Introduction

The dissemination of Islam has profoundly influenced societies, cultures, and civilizations of human history since its advent, rendering early Islam an epoch with far-reaching historical significance. This transformative impact has spurred scholars and researchers to engage in extensive inquiries into Islamic literature over the last centuries (Hoyland 1997; Nigosian 2004; Hoyland 2012; Buskens and Van Sandwijk 2016; Salvatore et al. 2018). Notably, the term “early Islam” has emerged as a recurrent theme in scholarly discourse, garnering considerable attention and citations (Watt 1990; Berg 2017; Sahner 2023). In recent times, the academic fascination with this term has experienced a notable upsurge across a diverse range of disciplines (Johns 2003, Milwright 2010; Fuks et al. 2020; Srigyan et al. 2022; Fadel and Al-Hendy 2022). This article delves into the extensive body of research indexed by the Web of Science (WoS) to reveal consistent applications of the term “early Islam” in chronological categorization.

While the specific temporal boundaries of early Islam can vary depending on the topic under consideration, a prevailing trend in the current literature on this subject encompasses the first four centuries of Islam, spanning c.600-1000 AD. Several studies have delved into the periodization of Islamic history.Footnote 1 Despite widespread use of the term “early Islam”, however, there is a notable absence of comprehensive or bibliometric literature reviews and critical assessments of its historical evolution. Given the rising interest among researchers and institutions in early Islamic history, and the consequently expanding body of literature relating thereto, the present study is prompted by the need for a bibliometric analysis to address this scholarly void. By undertaking a meticulous examination of articles indexed in the Web of Science Core Collection database (maintained by Clarivate Analytics) between 2000 and 2022, this article offers a substantive contribution to existing knowledge on this topic. In this way, it aims to enhance understanding of the dynamics surrounding scholarship on early Islam.

Main purpose of this paper is to identify scholarly publications on early Islam and understand the broad intellectual structure of this body of research. It aims to offer a comprehensive bibliometric analysis of the term “early Islam”, as often used by modern scholars, but will neither determine the boundaries of the period defined as early Islam nor propose any periodization. Instead, the study will use this analysis to identify research fields (and gaps), leading authors, countries, fields of study, journals, and most cited articles centred on the term “early Islam” in the third millennium.

This study seeks to answer the research questions (RQs) found in Fig. 1. Each of these questions correlates to a respective subheading presented in the Results section:

Fig. 1: The relationship between research questions and subheadings in the Results section.
figure 1

This figure maps each research question to relevant analytical categories, illustrating the structure of the Results section.

Full size image

This study is composed of three parts, as follows: Section 1 (Methodological analysis) details the research methodology and data; Section 2 (Results) presents and discusses the findings from the collected data; and Section 3 (Conclusion) concludes the study by offering some recommendations for future research.

Methodological analysis

This study mined the Web of Science (WoS) Core Collection database for documents on which to find its bibliometric analyses. The Web of Science (WoS) platform, by aggregating international scientific publications, serves as an ideal data source for such analyses, enabling comprehensive evaluations that span diverse research domains (Birkle et al. 2020).

The documents for this study were selected in accordance with the research process shown in Fig. 1. Before data collection, the topic search term “early Islam*” was first identified, being the most preferred term in contemporary Islamic history research. This term was then searched for in the title fields, abstracts and keywords of WoS articles on 24 June 2023. To prevent data variations, only the results of the search performed on this date were taken into account. The search was conducted in this way because some researchers use the term “early Islam” synonymously with the terms “early Islamic” and “early Islamicate.”Footnote 2 The first search with the term “early Islam*” yielded 1642 results. During the second stage, these results were further refined by restricting the document type to take only articles, review articles and early access articles into account. This process removed 760 studies classified as books, book reviews, proceeding papers, and editorial materials. Thus, 882 results remained. After further refining the temporal range to 2000–2022, 806 results remained. We have two reasons for making such a limitation. Firstly, there is not enough data for analysis before 2000 (76 records). Secondly, these studies are mostly inaccurately recorded in the WoS database. This data was then extracted, and a cleaning of individual results began. Cleaning was carried out in two stages: first, results included due to technical errors were eliminated; then, results that were irrelevant to the content analysis were removed. The first stage involved separating duplicates from document types that were incorrectly registered in WoS, as well as articles that were first published as early access articles and then later assigned to a completed journal. The second stage involved identifying results that were irrelevant because they did not consciously use the term “early Islam” in their titles, abstracts, or keywords (i.e., they did not in fact involve the early Islamic period). Moreover, studies that used the term “early Islam” to mean Islam’s geographical advent rather than the broader period spanning c.600 to c.1000 AD were also excluded from analysis. These sorting and screening procedures left 764 results remaining. These results were then analysed as this study’s data set (for the detailed process, see Table 1 and Fig. 2).

