity: Dutch ooit. In R. Mogg, & L. van Bergen (Eds.), Historical Linguistics 1995: Volume 2: Germanic linguistics (pp. 101–114). Benjamins. [Google Scholar]
  39. Hoeksema, J. (2010). Dutch ENIG: From nonveridicality to downward entailment. Natural Language & Linguistic Theory, 28, 837–859. [Google Scholar]
  40. Holmberg, A. (2013). The syntax of answers to polar questions in English and Swedish. Lingua, 128, 31–50. [Google Scholar] [CrossRef]
  41. Horn, L. R. (1972). On the semantic properties of logical operators in English. University of California, Los Angeles. [Google Scholar]
  42. Horn, L. R. (1989). A natural history of negation (Vol. 24). University of Chicago Press. [Google Scholar]
  43. Horn, L. R. (2000). Any and (-) ever: Free choice and free relatives. In A. Z. Wyner (Ed.), Proceedings of the 15th annual conference of the Israeli association for theoretical linguistics (Vol. 15, pp. 71–111). Israeli Association for Theoretical Linguistics. [Google Scholar]
  44. Iatridou, S., & Zeijlstra, H. (2013). Negation, polarity, and deontic modals. Linguistic Inquiry, 44(4), 529–568. [Google Scholar] [CrossRef]
  45. Kadmon, N., & Landman, F. (1993). Any. Linguistics and Philosophy, 16, 353–422. [Google Scholar] [CrossRef]
  46. Krifka, M. (2015). Bias in commitment space semantics: Declarative questions, negated quetions, and question tags. In S. D’Antonio, M. Moroney, & C. R. Littl (Eds.), Proceedings of SALT 25 (pp. 328–345). LSA Open Journal Systems. [Google Scholar]
  47. Krika, M. (1995). The semantics and pragmatics of polarity items. Linguistic Analysis, 25, 209–257. [Google Scholar]
  48. Ladusaw, W. A. (1979). Polarity sensitivity as inherent scope relations [Dotoral Dissertation, The University of Texas]. [Google Scholar]
  49. Ladusaw, W. A. (1983). Logical form and conditions on grammaticality. Linguistics and Philosophy, 6(3), 373–392. [Google Scholar] [CrossRef]
  50. Lahiri, U. (1998). Focus and negative polarity in Hindi. Natural Language Semantics, 6(1), 57–123. [Google Scholar] [CrossRef]
  51. Laka Mugarza, M. I. (1990). Negation in syntax—On the nature of functional categories and projections [Dotoral Dissertation, Massachusetts Institute of Technology]. [Google Scholar]
  52. Lee, C. (2003). Negative polarity items and free choice in Korean and Japanese: A contrastive study. Korean Society of Bilingualism, 22, 1–48. [Google Scholar]
  53. Linebarger, M. C. (1980). The grammar of negative polarity [Dotoral Dissertation, Massachusetts Institute of Technology]. [Google Scholar]
  54. Linebarger, M. C. (1987). Negative polarity and grammatical representation. Linguistics and Philosophy, 10(3), 325–387. [Google Scholar] [CrossRef]
  55. Linebarger, M. C. (1991). Negative polarity as linguistic evidence. Chicago Linguistic Society, 27(2), 165–188. [Google Scholar]
  56. Newman, M. E. (2006). Modularity and community structure in networks. Proceedings of the National Academy of Sciences, 103(23), 8577–8582. [Google Scholar] [CrossRef]
  57. Orth, W., Yoshida, M., & Sloggett, S. (2021). Negative polarity item (NPI) illusion is a quantification phenomenon. Journal of Experimental Psychology: Learning, Memory, and Cognition, 47(6), 906. [Google Scholar] [CrossRef]
  58. Progovac, L. (1992). Negative polarity: A semantico-syntactic approach. Lingua, 86(4), 271–299. [Google Scholar] [CrossRef]
  59. Rousseeuw, P. J. (1987). Silhouettes: A graphical aid to the interpretation and validation of cluster analysis. Journal of Computational and Applied Mathematics, 20, 53–65. [Google Scholar] [CrossRef]
  60. Schwab, J., Liu, M., & Mueller, J. L. (2021). On the acquisition of polarity items: 11-to 12-year-olds’ comprehension of German NPIs and PPIs. Journal of Psycholinguistic Research, 50(6), 1487–1509. [Google Scholar] [CrossRef]
  61. Sitaridou, I. (2014). Modality, antiveridicality and complementation: The Romeyka infinitive as a negative polarity item. Lingua, 148, 118–146. [Google Scholar] [CrossRef]
