The presence of theory is an indication of research eminence and respectability (Van Maanen, 1998), as well as a feature of the discipline’s maturity (Brookes, 1980; Hauser, 1988). The development of theory is the central goal; the ‘jewel in the crown’ of research (Eisenhardt, 1989). Theory plays a vital role in research (Ngulube, 2020; Thomas, 1997). A good theory continually advances knowledge, directs researchers to essential questions, and provides knowledge and understanding about a research topic and the discipline (Sonnenwald, 2016; van de Ven, 1989). Moreover, in any research, the effective and practical application of theory has always been critical to developing new knowledge or interpretations of existing knowledge. Research without the use of theories is poor and lacks a sound foundation and limited usefulness to the particular domain (Sarter, 2006). Theories are essential in research, and the importance of theories in research cannot be denied (Doherty, 2012; Gregor, 2006; Hall, 2003; Hider & Pymm, 2008; Jeong & Kim, 2005; Lee et al., 2004; Neuman, 2000; Ngulube, 2018; Van de Ven, 1989).

Research in LIS during the 1980s and 1990s produced several foundational theories in this domain. (Chatman, 1999; Cole, 2011; Dervin, 1998; Kuhlthau, 1991). For example, Dervin’s (1983) sense-making theory, Mellon’s (1986) library anxiety theory, Ellis’s (1987) model of information-seeking behaviour (ISB), Bates’ (1989) berry-picking theory, Kuhlthau’s (1991) information search process (ISP), Chatman’s (1996) theory of information poverty, and others. As a result, the use and application of theory in LIS research increased during the 1990s and early 2000s (Kim & Jeong, 2006; McKechnie et al., 2001; Pettigrew & McKechnie, 2001). In the LIS domain, theories are used to guide the analysis, explanation, and prediction of phenomena and to provide design and action guidelines (Gregor, 2006). It tells us why they are correlated (Sutton & Staw, 1995). It is necessary, in research writing, to explain relationships among concepts. Hence, it is an essential and crucial task for LIS professionals to form a clear picture of the application and use of the theories in LIS research.

Nevertheless, there is a lack of theoretical research in LIS (Kim & Jeong, 2006), and there is little discussion in different platforms and forums of what theory means in LIS. Indeed, the need to develop, teach, and apply theory in LIS research remains acute (Buckland, 2003; Hjørland, 2000; Thompson, 2009). Although a few LIS researchers and practitioners have created many useful conceptual frameworks, models, and theories (Fisher et al., 2005), they are restricted in their use in LIS. For example, Hartel (2009) and Bates (2006) reported on the development of meta-theories in this domain of LIS. In a more in-depth analysis of theory use in LIS, Kumasi et al. (2013) qualitatively analysed the extent to which theory is meaningfully used in the scholarly literature by developing a theory talk coding scheme, that included six analytical categories, describing how theory is discussed in a study. The intensity of theory talk in the articles was described across a continuum from minimal (e.g., a theory is discussed in the literature review and not mentioned later) through moderate (e.g., multiple theories are introduced but without discussing their relevance to the study) to major (e.g., a theory is employed throughout the study). Finally, they concluded that “LIS discipline has been focused on the application of specific theoretical frameworks rather than the generation of new theories”.

In the same vein, Pettigrew and McKechnie (2001) reported the limited application of theory and the failure of LIS research to address the practical problems of the profession. The same observation was made by Ukwoma and Ngulube (2021), who reported that much research works, such as theses and dissertations in LIS, needed more theory. Some studies used concepts like theory, theoretical framework, and conceptual framework interchangeably. This may be due to the fact that LIS researchers are not aware of the role of theory, including different components of a theory in the research process (Dankasa, 2015; Kim & Jeong, 2006; Pettigrew & McKechnie, 2001), or they have a lack of knowledge about the utility of theory in LIS research or have misconceptions about the theory and theoretical contribution in the LIS field (Ngulube, 2018; Ocholla & Roux, 2011).