Table 1 Retrieved results and data acquisition methods.
Full size table
Fig. 2: Flowchart of the research strategy.
figure 2

This flowchart outlines the stages of data acquisition, preparation, and analysis for the bibliometric study.

Full size image

The collection of 764 articles identified as the final dataset was saved as a marked list within the WoS. To facilitate subsequent scientific mapping using the VOSviewer software program (Van Eck and Waltman 2023), this dataset was then extracted in a tab-delimited format, capturing both the complete record of information as well as the references cited therein. VOSviewer is a freely available bibliometric visualization tool acknowledged for its user-friendly interface and was deliberately selected due to its capacity to handle substantial datasets. This program provides an array of analytical and exploratory features, generating visual representations that aid the evaluative examination of data patterns (Van Eck and Waltman 2010). However, it identifies distinct clusters within the dataset, encompassing various aspects, such as authors, institutions, countries, and keywords. These clusters were represented visually using different colours, while the size of individual data nodes within the figures denoted the frequency or prevalence of a specific item—whether it be prolific authors, journals, institutions, countries, or frequently co-occurring keywords. The examination focused on two primary variables: first, we delved into co-authorship connections across researchers, organizations, and countries. Secondly, we explored patterns of author-provided keywords. In this context, the nodes within the visualizations represented authors, institutions, countries, and keywords. In the context of data visualization, Microsoft Excel 365 was employed for creating graphs, maps, and pie charts, while Microsoft Visio was utilized for generating flowcharts and relational tables.

Results

Analysis of publication time trend

Figure 3 illustrates changes in the trend of annual publications on early Islam between 2000 and 2022. Although the general trend reflects a gradual rise in annual publications over the years, the increments of this increase have fluctuated rather than remained stable. Notably, of all 764 articles in the dataset, 348 (45.53%) were authored in the last five years. There are several potential reasons for this growth. The first is a remarkable increase in archaeological research (see Fig. 4). In addition, this notable rise could be the result of significant funding provided by supranational bodies, such as the European Research Council (ERC), and other national funders, like the Spanish Government and the Israel Science Foundation. Within this study’s timeframe, an especially noteworthy peak took place in 2021, when 93 articles were published—12.17% of the entire dataset over two decades. This growth in publications is anticipated to persist and to increase further.

Fig. 3: Number of dataset records on early Islam by year of publication.
figure 3

This bar chart illustrates the distribution of published articles on early Islam by year from 2000 to 2022.

Full size image
Fig. 4: Research areas addressing early Islam.
figure 4

Record count = n > 15.

Full size image

Research areas

The research fields with a minimum of 15 or more publications on early Islam are Archaeology (n = 239, 31.28%), Religion (n = 163, 21.33%), Asian Studies (n = 121, 15.83%), History (n = 84, 10.99%), Arts Humanities-Other Topics (n = 71, 9.29%), Geology (n = 64, 8.37%), Anthropology (n = 57, 7.46%), Area Studies (n = 24, 3.14%), Business Economics (n = 19, 2.48%), Art (n = 16, 2.09%), and Chemistry (n = 16, 2.09%).