  62. Steinhauer, K., Drury, J. E., Portner, P., Walenski, M., & Ullman, M. T. (2010). Syntax, concepts, and logic in the temporal dynamics of language comprehension: Evidence from event-related potentials. Neuropsychologia, 48(6), 1525–1542. [Google Scholar] [CrossRef] [PubMed]
  63. Tieu, L., & Lidz, J. (2016). NPI licensing and beyond: Children’s knowledge of the semantics of any. Language Acquisition, 23(4), 311–332. [Google Scholar] [CrossRef]
  64. Van Eck, N., & Waltman, L. (2010). Software survey: VOSviewer, a computer program for bibliometric mapping. Scientometrics, 84(2), 523–538. [Google Scholar] [CrossRef]
  65. Van Rooy, R. (2003). Negative polarity items in questions: Strength as relevance. Journal of Semantics, 20(3), 239–273. [Google Scholar] [CrossRef]
  66. Vasishth, S., Brüssow, S., Lewis, R. L., & Drenhaus, H. (2008). Processing polarity: How the ungrammatical intrudes on the grammatical. Cognitive Science, 32(4), 685–712. [Google Scholar] [CrossRef]
  67. Xiang, M., Dillon, B., & Phillips, C. (2009). Illusory licensing effects across dependency types: ERP evidence. Brain and Language, 108(1), 40–55. [Google Scholar] [CrossRef]
  68. Yanilmaz, A., & Drury, J. E. (2018). Prospective NPI licensing and intrusion in Turkish. Language, Cognition and Neuroscience, 33(1), 111–138. [Google Scholar] [CrossRef]
  69. Zwarts, F. (1996). A hierarchy of negative expressions. In H. Wansing (Ed.), Negation: A notion in focus (pp. 169–194). Walter de Gruyter GmbH. [Google Scholar]
Languages 10 00119 g001
Figure 1. The essential information and collection steps of data.
Figure 1. The essential information and collection steps of data.
Languages 10 00119 g001
Languages 10 00119 g002
Figure 2. The data analysis flow chart.
Figure 2. The data analysis flow chart.
Languages 10 00119 g002
Languages 10 00119 g003
Figure 3. The annual publication of polarity based on bibliometrics.
Figure 3. The annual publication of polarity based on bibliometrics.
Languages 10 00119 g003
Languages 10 00119 g004
Figure 4. Average article citation per year from 1980 to 2023.
Figure 4. Average article citation per year from 1980 to 2023.
Languages 10 00119 g004
Languages 10 00119 g005
Figure 5. The co-authorship map of countries and regions. Gray: no data; blue: publishing countries (darker = more collaboration); red lines: collaboration links.
Figure 5. The co-authorship map of countries and regions. Gray: no data; blue: publishing countries (darker = more collaboration); red lines: collaboration links.
Languages 10 00119 g005
Languages 10 00119 g006
Figure 6. The literature co-citation network in polarity sensitivity research. Some references appear as “v0, p0” due to incomplete metadata. This does not affect the co-citation analysis, as the network is based on citation links, not publication details.
Figure 6. The literature co-citation network in polarity sensitivity research. Some references appear as “v0, p0” due to incomplete metadata. This does not affect the co-citation analysis, as the network is based on citation links, not publication details.
Languages 10 00119 g006
Languages 10 00119 g007
Figure 7. Keyword co-occurrence network for documents of polarity sensitivity research.
Figure 7. Keyword co-occurrence network for documents of polarity sensitivity research.
Languages 10 00119 g007
Languages 10 00119 g008
Figure 8. Keyword co-occurrence-based clustering of polarity sensitivity research2.
Figure 8. Keyword co-occurrence-based clustering of polarity sensitivity research2.
Languages 10 00119 g008
Languages 10 00119 g009
Figure 9. Timeline of the temporal development of the top 20 keyword-based clusters in polarity sensitivity research. Note: To enhance readability, only the most representative keywords s each cluster are displayed. Node sizes correspond to keyword frequency, and clusters are distinguished by colour. Red circles denote burst keywords identified through burstness analysis, marking emergent topics characterized by rapid increases in citation activity. The timeline reflects the temporal emergence and evolution of research themes related to polarity sensitivity across the period from 1980 to 2023.