LIS theories are usually vague and conceptually unclear, and basic concepts must be defined in the literature (Jarvelin & Vakkari, 1990; Poole, 1985; Schrader, 1986) and research in LIS has been dominated by a paradigm that “has made little use of such traditional scientific approaches as foundations and conceptual analysis, or of scientific explanation and theory formulation” (Jarvelin & Vakkari, 1990). Our field, LIS, depends on other disciplines for theoretical work (Ferraro et al., 2009; Kumasi et al., 2013; Weber, 2003), and LIS researchers borrow theories mainly from several branches of social science such as psychology, sociology, or management (Hjørland, 1998; Morgan & Wildemuth, 2009; Poole, 1985). This is due to lack of qualitative scholarly literature on theory use in LIS (Kumasi et al., 2013). On the contrary, Ferraro et al. (2009) reported that theory borrowing like other disciplines is the tradition of information science. And, importing theories from other disciplines is not an indication of weakness in the discipline (Truex et al., 2006). Actually, researchers borrow a concept or a theory from other disciplines out of its original context to explain the same or a different phenomenon (Murray & Evers, 1989) for the purpose of investigating the questions of the form, structure, and organization of information and the social impacts of information technologies (Bates, 1999).

Many researchers (Hall, 2003; Oswick et al., 2011) argued that theory borrowing is a common practice prevalent in academic research and is an inevitable and integral part of theory development in all disciplines (Kenworthy & Verbeke, 2015; Truex et al., 2006; Whetten et al., 2009), especially those with a broad disciplinary base, including LIS (Mutula & Majinge, 2017; Ukwoma & Ngulube, 2021) as a result of a shortage of applicable home-based or native theories and core theories. Scholars borrow, adapt, extend and at times generate new theories when conducting research (Doherty, 2012). Theories are adopted and adapted or some variables are excluded to suit the LIS context. LIS has an interdisciplinary foundation that results in its borrowing from other disciplines. LIS domesticate theories mainly from sociology, management, psychology and information systems (Ukwoma & Ngulube, 2021). Incorporating theories from another discipline strengthens and enhances the linkages or connections between the two disciplines (Stock, 1997). So, there is a continuous call for developing “good theories” (Watson, 2001), especially home-based (here LIS) theories, e.g., theories of our “own” (Lee et al., 2004; Neuman, 2000; Weber, 2003). The paper reports the extent of theory use in LIS research and how and where researchers use theories in their papers, whether in their original format or otherwise modified.

The paper is structured as follows: Section 2 discusses the meaning and role of theories in research. Section 3 describes similar works related to theories, including the importance of theorising in LIS research. Section 4 elaborately discusses the scope and limitations of the study. Section 5 focuses on the research questions formulated for this study. Section 6 deals with the methodology used to fulfill the research questions. Section 7 shows the results from different perspectives. Section 8 provides an overall assessment of the study and proposes several critical directions, and Section 9 concludes the paper with avenues for future exploration.

The term “theory” derives from the Greek “theoria” and, in modern usage, from the Latin “teoria,” which means “a looking at, viewing, contemplation, or speculation”. It may have derived from the philosophy of science, particularly from Kuhn’s account of theory change and scientific revolutions and has a complicated origin story with roots in several philosophical and psychological doctrines (https://iep.utm.edu/theory-theory-of-concepts/).

Still, there is no agreed-upon definition of “theory”. Different experts have defined theories from different perspectives (Babbie, 1992; Merton, 1957; Odi, 1982; Schwandt, 1997; Vogt, 1993), and there is a considerable level of disagreement among experts about what constitutes a theory (Sutton & Staw, 1995). Sutton and Staw (1995) further stated that a consensus on exactly what theory is and why it is so challenging to develop a robust theory had not been reached.

Not everything is a theory, and it is quite difficult to judge what is actually a theory irrespective of disciplines as theory work may take a variety of forms. The term “theory” is almost a common phenomenon and is applied to all types of research both quantitative and qualitative, and to mixed methods irrespective of disciplines or domains. Experts have already admitted the essence of using theory in research and learning. It is used for giving meaning to abstractions and/or concepts to explain and aid understanding of phenomena by making generalisations about proven facts (Chijioke et al., 2021). Bhattacherjee (2012) clarifies that theory explains and predicts through building correlations and causations-cause-effect) relationships, respectively. It helps us understand things, events, activities, behaviours, and/or situations (Scott et al., 2008). In general, theory answers a human need to make sense of the world and to accumulate a body of knowledge that will aid in understanding, explaining, and predicting the things we see around us, as well as providing a basis for action in the real world (Gregor, 2002). Babbie (2013) defined it as an organised explanation for the purpose of making observations regarding a specific aspect of life. Corley and Gioia (2011) concluded that theory shows relationships among different concepts that show how and why a phenomenon occurs.