The data indicates that studies on early Islam in the last quarter of a century or so have seen archaeological enquiry dominate publication output. This is clearly due to excavations that are currently being carried out in the geography of early Islam.Footnote 4 The findings from these excavations are published in the form of excavation reports, in which the relevant region’s relationship with early Islam is also mentioned. Archaeological studies are also observed to have increased in the recent years. This seems to be due to the impact of research funding for archaeological excavations.

Countries

Researchers based in 60 different countries published research on early Islam between 2000 and 2022. However, researchers in some countries have contributed more to existing research than others. In terms of numbers of publications and citations on “early Islam”, Fig. 5 shows the countries and regions in which researchers have been most active. Figure 6 shows collaborative relations between countries regarding the early Islam studies included in the dataset. The USA provided the largest number of contributors to the field, having the highest number of publications (with 166; 21.728%). This was followed by the UK (with 119), Israel (with 114), Germany (with 66) and France (with 51). The USA ranks first in terms of number of publication contributions, WoS citations, citations per article, and total link strength, highlighting that its academic cohort has been quite dominant in the early Islam field. Some 670 articles were published in the ten most active countries. Of these top ten countries, one is in North America, five in Europe, three in the Middle East, and one in Asia. Archaeological studies are one of the factors enabling Israel to rank third.

Fig. 5: Countries/regions with the most active publication output on early Islam.
figure 5

Record count = n ≥ 20.

Full size image
Fig. 6: Network visualization for international collaboration between countries regarding articles on early Islam.
figure 6

Record count = n ≥ 1.

Full size image

Academic affiliation

In terms of publication numbers and total citation counts, Table 2 presents the top five institutions that have published research on early Islam, including their countries of origin. An analysis of these institutions, along with influential figures in the field, aids understanding and staying abreast of the latest developments and most authoritative perspectives in this area. Israel stands out as the country with the most contributing institutions with 114 articles between 2000 and 2022. Table 2 highlights the University of Haifa as the affiliate with the most publications (n = 48) on early Islam. The Centre National de la Recherche Scientifique (CNRS) is in second place with 38 publications, and the Israel Antiquities Authority is in third place with 35 publications. Israel stands out in this dataset with three institutions. The data reveals that the top five affiliates contribute over a quarter of the total literature output on early Islam, thus signifying a distribution of research on this topic across multiple academic establishments, rather than such work being concentrated predominantly within a select few (for a network of authors’ affiliations, see Fig. 7).

Table 2 Top five author affiliated institutions.
Full size table
Fig. 7: Network visualization of authors’ affiliations.
figure 7

Record count = n ≥ 1.

Full size image

Research funding organizations

An analysis of the top ten institutions funding articles on early Islam between 2000 and 2022 offers valuable insight into the scholarly emphasis and support for research in this specific area (Table 3). The European Research Council (ERC) emerges as the foremost contributor, funding as many as 40 projects. This remarkable investment underscores the ERC’s commitment to advancing understandings of early Islam. The substantial number of ERC funded projects suggests a dedicated effort to unravel the complexities of the formative period of Islamic civilization. The Spanish Government’s allocation of resources to 28 projects reflects a considerable focus on research pertaining to early Islam, highlighting the importance of understanding the historical and cultural interactions between Islam and Spain, given the region’s rich history of Islamic influence. The funding supports studies that shed light on the impact that early Islam has had on Spanish society, architecture, language and intellectual heritage. Spain’s significant contribution underscores its dedication to unravelling the historical narratives and legacies of early Islamic civilization within its own context. The Israel Science Foundation’s support for 20 projects indicates a notable interest in interdisciplinary studies that examine the interactions between early Islam and other cultural and religious contexts. In summary, the top three funding institutions in the field of early Islam have collectively made significant contributions to advancing knowledge about this crucial period in Islamic history. Their investments underscore a multifaceted exploration of early Islamic civilization, spanning its impact on different regions, cultures, and intellectual traditions.