Figure 9. Timeline of the temporal development of the top 20 keyword-based clusters in polarity sensitivity research. Note: To enhance readability, only the most representative keywords s each cluster are displayed. Node sizes correspond to keyword frequency, and clusters are distinguished by colour. Red circles denote burst keywords identified through burstness analysis, marking emergent topics characterized by rapid increases in citation activity. The timeline reflects the temporal emergence and evolution of research themes related to polarity sensitivity across the period from 1980 to 2023.
Languages 10 00119 g009
Table 1. The top 10 average citations of countries, i.e., the most productive countries and affiliations.
Table 1. The top 10 average citations of countries, i.e., the most productive countries and affiliations.
CountryAverage CitationRegionArticleAffiliationArticle
USA17.6USA298University of Chicago21
Germany9Germany155University of Groningen20
United Kingdom7.1United Kingdom99University of Amsterdam19
Spain10.2Netherlands96University of Connecticut18
Netherlands7.8France71Univ Autonoma Barcelona15
Canada12.1Japan65Hebrew Univ Jerusalem14
Italy8.7Spain61New York University14
China4.2China58Harvard University13
France4.3Israel43Macquarie University12
Israel7.2Australia42University of Tübingen12
Table 2. The top 10 journals in the number of publications related to polarity sensitivity.
Table 2. The top 10 journals in the number of publications related to polarity sensitivity.
SourcesArticlesh-Indexm-IndexPY Start
Lingua86100.2941990
Natural Language and Linguistic Theory5080.2881994
Natural Language Semantics4880.2861996
Journal of Semantics4240.1431996
Linguistics3730.071981
Linguistics And Philosophy3750.1351987
Glossa—A Journal of General Linguistics3140.6672018
Journal of Psycholinguistic Research2640.6672018
Journal of Pragmatics2240.1541998
Journal of Neurolinguistics1940.2672009
Table 3. The top 10 most productive authors and their local citations.
Table 3. The top 10 most productive authors and their local citations.
DisciplineAuthorArticleArticles FractionalisedDisciplineAuthorLocal Citation
Theoretical LinguisticsGiannakidou, A3124Theoretical LinguisticsGiannakidou, A125
Hoeksema, J2623.33Gajewski, Jr105
Gajewski, J1614.5De Swart, H56
Espinal, M145.75PsycholinguisticsChierchia, G102
Homer, V147.33Drenhaus, H90
Nishiguchi, S1414Phillips, C84
Collins, C128Frisch, S78
Crain, S249.26Saddy, D78
PsycholinguisticsZeijlstra, H2313Xiang, M64
Liu, M158.08Zeijlstra, H59
Table 4. Distribution of languages investigated in polarity item research (based on title and abstract indexing).
Table 4. Distribution of languages investigated in polarity item research (based on title and abstract indexing).
RegionLanguageCountRegionLanguageCount
Europe/GlobalEnglish197Northern EuropeSwedish5
Southern EuropeGreek36Finnish4
Spanish32Danish3
Italian32Norwegian2
Portuguese20East AsiaJapanese49
Catalan19Chinese48
Basque3Korean25
Galician1Western Asia/CaucasusTurkey4
Western EuropeFrench56Azerbaijani2
German37Armenian1
Eastern EuropeRomanian17South AsiaHindi11
Hungarian11Urdu4
Russian11Telugu1
Croatian5Southeast AsiaVietnamese2
Polish4Thai1
Bulgarian2Malay1
Czech2Middle EastArabic13
Slovenian2Hebrew12
Serbian1AfricaZulu2
OtherEstonian4Swahili1
Table 5. The top eight most globally cited documents.
Table 5. The top eight most globally cited documents.
PaperTotal CitationNormalised Total Citation
Xiang et al. (2009), Brain Lang.1029.7541
Vasishth et al. (2008), Cognitive Sci.8518.9402
Giannakidou (2009), Lingua807.65027
Ernst (2009), Nat. Lang. Linguist. Theo.636.02459
Gajewski (2011), Nat. Lang. Semant.4613.2692