So, it is difficult for the authors to give a solid definition of theory in the context of LIS research and the role of theory in the domain of LIS has been a subject for debate for several decades (Pinfield et al., 2021). Here, the authors have tried to focus on theory by citing experts’ views. For example, Boyce and Kraft (1985) regarded theory as ‘a body of principles: fundamental laws or empirical regularities. Sugimoto (2016) defined it as “a set of statements, systems, or principles, used to describe or explain phenomena”. Gregor (2002) opined that theory should be considered from many different dimensions and it could be classified in such a way that would fulfill its purposes. In the same line, Neuman (2000) proposed five factors of a theory, such as (i) the direction (deductive or inductive), (ii) the level of the theory, (iii) whether it is formal or substantive, (iv) the forms of explanations it employs, and (v) the overall framework of assumptions and concepts in which it is embedded, by which it could be classified. On the other hand, Wacker (2004) identified four key properties that constitute a good theory: “formal conceptual definitions, theory domain, explained relationships, and predictions”. He also considered theory as a link that creates relationships between concepts. Whereas Garver (2008) suggested that theories vary in their specifications. Another researcher (Buckland, 1991) said that a “good” theory is one that matches well our perception of whatever the theory is about. The closer the match, the better the theory is. Van Maanen (1998) suggested that theory must be convenient and should help and support to organise and communicate unwieldy data and simplify the terrible complexities of the social world, matters that may well be more important to the field than whether or not a given theory is true of false. Buckland (1991) defined theory “in the broad sense of a description or explanation of the nature of things, not in the more restricted sense, used in some sciences, of denoting fundamental laws formally stated and falsifiable.” Some other researchers have given the basic definitions of theory precisely such as Smit (1995) ‘a set of principles that is used to explain a certain phenomenon or phenomena; Silverman (2006) defines ‘a set of explanatory concepts’; Vogt (1993) defines ‘ a statement or group of statements about how some part of the world works—frequently explaining relations among phenomena’; Odi (1982) defines ‘an internally connected and logically consistent proposition about relationships among phenomena’; Welman and Kruger (1999) define ‘a group of logical, related statements, which is presented as an explanation of a phenomenon’; Babbie (1992) defines ‘a systematic explanation for the observed facts and laws that relate to a particular aspect of life’; Kaplan (1964) defines ‘a way of making sense of distributing situation’; Schwandt (1997) defines ‘a unified, systematic explanation of a diverse range of social phenomena’; Grover and Glazier (1986) define ‘generalisations which seek to explain relationships among phenomena’; and ‘a set of statements about the relationship(s) between two or more concepts or constructs’ (De Vos et al., 2005; Jaccard & Jacoby, 2010; Swanson, 2013). Sutton and Staw (1995) rightly said that the lack of a unified definition among scholars of what a theory is has often made it difficult to develop a strong theory for any discipline. But they all agreed that theory develops as an explanation to advance knowledge in the particular field (Thomas, 1997). Most of these definitions of ‘theory’ mention the relationships between or amongst several variables.

According to Babbie (1995), social science theory is ‘a systematic explanation for the observed facts and laws that relate to a specific aspect of life’. The role of theory in the social sciences is, among other things, to explain and predict behaviour, be usable in practical applications, and guide research (Glaser & Strauss, 1999). Some other experts have tried to give a formal definition of theory in the context of information science. Lee et al. (1997) discussed theory in this context in terms of underlying causal relationships, but primarily from a statistical viewpoint. Gregor (2002) identified and distinguished five different types of theory important for the discipline of LIS: (i) theory for analysing and describing, (ii) theory for understanding, (iii) theory for predicting, (iv) theory for explaining and predicting, and (v) theory for design and action. In the same study, she reported that many LIS researchers had failed to give any explicit definition of their own view of theory. Walster (1995) examined five instructional design theories valuable to LIS education and also described the basic components of the theories and their application in this domain.

Based on the discussion, determining the scope of the theory and suggesting a comprehensive definition of the theory is quite difficult. These vary depending on the type of research, academic field, and researcher. Hjørland (2013) correctly opined that the situation is somewhat chaotic and that it is difficult to get a clear overview of the theoretical landscape of the field as a whole. So, a theory, built from concepts, variables, or phenomena, is a mental activity and an interrelated set of constructs that seeks to explain an object or things, explains observed regularities or relationships between two or more variables, and shows how and why events occur.