Table 3 Top ten funding organizations (record count ≥ 7).
Full size table

Authorship and co-authorship analysis

In total, 1185 different authors have written articles on early Islam. As shown in Table 4, N. Schibille and I. Taxel have been by far the most active contributors (n = 13; 1.702%). G. Bar-Oz and Y. Tepper are joint second (n = 12; 1.571%), and I. C. Freestone is third (n = 11; 1.440%). In addition, Freestone (492 citations), Henderson (291 citations) and Molera (214 citations) appear to be among the most influential authors on this topic in terms of citations. The authors with the highest number of articles on early Islam in the study’s dataset specialize in archaeology and archaeometry. Figure 8 represents co-authorships on early Islam of authors with at least one publication and one citation. It depicts co-authorship links between these authors over time.

Table 4 Top ten authors.
Full size table
Fig. 8: Co-authorship networks.
figure 8

Record count = n ≥ 1.

Full size image

Influential journals

The retrieved documents (n = 764) were published in a total of 310 journals from different fields of study. Table 5 shows the top ten journals (record count = n > 10) identified as sources of literature in this field from 2000 to 2022. These journals published a total of 182 documents, accounting for 23.79% of the entire literature. Each had at least ten publications on the topic, with the top positions going to (Der) Islam-Zeitschrift für Geschichte und Kultur des Islamischen Orients with 26 articles (3.4%), the Journal of Archaeological Science Reports with 26 articles (3.4%), and the Journal of Archaeological Science with 24 articles (3.14%). Half of the active journals in Table 5 were based in the UK. In addition, half were journals of archaeology. Among these, the Journal of Archaeological Science had the highest impact factor (5-year JIF = 3.0), followed by Archaeological and Anthropological Sciences (2.2) and Antiquity (2.0).

Table 5 Top ten journals for early Islam research.
Full size table

Language of publications

The diagram presented in Fig. 9 offers an overview of the linguistic distribution of published material concerning early Islam. In total, the analysed publications were written in 12 distinct languages. The data clearly indicates English to be the overwhelmingly dominant publication language for academic output on early Islam, constituting 683 (89.39%) of all papers. After this, Turkish occupies second position, contributing 23 (3.01%) of the documented works. This is followed by German (16, 2.09%); French (9, 1.17%), Russian (9, 1.17%) and Spanish (9, 1.17%); Arabic (6, 0.78%), Malay (3, 0.39%), Portuguese (3, 0.39%), Italian (2, 0.26%), and Croatian (1, 0.13%) and Slovenian (1, 0.13%). Thus, while scholarly articles have emerged from numerous countries worldwide, English unmistakably functions as the prevailing language for academic dissemination.

Fig. 9: Language of publication.
figure 9

Record count = n ≥ 1.

Full size image

Citation analysis

Citation analysis of the dataset was conducted using VOSviewer. A citation analysis for authors was performed first, and followed by a citation analysis of the publications themselves to identify the most cited articles. Finally, a citation analysis was performed based on the journals in which these publications were published.

Figure 10 shows the relationship networks between authors with at least one publication and at least one citation. Of 764 publications, 525 were cited at least once, while 239 were not cited at all. The ten most cited articles and their authors are outlined in Table 6. Here, papers by I. C. Freestone et al. (2003), J. Henderson et al. (2004), and Y. Avni et al. (2006) have 100 or more citations. Figure 11 represents the citation networks between articles with at least five citations, highlighting the mutual citation relationships between authors. The citation relationships between journals with a minimum of 1 citation is shown in Fig. 12. Journals in the same or similar fields naturally appear to have a citation relationship with one another.

Fig. 10: Citation network visualization between authors with at least one publication and one citation.
figure 10

Record count = n ≥ 1; citation count = n ≥ 1.

Full size image
Table 6 Top ten most cited articles on early Islam (Total number of citations = n > 55).
Full size table
Fig. 11: Network visualization of a citation between articles with at least five citations.
figure 11

Citation count = n ≥ 5.

Full size image
Fig. 12: Network visualization of citation relationships between journals with at least one citation.
figure 12

Citation count = n ≥ 1.