Iatridou and Zeijlstra (2013), Linguist. Inq.459.60191
Holmberg (2013), Lingua.449.38854
Steinhauer et al. (2010), Neuropsychologia414.99045
Table 6. The top 10 co-occurrence times of keywords. Centrality (betweenness centrality) measures a keyword’s bridging role in the co-occurrence network. Values ≥ 0.1 indicate key connectors or thematic hubs.
Table 6. The top 10 co-occurrence times of keywords. Centrality (betweenness centrality) measures a keyword’s bridging role in the co-occurrence network. Values ≥ 0.1 indicate key connectors or thematic hubs.
KeywordsCountCentrality
negation920.26
polarity850.3
negative polarity item720.12
negative polarity710.11
semantics410.2
human390.25
polarity item360.19
language340.04
negative concord320.05
comprehension310.15
Table 7. The high-frequency co-words for the top 10 co-occurrence keywords clustering.
Table 7. The high-frequency co-words for the top 10 co-occurrence keywords clustering.
SizeSilhouetteMean Number of YearsTop Terms (LLR, p-Level)
480.7712008language (30.56, 0.0001); human (28.74, 0.0001); comprehension (26.62, 0.0001); child (22.19, 0.0001); reading (21.68, 0.0001)
470.9332007male (11.9, 0.001); female (11.9, 0.001); temporal lobe (10.23, 0.005); executive function (10.23, 0.005); middle aged (10.23, 0.005)
440.9272009polarity items (24.46, 0.0001); polarity item (19.74, 0.0001); negative polarity items (NPIs) (17.8, 0.0001); intervention effect (17.8, 0.0001); xml (11.85, 0.001)
440.942008polarity (58.68, 0.0001); negation (50.51, 0.0001); human (12.68, 0.001); gradable adjective (12.48, 0.001); verb movement (12.48, 0.001)
330.9562006article (10.12, 0.005); analysis of variance (9.97, 0.005); electroretinography (9.51, 0.005); optic nerve (9.51, 0.005); retina (9.51, 0.005)
270.9982007even (25.07, 0.0001); free choice (14.58, 0.001); sub triggering (14.1, 0.001); locality (14.1, 0.001); referential deficiency (14.1, 0.001)
260.9752011information structure (37.61, 0.0001); focus (30.18, 0.0001); topic comment (12.46, 0.001); polarity correction (12.46, 0.001); polarity contrast (12.46, 0.001)
250.9182009negative polarity (57.28, 0.0001); negative concord (51.27, 0.0001); double negation (11.45, 0.001); indefinite pronoun (10.52, 0.005); Russian (10.52, 0.005)
250.8942010knowledge (34.16, 0.0001); computational modelling (34.16, 0.0001); category learning (27.27, 0.0001); judgment (17.91, 0.0001); decision making (16.44, 0.0001)
240.9252010corpus analysis (31.32, 0.0001); enantiomers (15.57, 0.0001); lexical pragmatics (15.57, 0.0001); connective (15.57, 0.0001); alternative (11.79, 0.001)
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Kong, L.; Li, Y.; Sun, Y.; Jiang, Y.; Jiang, X. From Theoretical Framework to Empirical Investigation: A Bibliometric Analysis of Research Evolution and Emerging Trends in Polarity Sensitivity Studies Between 1980 and 2023. Languages 2025, 10, 119. https://doi.org/10.3390/languages10060119

AMA Style

Kong L, Li Y, Sun Y, Jiang Y, Jiang X. From Theoretical Framework to Empirical Investigation: A Bibliometric Analysis of Research Evolution and Emerging Trends in Polarity Sensitivity Studies Between 1980 and 2023. Languages. 2025; 10(6):119. https://doi.org/10.3390/languages10060119

Chicago/Turabian Style

Kong, Lingda, Yi Li, Yanting Sun, Yong Jiang, and Xiaoming Jiang. 2025. "From Theoretical Framework to Empirical Investigation: A Bibliometric Analysis of Research Evolution and Emerging Trends in Polarity Sensitivity Studies Between 1980 and 2023" Languages 10, no. 6: 119. https://doi.org/10.3390/languages10060119

APA Style

Kong, L., Li, Y., Sun, Y., Jiang, Y., & Jiang, X. (2025). From Theoretical Framework to Empirical Investigation: A Bibliometric Analysis of Research Evolution and Emerging Trends in Polarity Sensitivity Studies Between 1980 and 2023. Languages, 10(6), 119. https://doi.org/10.3390/languages10060119

Article Metrics

Back to TopTop