The use and application of theory are common in any academic discipline, and LIS is no exception. Apart from its own theory, many theories from other disciplines have also been used in the LIS field for the development of research productivity in this domain. Many LIS researchers, particularly in the field of information science, have developed theories about information-seeking behaviour. Initially, a theory was developed for a particular discipline and has since been modified and utilised in other disciplines or for any set of phenomena. For example, management theory is now being taught in different library schools (Trosow, 2000).

The LIS literature, according to Feehan et al. (1987), has not evolved sufficiently to support a rigid body of its own theoretical basis. Chatman (1996) is indeed correct when she claims that using and developing a theory is hard work in LIS. Vakkari and Kuokkanen (1997) attempted to analyse theory development in LIS using a case study from information seeking studies. Hjørland (2000) reported that LIS lacks good theories because there are no explicit theories in LIS. Many of the theories used in LIS are from other fields such as psychology, sociology, or management (Dillon, 2007). In the same vein, Ocholla and Roux (2011) opined that LIS largely relies on theories from other disciplines. They also presented a theoretical framework model used in LIS research and clustered it by research themes. Pierce (1992), citing the work conducted by Pettigrew and McKechnie (2001), also reported that LIS researchers tend to borrow theories from other disciplines. Furthermore, most LIS researchers borrowed theories from the social sciences (Oswick et al., 2011). Onyancha and Kwanya (2019) were in support. Again, Pettigrew and McKechnie (2001) found that 45.4% of theories used in LIS came from the social sciences, followed by LIS (29.9%), sciences (19.3%), and humanities (5.4%). Besides, more than 70% of the theories applied in Chinese LIS journals (Wang et al., 2016) and 57.5% of the theories used in Korean LIS journals (Kim & Jeong, 2006) were borrowed from other disciplines.

There have been many efforts by researchers to analyse the state of theoretical research in LIS. Here, authors have tried to give an overview of theories used in the LIS domain. Many authors have advocated for the development and application of theory in LIS (Buckland, 2003; Hjørland, 2000; Thompson, 2009). Many authors have developed many conceptual frameworks, models, and theories (Fisher et al., 2005; Pettigrew & McKechnie, 2001). dos Santos Maculan and de Oliveira Lima (2017) reported two theories, viz., Dahlberg’s analytical concept theory and Ranganathan’s faceted classification theory on concepts that are commonly taken for granted and discussed in the LIS literature. Some other experts reported the use of different theories in several branches of LIS research. For example, Mellon (1986), Mansourian (2006) and Ellis (1993) used grounded theory; Aluri (1981) applied learning theory; Middleton et al. (2019) used social cognitive theory; Bossaler et al. (2010) applied critical theory; Rogers (1995) used diffusion of innovation theory; Fishbein and Ajzen (1975) used the theory of reasonable action; Ajzen (1991) used the theory of planned behaviour in different areas of the LIS domain. Oliveira Machado et al. (2019) discussed concept theory in LIS from an epistemological perspective. Benoit (2007), on the other hand, provided an overview of major critical theories from a variety of disciplines, including the humanities, social sciences, and education. Michell and Dewdney (1998) provided a brief explanation of another social science theory i.e., the mental models theory, and the use of it in LIS research. There are many meta-theories operating in the field currently. Vickery (1997) discussed the meta-theory of information science research. Grover and Glazier (1986) proposed a model for theory building in LIS called “circuits of theory”. The model includes a taxonomy of theories, developed earlier by the authors. The purpose of the taxonomy was to demonstrate the relationships among the concepts of research, theory, paradigms, and phenomena. Rioux (2010) explored the use of meta-theory as an integrative conceptual tool that can help analyse, direct, and enhance theory building, professional practice, and professional preparation in LIS. Kaijun et al. (2019) discussed fractal theory in information science (Parsa et al., 2016) including other domains such as mechanical science (Rinaldo et al., 1993), astronomical meteorology (Fossum et al., 2013), and life sciences (Puetz & Borchardt, 2015).