Full size image

Keyword analysis

Analysing a particular scientific discipline, keywords assume significant importance by encapsulating the core content of articles. Examining patterns and trends in keyword compilations can shed light on specific academic research endeavours. Specific visualization techniques, like network maps and heat maps, can help to deepen understanding of the academic landscape on early Islam. The proportional frequency and link strength of terminology used in works on early Islam is depicted in Figs. 13 and 14. Islam, early Islam, early Islamic period, and Qur’ān are the most frequently used keywords. A closer analysis of these keywords sheds light on research trends and possible future research topics in the field. To this end, a keyword co-occurrence analysis was conducted using VOSviewer. As illustrated in Fig. 15, by setting a minimum occurrence threshold of 5 mentions per word, a total of 58 keywords emerged and were grouped into seven clusters. Figure 15 displays the distribution of these keywords, highlighting the prominence of certain terms such as “Islam” with 49 occurrences, followed by “early Islam” with 31 occurrences, “early Islamic period” with 30 occurrences, and “early Islamic” with 24 occurrences.

Fig. 13: Network visualization of co-occurrence of keywords.
figure 13

“minimum occurrence frequency = n ≥ 2”.

Full size image
Fig. 14: Heat map of keywords.
figure 14

“minimum frequency of occurrence = n ≥ 2”.

Full size image
Fig. 15: A more detailed network visualization of the co-occurrence of keywords.
figure 15

“minimum frequency of occurrence = n ≥ 5”.

Full size image

Cluster 1 is represented in red in the visualization and identifies one cluster of concepts with frequent inter-related occurrence. It has 13 keywords that stand out: Islam, early Islamic history, historiography, history, hadith, Muhammad, Palestine, Sufism, archaeology, and some of the related regions. Cluster 2 is green and includes topics related to early Islam, the Qur’ān, Late Antiquity, Zoroastrianism, asceticism, conversion, law, monasticism, Arabic, Pahlavi literature, and the Persian Gulf. Cluster 3 is blue and contains keywords more related to archaeology. These are early Islamic period, Islamic glass, Byzantine glass, Byzantine period, Jerash, Jordan, IA-ICP-MS, pottery, and Sasanian. Cluster 4 is yellow represents a set of keywords where the sub-disciplines of archaeology stand out. Here, related keywords are early Islamic, archaeobotany, zooarchaeology, Islamic archaeology, agriculture, Abbasid, trade, Jerusalem, and Israel. Cluster 5 is purple and is another archaeology-focused grouping more clearly dominated by Islamic and Byzantine glass. Here, prominent keywords here are Islamic, Byzantine, Roman, Umayyad, shipwreck, glass, mosque, and technology. Group 6 is light blue and clustered with the keywords of Islamic history, Prophet Muhammad, Iran, Central Asia. Cluster 7 is orange and contains such topics as the Qur’ān, religion, and ethics.

Conclusion

This study highlights several critical outcomes based on the bibliometric analysis of publications related to early Islam from the WoS database. It reveals the growing prominence of archaeology in early Islamic studies, reflecting a shift toward interdisciplinary approaches that integrate archaeological findings with traditional historical and religious scholarship. Another key outcome is the broad adoption of the term “early Islam” which enables a more inclusive examination of the period and facilitates the exploration of diverse aspects and socio-political contexts beyond traditional periodizations.

This study has several limitations. First, data prior to 2000 is excluded due to higher error rates and fewer records, which affect the accuracy and reliability of the analysis. The focus on data from 2000–2022 ensures greater precision and robustness. Second, the WoS database limits the study, as it does not include articles from local journals in Arabic, Turkish, Persian, Malay, and Urdu. While the WoS database is comprehensive, including the Arabic Citation Index (ARCI) in the future may address this gap and improve bibliometric analyses of early Islamic studies.

Finally, another bibliometric analysis could be conducted on leading journals from the dataset, which publish research on early Islam. Such analysis could assess the impact and status of these journals in the academic community and provide more precise insights into the field’s dynamics and trends.