In reality, little research has been conducted to investigate the use of theory in LIS, and thus has been often criticised as being fragmentary, narrowly focused, and oriented to practical problems (Grover & Glazier, 1986). Many authors have noticed limited use of theory in published research and have advocated greater use of theory as a conceptual basis in LIS research (Boyce & Kraft, 1985; Feehan et al., 1987; Grover & Glazier, 1986; Hjørland, 1998; Spink, 1997). Some quantitative studies have been conducted on the theory use in LIS. A number of studies (Feehan et al., 1987; Jarvelin & Vakkari, 1990; Julien, 1996; Julien & Duggan, 2000; Nour, 1985; Peritz, 1980) concluded that most LIS research is atheoretical, with the rate of theory use in LIS ranging from 10 to 21 percent. For example, Peritz (1980) reported that only 14% of sample articles from 1950 to 1975 could be considered theoretical research. Whereas Nour (1985) reported 21.2%, Feehan et al. (1987) reported 13%, and Järvelin and Vakkari (1990) reported 10% of the literature published in 1980 using theory. Julien (1996) found that theoretical studies occupied 32% of the 241 randomly selected papers on information needs and use from 1990 to 1994. Gonzalez-Teruel and Abad-Garcia (2007) found that theories were used in only 14% of the papers on information needs published in Spanish journals from 1990 to 2004. Even so, variations in the use of theory may also be regional. Wang et al. (2015) found that at least one theory was mentioned in the full-texts of 30.2% of papers from the Journal of the China Society for Scientific and Technical Information (JCSSTI) from 2000 to 2013, which is taken as the top one journal of information science by Chinese universities. Then, Wang et al. (2018) further analysed the LIS papers published in all 52 Chinese LIS journals from 2008 to 2017 and found that 18.97% of them mentioned at least one theory in their abstracts. Wu et al. (2017) found that 49.9% of articles published in Taiwanese LIS journals between 2010 and 2015 use theory. In Korean LIS research journals, Jeong and Kim (2005) found only 10% of studies applied a theory. The authors observe that these variations in the use of theory among different LIS journals are due to the number of journals as well as the journal selection process used by the researchers.

According to another study (Mckechnie & Pettigrew, 2002), a theory was discussed in 34.2 percent of the articles. The study covered almost 1,160 articles published in six prominent LIS journals from 1993 to 1998. The same type of work has been conducted by Järvelin and Vakkari (1990). Authors reported 10% use of theories in LIS, whereas it was 18.3% in another work (Julien & Duggan, 2000). Kim and Jeong (2006) reported that the use of theory was only 17.57%. Nour (1985) reported that while 21.2% of articles used theory in an analytical sense, less than 3% of the articles were about information science theory. Feehan et al. (1987) reported that only 13% of the 123 research articles sampled from 91 LIS journals either discussed or applied theory in the study design or attempted to formulate theories or principles that could provide a theoretical basis for LIS.

As previously stated, (see Section 2), the term “theory” refers to ideas, works, opinions, models, hypotheses, and so on (see also Section 6). The sample the authors have analysed does not claim to be exhaustive, as the search operators or the search syntax (part of our search strategy) were limited to those articles that contained the term ‘theory’ or its equivalent terms, as stated above, in the title and in the abstract, along with the other two terms, viz. ‘library’ and ‘librarian’. The authors face specific difficulties where researchers have used such parallel terms in place of the ‘theory’ or a different name but not the word ‘theory’. The authors have considered articles where these terms were not present/used or were used in other ways, but the creator of the theory was mentioned, or researchers did mention the name of the theory. The keywords provided by the researchers were not used as such words were manually added and might not capture all the topics being discussed in the papers. The authors have identified many retrieved papers that do not directly relate to the development of the LIS theory. Moreover, these papers were rejected from our analysis for not covering any potential theoretical work promoting LIS research. Another problem relates to suggesting the definition of the ‘theory’ and determining the scope of the ‘theory’. The authors do not claim that the journals covered by Scopus or the articles the authors have studied exclusive representatives of LIS research. Our classification of LIS sub-fields (Section 7.4) and grouping of all theories based on originating domain (Section 7.7) needs to be more foolproof.

In many cases, extended abstracts were not provided by the researchers, and authors could not identify the areas where theories were used in the article, i.e., the introduction, method, results/discussion, or at what level or extent researchers had used the theory with fidelity. The authors could not identify the type of research for which theories were tested or employed. Moreover, determining the quality of the theories used differs from our study’s objective.

Summarising the above, we have set the following research questions to fulfill our objectives for this study:

What were the theories applied in LIS research? Or which were the most important theories discussed in the corpus data from 1970–2021?

Traditional quantitative content analysis, with natural language processing (NLP) and text mining technology, is taken as the research method. Data were extracted only from LIS journals covered by Scopus (https://www.scopus.com/) against carefully crafted search queries: library, librarian, theory, and synonymous/equivalent terms related to ‘theory’. For example, sometimes, keywords such as ‘framework’, ‘model’, ‘pattern’, ‘paradigm’ ‘method’ were used interchangeably by scholars in place of ‘theory’ and thus were used and considered to search for information. The authors have selected the Scopus database as the primary data source and limited the search to only LIS research articles published from 1971 to December 2021. A total of 14,294 raw articles (from 1971 to 2021) were downloaded or extracted using various search terms like ‘library’, ‘librarian’, ‘theory’, and parallel/synonymous terms of ‘theory’, all of which appeared in the title and the abstract of the paper. The authors have excluded 368 titles or papers for not having an abstract. A total of 61 articles were also removed from three unrelated journals. Only full-length articles written in English were collected where the specific application of theories was made. Editorials, book reviews, letters, interviews, commentaries, non-research articles, and news items were also excluded from the analysis.

A total of 13,865 articles were considered for this paper, and all the relevant bibliographic data, such as titles, abstracts, authors, names of the journals, etc., were recorded in a single Excel file. The sample data were then ranked and curated using ASReview (https://asreview.nl/) to examine whether the articles were relevant to the development of LIS theory. It is used to rank a set of selected abstracts based on their context and relevance. Finally, 13,225 (95.4%) articles were removed for not having a direct link with the LIS theory. The sample size for this study was 640 (4.6%) research articles directly related to the development of LIS theory. In the next step, OpenRefine (https://openrefine.org/), an open-source data rackling tool, is used to obtain the results as reflected in Table 1. The CSV file of sample data is transformed into ‘OpenRefine’ for further processing, and Table 1 reflects the dataset generated by ‘OpenRefine’ after analysing the corpus data. This tool helps quantify and deep faceting the text corpora to identify categories of the theories used in LIS research.

The authors have taken only those papers where theories of LIS or theories of other disciplines were applied. The ‘title’ and the ‘abstract’ and ‘index terms’ of the papers were used as chief sources of information. All the articles were checked and validated by domain experts other than the authors to improve the sample data’s validity and reliability. The authors have classified all papers using a descriptive framework that considers the level of theories used and the stages at which theories are used as it reflects the intensiveness of the use of theories within the studies. Again, the authors have critically reviewed and analysed each theory against selected criteria and consulted different documentary sources to determine its originating discipline. Furthermore, for this purpose, the paper’s focus and the background information of the creator of the particular theory are considered.

This section presents our results, organised by analysis of research questions as stated in Section 5. Keeping in mind the study’s objectives, the following aspects of importation are examined considering the stated parameters (Table 1).

The essence of using theory from other domains for conducting LIS research has been felt since the beginning of the 21st century (Hall, 2003; Kenworthy & Verbeke, 2015; Oswick et al., 2011; Ukwoma & Ngulube, 2021). Moreover, establishing a link between theory and research has become a hot topic among LIS researchers (Mueller & Urbach, 2013). However, the development of theory or the range of theories used in LIS research has expanded over the past forty or fifty years (Sonnenwald, 2016), and the use and application of socio-cultural theories to uncover or underpin LIS research continue to increase. Furthermore, this is reflected in our study (Section 7.1), where the rate of use of theories in LIS research was 64.21%.

As discussed in Sections 4 and 6, there were misconceptions, a lack of awareness regarding theoretical discussions, and misuse of the term ‘theory’ among researchers (Ngulube, 2018). Only 640 (4.83%) core articles were finally selected for this study, which indicates that researchers have used unintentionally theories without having a clear idea of theory regarding their application and utilisation in research. This conceptual misunderstanding regarding the ‘theory’ among scholars may misdirect the focus of the research and affect the results. Bibliometric laws, classification rules, and cataloguing rules were considered theories in many cases. The theories included -

Researchers disagree and hold opposing views on whether these should be considered theories (Ngulube, 2020; Ocholla & Roux, 2011). In a few cases, two or more theories were used or applied in the same paper. However, there was no such application of said theories, and it was difficult to measure at what level or for which purpose a theory was used in the study. As a result, the relevance of the theory to the study could not be identified and was unclear as it was not used consistently throughout the study.

In many cases, the purpose of using a theory in the paper was unclear and misleading. So, researchers should have mentioned their roles in the article or the extent to which they employed the theory with fidelity. Even so, some articles used multiple or one theory in multiple ways. It is unclear how a particular theory was used in the work. In most cases, the researchers did not mention the origin of the single theory (e.g., the discipline of origin). Additionally, the researchers should have mentioned the name of the particular theory with further explanation including the theory’s originator (Section 7, Table 1). So, these issues must be reworked and clarified adequately in the text.

Another problem was identifying the original subject or discipline from which a theory had come (Section 7.7). The authors found that the same theory was simultaneously considered as ‘science’ and ‘social science’ theory by the experts and was kept in both places. The same theory may be applied to science and social science or management studies depending on the context of the research work with any necessary modifications or extensions to the original theory. It is important to remember that the authors did not find any best practices or formal guidelines to identify the emerging discipline of the theories, which appeared to belong to more than one discipline. The authors have noticed that, in many cases, some theories originating from multiple disciplines had no option but to fit into a particular discipline considering the focus of the paper and the background information of the originator. The situation mostly happened in the ‘social sciences’. For example, it can be argued that the ‘resource-based-view’ (RBV) theory may be placed in ‘management studies’ in a broader context or in ‘economics’ under ‘social sciences’. Moreover, both sides/cases have compelling logic. Even the discipline of ‘management studies’ could be treated as a ‘social science’ subject rather than a separate discipline. This aspect may be one of the drawbacks of this study.

The assessment of the past studies by different scholars has criticised LIS for its lack of theoretical research and may be trending downward. Regarding theory, Orlikowski and Iacono (2001) argued that LIS research is under-theorised. Rayward (2004) also supports this view. It is a hard fact that the LIS domain has yet to produce grand theories, and LIS researchers have used theories in their studies mainly to frame, design, and interpret findings. Many commentators have also suggested that the LIS domain needs to make more use of theory (Pinfield et al., 2021) and has thus been criticised for relying on theories imported from other disciplines rather than applying or developing theories from within (Park et al., 2022). The same view was also expressed by McKechnie and Pettigrew (2002). Although Kim (2004) disagreed with previous studies that reported that theoretical research was insufficient in LIS research. He found that 41.4% of the studies dealt with theoretical development and utilisation.

Still, research in LIS is confined to theories developed and used in complementary disciplines. In this study, it was found that, out of 411 theories, only 45 were from the LIS domain, and the majority of the theories (366) were from adjacent disciplines. This is due to the lack of sound or home-based theories of LIS. Even the home-based theories are pragmatic and descriptive. This lack of theoretical contributions may be associated with the fact that LIS emanates from professional practice and is therefore closely linked to practical problems such as the processing and organisation of library materials, documentation, and information retrieval (Jarvelin & Vakkari, 1990; Kim & Jeong, 2006). As previously stated, LIS has borrowed many theories from other disciplines due to a scarcity of theorists and theory-illuminated practitioners (Schrader, 1986).

The authors think that there is considerable scope for further research in this area of theorising LIS research and identified mainly two areas: the development, use, and application of home-based theory; and general conceptual and methodological awareness among LIS researchers regarding theories, particularly the specific use of the meta-theoretical assumptions inherent in LIS research. However, the advancement of information technology and the use of such technologies in the library environment have significantly contributed to the domain of LIS research. Feather (2008) urged LIS researchers to prioritise various aspects of LIS research. He correctly stated that now is the time for LIS researchers to engage in and focus on several important issues that are more important in the LIS domain than today’s management theories or tomorrow’s technological miracle.

However, many articles are published, and many practical works are done without explicating any theoretical or meta-theoretical assumptions. These works were not tested for how the term ‘theory’ is operationalised in the study, at what level it is used or generalised, or how the theory is proposed to operate within the study. In most cases, theories were discussed and talked about marginally or minimally. Freehan et al. (1987) correctly observed that LIS research had not matured sufficiently to support a cohesive body of its theoretical foundation and was instead built on theories from other disciplines (Gregor, 2006). However, the authors disagree with previous researchers who have treated our field as under-theorised because more than forty-five LIS theories have been identified (see Appendix C). Moreover, we fully support the view of Gregor (2006), who rightly realised the potential benefits of using theories in the LIS domain by saying that LIS professionals use both home-based and borrowed theories in a new look/way to make sense of their data.

The authors claim that LIS is facing a ‘theory crisis’ and is still in a grey area. Existing findings of our study focus on essential insights for the LIS research community, which raises different questions about LIS research. Does this mean that LIS research has no systematic theoretical base? Do we have our own or native theories? Or to what extent does LIS depend on other neighbouring disciplines for its theoretical base? How will LIS respond to this unmet situation? What will be at the forefront of LIS research in the near future?

The process of theory building in LIS was not so strong and is evident through the tremendous borrowing of various theoretical concepts and the use of theoretical frameworks from other disciplines to address various LIS issues. Rawson & Hughes-Hassell (2015) rightly opined that the treatment of theory in LIS research covers a spectrum of intensity, from marginal mentions to theory revising, expanding, or building. He further stated that the field of LIS has not been very successful in contributing to existing theory or producing new theory. In spite of this, one may still assert that LIS research employs theory, and, in fact, there are many theories that have been used or generated by LIS scholars. However, “calls for additional and novel theory development work in LIS continue, particularly for theories that might help to address the research practice gap”.

As stated, using theory in research is essential as it helps produce transformative knowledge (Ngulube, 2020). As is evident in this study, the results focus on the fact that LIS researchers could not develop home-based theories and make good use of social science theories due to its (LIS) interdisciplinary nature or being transdisciplinary. As a result of the absence of its own or discipline-based theories, researchers have borrowed from other disciplines (see Appendix B). However, it would be improper to say that LIS research is conducted in collaboration but is widely connected with other disciplines. Even so, authors have noted a decreasing trend among LIS researchers to use existing theories rather than develop new ones. As there is no “best theory” (Weick, 1985), we should develop and improve theory-building skills (e.g. theorising of theories) to develop more home-based theories (Weick, 1989). Doherty (2012) rightly said that theory-building research is important to ensure the development of LIS research. LIS professionals should invest more in theory building and conducting such studies on the trends of theoretical and conceptual frameworks in LIS research. Researchers should focus more on the theoretical ties between LIS research and research in other neighbouring disciplines, and a meta-analysis of LIS theories could be a solution. Boyce and Kraft (1985) and Buckland (1991) have suggested a strictly defined standard would not allow for the finding of many theories, even those considered theories within the bounds of LIS, because theories in this field may have had the status of “quasi-theories” (Boyce & Kraft, 1985). Moreover, the lack of a clear road map for theory development in LIS makes the process ‘one of the most frustrating and arduous tasks in which a scholar engages’ (Cunningham, 2013).

Active Learning for Systematic Reviews (ASReview) is a free (Libre) open-source machine learning tool for screening and systematically labeling a large collection of textual data. It is designed to accelerate the step of screening abstracts and titles with a minimum of papers to be read by a human with no or very few false negatives. It employs a machine learning technique known as active learning to help with screening for efficient and transparent systematic reviews for academia and beyond. The goal of ASReview is to help scholars and practitioners to get an overview of the most relevant papers for their work as efficiently as possible, while being transparent in the process.

OpenRefine is a free, open source, standalone Java application which visualises and manipulates large quantities of data all at once. It is used for exploring, cleaning, linking and transforming data (an activity commonly known as data wrangling) on a large scale. The functions include are data normalisation, column reorganisation, faceting/clustering, tracking operations, exporting data and so on. It is similar to spreadsheet applications, and can handle spreadsheet file formats such as CSV, but it behaves more like a database. It is more powerful than a spreadsheet; more interactive and visual than scripting; more provisional/exploratory/experimental/playful than a database.

Scopus is an abstracting, indexing and citation database with enriched data and linked scholarly literature across a wide variety of disciplines. It indexes content that is rigorously vetted and selected by an independent review board of experts in their fields. With comprehensive content coverage, high-quality data, and precise search and analytical tools. Scopus gives researchers, librarians, research managers, and R&D professionals the insights to drive better decisions, actions, and outcomes. It empowers users to discover critical information, monitor trends, and identify subject matter experts. It also helps users visualise, compare and export data to evaluate research output and trends.